Practice Guidelines Chest Physicians Evidence-Based Clinical Thrombosis, 9Th Ed: American College of Antithrombotic Therapy

Home , Thrombosis prevention

Introduction to the Ninth Edition : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Gordon H. Guyatt, Elie A. Akl, Mark Crowther, Holger J. Schünemann, David D. Gutterman and Sandra Zelman Lewis

Chest 2012;141;48S-52S DOI 10.1378/chest.11-2286

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/141/2_suppl/48S.full.html

Chest is the official journal of the American College of Chest Physicians. It has been published monthly since 1935. Copyright2012by the American College of Chest Physicians, 3300 Dundee Road, Northbrook, IL 60062. All rights reserved. No part of this article or PDF may be reproduced or distributed without the prior written permission of the copyright holder. (http://chestjournal.chestpubs.org/site/misc/reprints.xhtml) ISSN:0012-3692

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians CHEST Supplement ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: ACCP GUIDELINES Introduction to the Ninth Edition Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Gordon H. Guyatt , MD , FCCP ; Elie A. Akl , MD , PhD, MPH ; Mark Crowther , MD ; Holger J. Schünemann , MD , PhD , FCCP ; David D. Gutterman , MD , FCCP ; and Sandra Zelman Lewis , PhD

The Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines differs substantially from the prior versions both in process and in content. In this introduction, we describe some of the differences and the rationale for the changes. CHEST 2012; 141(2)(Suppl):48S–52S

Abbreviations: A C C P 5 American College of Chest Physicians; AT6 5 Sixth ACCP Consensus Conference on Antithrombotic Therapy; AT7 5 Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy: Evidence- Based Guidelines; AT8 5 Antithrombotic and Thrombolytic Therapy: American College of Chest Physicians Evidence- Based Clinical Practice Guidelines (8th Edition); AT9 5 Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines; GRADE 5 Grades of Recommen- dations, Assessment, Development, and Evaluation

e would like to begin by acknowledging the and researcher, while President of the American W contributions of the visionaries whose work on College of Chest Physicians (ACCP), had the foresight past editions of these guidelines have allowed the not only to propose the original consensus conference current panel to develop this edition using the changes on the controversial issues of the indications for anti- noted herein. First, James E. Dalen, respected clinician thrombotics, antiplatelet agents, and thrombolytics for the prevention and treatment of cardiovascular Revision accepted August 31, 2011. disorders but also to invite Jack Hirsh, an extremely Affi liations: From the Department of Clinical Epidemiology and Biostatistics (Drs Guyatt, Akl, and Schünemann) and Depart- productive scientist and leader in the fi eld of throm- ment of Medicine (Drs Guyatt, Crowther, and Schünemann), bosis, to join him in leading this important project. McMaster University Faculty of Health Sciences, Hamilton, ON, Drs Hirsh and Dalen brought a panel of leading Canada; Departments of Medicine and Family Medicine (Dr Akl), State University of New York, Buffalo, NY; Cardiovascular experts together for the fi rst antithrombotic guide- Research Center (Dr Gutterman), Medical College of Wisconsin, line1 in 1986. Dr Dalen was co-editor of the fi rst six Milwaukee, WI; and Evidence-Based Clinical Practice Guidelines and Clinical Standards (Dr Lewis), American College of Chest guidelines from 1986 to 2001. Physicians, Northbrook, IL. Dr Hirsh is, to an extraordinary extent, responsible Funding/Support: The Antithrombotic Therapy and Prevention not only for creating the platform that has made Anti- of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines received support from thrombotic Therapy and Prevention of Thrombo- the National Heart, Lung, and Blood Institute [R13 HL104758] sis, 9th ed: American College of Chest Physicians and Bayer Schering Pharma AG. Support in the form of educational grants was also provided by Bristol-Myers Squibb; Pfi zer, Correspondence to: Gordon H. Guyatt, MD, FCCP, Department Inc; Canyon Pharmaceuticals; and sanofi -aventis US. of Clinical Epidemiology and Biostatistics, McMaster University, Disclaimer: American College of Chest Physician guidelines Hamilton, ON, L8N 3Z5, Canada; e-mail: [email protected] are intended for general information only, are not medical advice, © 2012 American College of Chest Physicians. Reproduction and do not replace professional medical care and physician of this article is prohibited without written permission from the advice, which always should be sought for any medical condition. American College of Chest Physicians ( http://www.chestpubs.org/ The complete disclaimer for this guideline can be accessed at site/misc/reprints.xhtml ). http://chestjournal.chestpubs.org/content/141/2_suppl/1S DOI: 10.1378/chest.11-2286

48S Introduction to the Ninth Edition

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (AT9) Table 1— Major Innovations in AT9 possible but also for the advances from Antithrom- 1. Unconfl icted methodologists as topic editors. Confl icted experts botic and Thrombolytic Therapy: American College did not participate in fi nal process of making recommendations. of Chest Physicians Evidence-Based Clinical Practice 2. Many evidence profi le and summary of fi nding tables. Guidelines (8th Edition) (AT8) to AT9. It was not 3. New insights into evidence (asymptomatic thrombosis, aspirin). only the creation of an expert panel—a standard fea- 4. Quantitative specifi cation of values and preferences based on ture of specialty society clinical practice guidelines systematic review of relevant evidence and formal preference from the outset—that made the antithrombotic rating exercise. guidelines extremely successful. Dr Hirsh, an accom- 5. Article addressing diagnosis of DVT. plished basic science researcher and a brilliant clin- AT9 5 Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: ical trialist, deeply understood the value of what was American College of Chest Physicians Evidence-Based Clinical Practice to become more than a decade later evidence-based Guidelines. medicine. He recruited one of the world leaders in the burgeoning fi eld of clinical epidemiology, David Sackett, to play a major role in the guidelines. responsibility for each article not to a thrombosis Drs Hirsh and Sackett developed and applied an expert but to a methodologist who, in almost all cases, innovative system of rating the quality of evidence and also is a practicing physician without important con- strength of recommendations that was at the time fl icts of interest 4,5 ( Table 1 ). These editors of each unique to specialty guidelines. The combined impact article had specifi c training in ACCP’s approach of the authoritative, carefully considered recommen- to rating the quality of evidence and grading strength dations and explicit acknowledgment of the quality of of recommendations (see “Evidence Summaries” evidence and strength of recommendations immedi- section). Further, building on the work of prior guide- ately made these guidelines the reference standard line groups, 6 the process stipulated that although for antithrombotic therapy around the world. confl icted thrombosis experts could engage in discus- Under Dr Hirsh’s leadership, the guidelines more sion and even draft evidence summaries, they would than kept pace with advances in the science of guide- be excluded from the fi nal decisions regard ing quality of line methodology and continued to improve with evidence and direction and strength of recommenda- each iteration. Each new edition provided a model tions. Intellectual confl icts received the same atten- incorporating not only the latest evidence regarding tion as fi nancial confl icts. Readers of the guidelines antithrombotic therapy but also advances in special- can fi nd in the online data supplements to AT9 articles ty-based guidelines and therefore maintained preem- a recommendation-by-recommendation accounting inence in the fi eld of thrombosis. of the intellectual and fi nancial confl icts of each panel As Dr Hirsh was stepping down from the leader- member. The most important changes in AT9 con- ship of the guidelines in 2007, his insight led him to tent have fl owed directly from this change in process. question the reliance on expert opinion that provided the basis for the fi rst eight iterations of the anti- Evidence Summaries thrombotic guidelines. Reviewing his experience—and in the process giving new life to an idea suggested The Sixth ACCP Consensus Conference on Anti- decades earlier but seldom applied2 —Dr Hirsh con- thrombotic Therapy (AT6), 7 published in 2001, cluded that the confl ict of interest of leading experts adopted an approach to rating quality of evidence was highly problematic.3 Furthermore, the problem and strength of recommendations 8 that presaged arose not only from their fi nancial but equally or per- that of the Grades of Recommendations, Assessment, haps more important, their intellectual confl ict of Development, and Evaluation (GRADE) working interest. This revelation and the changes in process group,9 itself adopted with minor modifi cations for that Dr Hirsh suggested as a solution to the problem AT8 10 and for all ACCP guidelines.11 AT8, like AT6 had a profound impact on the leadership to whom he and the Seventh ACCP Conference on Antithrom- was passing the torch. These changes underlie all the botic and Thrombolytic Therapy: Evidence-Based innovations in AT9. Guidelines (AT7),12 underwent thorough editing and review of the quality of evidence, including in many cases a careful assessment of single studies by topic The New Process in AT9—Dealing editors and the guideline executive committee. With Conflict of Interest Nevertheless, many of the topic editors did not have methodologic training or, as was the case in AT8, spe- The solution that Dr Hirsh proposed was endorsed cifi c training in the ACCP-GRADE approach. The by the ACCP leadership and implemented in AT9. result was that in AT8, the ACCP-GRADE approach That solution is to give primary leadership and often was not applied with optimal rigor, and authors www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 49S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians produced very few of the summary tables that are the trials—asymptomatic, screening-detected throm- hallmark end product of the GRADE process. bosis. Use of this surrogate creates major problems There are two types of such tables: evidence pro- in making the trade-off between patient-important fi les and summary of fi ndings. Evidence profi les sum- outcomes (thrombosis and serious bleeding). For marize the quality of a body of evidence for each instance, if one knows that an intervention increases relevant outcome and, when evidence comes from serious bleeding by 20 events in 1,000 patients but randomized trials, include a presentation of reviewers’ reduces asymptomatic thrombosis by 100 in 1,000, assessment of risk of bias, precision, consistency, what is the net benefi t? Establishing net benefi t in directness, and publication bias.5 Readers of AT9 can outcomes important to patients requires knowing fi nd the evidence profi les in the online data supple- the symptomatic DVT and symptomatic pulmo- ments. The text in the main AT9 articles includes nary embolism reduction represented by the reduc- the more succinct summary of fi ndings tables, which tion in 100 asymptomatic events. Dr Hirsh and include the overall quality assessment as well as John Eikelboom led the prevention topic editors in relative and absolute effect sizes for each outcome. developing innovative approaches to dealing with We did not succeed in the lofty goal of producing a the problem of inferring the impact of thrombopro- summary of fi ndings table for each recommendation, phylaxis on symptomatic thrombosis from studies but readers will fi nd these for most major and poten- that relied to a considerable extent on detection of tially controversial recommendations. Readers will asymptomatic events.13 The available approaches, fi nd . 10 such tables in most articles and . 20 in although representing a step forward, all have limita- some articles. tions and highlight the need for studies that directly Producing these tables forces a rigor of thinking measure symptomatic thrombosis without veno- achievable in few other ways. Creating a large number graphic or ultrasound surveillance. of evidence profi les provides deep insight into the As a result of a reevaluation led by Dr Eikelboom, ACCP-GRADE approach to assessing the quality of one consequence of the recognition that measure- evidence and strength of recommendations. Along ment of patient-important events in a naturalistic with recruitment of GRADE-expert topic editors, clinical setting (as opposed to in the context of veno- production of evidence profi les and summary of fi nd- graphic or ultrasonographic screening for asymptom- ings tables is responsible for the increased method- atic thrombosis) was a differing perspective on the ologic rigor of AT9. use of aspirin in thromboprophylaxis in orthopedic We also tried to apply a rigorous approach to surgery. The authors of AT8 had concluded that there choosing the format of these tables, an issue that has was high-quality evidence justifying a strong recom- generated some controversy within the GRADE work- mendation against aspirin as the sole agent for throm- ing group. Per Olav Vandvik, Holger Schünemann, boprophylaxis in surgical patients. Authors of AT9 and Nancy Santesso led the group in a formal study focused on results from a very large trial using con- in which AT9 panelists expressed their view of the cealed randomization and blinding and achieving optimal presentation of the tables.5 The results have near-complete follow-up.14 After an exhaustive and not only guided the presentation of evidence profi les repeated discussion involving the authors of all the and summary of fi ndings in AT9 but also provided prevention articles, and ultimately the entire panel, one of the recommended options for the GRADE AT9 authors concluded that the trial provides moder- working group. ate-quality evidence supporting the use of aspirin, which is now offered as an option for thromboprophylaxis in patients undergoing major orthopedic 15 Reevaluation of Evidence procedures. AT9 authors concluded that there is low-quality evidence supporting a weak recommen- Relying on the perspective of unconfl icted meth- dation of low-molecular-weight heparin over aspirin odologists, rigorously applying the GRADE approach, in these patients. and excluding those with fi nancial and intellectual The GRADE approach defi nes quality of evidence confl ict of interests from bottom-line decisions as our level of confi dence in estimates of diagnostic regarding the quality of evidence and strength of or treatment effect to support a particular recom- recommendations led to reevaluations of previously mendation.16 In general, the changing perspectives existing evidence ( Table 1 ). For instance, application on evidence led to the conclusion that we often could of the ACCP-GRADE approach requires the distinc- not be as confi dent in estimates of effect as previously tion between patient-important and surrogate out- believed. Readers of AT9 will often fi nd, therefore, comes. The fi rst eight editions of the antithrombotic that some of the evidence previously rated as high guidelines failed to fully recognize the implications quality is now moderate, and evidence previously of a surrogate widely used in thrombosis prevention rated as moderate quality is now low.

50S Introduction to the Ninth Edition

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Values and Preferences environment and to support efforts to make the phrasing of recommendations more user friendly and Serial iterations of AT9 dating back to AT6 have implementable. gradually put increasing emphasis on patient values A limitation of AT8 was the very inconsistent and preferences. For the fi rst time, the AT7 panel approaches to assessing bleeding risk. Sam Schulman, made the assumptions about values and preferences author of the bleeding risk article in AT8, 19 took explicit. The AT9 panel has accelerated that process responsibility for developing the AT9 approach to by conducting a systematic review17 of the relevant bleeding and ensuring that it was consistently applied research of empirical investigations of values and pref- across chapters.5 erences of patients regarding antithrombotic therapy. To address issues of economic effi ciency, we Based on that review, AT9 panelists conducted a included “resource use consultants” on the AT9 article value rating exercise that provided the basis for values panels charged with making recommendations. They and preference judgments within AT9, judgments developed a transparent and systematic methodology that are summarized in the introductory section of to address questions for which resource use might each article.13 The judgments are more explicit and change the direction or strength of recommendations.5 quantitative than any previous guideline. For example, We made an intensive effort to remove duplication we estimated that on average, patients experience the between articles and to harmonize recommendations disutility of a GI bleed more or less equally to that of between related articles. An important strategy was VTE but only one-third of the disutility of a stroke. to include topic editors and deputy editors as panel- Among the fi ndings of the systematic review of ists from both articles when two had overlapping patient values and preferences regarding antithrom- issues. botic treatment are the heterogeneity of results Finally, for the fi rst time, the antithrombotic guide- between studies and the wide variability in values and lines have addressed issues of diagnosis. Shannon Bates preferences among patients. Because the core char- and Roman Jaeschke led a panel that took on the acteristic of a strong recommendation is the belief challenging task of applying the newly developed that across the range of values and preferences, virtu- GRADE methodology for recommendations regarding ally all informed patients will make the same choice, the diagnosis of DVT.20 the wide variability in patient values and preferences makes strong recommendations less likely.18 Conclusion Impact of Innovations on Building on the seminal work of Drs Hirsh and the Recommendations Dalen and their colleagues over the 20-year history of Readers of AT9 will fi nd many weak recommenda- the ACCP antithrombotic guidelines, AT9 has made tions replacing the strong recommendations of AT8. a number of changes in process, resulting in differ- One major reason for this change is the more critical ences in the approach to making recommendations look at the evidence and the resulting inferences that and their content. Past iterations of these guidelines some evidence is lower quality than previously believed. have celebrated new high-quality evidence and the A second is the recognition of variability in values and strong recommendations that such evidence war- preferences. Third, in the small number of controversial rants. The insights from AT9 include the persisting recommendations that came to a formal vote using limitations in evidence quality (particularly with anonymous electronic voting, we required the endorse- respect to the use of surrogate outcomes in prophy- ment of . 80% of panelists to make a strong recom- laxis trials) and the appropriateness of weak recommendation. Finally, the exclusion of confl icted experts, mendations that refl ect our lack of confi dence in who often hold strong opinions about optimal man- effect estimates and the variability in patient values agement approaches, from fi nal decisions regarding and preferences. We believe that the objective, rig- quality of evidence and strength of recommendations orous application of the science of guideline develop- also may have contributed. ment will ultimately best serve our patients.

Other Innovations in AT9 Acknowledgments Financial/nonfi nancial disclosures: In summary, the authors For the fi rst time, each panel included a frontline have reported to CHEST the following confl icts of interest: clinician not involved in thrombosis research. The Dr Crowther has served on various advisory boards, has assisted goal of including these individuals was to ensure that in the preparation of educational materials, and has sat on data safety and monitoring boards. His institution has received research recommendations considered the full realities of clin- funds from the following companies: Leo Pharma A/S, Pfi zer Inc, ical practice as viewed by those outside the research Boerhinger Ingelheim GmbH, Bayer Healthcare Pharmaceuticals, www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 51S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Octapharm AG, CSL Behring, and Artisan Pharma. Personal total 9 . Guyatt GH , Oxman AD , Vist GE , et al ; GRADE Working compensation for these activities over the past 3 years totals less Group . GRADE: an emerging consensus on rating quality than US $10,000. Dr Gutterman has had the following relation- of evidence and strength of recommendations . BMJ . 2008 ; ships that are entirely unrelated to the AT9 guidelines: ACCP 336 ( 7650 ): 924 - 926 . President, GlaxoSmithKline plc grant to study vasodilation in adipose tissue , National Institutes of Health grant to study human 10 . Guyatt GH , Cook DJ , Jaeschke R , Pauker SG , Schünemann coronary dilation, and GE Healthcare consultation on a study HJ . Grades of recommendation for antithrombotic agents: for ECG evaluation of chronic heart disease. Drs Guyatt and American College of Chest Physicians evidence-based clinical Schünemann are co-chairs of the GRADE Working Group, and practice guidelines (8th edition). Chest . 2 0 0 8 ; 1 3 3 ( s u p p l 6 ) : Dr Akl is a member and prominent contributor to the GRADE 123S - 131S . Working Group. Dr Lewis is a full-time employee of the ACCP . 11 . Guyatt G , Gutterman D , Baumann MH , et al . Grading Role of sponsors: The sponsors played no role in the develop- strength of recommendations and quality of evidence in clin- ment of these guidelines. Sponsoring organizations cannot ical guidelines: report from an American College of Chest recommend panelists or topics, nor are they allowed prepublication access to the manuscripts and recommendations. Guideline Physicians task force . Chest . 2006 ; 129 ( 1 ): 174 - 181 . panel members, including the chair, and members of the Health & 12 . American College of Chest Physicians . The seventh ACCP Science Policy Committee are blinded to the funding sources. conference on antithrombotic and thrombolytic therapy: Further details on the Confl ict of Interest Policy are available evidence-based guidelines . Chest . 2004 ; 126 ( suppl 3 ): 1S - 703S . online at http://chestnet.org. 13. Guyatt GH , Eikelboom JW , Gould MK , et al. Approach to Other contributions: The authors thank Rachel Gutterman, BA, outcome measurement in the prevention of thrombosis in and Joe Ornelas, DC, for their assistance in managing this complex project. surgical and medical patients: antithrombotic therapy and Endorsements: This guideline is endorsed by the American prevention of thrombosis, 9th ed: American College of Chest Association for Clinical Chemistry, the American College of Clinical Physicians evidence-based clinical practice guidelines. Chest . Pharmacy, the American Society of Health-System Pharmacists, 2012 ;141(2)(suppl):e185S-e194S. the American Society of Hematology, and the International Society 14 . Pulmonary Embolism Prevention (PEP) trial Collaborative of Thrombosis and Hematosis. Group. Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: Pulmonary Embolism References Prevention (PEP) trial . Lancet . 2000 ; 355 ( 9212 ): 1295 - 1302 . 15 . Eikelboom JW , Hirsh J , Spencer FA , Baglin TP, Weitz JI . 1 . American College of Chest Physicians; National Heart, Lung, Antiplatelet drugs: antithrombotic therapy and prevention of and Blood Institute . ACCP-NHLBI National Conference on throm bosis, 9th ed: American College of Chest Physicians Antithrombotic Therapy . Chest . 1986 ; 89 ( suppl 2 ): 1S - 106S . evidence-based clinical practice guidelines. Chest . 2 0 1 2 ; 2 . Battista RN . The wizard of WOS—or the art of the task force . 141(2)(suppl):e89S-e119S. J Clin Epidemiol . 2009 ; 62 ( 6 ): 571 - 573 . 16 . Guyatt GH , Oxman AD , Kunz R , Vist GE , Falck-Ytter Y , 3 . Hirsh J , Guyatt G . Clinical experts or methodologists to write Schünemann HJ ; GRADE Working Group . What is “quality clinical guidelines? Lancet . 2009 ; 374 ( 9686 ): 273 - 275 . of evidence” and why is it important to clinicians? BMJ . 4 . Guyatt G , Akl EA , Hirsh J , et al . The vexing problem of 2008 ; 336 ( 7651 ): 995 - 998 . guidelines and confl ict of interest: a potential solution . Ann Intern Med . 2010 ; 152 ( 11 ): 738 - 741 . 17. MacLean S , Mulla S , Akl EA , et al. Patient values and pref- 5 . Guyatt GH , Norris SL , Schulman S , et al . Methodology for erences in decision making for antithrombotic therapy: a the development of antithrombotic therapy and prevention systematic review: antithrombotic therapy and prevention of of thrombosis guidelines: antithrombotic therapy and pre- thrombosis, 9th ed: American College of Chest Physicians vention of thrombosis, 9th ed: American College of Chest evidence-based clinical practice guidelines. Chest . 2 0 1 2 ; Physicians evidence-based clinical practice guidelines . Chest . 141(2)(suppl):e1S-e23S. 2012 ;141(2)(suppl):53S-70S. 18 . Guyatt GH , Oxman AD , Kunz R , et al ; GRADE Working 6. Schünemann H , Osborne M , Moss J , et al. An offi cial Group . Going from evidence to recommendations . BMJ . American Thoracic Society Policy Statement: managing con- 2008 ; 336 ( 7652 ): 1049 - 1051 . fl ict of interest in professional societies. Am J Respir Crit 19 . Schulman S , Beyth RJ , Kearon C , Levine MN ; American Care Med . 2009 ;180(6):564-580. College of Chest Physicians . Hemorrhagic complications of 7 . American College of Chest Physicians . Sixth ACCP consensus anticoagulant and thrombolytic treatment: American College conference on antithrombotic therapy . Chest . 2 0 0 1 ; 1 1 9 ( s u p p l 1 ) : of Chest Physicians evidence-based clinical practice guide- 1S - 370S . lines (8th edition). Chest . 2008 ; 133 ( suppl 6 ): 257S - 298S . 8 . Guyatt G , Schunëmann H , Cook D , Jaeschke R , Pauker S , 20 . Bates SM , Jaeschke R , Stevens SM , et al . Diagnosis of DVT: Bucher H ; American College of Chest Physicians . Grades of antithrombotic therapy and prevention of thrombosis, 9th ed: recommendation for antithrombotic agents. Chest . 2 0 0 1 ; American College of Chest Physicians evidence-based clinical 119 ( suppl 1 ): 3S - 7S . practice guidelines . Chest . 2012 ;141(2)(suppl):e351S-e418S.

52S Introduction to the Ninth Edition

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Introduction to the Ninth Edition : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Gordon H. Guyatt, Elie A. Akl, Mark Crowther, Holger J. Schünemann, David D. Gutterman and Sandra Zelman Lewis Chest 2012;141; 48S-52S DOI 10.1378/chest.11-2286

This information is current as of February 7, 2012

Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/141/2_suppl/48S.full.html References This article cites 20 articles, 16 of which can be accessed free at: http://chestjournal.chestpubs.org/content/141/2_suppl/48S.full.html#ref-list-1 Cited Bys This article has been cited by 2 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/141/2_suppl/48S.full.html#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Gordon H. Guyatt, Elie A. Akl, Mark Crowther, David D. Gutterman, Holger J. Schuünemann and for the American College of Chest Physicians Antithrombotic Therapy and Prevention of Thrombosis Panel

Chest 2012;141;7S-47S DOI 10.1378/chest.1412S3

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/141/2_suppl/7S.full.html

Supplemental material related to this article is available at: http://chestjournal.chestpubs.org/content/suppl/2012/02/06/141.2_suppl. 7S.DC1.html

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians CHEST Supplement ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: ACCP GUIDELINES Executive Summary Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Gordon H. Guyatt , MD, FCCP ; Elie A. Akl , MD, PhD, MPH ; Mark Crowther , MD ; D a v i d D . G u t t e r m a n , M D , F C C P ; H o l g e r J . S c h ü n e m a n n , M D , P h D , F C C P ; f o r t h e American College of Chest Physicians Antithrombotic Therapy and Prevention of Thrombosis Panel*

CHEST 2012; 141(2)(Suppl):7S–47S evidence, and some articles with quite extensive summary tables of primary studies. In total, this Abbreviations: A C S 5 acute coronary syndrome; AF 5 atrial represented 600 recommendations summarized in fi brillation; AIS 5 arterial ischemic stroke; APLA 5 antiphospolipid 968 pages of text. Many readers responded that the antibodies; ASA 5 acetylsalicylic acid; AT9 5 Antithrombotic result was too voluminous for their liking or prac- Therapy and Prevention of Thrombosis, 9th ed: American tical use. College of Chest Physicians Evidence-Based Clinical Practice Cognizant of this feedback, we worked hard to Guidelines; BMS 5 bare-metal stent; CABG 5 coronary artery bypass graft; CAD 5 c o r o n a r y a r t e r y disease; CDT 5 catheter- minimize the length of the text for the ninth iteration directed thrombosis; CHADS 2 5 congestive heart failure, hyperten- of the guidelines Antithrombotic Therapy and Pre- sion, age Ն 75 years, diabetes mellitus, prior stroke or transient vention of Thrombosis, 9th ed: American College of ischemic attack; CSVT 5 cerebral sinovenous thrombosis; CTPH 5 Chest Physicians Evidence-Based Clinical Practice chronic thromboembolic pulmonary hypertension; CUS 5 com- Guidelines (AT9) without sacrifi cing key content. A pression ultrasound; CVAD 5 central venous access device; DES 5 drug-eluting stent; GCS 5 graduated compression stockings; number of topic editors found our shortening edits HFS 5 hip fracture surgery; HIT 5 heparin-induced thrombocy- draconian, but we were determined to produce the topenia; HITT 5 heparin-induced thrombocytopenia complicated leanest product possible. by thrombosis; IA 5 intraarterial; ICH 5 intracerebral hemor- There were, however, a number of obstacles. In rhage; IE 5 infective endocarditis; INR 5 international normalized what we believe is a key advance in AT9, we con- ratio; IPC 5 intermittent pneumatic compression; IPCD 5 intermittent pneumatic compression device; IVC 5 inferior vena cava; ducted a systematic review of what is known about LDUH 5 low-dose unfractionated heparin; LMWH 5 low-molecular- patients’ values and preferences regarding antithrom- weight heparin; LV 5 left ventricular; MBTS 5 modifi ed Blalock- botic therapy and included the results as an article Taussig shunt; MR 5 magnetic resonance; PAD 5 peripheral artery in AT9. In another forward step, we recognized the disease; PCI 5 percutaneous coronary inter vention; PE 5 pul- problems with asymptomatic thrombosis as a surro- monary embolism; PFO 5 patent foramen ovale; PMBV 5 percuta - neous mitral balloon valvotomy; PTS 5 postthrombotic syndrome; gate outcome, and devised strategies to estimate PVT 5 prosthetic valve thrombosis; r-tPA 5 recombinant tissue plas- reductions in symptomatic DVT and pulmonary minogen activator; RVT 5 renal vein thrombosis; SC 5 subcuta- embolism with antithrombotic prophylaxis. We felt it neous; TEE 5 transesophageal echocardiography; THA 5 total important to explain this innovation to users of AT9, hip arthroplasty; TIA 5 transient ischemic attack; TKA 5 total knee and this meant another article. arthroplasty; UAC 5 umbilical arterial catheter; UEDVT 5 upper- extremity DVT; UFH 5 unfractionated heparin; US 5 ultrasound; We included, for the fi rst time, an article on diag- UVC 5 umbilical venous catheter; VAD 5 ventricular assist device; nosis addressing patients with symptoms and signs VKA 5 vitamin K antagonist suggesting DVT. We increased the range of interventions we have covered, resulting in additional recommendations. Finally, we produced many summary he eighth iteration of the American College of of fi ndings tables, which offer extremely succinct and T Chest Physicians Antithrombotic Guidelines pre- informative presentations of best estimates of effect sented, in a paper version, a narrative evidence sum- and the confi dence associated with those estimates. mary and rationale for the recommendations, a small If published in the same fashion as the Antithrom- number of evidence profi les summarizing bodies of botic and Thrombolytic Therapy, 8th ed: American www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 7S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians College of Chest Physicians Antithrombotic Guide- tables) and some tables summarizing the methods lines, this would have resulted in a document and results, and the risk of bias, associated with the with . 850 pages of paper text, an unacceptable individual studies that contributed to the evidence length. Given this and with the advice of the journal, profi les and summary of fi ndings tables. we decided to adopt a highly focused print version The world of medical information is rapidly becom- that includes only this executive summary and the ing a world of electronic storage and presentation following articles: of primary studies, recommendations, and a wide variety of other information of interest to health care • An introduction describing the major innovations prac titioners. Although our abbreviated paper copy in AT9 presentation represents a necessary response to a • A methods article explaining how we devel- challenging situation, it is also a harbinger of the oped the guidelines (a potential model for other increasingly electronic world of medical information guideline groups interested in optimal rigor) into which future editions of guidelines are destined • Recommendations and grading from each arti- to move. cle embedded in the table of contents of each article Summary of Recommendations Those seeking the rationale for the recommenda- Note on Shaded Text: Throughout this guideline, tions, including the supporting evidence, should shading is used within the summary of recommenda- access the online version of the guideline (http:// tions sections to indicate recommendations that are http://chestjournal.chestpubs.org/content/141/2_suppl ) newly added or have been changed since the publica- that includes a narrative summaries and support- tion of Antithrombotic and Thrombolytic Therapy: ing summary of fi ndings tables. The numbering indi- American College of Chest Physicians Evidence- cated beside the recommendations in this summary Based Clinical Practice Guidelines (8th Edition). Rec- is aligned with the sections and tables found in the ommendations that remain unchanged are not shaded. full articles. Those interested in a deeper understanding of the evidence can turn to online data supplements for each of the articles that include rec- Evidence-Based Management of ommendations. There, they will fi nd evidence pro- Anticoagulant Therapy fi les (expanded versions of the summary of fi ndings For further details, see Holbrook et al.1

Revision accepted August 31, 2011. 2.1 Loading Dose for Initiation of Vitamin K Antagonist Affi liations: From the Department of Clinical Epidemiology (VKA) Therapy and Biostatistics (Drs Guyatt, Akl, and Schü nemann) and Depart- ment of Medicine (Drs Guyatt, Crowther, and Schü nemann), 2.1. For patients suffi ciently healthy to be McMaster University Faculty of Health Sciences, Hamilton, treated as outpatients, we suggest initiating VKA ON, Canada; Departments of Medicine and Family Medicine (Dr Akl), State University of New York, Buffalo, NY; Cardiovascular therapy with warfarin 10 mg daily for the fi rst Research Center (Dr Gutterman), Medical College of Wisconsin, 2 days followed by dosing based on international Milwaukee, WI. normalized ratio (INR) measurements rather than *For complete panel list, see: http://chestjournal.chestpubs.org/ content/141/2_suppl/2S starting with the estimated maintenance dose Funding/Support : The Antithrombotic Therapy and Prevention (Grade 2C) . of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines received support from 2.2 Initial Dose Selection and Pharmacogenetic the National Heart, Lung, and Blood Institute [R13 HL104758] Testing and Bayer Schering Pharma AG. Support in the form of educational grants were also provided by Bristol-Myers Squibb; Pfi zer, 2.2. For patients initiating VKA therapy, we Inc; Canyon Pharmaceuticals; and sanofi -aventis US. recommend against the routine use of pharma- Disclaimer: American College of Chest Physician guidelines are intended for general information only, are not medical advice, cogenetic testing for guiding doses of VKA and do not replace professional medical care and physician advice, (Grade 1B). which always should be sought for any medical condition. The complete disclaimer for this guideline can be accessed at http:// 2.3 Initiation Overlap for Heparin and VKA chestjournal.chestpubs.org/content/141/2_suppl/1S Correspondence to: Gordon H. Guyatt, MD, FCCP, Department 2.3. For patients with acute VTE, we suggest of Clinical Epidemiology and Biostatistics, McMaster University, that VKA therapy be started on day 1 or 2 Hamilton, ON, L8N 3Z5, Canada; e-mail: [email protected] © 2012 American College of Chest Physicians. Reproduction of low-molecular-weight heparin (LMWH) or of this article is prohibited without written permission from the low-dose unfractionated heparin (UFH) therapy American College of Chest Physicians ( http://www.chestpubs.org/ rather than waiting for several days to start site/misc/reprints.xhtml ). DOI: 10.1378/chest.1412S3 (Grade 2C) .

8S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.1 Monitoring Frequency for VKAs 3.8 VKA Drug Interactions to Avoid 3.1. For patients taking VKA therapy with con- 3.8. For patients taking VKAs, we suggest avoid- sistently stable INRs, we suggest an INR testing ing concomitant treatment with nonsteroidal frequency of up to 12 weeks rather than every antiinfl ammatory drugs, including cyclooxyge- 4 weeks (Grade 2B). nase-2-selective nonsteroidal antiinfl ammatory drugs, and certain antibiotics (see Table 8 in main 3.2 Management of the Single Out-of-Range INR article1 ) (Grade 2C). 3.2. For patients taking VKAs with previously For patients taking VKAs, we suggest avoiding stable therapeutic INRs who present with a concomitant treatment with antiplatelet agents single out-of-range INR of Յ 0.5 below or above except in situations where benefi t is known or is therapeutic, we suggest continuing the current highly likely to be greater than harm from dose and testing the INR within 1 to 2 weeks bleeding, such as patients with mechanical valves, (Grade 2C). patients with acute coronary syndrome, or patients 3.3 Bridging for Low INRs with recent coronary stents or bypass surgery (Grade 2C). 3.3. For patients with stable therapeutic INRs presenting with a single subtherapeutic INR 4.1 Optimal Therapeutic INR Range value, we suggest against routinely administering bridging with heparin (Grade 2C) . 4.1. For patients treated with VKAs, we recommend a therapeutic INR range of 2.0 to 3.0 (tar- 3.4 Vitamin K Supplementation get INR of 2.5) rather than a lower (INR , 2) or higher (INR 3.0-5.0) range (Grade 1B). 3.4. For patients taking VKAs, we suggest against routine use of vitamin K supplementa- 4.2 Therapeutic Range for High-Risk Groups tion (Grade 2C). 4.2. For patients with antiphospholipid syndrome 3.5 Anticoagulation Management Services for VKAs with previous arterial or venous thromboembolism, 3.5. (Best Practices Statement) We suggest that we suggest VKA therapy titrated to a moderate- health-care providers who manage oral antico- intensity INR range (INR 2.0-3.0) rather than agulation therapy should do so in a systematic higher intensity (INR 3.0-4.5) ( G r a d e 2 B ). and coordinated fashion, incorporating patient 5.0 Discontinuation of Therapy education, systematic INR testing, tracking, follow-up, and good patient communication of 5.0. For patients eligible to discontinue treat- results and dosing decisions. ment with VKA, we suggest abrupt discontinuation rather than gradual tapering of the dose to 3.6 Patient Self-Testing and Self-Management discontinuation (Grade 2C). 3.6. For patients treated with VKAs who are motivated and can demonstrate competency 6.1 Unfractionated Heparin (UFH) Dose Adjustment in self-management strategies, including the by Weight self-testing equipment, we suggest patient self- 6.1. For patients starting IV UFH, we suggest management rather than usual outpatient INR that the initial bolus and the initial rate of the monitoring (Grade 2B). For all other patients, continuous infusion be weight adjusted (bolus we suggest monitoring that includes the safe- 80 units/kg followed by 18 units/kg per h for VTE; guards in our best practice statement 3.5. bolus 70 units/kg followed by 15 units/kg per h for cardiac or stroke patients) or use of a fi xed 3.7 Dosing Decision Support dose (bolus 5,000 units followed by 1,000 units/h) 3.7. For dosing decisions during maintenance rather than alternative regimens (Grade 2C). VKA therapy, we suggest using validated decision support tools (paper nomograms or com- 6.2 Dose Management of Subcutaneous (SC) UFH puterized dosing programs) rather than no 6.2. For outpatients with VTE treated with SC decision support (Grade 2C) . UFH, we suggest weight-adjusted dosing (fi rst Remarks: Inexperienced prescribers may be more dose 333 units/kg, then 250 units/kg) with- likely to improve prescribing with use of decision sup- out monitoring rather than fi xed or weight- port tools than experienced prescribers. adjusted dosing with monitoring (Grade 2C) . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 9S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 7.1 Therapeutic Dose of LMWH in Patients With unfractionated heparin (LDUH) bid, LDUH Decreased Renal Function tid, or fondaparinux (Grade 1B). 7.1. For patients receiving therapeutic LMWH Remarks: In choosing the speciﬁ c anticoagulant drug who have severe renal insufﬁ ciency (calculated to be used for pharmacoprophylaxis, choices should creatinine clearance , 30 mL/min), we suggest be based on patient preference, compliance, and ease a reduction of the dose rather than using stan- of administration (eg, daily vs bid vs tid dosing), as dard doses (Grade 2C). well as on local factors affecting acquisition costs (eg, prices of various pharmacologic agents in individual 8.1 Fondaparinux Dose Management by Weight hospital formularies). 8.1. For patients with VTE and body weight over 100 kg, we suggest that the treatment dose 2.4. For acutely ill hospitalized medical patients of fondaparinux be increased from the usual at low risk of thrombosis, we recommend against 7.5 mg to 10 mg daily SC (Grade 2C). the use of pharmacologic prophylaxis or mechanical prophylaxis (Grade 1B). 9.1 Vitamin K for Patients Taking VKAs With High 2.7.1. For acutely ill hospitalized medical INRs Without Bleeding patients who are bleeding or at high risk for 9.1. bleeding, we recommend against anticoagulant thromboprophylaxis (Grade 1B). (a) For patients taking VKAs with INRs between 4.5 and 10 and with no evidence of bleeding, we 2.7.2. For acutely ill hospitalized medical patients at suggest against the routine use of vitamin K increased risk of thrombosis who are bleeding or (Grade 2B). at high risk for major bleeding, we suggest the optimal use of mechanical thromboprophylaxis (b) For patients taking VKAs with INRs . 10.0 with graduated compression stockings (GCS) and with no evidence of bleeding, we suggest ( G r a d e 2 C ) or intermittent pneumatic compres- that oral vitamin K be administered (Grade 2C). sion (IPC) (Grade 2C) , rather than no mechanical thromboprophylaxis. When bleeding risk 9.2 Clinical Prediction Rules for Bleeding While decreases, and if VTE risk persists, we sug- Taking VKA gest that pharmacologic thromboprophylaxis 9.2. For patients initiating VKA therapy, we be substituted for mechanical thromboprophy- suggest against the routine use of clinical pre- laxis ( G r a d e 2 B ) . diction rules for bleeding as the sole criterion to Remarks: Patients who are particularly averse to the withhold VKA therapy (Grade 2C). potential for skin complications, cost, and need for 9.3 Treatment of Anticoagulant-Related Bleeding clinical monitoring of GCS and IPC use are likely to decline mechanical prophylaxis. 9.3. For patients with VKA-associated major bleeding, we suggest rapid reversal of antico- 2.8. In acutely ill hospitalized medical patients agulation with four-factor prothrombin complex who receive an initial course of thrombopro- concentrate rather than with plasma. ( G r a d e 2 C ). phylaxis, we suggest against extending the duration of thromboprophylaxis beyond the period We suggest the additional use of vitamin K 5 to of patient immobilization or acute hospital stay 10 mg administered by slow IV injection rather (Grade 2B). than reversal with coagulation factors alone (Grade 2C). 3.0 Critically Ill Patients 3.2. In critically ill patients, we suggest against Prevention of VTE in Nonsurgical Patients routine ultrasound screening for DVT ( G r a d e 2 C ). For further details, see Kahn et al.2 3.4.3. For critically ill patients, we suggest using LMWH or LDUH thromboprophylaxis over no 2.0 Hospitalized Acutely Ill Medical Patients prophylaxis (Grade 2C) .

2.3. For acutely ill hospitalized medical patients 3.4.4. For critically ill patients who are bleeding, at increased risk of thrombosis, we recommend or are at high risk for major bleeding, we anticoagulant thromboprophylaxis with low- suggest mechanical thromboprophylaxis with molecular-weight heparin [LMWH], low-dose GCS ( G r a d e 2 C ) or IPC ( G r a d e 2 C ) until the

10S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians bleeding risk decreases, rather than no mechan- knee GCS providing 15 to 30 mm Hg of pressure ical thromboprophylaxis. When bleeding risk at the ankle during travel ( G r a d e 2 C ). For all decreases, we suggest that pharmacologic throm- other long-distance travelers, we suggest against boprophylaxis be substituted for mechanical the use of GCS (Grade 2C). thromboprophylaxis (Grade 2C). 6.1.3. For long-distance travelers, we suggest 4.0 Patients With Cancer in the Outpatient Setting against the use of aspirin or anticoagulants to prevent VTE (Grade 2C). 4.2.1. In outpatients with cancer who have no additional risk factors for VTE, we suggest 7.0 Persons With Asymptomatic Thrombophilia against routine prophylaxis with LMWH or LDUH ( G r a d e 2 B ) and recommend against 7.1. In persons with asymptomatic thrombo- the prophylactic use of VKAs (Grade 1B). philia (ie, without a previous history of VTE), we recommend against the long-term daily use Remarks: Additional risk factors for venous throm- of mechanical or pharmacologic thrombopro- bosis in cancer outpatients include previous venous phylaxis to prevent VTE (Grade 1C). thrombosis, immobilization, hormonal therapy, angio- genesis inhibitors, thalidomide, and lenalidomide. Prevention of VTE in Nonorthopedic 4.2.2. In outpatients with solid tumors who have Surgical Patients additional risk factors for VTE and who are at low risk of bleeding, we suggest prophylactic- For further details, see Gould et al.3 dose LMWH or LDUH over no prophylaxis (Grade 2B). 3.6 Patients Undergoing General, GI, Urological, Gynecologic, Bariatric, Vascular, Plastic, or Recon- Remarks: Additional risk factors for venous thrombo- structive Surgery sis in cancer outpatients include previous venous 3.6.1. For general and abdominal-pelvic sur- thrombosis, immobilization, hormonal therapy, angio- gery patients at very low risk for VTE (, 0.5%; genesis inhibitors, thalidomide, and lenali domide. Rogers score, , 7; Caprini score, 0), we recom- 4.4. In outpatients with cancer and indwelling mend that no speciﬁ c pharmacologic ( G r a d e 1 B ) central venous catheters, we suggest against or mechanical (Grade 2C) prophylaxis be used routine prophylaxis with LMWH or LDUH other than early ambulation. (Grade 2B) and suggest against the prophylactic 3.6.2. For general and abdominal-pelvic sur- -Rog ;%1.5 ف ) use of VKAs (Grade 2C). gery patients at low risk for VTE 5.0 Chronically Immobilized Patients ers score, 7-10; Caprini score, 1-2), we suggest mechanical prophylaxis, preferably with inter- 5.1. In chronically immobilized persons residing mittent pneumatic compression (IPC), over no at home or at a nursing home, we suggest against prophylaxis (Grade 2C) . the routine use of thromboprophylaxis ( G r a d e 2 C ) . 3.6.3. For general and abdominal-pelvic sur- ;%3.0 ف) Persons Traveling Long-Distance gery patients at moderate risk for VTE 6.0 6.1.1. For long-distance travelers at increased Rogers score, . 10; Caprini score, 3-4) who are risk of VTE (including previous VTE, recent not at high risk for major bleeding complica- surgery or trauma, active malignancy, preg- tions, we suggest LMWH ( G r a d e 2 B), LDUH nancy, estrogen use, advanced age, limited (Grade 2B) , or mechanical prophylaxis, prefer- mobility, severe obesity, or known thrombo- ably with IPC (Grade 2C), over no prophylaxis. philic disorder), we suggest frequent ambula- Remarks: Three of the seven authors favored a strong tion, calf muscle exercise, or sitting in an aisle (Grade 1B) recommendation in favor of LMWH or seat if feasible (Grade 2C). LDUH over no prophylaxis in this group. 6.1.2. For long-distance travelers at increased 3.6.4. For general and abdominal-pelvic sur- risk of VTE (including previous VTE, recent gery patients at moderate risk for VTE (3.0%; surgery or trauma, active malignancy, pregnancy, Rogers score, . 10; Caprini score, 3-4) who are estrogen use, advanced age, limited mobility, at high risk for major bleeding complications severe obesity, or known thrombophilic disor- or those in whom the consequences of bleed- der), we suggest use of properly ﬁ tted, below- ing are thought to be particularly severe, we www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 11S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians suggest mechanical prophylaxis, preferably with gest use of mechanical prophylaxis, preferably IPC, over no prophylaxis (Grade 2C) . with optimally applied IPC, over either no prophylaxis (Grade 2C) or pharmacologic prophy- 3.6.5. For general and abdominal-pelvic sur- laxis (Grade 2C) . ;%6.0 ف) gery patients at high risk for VTE Caprini score, Ն 5) who are not at high risk for 4.4.2. For cardiac surgery patients whose hos- major bleeding complications, we recommend pital course is prolonged by one or more non- pharmacologic prophylaxis with LMWH (Grade hemorrhagic surgical complications, we suggest 1B) or LDUH (Grade 1B) over no prophylaxis. adding pharmacologic prophylaxis with LDUH We suggest that mechanical prophylaxis with or LMWH to mechanical prophylaxis ( G r a d e 2 C ). elastic stockings or IPC should be added to pharmacologic prophylaxis (Grade 2C). 5.0 Patients Undergoing Thoracic Surgery 3.6.6. For high-VTE-risk patients undergoing 5.4.1. For thoracic surgery patients at mod- abdominal or pelvic surgery for cancer who are erate risk for VTE who are not at high risk for not otherwise at high risk for major bleeding perioperative bleeding, we suggest LDUH complications, we recommend extended-duration (Grade 2B), LMWH (Grade 2B), or mechanical pharmacologic prophylaxis (4 weeks) with LMWH prophylaxis with optimally applied IPC (Grade 2C) over limited-duration prophylaxis (Grade 1B). over no prophylaxis. Remarks: Patients who place a high value on mini- Remarks: Three of the seven authors favored a strong mizing out-of-pocket health-care costs might prefer (Grade 1B) recommendation in favor of LMWH or limited-duration over extended-duration prophylaxis LDUH over no prophylaxis in this group. in settings where the cost of extended-duration prophylaxis is borne by the patient. 5.4.2. For thoracic surgery patients at high risk for VTE who are not at high risk for periopera- 3.6.7. For high-VTE-risk general and abdominal- tive bleeding, we suggest LDUH (Grade 1B) or pelvic surgery patients who are at high risk for LMWH (Grade 1B) over no prophylaxis. In addi- major bleeding complications or those in whom the tion, we suggest that mechanical prophylaxis consequences of bleeding are thought to be with elastic stockings or IPC should be added to particularly severe, we suggest use of mechan- pharmacologic prophylaxis (Grade 2C). ical prophylaxis, preferably with IPC, over no prophylaxis until the risk of bleeding diminishes 5.4.3. For thoracic surgery patients who are at and pharmacologic prophylaxis may be initiated high risk for major bleeding, we suggest use of (Grade 2C). mechanical prophylaxis, preferably with optimally applied IPC, over no prophylaxis until the 3.6.8. For general and abdominal-pelvic sur- risk of bleeding diminishes and pharmacologic gery patients at high risk for VTE (6%; Caprini prophylaxis may be initiated (Grade 2C) . score, Ն 5) in whom both LMWH and unfractionated heparin are contraindicated or unavail- 6.0 Patients Undergoing Craniotomy able and who are not at high risk for major 6.4.1. For craniotomy patients, we suggest that bleeding complications, we suggest low-dose mechanical prophylaxis, preferably with IPC, aspirin (Grade 2C), fondaparinux (Grade 2C), or be used over no prophylaxis (Grade 2C) or phar- mechanical prophylaxis, preferably with IPC macologic prophylaxis (Grade 2C) . (Grade 2C), over no prophylaxis. 6.4.2. For craniotomy patients at very high risk 3.6.9. For general and abdominal-pelvic sur- for VTE (eg, those undergoing craniotomy for gery patients, we suggest that an inferior vena malignant disease), we suggest adding pharma- cava (IVC) ﬁ lter should not be used for primary cologic prophylaxis to mechanical prophylaxis VTE prevention (Grade 2C) . once adequate hemostasis is established and the 3.6.10. For general and abdominal-pelvic surgery risk of bleeding decreases (Grade 2C). patients, we suggest that periodic surveillance with venous compression ultrasound should 7.0 Patients Undergoing Spinal Surgery not be performed (Grade 2C) . 7.4.1. For patients undergoing spinal surgery, we suggest mechanical prophylaxis, prefer- 4.0 Patients Undergoing Cardiac Surgery ably with IPC, over no prophylaxis ( G r a d e 2 C ) , 4.4.1. For cardiac surgery patients with an unfractionated heparin ( G r a d e 2 C ) , or LMWH uncomplicated postoperative course, we sug- (Grade 2C) .

12S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 7.4.2. For patients undergoing spinal surgery at igatran, rivaroxaban, low-dose unfractionated high risk for VTE (including those with malig- heparin (LDUH), adjusted-dose VKA, aspirin nant disease or those undergoing surgery with (all Grade 1B), or an intermittent pneumatic com- a combined anterior-posterior approach), we pression device (IPCD) (Grade 1C) . suggest adding pharmacologic prophylaxis to mechanical prophylaxis once adequate hemo- Remarks: We recommend the use of only portable, stasis is established and the risk of bleeding battery-powered IPCDs capable of recording and decreases (Grade 2C). reporting proper wear time on a daily basis for inpatients and outpatients. Efforts should be made to 8.0 Patients With Major Trauma: Traumatic Brain achieve 18 h of daily compliance. One panel member Injury, Acute Spinal Injury, and Traumatic Spine believed strongly that aspirin alone should not be Injury included as an option. 8.4.1. For major trauma patients, we suggest 2.1.2. In patients undergoing HFS, we recom- use of LDUH (Grade 2C), LMWH (Grade 2C), or mend use of one of the following rather than no mechanical prophylaxis, preferably with IPC antithrombotic prophylaxis for a minimum of (Grade 2C), over no prophylaxis. 10 to 14 days: LMWH, fondaparinux, LDUH, 8.4.2. For major trauma patients at high risk for adjusted-dose VKA, aspirin (all Grade 1B), or an VTE (including those with acute spinal cord IPCD (Grade 1C) . injury, traumatic brain injury, and spinal surgery for trauma), we suggest adding mechan- Remarks: We recommend the use of only portable, ical prophylaxis to pharmacologic prophylaxis battery-powered IPCDs capable of recording and (Grade 2C) when not contraindicated by lower- reporting proper wear time on a daily basis for inpa- extremity injury. tients and outpatients. Efforts should be made to achieve 18 h of daily compliance. One panel member 8.4.3. For major trauma patients in whom believed strongly that aspirin alone should not be LMWH and LDUH are contraindicated, we sug- included as an option. gest mechanical prophylaxis, preferably with IPC, over no prophylaxis (Grade 2C) when not 2.2. For patients undergoing major orthopedic contraindicated by lower-extremity injury. We surgery (THA, TKA, HFS) and receiving LMWH suggest adding pharmacologic prophylaxis with as thromboprophylaxis, we recommend starting either LMWH or LDUH when the risk of either 12 h or more preoperatively or 12 h or more bleeding diminishes or the contraindication to postoperatively rather than within 4 h or less pre- heparin resolves (Grade 2C) . operatively or 4 h or less postoperatively ( G r a d e 1 B ). 8.4.4. For major trauma patients, we suggest 2.3.1. In patients undergoing THA or TKA, that an IVC ﬁ lter should not be used for pri- irrespective of the concomitant use of an IPCD mary VTE prevention (Grade 2C) . or length of treatment, we suggest the use of LMWH in preference to the other agents 8.4.5. For major trauma patients, we suggest that we have recommended as alternatives: fonda- periodic surveillance with venous compression parinux, apixaban, dabigatran, rivaroxaban, ultrasound should not be performed ( G r a d e 2 C ) . LDUH (all Grade 2B), adjusted-dose VKA, or aspirin (all Grade 2C).

Prevention of VTE in Orthopedic Remarks: If started preoperatively, we suggest admin- Surgery Patients istering LMWH Ն 12 h before surgery. Patients who place a high value on avoiding the inconvenience For further details, see Falck-Ytter et al.4 of daily injections with LMWH and a low value on 2.0 Patients Undergoing Major Orthopedic Surgery: the limitations of alternative agents are likely to Total Hip Arthroplasty (THA), Total Knee Arthroplasty choose an alternative agent. Limitations of alter- (TKA), Hip Fracture Surgery (HFS) native agents include the possibility of increased bleeding (which may occur with fondaparinux, rivar- 2.1.1. In patients undergoing THA or TKA, we oxaban, and VKA), possible decreased efﬁ cacy (LDUH, recommend use of one of the following for a VKA, aspirin, and IPCD alone), and lack of long-term minimum of 10 to 14 days rather than no anti- safety data (apixaban, dabigatran, and rivaroxaban). thrombotic prophylaxis: low-molecular-weight Furthermore, patients who place a high value on heparin (LMWH), fondaparinux, apixaban, dab- avoiding bleeding complications and a low value on www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 13S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians its inconvenience are likely to choose an IPCD over nience of IPCD and a low value on avoiding a small the drug options. absolute increase in bleeding with pharmacologic agents when only one bleeding risk factor is present 2.3.2. In patients undergoing HFS, irrespective (in particular the continued use of antiplatelet agents) of the concomitant use of an IPCD or length of are likely to choose pharmacologic thromboprophy- treatment, we suggest the use of LMWH in pref- laxis over IPCD. erence to the other agents we have recommended as alternatives: fondaparinux, LDUH ( G r a d e 2 B ), 2.7. In patients undergoing major orthopedic adjusted-dose VKA, or aspirin ( a l l G r a d e 2 C ). surgery and who decline or are uncooperative with injections or an IPCD, we recommend Remarks: For patients in whom surgery is likely to using apixaban or dabigatran (alterna tively be delayed, we suggest that LMWH be initiated rivaroxaban or adjusted-dose VKA if apixa- during the time between hospital admission and sur- ban or dabigatran are unavailable) rather than gery but suggest administering LMWH at least 12 h alternative forms of prophylaxis (all Grade 1B). before surgery. Patients who place a high value on avoiding the inconvenience of daily injections with 2.8. In patients undergoing major orthopedic LMWH and a low value on the limitations of alterna- surgery, we suggest against using IVC ﬁ lter tive agents are likely to choose an alternative agent. placement for primary prevention over no throm- Limitations of alternative agents include the possi- boprophylaxis in patients with an increased bility of increased bleeding (which may occur with bleeding risk or contraindications to both phar- fondaparinux) or possible decreased efﬁ cacy (LDUH, macologic and mechanical thromboprophylaxis VKA, aspirin, and IPCD alone). Furthermore, patients (Grade 2C) . who place a high value on avoiding bleeding compli- 2.9. For asymptomatic patients following major cations and a low value on its inconvenience are likely orthopedic surgery, we recommend against to choose an IPCD over the drug options. Doppler (or duplex) ultrasound screening before 2.4. For patients undergoing major orthopedic hospital discharge (Grade 1B) . surgery, we suggest extending thromboprophy- 3.0 Patients With Isolated Lower-Leg Injuries Distal laxis in the outpatient period for up to 35 days to the Knee from the day of surgery rather than for only 10 to 14 days (Grade 2B). 3.0. We suggest no prophylaxis rather than pharmacologic thromboprophylaxis in patients 2.5. In patients undergoing major orthopedic with isolated lower-leg injuries requiring leg surgery, we suggest using dual prophylaxis with immobilization (Grade 2C). an antithrombotic agent and an IPCD during 4.0 Patients Undergoing Knee Arthroscopy the hospital stay (Grade 2C). 4.0. For patients undergoing knee arthroscopy Remarks: We recommend the use of only portable, without a history of prior VTE, we suggest no battery-powered IPCDs capable of recording and thromboprophylaxis rather than prophylaxis reporting proper wear time on a daily basis for inpa- (Grade 2B). tients and outpatients. Efforts should be made to achieve 18 h of daily compliance. Patients who place a high value on avoiding the undesirable consequences Perioperative Management of associated with prophylaxis with both a pharmacologic Antithrombotic Therapy agent and an IPCD are likely to decline use of dual prophylaxis. For further details, see Douketis et al.5 2.6. In patients undergoing major orthopedic 2.1 Interruption of VKAs Before Surgery surgery and increased risk of bleeding, we 2.1. In patients who require temporary inter- suggest using an IPCD or no prophylaxis rather ruption of a VKA before surgery, we recom- than pharmacologic treatment (Grade 2C) . mend stopping VKAs approximately 5 days Remarks: We recommend the use of only portable, before surgery instead of stopping VKAs a battery-powered IPCDs capable of recording and shorter time before surgery (Grade 1C). reporting proper wear time on a daily basis for inpa- 2.2 Resumption of VKAs After Surgery tients and outpatients. Efforts should be made to achieve 18 h of daily compliance. Patients who place 2.2. In patients who require temporary interrup- a high value on avoiding the discomfort and inconve- tion of a VKA before surgery, we recommend

14S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians resuming VKAs approximately 12 to 24 h after time of the procedure instead of stopping ASA surgery (evening of or next morning) and when 7 to 10 days before the procedure (Grade 2C) . there is adequate hemostasis instead of later resumption of VKAs (Grade 2C) . 3.5. In patients at moderate to high risk for cardiovascular events who are receiving ASA 2.4 Bridging Anticoagulation During Interruption of therapy and require noncardiac surgery, we VKA Therapy suggest continuing ASA around the time of surgery instead of stopping ASA 7 to 10 days before 2.4. In patients with a mechanical heart valve, surgery (Grade 2C). In patients at low risk for atrial fi brillation, or VTE at high risk for throm- cardiovascular events who are receiving ASA boembolism, we suggest bridging anticoagula- therapy, we suggest stopping ASA 7 to 10 days tion instead of no bridging during interruption before surgery instead of continuation of ASA of VKA therapy (Grade 2C) . (Grade 2C). Remarks: Patients who place a higher value on avoid- 3.6 Patients Undergoing Coronary Artery Bypass ing perioperative bleeding than on avoiding peri- Graft Surgery operative thromboembolism are likely to decline 3.6. In patients who are receiving ASA and heparin bridging. require coronary artery bypass graft (CABG) In patients with a mechanical heart valve, atrial surgery, we suggest continuing ASA around the fi brillation, or VTE at low risk for thrombo- time of surgery instead of stopping ASA 7 to embolism, we suggest no bridging instead of 10 days before surgery (Grade 2C). In patients bridging anticoagulation during interruption who are receiving dual antiplatelet drug therapy of VKA therapy (Grade 2C) . and require CABG surgery, we suggest continuing ASA around the time of surgery and In patients with a mechanical heart valve, atrial fi bril- stopping clopidogrel/prasugrel 5 days before lation, or VTE at moderate risk for thromboembo- surgery instead of continuing dual antiplatelet lism, the bridging or no-bridging approach chosen is, therapy around the time of surgery (Grade 2C) . as in the higher- and lower-risk patients, based on an 3.7 Surgical Patients With Coronary Stents assessment of individual patient- and surgery-related factors. 3.7. In patients with a coronary stent who are receiving dual antiplatelet therapy and require 2.5 Perioperative Management of VKA-Treated surgery, we recommend deferring surgery for Patients Who Require Minor Procedures at least 6 weeks after placement of a bare-metal 2.5. In patients who require a minor dental pro- stent and for at least 6 months after placement cedure, we suggest continuing VKAs with coad- of a drug-eluting stent instead of undertaking ministration of an oral prohemostatic agent or surgery within these time periods (Grade 1C). In stopping VKAs 2 to 3 days before the procedure patients who require surgery within 6 weeks instead of alternative strategies (Grade 2C). In of placement of a bare-metal stent or within patients who require minor dermatologic pro- 6 months of placement of a drug-eluting stent, cedures and are receiving VKA therapy, we sug- we suggest continuing dual antiplatelet therapy gest continuing VKAs around the time of the around the time of surgery instead of stopping procedure and optimizing local hemostasis dual antiplatelet therapy 7 to 10 days before instead of other strategies (Grade 2C). In patients surgery (Grade 2C) . who require cataract surgery and are receiving Remarks: Patients who are more concerned about VKA therapy, we suggest continuing VKAs avoiding the unknown, but potentially large increase around the time of the surgery instead of other in bleeding risk associated with the perioperative strategies (Grade 2C). continuation of dual antiplatelet therapy than avoiding the risk for coronary stent thrombosis are unlikely 3.4 Patients Undergoing a Minor Dental, Dermatologic, to choose continuation of dual antiplatelet therapy. or Ophthalmologic Procedure 4.2 Perioperative Use of IV UFH 3.4. In patients who are receiving acetylsalicylic acid (ASA) for the secondary prevention 4.2. In patients who are receiving bridging anti- of cardiovascular disease and are having minor coagulation with therapeutic-dose IV UFH, we dental or dermatologic procedures or cataract suggest stopping UFH 4 to 6 h before surgery surgery, we suggest continuing ASA around the instead of closer to surgery (Grade 2C). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 15S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 4.3 Preoperative Interruption of Therapeutic-Dose DVT is not present. Initial testing with US would be Bridging LMWH preferred if the patient has a comorbid condition associated with elevated D-dimer levels and is likely to have 4.3. In patients who are receiving bridging anti- a positive D-dimer result, even if DVT is absent. In coagulation with therapeutic-dose SC LMWH, patients with suspected ﬁ rst lower extremity DVT in we suggest administering the last preoperative whom US is impractical (eg, when leg casting or dose of LMWH approximately 24 h before sur- excessive SC tissue or ﬂ uid prevent adequate assess- gery instead of 12 h before surgery (Grade 2C). ment of compressibility) or nondiagnostic, we sug- 4.4 Postoperative Resumption of Therapeutic-Dose gest CT scan venography or magnetic resonance Bridging LMWH (MR) venography, or MR direct thrombus imaging could be used as an alternative to venography. 4.4. In patients who are receiving bridging anticoagulation with therapeutic-dose SC LMWH If the D-dimer is negative, we recommend no and are undergoing high-bleeding-risk surgery, further testing over further investigation with we suggest resuming therapeutic-dose LMWH (i) proximal CUS, (ii) whole-leg US, or (iii) venog- 48 to 72 h after surgery instead of resuming raphy (Grade 1B for all comparisons). If the prox- LMWH within 24 h after surgery (Grade 2C). imal CUS is negative, we recommend no further testing compared with (i) repeat proximal CUS after 1 week, (ii) whole-leg US, or (iii) venog- Diagnosis of DVT raphy (Grade 1B for all comparisons).

For further details, see Bates et al.6 If the D-dimer is positive, we suggest further testing with CUS of the proximal veins rather 3.0 Diagnosis of Suspected First Lower Extremity than (i) whole-leg US (Grade 2C) or (ii) venog- DVT raphy (Grade 1B) . If CUS of the proximal veins is 3.1. In patients with a suspected fi rst lower positive, we suggest treating for DVT and per- extremity DVT, we suggest that the choice of forming no further testing over performing diagnostic tests process should be guided by the confi rmatory venography (Grade 2C). clinical assessment of pretest probability rather than by performing the same diagnostic tests in Remarks: In circumstances when high-quality venog- all patients (Grade 2B). raphy is available, patients who are not averse to the discomfort of venography, are less concerned about Remarks: In considering this recommendation, fi ve the complications of venography, and place a high panelists voted for a strong recommendation and four value on avoiding treatment of false-positive results voted for a weak recommendation (one declined are likely to choose confi rmatory venography if fi nd- to vote and two did not participate). According ings for DVT are less certain (eg, a short segment of to predetermined criteria, this resulted in weak venous noncompressibility). recommendation. 3.3. In patients with a moderate pretest proba- 3.2. In patients with a low pretest probability bility of fi rst lower extremity DVT, we recom- of fi rst lower extremity DVT, we recommend mend one of the following initial tests: (i) a highly one of the following initial tests: (i) a moder- sensitive D-dimer or (ii) proximal CUS, or (iii) ately sensitive D-dimer, (ii) a highly sensitive whole-leg US rather than (i) no testing (Grade 1B D-dimer, or (iii) compression ultrasound (CUS) for all comparisons) or (ii) venography (Grade 1B for of the proximal veins rather than (i) no diagnos- all comparisons). We suggest initial use of a highly tic testing (Grade 1B for all comparisons), (ii) venog- sensitive D-dimer rather than US ( G r a d e 2 C ). raphy (Grade 1B for all comparisons), or (iii) whole-leg ultrasound (US) (Grade 2B for all com- Remarks: The choice between a highly sensitive parisons). We suggest initial use of a moderately D-dimer test or US as the initial test will depend on sensitive (Grade 2C) or highly sensitive (Grade local availability, access to testing, costs of testing, 2B) D-dimer rather than proximal CUS. and the probability of obtaining a negative D-dimer result if DVT is not present. Initial testing with US Remarks: The choice between a moderately sensitive may be preferred if the patient has a comorbid condi- D-dimer test, a highly sensitive D-dimer test, or prox- tion associated with elevated D-dimer levels and is imal CUS as the initial test will depend on local avail- likely to have a positive D-dimer result even if DVT ability, access to testing, costs of testing, and the is absent. Whole-leg US may be preferred in patients probability of obtaining a negative D-dimer result if unable to return for serial testing and those with

16S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians severe symptoms consistent with calf DVT. In patients 3.4. In patients with a high pretest probability with suspected fi rst lower extremity DVT in whom of fi rst lower extremity DVT, we recommend US is impractical (eg, when leg casting or excessive either (i) proximal CUS or (ii) whole-leg US over SC tissue or fl uid prevent adequate assessment of no testing (Grade 1B for all comparisons) or venog- compressibility) or nondiagnostic, we suggest CT scan raphy (Grade 1B for all comparisons). venography, MR venography, or MR direct thrombus imaging could be used as an alternative to venography. Remarks: Whole-leg US may be preferred to proximal CUS in patients unable to return for serial test- If the highly sensitive D-dimer is negative, we ing and those with severe symptoms consistent with recommend no further testing over further calf DVT. In patients with extensive unexplained investigation with (i) proximal CUS, (ii) whole- leg swelling, if there is no DVT on proximal CUS or leg US, or (iii) venography (Grade 1B for all whole-leg US and d-dimer testing has not been comparisons). If the highly sensitive D-dimer performed or is positive, the iliac veins should be is positive, we recommend proximal CUS or imaged to exclude isolated iliac DVT. In patients with whole-leg US rather than no testing (Grade 1B suspected fi rst lower extremity DVT in whom US is for all comparisons) or venography (Grade 1B for impractical (eg, when leg casting or excessive SC all comparisons). tissue or fl uid prevent adequate assessment of compressibility) or nondiagnostic, we suggest CT scan If proximal CUS is chosen as the initial test and venography, MR venography, or MR direct thrombus is negative, we recommend (i) repeat proximal imaging could be used as an alternative to venography. CUS in 1 week or (ii) testing with a moderate or highly sensitive D-dimer assay over no further If proximal CUS or whole-leg US is positive for testing (Grade 1C) or venography (Grade 2B). In DVT, we recommend treatment rather than patients with a negative proximal CUS but a confi rmatory venography (Grade 1B). positive D-dimer, we recommend repeat proximal CUS in 1 week over no further testing In patients with a negative proximal CUS, we (Grade 1B) or venography (Grade 2B). recommend additional testing with a highly sensitive D-dimer or whole-leg US or repeat In patients with (i) negative serial proximal CUS proximal CUS in 1 week over no further or (ii) a negative single proximal CUS and nega- testing (Grade 1B for all comparisons) or venog- tive moderate or highly sensitive D-dimer, we raphy (Grade 2B for all comparisons). We rec- recommend no further testing rather than fur- ommend that patients with a single negative ther testing with (i) whole-leg US or (ii) venog- proximal CUS and positive D-dimer undergo raphy (Grade 1B for all comparisons). whole-leg US or repeat proximal CUS in 1 week If whole-leg US is negative, we recommend over no further testing ( G r a d e 1 B ) or venography no further testing over (i) repeat US in one (Grade 2B) . In patients with negative serial prox- week, (ii) D-dimer testing, or (iii) venography imal CUS, a negative single proximal CUS and (Grade 1B for all comparisons). If proximal CUS negative highly sensitive D-dimer, or a negative is positive, we recommend treating for DVT whole-leg US, we recommend no further testing rather than confi rmatory venography ( G r a d e 1 B ). over venography or additional US (Grade 1B for If isolated distal DVT is detected on whole- negative serial proximal CUS and for negative single leg US, we suggest serial testing to rule out proximal CUS and highly sensitive D-dimer; Grade proximal extension over treatment (Grade 2C) . 2B for negative whole-leg US).

Remarks: Patients with abnormal isolated distal US We recommend that in patients with high pre- fi ndings on whole-leg US who place a high value on test probability, moderately or highly sensitive avoiding the inconvenience of repeat testing and a D-dimer assays should not be used as stand- low value on avoiding treatment of false-positive alone tests to rule out DVT (Grade 1B) . results are likely to choose treatment over repeat US. Patients with severe symptoms and risk factors for 3.5. If risk stratifi cation is not performed in extension as outlined in Perioperative Management patients with suspected fi rst lower extremity of Antithrombotic Therapy. Antithrombotic Therapy DVT, we recommend one of the following and Prevention of Thrombosis, 9th ed: American initial tests: (i) proximal CUS or (ii) whole- College of Chest Physicians Evidence-Based Clinical leg US rather than (i) no testing (Grade 1B), Practice Guidelines are more likely to benefi t from (ii) venography (Grade 1B), or D-dimer testing treatment over repeat US. (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 17S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Remarks: Whole-leg US may be preferred to prox- 4.1 Venography in Patients With Suspected Recur- imal CUS in patients unable to return for serial rent DVT testing and those with severe symptoms consistent 4.1. In patients suspected of having recurrent with calf DVT or risk factors for extension of distal lower extremity DVT, we recommend initial DVT. In patients with suspected fi rst lower extremity evaluation with proximal CUS or a highly sensi- DVT in whom US is impractical (eg, when leg cast- tive D-dimer over venography, CT venography, ing or excessive SC tissue or fl uid prevent adequate or MRI (all Grade 1B). assessment of compressibility) or nondiagnostic, we suggest that CT scan venography, MR venography, Remarks: Initial D-dimer testing with a high-sensitivity or MR direct thrombus imaging could be used as an assay is preferable if prior US is not available for alternative to venog raphy. comparison. We recommend that patients with a negative If the highly sensitive D-dimer is positive, we proximal CUS undergo testing with a mod- recommend proximal CUS over venography, CT erate- or high-sensitivity D-dimer, whole-leg venography, or MRI (Grade 1B for all com parisons). US, or repeat proximal CUS in 1 week over no further testing ( G r a d e 1 B ) or venography In patients with suspected recurrent lower (Grade 2B). In patients with a negative prox- extremity DVT in whom initial proximal CUS is imal CUS, we suggest D-dimer rather than negative (normal or residual diameter increase routine serial CUS ( G r a d e 2 B ) or whole-leg US of , 2 mm), we suggest at least one further (Grade 2C) . We recommend that patients with proximal CUS (day 7 Ϯ 1) or testing with a mod- a single negative proximal CUS and positive erately or highly sensitive D-dimer (followed by D-dimer undergo further testing with repeat repeat CUS [day 7 Ϯ 1] if positive) rather than proximal CUS in 1 week or whole-leg US no further testing or venography (Grade 2B). rather than no further testing (Grade 1B for Remarks: In patients with an abnormal proximal CUS both comparisons). at presentation that does not meet the criteria for the We recommend that in patients with (i) nega- diagnosis of recurrence, an additional proximal CUS Ϯ Ϯ tive serial proximal CUS, (ii) a negative D-dimer on day 2 1 in addition to that on (day 7 1) may be following a negative initial proximal CUS, or (iii) preferred. Patients who place a high value on an negative whole-leg US, no further testing be accurate diagnosis and a low value on avoiding the performed rather than venography (Grade 1B). inconvenience and potential side effects of a venography are likely to choose venography over missed If proximal US is positive for DVT, we recom- diagnosis (in the case of residual diameter increase mend treatment rather than confi rmatory venog- of , 2 mm). raphy ( G r a d e 1 B ). If isolated distal DVT is We recommend that patients with suspected detected on whole-leg US, we suggest serial recurrent lower extremity DVT and a negative testing to rule out proximal extension over treat- highly sensitive D-dimer or negative proximal ment (Grade 2C). CUS and negative moderately or highly sensitive D-dimer or negative serial proximal CUS Remarks: Patients with abnormal isolated distal US undergo no further testing for suspected recur- fi ndings on whole-leg US who place a high value on rent DVT rather than venography (Grade 1B). avoiding the inconvenience of repeat testing and a low value on avoiding treatment of false-positive If CUS of the proximal veins is positive, we rec- results are likely to choose treatment over repeat US. ommend treating for DVT and performing no Patients with severe symptoms and risk factors for further testing over performing confi rmatory extension as outlined in “Perioperative Management venography (Grade 1B for the fi nding of a new non- of Antithrombotic Therapy. Antithrombotic Therapy compressible segment in the common femoral or and Prevention of Thrombosis, 9th ed: American popliteal vein, Grade 2B for a Ն 4-mm increase in College of Chest Physicians Evidence-Based Clinical venous diameter during compression compared with Practice Guidelines” are more likely to benefi t from that in the same venous segment on a previous result). treatment over repeat US. Remarks: Patients with US abnormalities at pre- 3.6. In patients with suspected first lower sentation that do not include a new noncompressible extremity DVT, we recommend against the rou- segment who place a high value on an accurate diag- tine use of CT venography or MRI (Grade 1C). nosis and a low value on avoiding the inconvenience

18S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians and potential side effects of a venography are likely D-dimer done at the time of presentation (Grade to choose venography over treatment (in the case 2B) over no further testing for DVT. We recom- of Ն 4-mm increase in venous diameter). mend that patients with an initial negative proximal CUS and a subsequent negative sensi- 4.2 Compression Ultrasonography in Patients With tive D-dimer or negative serial proximal CUS Suspected Recurrent DVT undergo no further testing for DVT (Grade 1B) 4.2. In patients with suspected recurrent lower and that patients with positive D-dimer have an extremity DVT and abnormal but nondiagnostic additional follow-up proximal CUS (day 3 and US results (eg, an increase in residual venous day 7) rather than venography (Grade 1B) or diameter of , 4 but Ն 2 mm), we recommend whole-leg US (Grade 2C). further testing with venography, if available (Grade 1B); serial proximal CUS (Grade 2B) or 5.3 Pretest Probability in Pregnancy-Related DVT testing with a moderately or highly sensitive 5.3. In pregnant patients with symptoms sug- D-dimer with serial proximal CUS as above if gestive of isolated iliac vein thrombosis (swelling the test is positive (Grade 2B), as opposed to of the entire leg, with or without ﬂ ank, buttock, other testing strategies or treatment. or back pain) and no evidence of DVT on standard proximal CUS, we suggest further testing 4.3 Pretest Probability Assessment in Patients With with either Doppler US of the iliac vein (Grade Suspected Recurrent DVT 2C) , venography ( G r a d e 2 C ) , or direct MRI 4.3. In patients with suspected recurrent ipsilat- (Grade 2C) , rather than standard serial CUS of eral DVT and an abnormal US without a prior the proximal deep veins. result for comparison, we recommend further testing with venography, if available (Grade 1B) 6.1 Ultrasonography in Patients With Upper- or a highly sensitive D-dimer ( G r a d e 2 B ) over Extremity DVT (UEDVT) serial proximal CUS. In patients with suspected 6.1. In patients suspected of having UEDVT, we recurrent ipsilateral DVT and an abnormal US suggest initial evaluation with combined modality without prior result for comparison and a nega- US (compression with either Doppler or color tive highly sensitive D-dimer, we suggest no Doppler) over other initial tests, including highly further testing over venography (Grade 2C). In sensitive D-dimer or venography (Grade 2C) . patients with suspected recurrent ipsilateral DVT and an abnormal US without prior result 6.2 Clinical Pretest Probability Assessment in Patients for comparison and a positive highly sensitive With UEDVT D-dimer, we suggest venography if available over 6.2. In patients with suspected UEDVT in whom empirical treatment of recurrence (Grade 2C). initial US is negative for thrombosis despite a Remarks: Patients who place a high value on avoid- high clinical suspicion of DVT, we suggest fur- ing the inconvenience and potential side effects of ther testing with a moderate or highly sensitive a venography are likely to choose treatment over D-dimer, serial US, or venographic-based imaging venography. (traditional, CT scan, or MRI), rather than no further testing ( G r a d e 2 C ). 5.1 Venography in Pregnancy-Related DVT In patients with suspected UEDVT and an ini- 5.1. In pregnant patients suspected of having tial negative combined-modality US and sub- lower extremity DVT, we recommend initial sequent negative moderate or highly sensitive evaluation with proximal CUS over other initial D-dimer or CT or MRI, we recommend no fur- tests, including a whole-leg US (Grade 2C), mod- ther testing, rather than conﬁ rmatory venog- erately sensitive D-dimer ( G r a d e 2 C ) , highly raphy (Grade 1C). We suggest that patients with sensitive D-dimer (Grade 1B), or venography an initial combined negative modality US and (Grade 1B). positive D-dimer or those with less than complete evaluation by US undergo venography 5.2 Compression Ultrasonography in Pregnancy- rather than no further testing, unless there is Related DVT an alternative explanation for their symptoms 5.2. In pregnant patients with suspected DVT in (Grade 2B), in which case testing to evaluate for whom initial proximal CUS is negative, we sug- the presence an alternative diagnosis should be gest further testing with either serial proximal performed. We suggest that patients with a pos- CUS (day 3 and day 7) (Grade 1B) or a sensitive itive D-dimer or those with less than complete www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 19S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians evaluation by US but an alternative explana- Remarks: Patients at high risk for bleeding are more tion for their symptoms undergo confi rmatory likely to benefi t from serial imaging. Patients who testing and treatment of this alternative expla- place a high value on avoiding the inconvenience of nation rather than venography (Grade 2C). repeat imaging and a low value on the inconvenience of treatment and on the potential for bleeding are Remarks: Further radiologic testing (serial US or likely to choose initial anticoagulation over serial venographic-based imaging or CT/MR to seek an imaging. alternative diagnosis) rather than d-dimer testing is preferable in patients with comorbid conditions typi- 2.3.3. In patients with acute isolated distal DVT cally associated with elevated D-dimer levels. of the leg who are managed with initial anticoagulation, we recommend using the same approach as for patients with acute proximal Antithrombotic Therapy for VTE Disease DVT (Grade 1B). For further details, see Kearon et al.7 2.3.4. In patients with acute isolated distal DVT 2.1 Initial Anticoagulation for Patients With Acute DVT of the leg who are managed with serial imaging, of the Leg we recommend no anticoagulation if the thrombus does not extend (Grade 1B); we suggest anti- 2.1. In patients with acute DVT of the leg treated coagulation if the thrombus extends but remains with VKA therapy, we recommend initial treat- confi ned to the distal veins (Grade 2C); we rec- ment with parenteral anticoagulation (LMWH, ommend anticoagulation if the thrombus fondaparinux, IV UFH, or SC UFH) over no extends into the proximal veins (Grade 1B). such initial treatment (Grade 1B). 2.2 Parenteral Anticoagulation Prior to Receipt of the 2.4 Timing of Initiation of VKA and Associated Results of Diagnostic Work-up for VTE Duration of Parenteral Anticoagulant Therapy 2.2.1. In patients with a high clinical suspicion of 2.4. In patients with acute DVT of the leg, we acute VTE, we suggest treatment with parenteral recommend early initiation of VKA (eg, same anticoagulants compared with no treatment while day as parenteral therapy is started) over awaiting the results of diagnostic tests ( G r a d e 2 C ). delayed initiation, and continuation of parenteral anticoagulation for a minimum of 5 days 2.2.2. In patients with an intermediate clinical and until the international normalized ratio suspicion of acute VTE, we suggest treatment (INR) is 2.0 or above for at least 24 h ( G r a d e 1 B ). with parenteral anticoagulants compared with no treatment if the results of diagnostic tests 2.5 Choice of Initial Anticoagulant Regimen in are expected to be delayed for more than 4 h Patients With Proximal DVT (Grade 2C). 2.5.1. In patients with acute DVT of the leg, we 2.2.3. In patients with a low clinical suspicion of suggest LMWH or fondaparinux over IV UFH acute VTE, we suggest not treating with paren- (Grade 2C) and over SC UFH (Grade 2B for teral anticoagulants while awaiting the results LMWH; Grade 2C for fondaparinux). of diagnostic tests, provided test results are Remarks: Local considerations such as cost, avail- expected within 24 h (Grade 2C). ability, and familiarity of use dictate the choice 2.3 Anticoagulation in Patients With Isolated Distal between fondaparinux and LMWH. LMWH and DVT fondaparinux are retained in patients with renal impairment, whereas this is not a concern with UFH . 2.3.1. In patients with acute isolated distal DVT of the leg and without severe symptoms or risk 2.5.2. In patients with acute DVT of the leg factors for extension, we suggest serial imaging treated with LMWH, we suggest once- over of the deep veins for 2 weeks over initial antico- twice-daily administration (Grade 2C) . agulation (Grade 2C). Remarks: This recommendation only applies when 2.3.2. In patients with acute isolated distal DVT the approved once-daily regimen uses the same of the leg and severe symptoms or risk factors daily dose as the twice-daily regimen (ie, the once- for extension (see text), we suggest initial antico- daily injection contains double the dose of each agulation over serial imaging of the deep veins twice-daily injection). It also places value on avoiding (Grade 2C). an extra injection per day.

20S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 2.7 At-Home vs In-Hospital Initial Treatment of Patients 2.13.1. In patients with acute DVT of the leg, we With DVT recommend against the use of an IVC fi lter in 2.7. In patients with acute DVT of the leg and addition to anticoagulants (Grade 1B). whose home circumstances are adequate, we 2.13.2. In patients with acute proximal DVT of the recommend initial treatment at home over leg and contraindication to anticoagulation, we treatment in hospital (Grade 1B). recommend the use of an IVC fi lter (Grade 1B). Remarks: The recommendation is conditional on the adequacy of home circumstances: well-maintained 2.13.3. In patients with acute proximal DVT of living conditions, strong support from family or friends, the leg and an IVC fi lter inserted as an alterna- phone access, and ability to quickly return to the hos- tive to anticoagulation, we suggest a conven- pital if there is deterioration. It is also conditional on tional course of anticoagulant therapy if their the patient feeling well enough to be treated at home risk of bleeding resolves (Grade 2B). (eg, does not have severe leg symptoms or comorbidity). Remarks: We do not consider that a permanent 2.9 Catheter-Directed Thrombolysis for Patients With IVC fi lter, of itself, is an indication for extended Acute DVT anticoagulation. 2.9. In patients with acute proximal DVT of the 2.14 Early Ambulation of Patients With Acute DVT leg, we suggest anticoagulant therapy alone over 2.14. In patients with acute DVT of the leg, we catheter-directed thrombolysis (CDT) ( G r a d e 2 C ). suggest early ambulation over initial bed rest Remarks: Patients who are most likely to benefi t from (Grade 2C) . CDT (see text), who attach a high value to prevention of postthrombotic syndrome (PTS), and a lower Remarks: If edema and pain are severe, ambulation value to the initial complexity, cost, and risk of may need to be deferred. As per section 4.1, we suggest bleeding with CDT, are likely to choose CDT over the use of compression therapy in these patients. anticoagulation alone. 3.0 Long-term Anticoagulation in Patients With Acute 2.10 Systemic Thrombolytic Therapy for Patients With DVT of the Leg Acute DVT 3.0. In patients with acute VTE who are treated 2.10. In patients with acute proximal DVT of with anticoagulant therapy, we recommend long- the leg, we suggest anticoagulant therapy alone term therapy (see section 3.1 for recommended over systemic thrombolysis (Grade 2C). duration of therapy) over stopping anticoagulant therapy after about 1 week of initial therapy Remarks: Patients who are most likely to benefi t from (Grade 1B). systemic thrombolytic therapy (see text), who do not have access to CDT, and who attach a high value 3.1 Duration of Long-term Anticoagulant Therapy to prevention of PTS, and a lower value to the initial 3.1.1. In patients with a proximal DVT of the leg complexity, cost, and risk of bleeding with systemic provoked by surgery, we recommend treatment thrombolytic therapy, are likely to choose systemic with anticoagulation for 3 months over (i) treat- thrombolytic therapy over anticoagulation alone. ment of a shorter period (Grade 1B) , (ii) treat- 2.11 Operative Venous Thrombectomy for Acute DVT ment of a longer time-limited period (eg, 6 or 12 months) ( G r a d e 1 B ) , or (iii) extended therapy 2.11. In patients with acute proximal DVT of the (Grade 1B regardless of bleeding risk). leg, we suggest anticoagulant therapy alone over operative venous thrombectomy (Grade 2C). 3.1.2. In patients with a proximal DVT of the leg 2.12 Anticoagulation in Patients Who Have Had Any provoked by a nonsurgical transient risk factor, Method of Thrombus Removal Performed we recommend treatment with anticoagulation for 3 months over (i) treatment of a shorter 2.12. In patients with acute DVT of the leg who period ( G r a d e 1 B ), (ii) treatment of a longer time- undergo thrombosis removal, we recommend limited period (eg, 6 or 12 months) ( G r a d e 1 B ) , the same intensity and duration of anticoagu- and (iii) extended therapy if there is a high lant therapy as in comparable patients who do bleeding risk (Grade 1B). We suggest treatment not undergo thrombosis removal (Grade 1B). with anticoagulation for 3 months over extended 2.13 Vena Cava Filters for the Initial Treatment of therapy if there is a low or moderate bleeding Patients With DVT risk (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 21S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.1.3. In patients with an isolated distal DVT Remarks (3.1.3, 3.1.4, 3.1.4.3): Duration of treatment of the leg provoked by surgery or by a nonsur- of patients with isolated distal DVT refers to patients gical transient risk factor (see remark), we sug- in whom a decision has been made to treat with anti- gest treatment with anticoagulation for 3 months coagulant therapy; however, it is anticipated that not over treatment of a shorter period (Grade 2C) all patients who are diagnosed with isolated distal and recommend treatment with anticoagula- DVT will be given anticoagulants (see section 2.3). In tion for 3 months over treatment of a longer time- all patients who receive extended anticoagulant therapy, limited period (eg, 6 or 12 months) (Grade 1B) the continuing use of treatment should be reassessed at or extended therapy (Grade 1B regardless of periodic intervals (eg, annually). bleeding risk). 3.2 Intensity of Anticoagulant Effect 3.1.4. In patients with an unprovoked DVT of 3.2. In patients with DVT of the leg who are the leg (isolated distal [see remark] or proximal), treated with VKA, we recommend a therapeutic we recommend treatment with anticoagulation INR range of 2.0 to 3.0 (target INR of 2.5) over for at least 3 months over treatment of a shorter a lower (INR , 2) or higher (INR 3.0-5.0) range duration ( G r a d e 1 B ) . After 3 months of treatment, for all treatment durations (Grade 1B). patients with unprovoked DVT of the leg should be evaluated for the risk-benefi t ratio of extended 3.3 Choice of Anticoagulant Regimen for Long-term therapy . Therapy 3.1.4.1. In patients with a fi rst VTE that is an 3.3.1. In patients with DVT of the leg and no can- unprovoked proximal DVT of the leg and who cer, we suggest VKA therapy over LMWH for have a low or moderate bleeding risk, we suggest long-term therapy ( G r a d e 2 C ). For patients with extended anticoagulant therapy over 3 months DVT and no cancer who are not treated with VKA of therapy (Grade 2B). therapy, we suggest LMWH over dabigatran or rivaroxaban for long-term therapy (Grade 2C). 3.1.4.2. In patients with a fi rst VTE that is an unprovoked proximal DVT of the leg and who 3.3.2. In patients with DVT of the leg and can- have a high bleeding risk, we recommend cer, we suggest LMWH over VKA therapy (Grade 3 months of anticoagulant therapy over extended 2B) . In patients with DVT and cancer who are therapy ( G r a d e 1 B ). not treated with LMWH, we suggest VKA over dabigatran or rivaroxaban for long-term therapy 3.1.4.3. In patients with a fi rst VTE that is an (Grade 2B). unprovoked isolated distal DVT of the leg (see remark), we suggest 3 months of anticoagulant Remarks (3.3.1-3.3.2): Choice of treatment in patients therapy over extended therapy in those with a low with and without cancer is sensitive to the individual or moderate bleeding risk (Grade 2B) and rec- patient’s tolerance for daily injections, need for labo- ommend 3 months of anticoagulant treatment ratory monitoring, and treatment costs. LMWH, rivar- in those with a high bleeding risk (Grade 1B). oxaban, and dabigatran are retained in patients with renal impairment, whereas this is not a concern with 3.1.4.4. In patients with a second unprovoked VKA. Treatment of VTE with dabigatran or rivaroxa- VTE, we recommend extended anticoagulant ban, in addition to being less burdensome to patients, therapy over 3 months of therapy in those who may prove to be associated with better clinical out- have a low bleeding risk (Grade 1B), and we sug- comes than VKA and LMWH therapy. When these gest extended anticoagulant therapy in those guidelines were being prepared (October 2011), with a moderate bleeding risk (Grade 2B). postmarketing studies of safety were not available. Given the paucity of currently available data and that 3.1.4.5. In patients with a second unprovoked new data are rapidly emerging, we give a weak rec- VTE who have a high bleeding risk, we sug- ommendation in favor of VKA and LMWH therapy gest 3 months of anticoagulant therapy over over dabigatran and rivaroxaban, and we have not extended therapy (Grade 2B). made any recommendations in favor of one of the 3.1.5. In patients with DVT of the leg and active new agents over the other. cancer, if the risk of bleeding is not high, we 3.4 Choice of Anticoagulant Regimen for Extended recommend extended anticoagulant therapy Therapy over 3 months of therapy (Grade 1B), and if there is a high bleeding risk, we suggest extended 3.4. In patients with DVT of the leg who receive anticoagulant therapy (Grade 2B). extended therapy, we suggest treatment with the

22S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians same anticoagulant chosen for the fi rst 3 months 5.2.2. In patients with an intermediate clinical (Grade 2C). suspicion of acute PE, we suggest treatment with parenteral anticoagulants compared with 3.5 Treatment of Patients With Asymptomatic DVT no treatment if the results of diagnostic tests of the Leg are expected to be delayed for more than 4 h 3.5. In patients who are incidentally found to have (Grade 2C) . asymptomatic DVT of the leg, we suggest the same 5.2.3. In patients with a low clinical suspicion of initial and long-term anticoagulation as for compa- acute PE, we suggest not treating with paren- rable patients with symptomatic DVT ( G r a d e 2 B ). teral anticoagulants while awaiting the results 4.1 Compression Stockings and Bandages to Prevent of diagnostic tests, provided test results are PTS expected within 24 h (Grade 2C). 4.1. In patients with acute symptomatic DVT of 5.3 Timing of Initiation of VKA and Associated the leg, we suggest the use of compression stock- Duration of Parenteral Anticoagulant Therapy ings (Grade 2B). 5.3. In patients with acute PE, we recommend early initiation of VKA (eg, same day as paren- Remarks: Compression stockings should be worn for teral therapy is started) over delayed initiation, 2 years, and we suggest beyond that if patients have and continuation of parenteral anticoagulation developed PTS and fi nd the stockings helpful. for a minimum of 5 days and until the INR is 2.0 Patients who place a low value on preventing PTS or or above for at least 24 h (Grade 1B) . a high value on avoiding the inconvenience and discomfort of stockings are likely to decline stockings. 5.4 Choice of Initial Parenteral Anticoagulant Regimen in Patients With PE 4.2 Physical Treatment of Patients With PTS 5.4.1. In patients with acute PE, we suggest 4.2.1. In patients with PTS of the leg, we sug- LMWH or fondaparinux over IV UFH (Grade 2C gest a trial of compression stockings ( G r a d e 2 C ). for LMWH; Grade 2B for fondaparinux) and over SC UFH (Grade 2B for LMWH; Grade 2C for 4.2.2. In patients with severe PTS of the leg that fondaparinux). is not adequately relieved by compression stockings, we suggest a trial of an intermittent com- Remarks: Local considerations such as cost, availability, pression device (Grade 2B). and familiarity of use dictate the choice between fondaparinux and LMWH. LMWH and fondaparinux 4.3 Pharmacologic Treatment of Patients With PTS are retained in patients with renal impairment, 4.3. In patients with PTS of the leg, we suggest that whereas this is not a concern with UFH. In patients venoactive medications (eg, rutosides, defi brot- with PE where there is concern about the adequacy ide, and hidrosmin) not be used (Grade 2C) . of SC absorption or in patients in whom thrombolytic therapy is being considered or planned, initial treat- Remarks: Patients who value the possibility of response ment with IV UFH is preferred to use of SC therapies. over the risk of side effects may choose to undertake a therapeutic trial. 5.4.2. In patients with acute PE treated with LMWH, we suggest once- over twice-daily admin- 5.1 Initial Anticoagulation for Patients With Acute istration (Grade 2C). Pulmonary Embolism (PE) Remarks: This recommendation only applies when 5.1. In patients with acute PE, we recommend the approved once-daily regimen uses the same initial treatment with parenteral anticoagula- daily dose as the twice-daily regimen (ie, the once- tion (LMWH, fondaparinux, IV UFH, or SC daily injection contains double the dose of each UFH) over no such initial treatment (Grade 1B). twice-daily injection). It also places value on avoiding an extra injection per day. 5.2 Parenteral Anticoagulation Prior to Receipt of the Results of Diagnostic Work-up for PE 5.5 Early vs Standard Discharge of Patients With Acute PE 5.2.1. In patients with a high clinical suspicion of acute PE, we suggest treatment with paren- 5.5. In patients with low-risk PE and whose teral anticoagulants compared with no treat- home circumstances are adequate, we suggest ment while awaiting the results of diagnostic early discharge over standard discharge (eg, tests (Grade 2C). after fi rst 5 days of treatment) (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 23S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Remarks: Patients who prefer the security of the hos- 5.9. Vena Cava Filters for the Initial Treatment of pital to the convenience and comfort of home are Patients With PE likely to choose hospitalization over home treatment. 5.9.1. In patients with acute PE who are treated 5.6 Systemic Thrombolytic Therapy for Patients With with anticoagulants, we recommend against the PE use of an IVC fi lter (Grade 1B). 5.6.1.1. In patients with acute PE associated 5.9.2. In patients with acute PE and contraindi- with hypotension (eg, systolic BP , 90 mm Hg) cation to anticoagulation, we recommend the who do not have a high bleeding risk, we sug- use of an IVC fi lter (Grade 1B). gest systemically administered thrombolytic therapy over no such therapy (Grade 2C). 5.9.3. In patients with acute PE and an IVC fi lter inserted as an alternative to anticoagu- 5.6.1.2. In most patients with acute PE not asso- lation, we suggest a conventional course of ciated with hypotension, we recommend against anticoagulant therapy if their risk of bleeding systemically administered thrombolytic therapy resolves (Grade 2B) . (Grade 1C). Remarks: We do not consider that a permanent 5.6.1.3. In selected patients with acute PE not IVC fi lter, of itself, is an indication for extended associated with hypotension and with a low bleed- anticoagulation. ing risk whose initial clinical presentation, or clinical course after starting anticoagulant ther- 6.0 Long-term Treatment of Patients With PE apy, suggests a high risk of developing hypotension, we suggest administration of throm bolytic 6.1. In patients with PE provoked by surgery, therapy (Grade 2C). we recommend treatment with anticoagulation for 3 months over (i) treatment of a shorter 5.6.2.1. In patients with acute PE, when a throm- period ( G r a d e 1 B ), (ii) treatment of a longer time- bolytic agent is used, we suggest short infusion limited period (eg, 6 or 12 months) (Grade 1B), times (eg, a 2-h infusion) over prolonged infu- or (iii) extended therapy (Grade 1B regardless of sion times (eg, a 24-h infusion) (Grade 2C) . bleeding risk). 5.6.2.2. In patients with acute PE when a throm- 6.2. In patients with PE provoked by a non- bolytic agent is used, we suggest administration surgical transient risk factor, we recommend through a peripheral vein over a pulmonary treatment with anticoagulation for 3 months artery catheter (Grade 2C) . over (i) treatment of a shorter period (Grade 1B) , (ii) treatment of a longer time-limited period 5.7 Catheter-Based Thrombus Removal for the Initial (eg, 6 or 12 months) ( G r a d e 1 B ), and (iii) extended Treatment of Patients With PE therapy if there is a high bleeding risk (Grade 5.7. In patients with acute PE associated with 1B) . We suggest treatment with anticoagulation hypotension and who have (i) contraindications for 3 months over extended therapy if there is a to thrombolysis, (ii) failed thrombolysis, or (iii) low or moderate bleeding risk (Grade 2B). shock that is likely to cause death before systemic thrombolysis can take effect (eg, within hours), if 6.3. In patients with an unprovoked PE, we rec- appropriate expertise and resources are avail- ommend treatment with anticoagulation for at able, we suggest catheter-assisted thrombus least 3 months over treatment of a shorter dura- removal over no such intervention (Grade 2C). tion (Grade 1B). After 3 months of treatment, patients with unprovoked PE should be evaluated 5.8 Surgical Embolectomy for the Initial Treatment for the risk-benefi t ratio of extended therapy. of Patients With PE 6.3.1. In patients with a fi rst VTE that is an 5.8. In patients with acute PE associated with unprovoked PE and who have a low or mod- hypotension, we suggest surgical pulmonary erate bleeding risk, we suggest extended anti- embolectomy over no such intervention if they coagulant therapy over 3 months of therapy have (i) contraindications to thrombolysis, (ii) (Grade 2B). failed thrombolysis or catheter-assisted embolectomy, or (iii) shock that is likely to cause 6.3.2. In patients with a fi rst VTE that is an death before thrombolysis can take effect (eg, unprovoked PE and who have a high bleeding within hours), provided surgical expertise and risk, we recommend 3 months of anticoagulant resources are available (Grade 2C). therapy over extended therapy (Grade 1B).

24S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 6.3.3. In patients with a second unprovoked 6.8. In patients with PE who receive extended VTE, we recommend extended anticoagulant therapy, we suggest treatment with the same therapy over 3 months of therapy in those who anticoagulant chosen for the ﬁ rst 3 months have a low bleeding risk (Grade 1B), and we sug- (Grade 2C) . gest extended anticoagulant therapy in those with a moderate bleeding risk (Grade 2B). 6.9. In patients who are incidentally found to have asymptomatic PE, we suggest the same 6.3.4. In patients with a second unprovoked initial and long-term anticoagulation as for VTE who have a high bleeding risk, we suggest comparable patients with symptomatic PE 3 months of therapy over extended therapy (Grade 2B). (Grade 2B). 7.1 Pulmonary Thromboendarterectomy, Anticoagulant 6.4. In patients with PE and active cancer, if Therapy, and Vena Cava Filter for the Treatment of there is a low or moderate bleeding risk, we Chronic Thromboembolic Pulmonary Hypertension recommend extended anticoagulant therapy (CTPH) over 3 months of therapy (Grade 1B), and if there 7.1.1. In patients with CTPH, we recommend is a high bleeding risk, we suggest extended anti- extended anticoagulation over stopping therapy coagulant therapy (Grade 2B). (Grade 1B).

Remarks: In all patients who receive extended anti- 7.1.2. In selected patients with CTPH, such as coagulant therapy, the continuing use of treatment those with central disease under the care of should be reassessed at periodic intervals (eg, annually). an experienced thromboendarterectomy team, we suggest pulmonary thromboendarterectomy 6.5. In patients with PE who are treated with over no pulmonary thromboendarterectomy VKA, we recommend a therapeutic INR range (Grade 2C) . of 2.0 to 3.0 (target INR of 2.5) over a lower (INR , 2) or higher (INR 3.0-5.0) range for all 8.1 Treatment of Patients With Superfi cial Vein treatment durations (Grade 1B). Thrombosis 6.6. In patients with PE and no cancer, we sug- 8.1.1. In patients with superfi cial vein thrombo- gest VKA therapy over LMWH for long-term sis of the lower limb of at least 5 cm in length, therapy (Grade 2C). For patients with PE and no we suggest the use of a prophylactic dose of cancer who are not treated with VKA therapy, fondaparinux or LMWH for 45 days over no we suggest LMWH over dabigatran or rivaroxa- anticoagulation (Grade 2B). ban for long-term therapy (Grade 2C) . Remarks: Patients who place a high value on avoiding 6.7. In patients with PE and cancer, we suggest the inconvenience or cost of anticoagulation and a LMWH over VKA therapy (Grade 2B). In patients low value on avoiding infrequent symptomatic VTE with PE and cancer who are not treated with are likely to decline anticoagulation. LMWH, we suggest VKA over dabigatran or 8.1.2. In patients with superfi cial vein throm- rivaroxaban for long-term therapy (Grade 2C). bosis who are treated with anticoagulation, we Remarks (6.6-6.7): Choice of treatment in patients suggest fondaparinux 2.5 mg daily over a pro- with and without cancer is sensitive to the individual phylactic dose of LMWH (Grade 2C). patient’s tolerance for daily injections, need for labo- 9.1 Acute Anticoagulation for Patients With UEDVT ratory monitoring, and treatment costs. Treatment of VTE with dabigatran or rivaroxaban, in addition to 9.1.1. In patients with UEDVT that involves being less burdensome to patients, may prove to be the axillary or more proximal veins, we recom- associated with better clinical outcomes than VKA mend acute treatment with parenteral anti- and LMWH therapy. When these guidelines were coagulation (LMWH, fondaparinux, IV UFH, being prepared (October 2011), postmarketing studies or SC UFH) over no such acute treatment of safety were not available. Given the paucity of cur- (Grade 1B). rently available data and that new data are rapidly emerging, we give a weak recommendation in favor 9.1.2. In patients with acute UEDVT that of VKA and LMWH therapy over dabigatran and involves the axillary or more proximal veins, we rivaroxaban, and we have not made any recommen- suggest LMWH or fondaparinux over IV UFH dation in favor of one of the new agents over the other. (Grade 2C) and over SC UFH (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 25S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 9.2 Thrombolytic Therapy for the Initial Treatment 9.4 Prevention of PTS of the Arm of Patients With UEDVT 9.4. In patients with acute symptomatic UEDVT, 9.2.1. In patients with acute UEDVT that we suggest against the use of compression involves the axillary or more proximal veins, we sleeves or venoactive medications (Grade 2C). suggest anticoagulant therapy alone over thrombolysis (Grade 2C). 9.5 Treatment of Patients With PTS of the Arm Remarks: Patients who (i) are most likely to beneﬁ t 9.5.1. In patients who have PTS of the arm, we from thrombolysis (see text); (ii) have access to CDT; suggest a trial of compression bandages or (iii) attach a high value to prevention of PTS; and (iv) sleeves to reduce symptoms (Grade 2C). attach a lower value to the initial complexity, cost, and risk of bleeding with thrombolytic therapy are 9.5.2. In patients with PTS of the arm, we sug- likely to choose thrombolytic therapy over anticoagu- gest against treatment with venoactive medica- lation alone. tions (Grade 2C). 9.2.2. In patients with UEDVT who undergo 10.0 Patients With Splanchnic Vein Thrombosis thrombolysis, we recommend the same inten- 10.1. In patients with symptomatic splanchnic sity and duration of anticoagulant therapy as in vein thrombosis (portal, mesenteric, and/or similar patients who do not undergo thromboly- splenic vein thromboses), we recommend anti- sis (Grade 1B). coagulation over no anticoagulation (Grade 1B). 9.3 Long-term Anticoagulation for Patients With UEDVT 10.2. In patients with incidentally detected splanchnic vein thrombosis (portal, mesen- 9.3.1. In most patients with UEDVT that is asso- teric, and/or splenic vein thromboses), we sug- ciated with a central venous catheter, we suggest no anticoagulation over anticoagulation gest that the catheter not be removed if it is (Grade 2C) . functional and there is an ongoing need for the catheter (Grade 2C). 11.0 Patients With Hepatic Vein Thrombosis 9.3.2. In patients with UEDVT that involves 11.1. In patients with symptomatic hepatic vein the axillary or more proximal veins, we suggest thrombosis, we suggest anticoagulation over no a minimum duration of anticoagulation of anticoagulation (Grade 2C) . 3 months over a shorter period (Grade 2B). 11.2. In patients with incidentally detected Remarks: This recommendation also applies if the hepatic vein thrombosis, we suggest no antico- UEDVT was associated with a central venous cath- agulation over anticoagulation (Grade 2C). eter that was removed shortly after diagnosis. 9.3.3. In patients who have UEDVT that is associated with a central venous catheter that is Treatment and Prevention of removed, we recommend 3 months of antico- Heparin-Induced Thrombocytopenia agulation over a longer duration of therapy in For further details, see Linkins et al.8 patients with no cancer (Grade 1B), and we suggest this in patients with cancer (Grade 2C). 2.1 Platelet Count Monitoring Combined With the 4Ts Score for Patients Receiving Heparin/LMWH 9.3.4. In patients who have UEDVT that is associated with a central venous catheter that is not 2.1.1. For patients receiving heparin in whom removed, we recommend that anticoagulation clinicians consider the risk of heparin-induced is continued as long as the central venous cath- thrombocytopenia (HIT) to be . 1%, we suggest eter remains over stopping after 3 months of that platelet count monitoring be performed treatment in patients with cancer (Grade 1C), every 2 or 3 days from day 4 to day 14 (or until and we suggest this in patients with no cancer heparin is stopped, whichever occurs ﬁ rst) (Grade 2C). (Grade 2C) . 9.3.5. In patients who have UEDVT that is not 2.1.2. For patients receiving heparin in whom associated with a central venous catheter or with clinicians consider the risk of HIT to be , 1%, cancer, we recommend 3 months of anticoagula- we suggest that platelet counts not be moni- tion over a longer duration of therapy ( G r a d e 1 B ). tored (Grade 2C).

26S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.1 Discontinuation of Heparin or Initiation of VKAs and until the INR is within the target range over vs Treatment With Nonheparin Anticoagulants shorter periods of overlap and that the INR be rechecked after the anticoagulant effect of the 3.1. In patients with HIT complicated by thrombosis nonheparin anticoagulant has resolved ( G r a d e 1 C ) . (HITT), we recommend the use of nonheparin anticoagulants, in particular lepirudin, arga- 4.1 Discontinuation of Heparin or Initiation of VKAs troban, and danaparoid, over the further use of vs Treatment With Nonheparin Anticoagulants heparin or LMWH or initiation/continuation of a VKA (Grade 1C). 4.1. In patients with isolated HIT (HIT without thrombosis), we recommend the use of lepirudin 3.2 Choice of Nonheparin Anticoagulants in Patients or argatroban or danaparoid over the further use With HITT of heparin or LMWH or initiation/continuation of a VKA (Grade 1C). 3.2.1. In patients with HITT who have normal renal function, we suggest the use of argatroban 4.2 Choice of Nonheparin Anticoagulants in Patients or lepirudin or danaparoid over other nonhepa- With Isolated HIT rin anticoagulants (Grade 2C). 4.2. In patients with isolated HIT (HIT without Remarks: Other factors not covered by our analysis, thrombosis) who have normal renal function, such as drug availability, cost, and ability to monitor the we suggest the use of argatroban or lepirudin anticoagulant effect, may infl uence the choice of agent. or danaparoid over other nonheparin anticoagulants (Grade 2C). 3.2.2. In patients with HITT and renal insuffi - ciency, we suggest the use of argatroban over Remarks: Other factors such as drug availability, cost, other nonheparin anticoagulants (Grade 2C). and ability to monitor the anticoagulant effect may infl uence the choice of agent. The dosing consider- 3.3 Platelet Transfusions ations are the same as for patients with HITT (see section 3.2). For a recommendation on choice of 3.3 In patients with HIT and severe thrombocy- nonheparin anticoagulants in the setting of renal topenia, we suggest giving platelet transfusions insuffi ciency, see Recommendation 3.2.2. only if bleeding or during the performance of an invasive procedure with a high risk of 5.1 Patients Who Require Urgent Cardiac Surgery bleeding (Grade 2C). 5.1.1. In patients with acute HIT (thrombocyto- 3.4 Starting VKAs Before Platelet Recovery penic, HIT antibody positive) or subacute HIT (platelets recovered but still HIT antibody pos- 3.4.1. In patients with strongly suspected or itive) who require urgent cardiac surgery, we confi rmed HIT, we recommend against starting suggest the use of bivalirudin over other non- VKA until platelets have substantially recovered heparin anticoagulants and over heparin plus 9 (ie, usually to at least 150 3 10 /L) over starting antiplatelet agents (Grade 2C). VKA at a lower platelet count and that the VKA be initially given in low doses (maximum, 5 mg 5.1.2. In patients with acute HIT who require of warfarin or 6 mg phenprocoumon) over using nonurgent cardiac surgery, we recommend higher doses (Grade 1C) . delaying the surgery (if possible) until HIT has resolved and HIT antibodies are negative (see 3.4.2. We further suggest that if a VKA has already section 6.1) (Grade 2C). been started when a patient is diagnosed with HIT, vitamin K should be administered ( G r a d e 2 C ) . Remarks: Other factors not covered by our analysis, such as drug availability, cost, and ability to monitor Remarks: We place a high value on the prevention the anticoagulant effect may infl uence the choice of of venous limb gangrene and a low value on the cost agent. For recommendations for patients with a past of the additional days of the parental nonheparin history of HIT (. 3 months previous) who require anticoagulant. cardiac surgery, see section 6.1. 3.5 Discontinuation of Thrombin Inhibitor After a 5.2 Patients Who Require Urgent Percutaneous Minimum of 5 Days of Overlap With VKAs Coronary Interventions 3.5. In patients with confi rmed HIT, we recom- 5.2. In patients with acute HIT or subacute HIT mend that that the VKA be overlapped with a non- who require percutaneous coronary interven- heparin anticoagulant for a minimum of 5 days tions, we suggest the use of bivalirudin ( G r a d e 2 B ) www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 27S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians or argatroban (Grade 2C) over other nonheparin absent who require cardiac catheterization or anticoagulants. percutaneous coronary interventions, the recommended treatment is the same as 5.2. Remarks: Other factors, such as drug availability, cost, and ability to monitor the anticoagulant effect, 6.3 Patients Who Require Prophylaxis or Treatment may infl uence the choice of agent. of Thrombosis 5.3 Patients Who Require Renal Replacement Therapy 6.3. In patients with a past history of HIT who have acute thrombosis (not related to HIT) and 5.3.1. In patients with acute or subacute HIT normal renal function, we suggest the use of who require renal replacement therapy, we fondaparinux at full therapeutic doses until suggest the use of argatroban or danaparoid transition to a VKA can be achieved (Grade 2C). over other nonheparin anticoagulants ( G r a d e 2 C ) . Remarks: We acknowledge that the cost of argatroban may be prohibitive at some clinical centers. We fur- Antithrombotic Therapy for ther suggest that if the prothrombotic state of HIT Atrial Fibrillation appears to have resolved (as seen by normalization of For further details, see You et al.9 the platelet count), saline fl ushes during dialysis would be a reasonable option. This suggestion is based on 2.1 Patients With Nonrheumatic Atrial Fibrillation the presumed pathogenesis of thrombosis in this (AF) condition and not on the results of clinical trials. 5.3.2. In patients with a past history of HIT who 2.1.8. For patients with AF, including those with require ongoing renal replacement therapy paroxysmal AF, who are at low risk of stroke (eg, CHADS [congestive heart failure, hyper- or catheter locking, we suggest the use of 2 Ն regional citrate over the use of heparin or tension, age 75 years, diabetes mellitus, prior 5 LMWH (Grade 2C). stroke or transient ischemic attack] score 0), we suggest no therapy rather than antithrom- 5.4 Pregnant Patients botic therapy (Grade 2B) . For patients who do choose antithrombotic therapy, we suggest aspi- 5.4. In pregnant patients with acute or subacute rin (75 mg to 325 mg once daily) rather than HIT, we suggest danaparoid over other non- oral anticoagulation (Grade 2B) or combination heparin anticoagulants (Grade 2C). We suggest therapy with aspirin and clopidogrel ( G r a d e 2 B ). the use of lepirudin or fondaparinux only if danaparoid is not available (Grade 2C) . Remarks : Patients who place an exceptionally high value on stroke reduction and a low value on avoiding Remarks: Other factors, such as drug availability, bleeding and the burden associated with antithrom- cost, and ability to monitor the anticoagulant effect, botic therapy are likely to choose antithrombotic may infl uence the choice of agent. therapy rather than no antithrombotic therapy. Other 6.1 Patients With a History of HIT Who Require factors that may infl uence the choices above are a Cardiac Surgery consideration of patient-specifi c bleeding risk and the presence of additional risk factors for stroke, 6.1.1. In patients with a history of HIT in whom including age 65 to 74 years and female gender, which heparin antibodies have been shown to be have been more consistently validated, and vascular absent who require cardiac surgery, we suggest disease, which has been less well validated (see sec- the use of heparin (short-term use only) over tion 2.1.12). The presence of multiple non-CHADS2 nonheparin anticoagulants (Grade 2C) . risk factors for stroke may favor oral anticoagulation therapy. 6.1.2. In patients with a history of HIT in whom heparin antibodies are still present who require 2.1.9. For patients with AF, including those with cardiac surgery, we suggest the use of nonhepa- paroxysmal AF, who are at intermediate risk of rin anticoagulants (see 5.1.1) over heparin or stroke (eg, CHADS2 score 5 1), we recommend LMWH (Grade 2C). oral anticoagulation rather than no therapy ( G r a d e 1 B ). We suggest oral anticoagulation 6.2 Patients Who Require PCI rather than aspirin (75 mg to 325 mg once 6.2. In patients with a history of HIT in whom daily) (Grade 2B) or combination therapy with heparin antibodies have been shown to be aspirin and clopidogrel (Grade 2B). For patients

28S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians who are unsuitable for or choose not to take an Grade 1B). For patients with AF and mitral ste- oral anticoagulant (for reasons other than con- nosis who are unsuitable for or choose not to cerns about major bleeding), we suggest com- take adjusted-dose VKA therapy (for reasons bination therapy with aspirin and clopidogrel other than concerns about major bleeding), we rather than aspirin (75 mg to 325 mg once daily) recommend combination therapy with aspirin (Grade 2B). and clopidogrel rather than aspirin (75 mg to 325 mg once daily) alone (Grade 1B). Remarks : Patients who place an exceptionally high value on stroke reduction and a low value on avoiding 3.1 Patients With AF and Stable Coronary Artery bleeding and the burden associated with anticoagu- Disease lant therapy are likely to choose oral anticoagulation rather than antiplatelet therapy. Other factors that 3.1. For patients with AF and stable coronary may infl uence the choice among antithrombotic artery disease (eg, no acute coronary syndrome therapies are a consideration of bleeding risk and within the previous year) and who choose oral the presence of additional risk factors for stroke, anticoagulation, we suggest adjusted-dose VKA including age 65 to 74 years and female gender, therapy alone (target international normalized which have been more consistently validated, and ratio [INR] range, 2.0-3.0) rather than the com- vascular disease, which has been less well validated bination of adjusted-dose VKA therapy and (see section 2.1.12). The presence of multiple addi- aspirin (Grade 2C). tional non-CHADS2 risk factors for stroke may favor oral anticoagulation therapy. 3.2 Patients With AF and Placement of an Intracoro- nary Stent 2.1.10. For patients with AF, including those with 3.2. For patients with AF at high risk of stroke paroxysmal AF, who are at high risk of stroke (eg, CHADS score of 2 or greater) during the 5 2 (eg, CHADS 2 score 2), we recommend oral anti- fi rst month after placement of a bare-metal coagulation rather than no therapy (Grade 1A), stent or the fi rst 3 to 6 months after place- aspirin (75 mg to 325 mg once daily) (Grade 1B), ment of a drug-eluting stent, we suggest triple or combination therapy with aspirin and clopid- therapy (eg, VKA therapy, aspirin, and clopidogrel ( G r a d e 1 B ). For patients who are unsuitable ogrel) rather than dual antiplatelet therapy for or choose not to take an oral anticoagulant (eg, aspirin and clopidogrel) ( G r a d e 2 C ). After (for reasons other than concerns about major this initial period of triple therapy, we suggest bleeding), we recommend combination therapy a VKA (INR 2.0-3.0) plus a single antiplatelet with aspirin and clopidogrel rather than aspirin drug rather than VKA alone ( G r a d e 2 C ). At (75 mg to 325 mg once daily) (Grade 1B). 12 months after intracoronary stent placement, antithrombotic therapy is suggested as 2.1.11. For patients with AF, including those for patients with AF and stable coronary artery with paroxysmal AF, for recommendations disease (see section 3.1). in favor of oral anticoagulation (including 2.1.9, 2.1.10, and excluding 2.2, 3.1, 3.2, 3.3), For patients with AF at low to intermediate risk we suggest dabigatran 150 mg twice daily rather of stroke (eg, CHADS score of 0 or 1) during than adjusted-dose VKA therapy (target INR 2 the fi rst 12 months after placement of an intra- range, 2.0-3.0) ( G r a d e 2 B ). coronary stent (bare metal or drug eluting), we suggest dual antiplatelet therapy rather than Remarks : Dabigatran is excreted primarily by the triple therapy (Grade 2C). At 12 months after kidney. It has not been studied and is contraindi- intracoronary stent placement, antithrombotic cated in patients with severe renal impairment (esti- therapy is suggested as for patients with AF and mated creatinine clearance of 30 mL/min or less). stable coronary artery disease (see section 3.1). Clinicians should be aware that there is no antidote for dabigatran. Remarks : Patients who place an exceptionally high 2.2 Patients With AF and Mitral Stenosis value on stroke reduction and a low value on avoiding bleeding and the burden associated with antico- 2.2. For patients with AF and mitral stenosis, we agulant therapy are likely to choose triple therapy recommend adjusted-dose VKA therapy (target rather than dual antiplatelet therapy. Other factors INR range, 2.0-3.0) rather than no therapy, that may infl uence this choice are a consideration of aspirin (75 mg to 325 mg once daily), or combi- bleeding risk and the presence of additional non- nation therapy with aspirin and clopidogrel (all CHADS2 risk factors for stroke (see section 2.1.12). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 29S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.3 Patients With AF and ACS Who Do Not Undergo elec trical or pharmacologic cardioversion, Intracoronary Stent Placement we rec ommend therapeutic anticoagulation (adjusted-dose VKA therapy, target INR range 3.3. For patients with AF at intermediate to high 2.0-3.0, low-molecular-weight heparin at full risk of stroke (eg, CHADS score of 1 or greater) 2 venous thromboembolism treatment doses, or who experience an acute coronary syndrome dabigatran) for at least 3 weeks before cardio- and do not undergo intracoronary stent place- version or a transesophageal echocardiography ment, we suggest for the fi rst 12 months, adjust- (TEE)-guided approach with abbreviated anti- ed-dose VKA therapy (INR 2.0-3.0) plus single coagulation before cardioversion rather than antiplatelet therapy rather than dual anti- no anticoagulation (Grade 1B). We recommend platelet therapy (eg, aspirin and clopidogrel) or therapeutic anticoagulation for at least 4 weeks triple therapy (eg, warfarin, aspirin, and clo- after successful cardioversion to sinus rhythm pidogrel) (Grade 2C). After the fi rst 12 months, rather than no anticoagulation, regardless of antithrombotic therapy is suggested as for the baseline risk of stroke (Grade 1B) . Decisions patients with AF and stable coronary artery about anticoagulation beyond 4 weeks should disease (see section 3.1). be made in accordance with our risk-based rec- For patients with AF at low risk of stroke (eg, ommendations for long-term antithrombotic therapy in section 2.1. CHADS2 score of 0), we suggest dual antiplatelet therapy (eg, aspirin and clopidogrel) rather 4.1.2. For patients with AF of documented than adjusted-dose VKA therapy (INR 2.0-3.0) duration of 48 h or less undergoing elective plus single antiplatelet therapy or triple therapy cardioversion (electrical or pharmacologic), we (eg, warfarin, aspirin, and clopidogrel) (Grade suggest starting anticoagulation at presentation 2C) . After the fi rst 12 months, antithrombotic (low-molecular-weight heparin or unfraction- therapy is suggested as for patients with AF and ated heparin at full venous thromboembolism stable coronary artery disease (see section 3.1). treatment doses) and proceeding to cardio- Remarks : Patients who place an exceptionally high version rather than delaying cardioversion value on stroke reduction and a low value on avoiding for 3 weeks of therapeutic anticoagulation or bleeding and the burden associated with anticoagu- a TEE-guided approach (Grade 2C). After suc- lant therapy are likely to choose adjusted-dose VKA cessful cardioversion to sinus rhythm, we rec- therapy plus single antiplatelet therapy rather than ommend therapeutic anticoagulation for at dual antiplatelet therapy. Other factors that may least 4 weeks rather than no anticoagulation, infl uence this choice are a consideration of bleeding regardless of baseline stroke risk ( G r a d e 2 C ). Decisions about long-term anticoagulation risk and the presence of additional non-CHADS2 risk factors for stroke (see section 2.1.12). after cardioversion should be made in accordance with our risk-based recommendations 3.4 Patients With AF Managed by a Rhythm Control for long-term antithrombotic therapy in sec- Strategy tion 2.1. 3.4. For patients with AF being managed with a 4.2 Patients Undergoing Urgent Cardioversion for rhythm control strategy (pharmacologic or cath- Hemodynamically Unstable AF eter ablation), we suggest that antithrombotic 4.2. For patients with AF and hemodynamic therapy decisions follow the general risk-based instability undergoing urgent cardioversion recommendations for patients with AF in sec- (electrical or pharmacologic), we suggest that tion 2.1, regardless of the apparent persistence therapeutic-dose parenteral anticoagulation of normal sinus rhythm (Grade 2C) . be started before cardioversion, if possible 3.5 Patients With Atrial Flutter ( G r a d e 2 C ), but that initiation of anticoagulation must not delay any emergency intervention 3.5. For patients with atrial fl utter, we suggest ( G r a d e 2 C ) . After successful cardioversion to that antithrombotic therapy decisions follow sinus rhythm, we suggest therapeutic anticoag- the same risk-based recommendations as for ulation for at least 4 weeks after successful AF. cardioversion to sinus rhythm rather than no anticoagulation, regardless of baseline stroke 4.1 Patients Undergoing Elective Cardioversion of AF risk ( G r a d e 2 C ). Decisions about anticoagula- 4.1.1. For patients with AF of greater than tion beyond 4 weeks should be made in accor- 48 h or unknown duration undergoing elective dance with our risk-based recommendations

30S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians for long-term antithrombotic therapy in sec- thrombus, if the left atrial thrombus does not tion 2.1. resolve with VKA therapy, we recommend that PMBV not be performed (Grade 1A). 4.3 Patients Undergoing Elective or Urgent Cardioversion for Atrial Flutter 6.2.1. In patients with asymptomatic patent foramen ovale (PFO) or atrial septal aneurysm, we 4.3. For patients with atrial fl utter undergoing suggest against antithrombotic therapy ( G r a d e 2 C ). elective or urgent pharmacologic or electrical cardioversion, we suggest that the same 6.2 Patients With PFO and Atrial Septal Aneurysm approach to thromboprophylaxis be used as for patients with atrial fi brillation undergoing 6.2.2. In patients with cryptogenic stroke and cardioversion. PFO or atrial septal aneurysm, we recommend aspirin (50-100 mg/d) over no aspirin (Grade 1A) . Antithrombotic and Thrombolytic 6.2.3. In patients with cryptogenic stroke and Therapy for Valvular Disease PFO or atrial septal aneurysm, who experience recurrent events despite aspirin therapy, we 10 For further details, see Whitlock et al. suggest treatment with VKA therapy (target 2.0 Patients With Rheumatic Mitral Valve Disease INR, 2.5; range, 2.0-3.0) and consideration of device closure over aspirin therapy (Grade 2C) . 2.0.1. In patients with rheumatic mitral valve disease and normal sinus rhythm with a left 6.2.4. In patients with cryptogenic stroke and atrial diameter , 55 mm we suggest not using PFO, with evidence of DVT, we recommend antiplatelet or VKA therapy (Grade 2C) . VKA therapy for 3 months (target INR, 2.5; range, 2.0-3.0) (Grade 1B) and consideration of 2.0.2. In patients with rheumatic mitral valve device closure over no VKA therapy or aspirin disease and normal sinus rhythm with a left therapy (Grade 2C) . atrial diameter . 55 mm, we suggest VKA therapy (target INR, 2.5; range, 2.0-3.0) over no 7.1 Role of Anticoagulants and Antiplatelet Agents in VKA therapy or antiplatelet (Grade 2C) . Patients With Native Valve Endocarditis 2.0.3. For patients with rheumatic mitral valve 7.1.1. In patients with infective endocarditis disease complicated by the presence of left (IE), we recommend against routine anticoagu- atrial thrombus, we recommend VKA therapy lant therapy, unless a separate indication exists (target INR, 2.5; range, 2.0-3.0) over no VKA (Grade 1C) . therapy (Grade 1A). 7.1.2. In patients with IE, we recommend 2.0.4. For patients with rheumatic mitral valve against routine antiplatelet therapy, unless a disease complicated singly or in combination by separate indication exists (Grade 1B). the presence of atrial fi brillation or previous 7.2 Role of Anticoagulants in Patients With Prosthetic systemic embolism, we recommend VKA therapy Valve Endocarditis (target INR, 2.5; range, 2.0-3.0) over no VKA therapy (Grade 1A). 7.2. In patients on VKA for a prosthetic valve who develop IE, we suggest VKA be discontin- 2.1 Patients With Rheumatic Mitral Valve Disease ued at the time of initial presentation until it Undergoing Percutaneous Mitral Balloon Valvotomy is clear that invasive procedures will not be (PMBV) required and the patient has stabilized without 2.1.1. For patients being considered for PMBV signs of CNS involvement. When the patient is with preprocedural TEE showing left atrial deemed stable without contraindications or neu- thrombus, we recommend postponement of rologic complications, we suggest reinstitution PMBV and that VKA therapy (target INR, 3.0; of VKA therapy (Grade 2C). range, 2.5-3.5) be administered until thrombus 7.3 Patients With Nonbacterial Thrombotic Endocarditis resolution is documented by repeat TEE over no VKA therapy (Grade 1A) . 7.3. In patients with nonbacterial thrombotic endocarditis and systemic or pulmonary emboli, 2.1.2. For patients being considered for PMBV we suggest treatment with full-dose IV UFH or with preprocedural TEE showing left atrial SC LMWH over no anticoagulation (Grade 2C). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 31S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 8.2 Antithrombotic Therapy in the First 3 Months 3.0 (range, 2.5-3.5) over lower INR targets After Surgery (Grade 2C) . 8.2.1. In patients with aortic bioprosthetic 9.5 Intensity of VKA Therapy in Patients With Double valves, who are in sinus rhythm and have no Mechanical Valve or With Additional Risk Factors other indication for VKA therapy, we suggest 9.5. In patients with mechanical heart valves in aspirin (50-100 mg/d) over VKA therapy in the both the aortic and mitral position, we suggest fi rst 3 months (Grade 2C). target INR 3.0 (range 2.5-3.5) over target INR 8.2.2. In patients with transcatheter aortic bio- 2.5 (range 2.0-3.0) (Grade 2C). prosthetic valves, we suggest aspirin (50-100 mg/d) 9.6 Antiplaelet Agent in Addition to VKA Therapy for plus clopidogrel (75 mg/d) over VKA therapy Patients With Mechanical Aortic or Mitral Valve and over no antiplatelet therapy in the fi rst Prostheses 3 months (Grade 2C) . 9.6. In patients with a mechanical mitral or aor- 8.2.3. In patients with a bioprosthetic valve in tic valve at low risk of bleeding, we suggest add- the mitral position, we suggest VKA therapy ing over not adding an antiplatelet agent such (target INR, 2.5; range, 2.0-3.0) over no VKA as low-dose aspirin (50-100 mg/d) to the VKA therapy for the fi rst 3 months after valve inser- therapy (Grade 1B). tion (Grade 2C). Remarks: Caution should be used in patients at 8.3 Long-term Antithrombotic Therapy for Patients increased bleeding risk, such as history of GI bleeding. With Bioprosthetic Valves 9.7 Antiplatelet Agent Therapy Instead of VKA Therapy 8.3. In patients with bioprosthetic valves in normal sinus rhythm, we suggest aspirin therapy 9.7. For patients with mechanical aortic or over no aspirin therapy after 3 months postop- mitral valves we recommend VKA over anti- erative (Grade 2C). platelet agents (Grade 1B). 9.1 Early Postoperative Bridging to Intermediate/ 10.1 Antithrombotic Therapy After Mitral Valve Long-term Therapy (Postoperative Day 0 to 5) Repair 9.1. In patients with mechanical heart valves, we 10.1. In patients undergoing mitral valve repair suggest bridging with unfractionated heparin with a prosthetic band in normal sinus rhythm, (UFH, prophylactic dose) or LMWH (prophy- we suggest the use of antiplatelet therapy for lactic or therapeutic dose) over IV therapeutic the fi rst 3 months over VKA therapy ( G r a d e 2 C ) . UFH until stable on VKA therapy (Grade 2C) . 10.2 Patients Undergoing Aortic Valve Repair 9.2 Long-term Antithrombotic Therapy for Patients 10.2. In patients undergoing aortic valve repair, With Mechanical Valves we suggest aspirin at 50 to 100 mg/d over VKA therapy (Grade 2C). 9.2. In patients with mechanical heart valves, we recommend VKA therapy over no VKA 11.1 Patients With Right-Sided Prosthetic Valve therapy for long-term management (Grade 1B). Thrombosis 9.3 Intensity of VKA Therapy for Patients With 11.1. For patients with right-sided prosthetic Mechanical Aortic Valve Prostheses valve thrombosis (PVT), in the absence of contraindications we suggest administration of 9.3.1. In patients with a mechanical aortic valve, fi brinolytic therapy over surgical intervention we suggest VKA therapy with a target of 2.5 (Grade 2C) . (range, 2.0-3.0) over lower targets (Grade 2C) . 11.2 Patients With Left-Sided Prosthetic Valve 9.3.2. In patients with a mechanical aortic Thrombosis valve, we recommend VKA therapy with a target of 2.5 (range 2.0-3.0) over higher targets 11.2.1. For patients with left-sided PVT and large ( G r a d e 1 B ). thrombus area (Ն 0.8 cm 2 ), we suggest early surgery over fi brinolytic therapy (Grade 2C). If con- 9.4 Intensity of VKA Therapy for Patients With traindications to surgery exist, we suggest the Mechanical Mitral Valve Prostheses use of fi brinolytic therapy (Grade 2C). 9.4. In patients with a mechanical mitral valve, 11.2.2. For patients with left-sided PVT and we suggest VKA therapy with a target of small thrombus area (, 0 . 8 c m2 ), we suggest

32S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians administration of fi brinolytic therapy over higher than the associated risks may choose this surgery. For very small, nonobstructive thrombus intervention. we suggest IV UFH accompanied by serial Dopp- ler echocardiography to document thrombus 2.4 Aspirin in Patients With Acute Ischemic Stroke resolution or improvement over other alterna- 2.4. In patients with acute ischemic stroke or tives (Grade 2C). transient ischemic attack (TIA), we recommend early (within 48 h) aspirin therapy at a dose of 160 to 325 mg over no aspirin therapy (Grade 1A). Antithrombotic and Thrombolytic Therapy for Ischemic Stroke 2.5 Anticoagulation in Patients With Acute Ischemic Stroke For further details, see Lansberg et al.11 2.5. In patients with acute ischemic stroke or 2.1 IV Recombinant Tissue Plasminogen Activator TIA, we recommend early (within 48 h) aspirin (r-tPA) for Patients With Acute Ischemic Stroke therapy with an initial dose of 160 to 325 mg over therapeutic parenteral anticoagulation 2.1.1. In patients with acute ischemic stroke in (Grade 1A). whom treatment can be initiated within 3 h of symptom onset, we recommend IV r-tPA over 3.1 VTE Prevention in Patients With Ischemic Stroke no IV r-tPA (Grade 1A). 3.1.1. In patients with acute ischemic stroke 2.1.2. In patients with acute ischemic stroke in and restricted mobility, we suggest prophylactic- whom treatment can be initiated within 4.5 but dose SC UFH or LMWH or intermittent pneu- not within 3 h of symptom onset, we suggest IV matic compression devices over no prophylaxis r-tPA over no IV r-tPA (Grade 2C). (Grade 2B). 2.1.3. In patients with acute ischemic stroke in 3.1.2. In patients with acute ischemic stroke and whom treatment cannot be initiated within 4.5 h restricted mobility, we suggest prophylactic-dose of symptom onset, we recommend against IV LMWH over prophylactic-dose UFH (Grade 2B). r-tPA (Grade 1B). 3.1.3. In patients with acute stroke and restricted 2.2 Intraarterial Thrombolysis in Patients With mobility, we suggest against elastic compression Acute Ischemic Stroke stockings (Grade 2B). 2.2.1. In patients with acute ischemic stroke due Remarks: Pharmacologic and mechanical prophy- to proximal cerebral artery occlusions who do laxis should be initiated as early as possible and should not meet eligibility criteria for treatment with be continued throughout the hospital stay or until IV r-tPA, we suggest intraarterial (IA) r-tPA the patient has regained mobility. Mechanical devices initiated within 6 h of symptom onset over no should be temporarily removed as often as needed IA r-tPA (Grade 2C). to allow for early mobilization and screening for skin complications. 2.2.2. In patients with acute ischemic stroke we suggest IV r-tPA over the combination IV/IA Combining pharmacologic therapy with intermittent r-tPA (Grade 2C). pneumatic compression devices may yield additional benefi t in prevention of VTEs compared with either Remarks: Carefully selected patients who value the method used alone. uncertain benefi ts of combination IV/IA thrombolysis higher than the associated risks may choose 3.2 VTE Prevention in Patients With Hemorrhagic this intervention. Patients who prefer to avoid risk in Stroke the setting of uncertain benefi ts are more likely to 3.2.1. In patients with acute primary intracere- choose IV r-tPA alone. bral hemorrhage and restricted mobility, we suggest prophylactic-dose SC heparin (UFH or 2.3 Mechanical Thrombectomy in Patients With LMWH) started between days 2 and 4 or inter- Acute Ischemic Stroke mittent pneumatic compression devices over 2.3. In patients with acute ischemic stroke, we no prophylaxis (Grade 2C). suggest against the use of mechanical throm- 3.2.2. In patients with acute primary intrac- bectomy (Grade 2C). erebral hemorrhage and restricted mobility, Remarks: Carefully selected patients who value we suggest prophylactic-dose LMWH over the uncertain benefi t of mechanical thrombectomy prophylactic-dose UFH (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 33S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.2.3. In patients with primary intracerebral recommend combination therapy with aspirin hemorrhage and restricted mobility, we suggest and clopidogrel over aspirin (Grade 1B). against elastic compression stockings (Grade 2B). Remarks: Patients should be treated (ie, bridged) Remarks: Patients who prefer to avoid a theoreti- with aspirin until anticoagulation has reached a cally increased risk of rebleeding with heparin therapeutic level. would favor mechanical prophylaxis with intermittent pneumatic compression devices over pharmaco- Oral anticoagulation should generally be initiated logic prophylaxis. within 1 to 2 weeks after stroke onset. Earlier anticoagulation can be considered for patients at low Combining pharmacologic therapy with intermit- risk of bleeding complications (eg, those with a small tent pneumatic compression devices may yield addi- infarct burden and no evidence of hemorrhage on tional benefi t in prevention of VTEs compared with brain imaging). Delaying anticoagulation should be either method used alone. considered for patients at high risk of hemorrhagic complications (eg, those with extensive infarct bur- 4.1 Antithrombotic Therapy for the Secondary den or evidence of signifi cant hemorrhagic transfor- Prevention of Noncardioembolic Stroke mation on brain imaging). 4.1.1. In patients with a history of noncardioem- Dabigatran is excreted primarily by the kidney. It has bolic ischemic stroke or TIA, we recommend not been studied and is contraindicated in patients long-term treatment with aspirin (75-100 mg once with severe renal impairment (estimated creatinine daily), clopidogrel (75 mg once daily), aspirin/ clearance of 30 mL/min or less). extended-release dipyridamole (25 mg/200 mg bid), or cilostazol (100 mg bid) over no anti- 4.3 Antithrombotic Therapy for Stroke Prevention in platelet therapy (Grade 1A), oral anticoagulants Patients With a History of Intracerebral Hemorrhage (Grade 1B), the combination of clopidogrel plus (ICH) aspirin (Grade 1B), or trifl usal (grade 2B). 4.3. In patients with a history of a symptomatic 4.1.2. Of the recommended antiplatelet regimens, primary ICH, we suggest against the long-term we suggest clopidogrel or aspirin/extended- use of antithrombotic therapy for the preven- release dipyridamole over aspirin (Grade 2B) or tion of ischemic stroke (Grade 2C). cilostazol (Grade 2C). Remarks: Patients who might benefi t from antithrom- Remarks: With long-term use (. 5 y), the benefi t of botic therapy are those at relatively low risk of recur- clopidogrel over aspirin in preventing major vascular rent ICH (eg, with deep hemorrhages) and relatively events may be offset by a reduction in cancer-related high risk (. 7% per year) of throm boembolic events mortality with regimens that contain aspirin. (eg, with mechanical heart valves or CHADS 2 (Con- gestive heart failure, Hypertension, Age . 75, Diabetes 4.2 Antithrombotic Therapy for the Secondary mellitus, Stroke or TIA) score . 4 points). Prevention of Cardioembolic Stroke 5.1 Anticoagulation for Patients With Symptomatic 4.2.1. In patients with a history of ischemic Cerebral Venous Sinus Thrombosis stroke or TIA and AF, including paroxysmal AF, we recommend oral anticoagulation over 5.1. In patients with cerebral venous sinus no antithrombotic therapy (Grade 1A), aspirin thrombosis, we suggest anticoagulation over (Grade 1B), or combination therapy with aspirin no anticoagulant therapy during the acute and and clopidogrel (Grade 1B). chronic phases (Grade 2C). 4.2.2. In patients with a history of ischemic Remarks: Patients with a history of ICH who might stroke or TIA and atrial fi brillation, including benefi t from antithrombotic therapy are those at rel- paroxysmal AF, we suggest oral anticoagulation atively low risk of recurrent ICH (eg, with deep hem- . with dabigatran 150 mg bid over adjusted-dose orrhages) and relatively high risk ( 7% per year) of VKA therapy (target range, 2.0-3.0) (Grade 2B). cardiac thromboembolic events (eg, with mechanical heart valves or CHADS2 score . 4 points). 4.2.3. In patients with a history of ischemic stroke or TIA and atrial fi brillation, including paroxysmal AF, who are unsuitable for or choose The Primary and Secondary Prevention not to take an oral anticoagulant (for reasons of Cardiovascular Disease other than concerns about major bleeding), we For further details, see Vandvik et al.12

34S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 2.0 Primary Prevention of Cardiovascular Disease daily plus low-dose aspirin, or prasugrel 10 mg daily plus low-dose aspirin over 2.1. For persons aged 50 years or older without single antiplatelet therapy) (Grade 1B) . symptomatic cardiovascular disease, we suggest low-dose aspirin 75 to 100 mg daily over no Remarks: Evidence suggests that prasugrel results in aspirin therapy (Grade 2B). no benefi t net harm in patients with a body weight of , 6 0 k g , a g e . 75 years, or with a previous stroke/ Remarks: Aspirin slightly reduces total mortality transient ischemic attack. regardless of cardiovascular risk profi le if taken over 10 years. In people at moderate to high risk of cardio- • We suggest ticagrelor 90 mg twice daily vascular events, the reduction in myocardial infarc- plus low-dose aspirin over clopidogrel tion (MI) is closely balanced with an increase in major 75 mg daily plus low-dose aspirin ( G r a d e 2 B ). bleeds. Whatever their risk status, people who are For patients with ACS who undergo PCI with averse to taking medication over a prolonged time period stent placement, we refer to sections 4.3.1 to for very small benefi ts will be disinclined to use aspirin 4.3.5 for recommendations concerning minimum for primary prophylaxis. Individuals who value pre- and prolonged duration of treatment. venting an MI substantially higher than avoiding a GI bleed will be, if they are in the moderate or high car- 3.2.6-3.2.7. For patients with anterior MI and left diovascular risk group, more likely to choose aspirin. ventricular (LV) thrombus, or at high risk for LV thrombus (ejection fraction , 40%, anteroapical 3.1 Choice of Long-term Antithrombotic Therapy in wall motion abnormality), who do not undergo Patients With Established Coronary Artery Disease (CAD) stenting: 3.1.1-3.1.5. For patients with established coronary • We recommend warfarin (INR 2.0-3.0) plus artery disease (CAD), defi ned as patients 1-year low-dose aspirin 75 to 100 mg daily over post-acute coronary syndrome (ACS), with prior single antiplatelet therapy or dual antiplatelet revascularization, coronary stenoses . 50% by therapy for the fi rst 3 months (Grade 1B). coronary angiogram, and/or evidence for cardiac Thereafter, we recommend discontinuation ischemia on diagnostic testing, (including patients of warfarin and continuation of dual anti- after the fi rst year post-ACS and/or with prior platelet therapy for up to 12 months as per coronary artery bypass graft [CABG] surgery): the ACS recommendations (see recommen- • We recommend long-term single antiplate- dations 3.2.1-3.2.5). After 12 months, single let therapy with aspirin 75 to 100 mg daily antiplatelet therapy is recommended as or clopidogrel 75 mg daily over no anti- per the established CAD recommendations platelet therapy (Grade 1A) . (see recommendations 3.1.1-3.1.5). • We suggest single over dual antiplatelet ther- For patients with anterior MI and LV thrombus, apy with aspirin plus clopidogrel ( G r a d e 2 B ). or at high risk for LV thrombus (ejection frac- 3.2 Choice of Antithrombotic Therapy Following ACS tion , 40%, anteroapical wall motion abnormality), who undergo bare-metal stent (BMS) 3.2.1-3.2.5. For patients in the fi rst year after an placement: ACS who have not undergone percutaneous coronary intervention (PCI): • We suggest triple therapy (warfarin [INR 2.0-3.0], low-dose aspirin, clopidogrel 75 mg • We recommend dual antiplatelet therapy daily) for 1 month over dual antiplatelet (ticagrelor 90 mg twice daily plus low-dose therapy (Grade 2C) . aspirin 75-100 mg daily or clopidogrel 75 mg • We suggest warfarin (INR 2.0-3.0) and single daily plus low-dose aspirin 75-100 mg daily) antiplatelet therapy for the second and third over single antiplatelet therapy ( G r a d e 1 B ). month post-BMS over alternative regimens • We suggest ticagrelor 90 mg daily plus low- and alternative time frames for warfarin dose aspirin over clopidogrel 75 mg daily use ( G r a d e 2 C ). Thereafter, we recommend plus low-dose aspirin (Grade 2B). discontinuation of warfarin and use of dual For patients in the fi rst year after an ACS who antiplatelet therapy for up to 12 months as have undergone PCI with stent placement: per the ACS recommendations (see recommendations 3.2.1-3.2.5). After 12 months, • We recommend dual antiplatelet therapy antiplatelet therapy is recommended as (ticagrelor 90 mg twice daily plus low-dose per the established CAD recommendations aspirin 75-100 mg daily, clopidogrel 75 mg (see recommendations 3.1.1-3.1.5). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 35S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians For patients with anterior MI and LV thrombus, as per the established CAD recommenda- or at high risk for LV thrombus (ejection fractions (see recommendations 3.1.1-3.1.5). tion , 40%, anteroapical wall motion abnor- For patients who have undergone elective BMS mality) who undergo drug-eluting stent (DES) or DES stent placement: placement: • We recommend using low-dose aspirin 75 • We suggest triple therapy (warfarin INR to 100 mg daily and clopidogrel 75 mg daily 2.0-3.0, low-dose aspirin, clopidogrel 75 mg alone rather than cilostazol in addition to daily) for 3 to 6 months over alternative these drugs (Grade 1B) . regimens and alternative durations of war- • We suggest aspirin 75 to 100 mg daily or farin therapy (Grade 2C). Thereafter, we clopidogrel 75 mg daily as part of dual anti- recommend discontinuation of warfarin and platelet therapy rather than the use of continuation of dual antiplatelet therapy for either drug with cilostazol (Grade 1B). up to 12 months as per the ACS recommen- • We suggest cilostazol 100 mg twice daily as dations (see recommendations 3.2.1-3.2.5). substitute for either low-dose aspirin 75 to After 12 months, antiplatelet therapy is 100 mg daily or clopidogrel 75 mg daily recommended as per the established CAD as part of a dual antiplatelet regimen in recommendations (see recommendations patients with an allergy or intolerance of 3.1.1-3.1.5). either drug class (Grade 2C). 4.0 Antithrombotic Therapy Following Elective PCI For patients with CAD undergoing elective PCI 4.1.1-4.3.5. For patients who have undergone but no stent placement: elective PCI with placement of BMS: • We suggest for the fi rst month dual antiplatelet therapy with aspirin 75 to 325 mg daily • For the fi rst month, we recommend dual and clopidogrel 75 mg daily over single anti- anti platelet therapy with aspirin 75 to platelet therapy ( G r a d e 2 C ). Single antiplate- 325 mg daily and clopidogrel 75 mg daily let therapy thereafter is recommended as over single antiplatelet therapy (Grade 1A). per the established CAD recommendations • For the subsequent 11 months, we suggest (see recommendations 3.1.1-3.1.5). dual antiplatelet therapy with combination of low-dose aspirin 75 to 100 mg daily and 5.0 Antithrombotic Therapy in Patients With Systolic clopidogrel 75 mg daily over single anti- LV Dysfunction platelet therapy (Grade 2C) . 5.1-5.3. For patients with systolic LV dysfunc- • After 12 months, we recommend single tion without established CAD and no LV throm- antiplatelet therapy over continuation of bus, we suggest not to use antiplatelet therapy dual antiplatelet therapy (Grade 1B). or warfarin (Grade 2C) . For patients who have undergone elective PCI Remarks: Patients who place a high value on an with placement of DES: uncertain reduction in stroke and a low value on avoiding an increased risk of GI bleeding are likely to • For the fi rst 3 to 6 months, we recommend choose to use warfarin. dual antiplatelet therapy with aspirin 75 to 325 mg daily and clopidogrel 75 mg daily For patients with systolic LV dysfunction with- over single antiplatelet therapy (Grade 1A). out established CAD with identifi ed acute LV thrombus (eg, Takotsubo cardiomyopathy), Remarks: Absolute minimum duration will vary based we suggest moderate-intensity warfarin (INR on stent type (in general, 3 months for -limus stents 2.0-3.0) for at least 3 months (Grade 2C) . and 6 months for -taxel stents). For patients with systolic LV dysfunction and • After 3 to 6 months, we suggest continua- established CAD, recommendations are as per tion of dual antiplatelet therapy with low- the established CAD recommendations (see rec- dose aspirin 75 to 100 mg and clopidogrel ommendations 3.1.1-3.1.5). (75 mg daily) until 12 months over single antiplatelet therapy (Grade 2C) . • After 12 months, we recommend single antiplatelet therapy over continuation of dual Antithrombotic Therapy in antiplatelet therapy ( G r a d e 1 B ) . Single anti- Peripheral Artery Disease platelet therapy thereafter is recommended For further details, see Alonso-Coello et al.13

36S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 2.0 Primary Prevention of Cardiovascular Events than avoidance of a high likelihood of drug-related in Patients with Asymptomatic PAD side effects will be disinclined to take prostanoids. 2.1. For persons with asymptomatic peripheral 6.0 Acute Limb Ischemia arterial disease (PAD), we suggest aspirin 75 to 100 mg daily over no aspirin therapy ( G r a d e 2 B ). 6.1-6.3. In patients with acute limb ischemia due to arterial emboli or thrombosis, we sug- Remarks: Aspirin slightly reduces total mortality gest immediate systemic anticoagulation with regardless of cardiovascular risk profi le if taken over unfractionated heparin over no anticoagulation 10 years. In people at moderate to high risk of cardio- (Grade 2C); we suggest reperfusion therapy (sur- vascular events, the reduction in myocardial infarc- gery or IA thrombolysis) over no reperfusion tion (MI) is closely balanced with an increase in major therapy ( G r a d e 2 C ); we recommend surgery bleeds. Whatever their risk status, people who are over IA thrombolysis ( G r a d e 1 B ). In patients averse to taking medication over a prolonged time undergoing IA thrombolysis, we suggest recom- period for very small benefi ts will be disinclined to binant tissue-type plasminogen activator (rt-PA) use aspirin for primary prophylaxis. Individuals who or urokinase over streptokinase (Grade 2C) . value preventing an MI substantially higher than avoiding a GI bleed, if they are in the moderate or 7.0 Endovascular Revascularization in Patients With high cardiovascular risk group, will be more likely to Symptomatic PAD choose aspirin. 7.1. For patients undergoing peripheral artery percutaneous transluminal angioplasty with 3.0 Secondary Prevention of Cardiovascular Events or without stenting, we recommend long- in Patients with Symptomatic PAD term aspirin (75-100 mg/day) or clopidogrel 3.1-3.4. For secondary prevention patients with (75 mg/day) ( G r a d e 1 A ). For patients under- symptomatic PAD, we recommend one of the going peripheral artery percutaneous translu- two following antithrombotic regimens to be minal angioplasty with stenting, we suggest continued long term over no antithrombotic single rather than dual antiplatelet therapy treatment: aspirin 75 to 100 mg daily or clopid- (Grade 2C) . ogrel 75 mg daily (all Grade 1A). We suggest not to use dual antiplatelet therapy with aspirin plus Values and preferences: Patients who place a high clopidogrel (Grade 2B) . We recommend not to value on an uncertain reduction in the risk of limb use an antiplatelet agent with moderate-inten- loss and a relatively low value on avoiding a defi nite sity warfarin (Grade 1B) . increased risk of bleeding are more likely to choose to use dual antiplatelet therapy. 4.0 Antithrombotic Therapy for the Management of Patients with Claudication 8.0 Antithrombotic Therapy Following Peripheral Artery Bypass Graft Surgery 4.1-4.4. For patients with intermittent claudication refractory to exercise therapy (and smoking 8.1-8.4. We recommend one of the following cessation), we suggest the use of cilostazol in antithrombotic regimens to be continued long- addition to previously recommended antithrom- term following peripheral artery bypass graft botic therapies (aspirin 75-100 mg daily or surgery over no antithrombotic treatment: clopidogrel 75 mg daily) (Grade 2C) ; we suggest aspirin 75 to 100 mg daily or clopidogrel 75 mg against the use of pentoxifylline, heparinoids, daily (all Grade 1A) . We recommend single anti- or prostanoids (Grade 2C) . platelet therapy over antiplatelet therapy and warfarin ( G r a d e 1 B ) . In patients undergoing 5.0 Patients With Critical Limb Ischemia below-knee bypass graft surgery with prosthetic grafts, we suggest clopidogrel 75 mg/d plus 5.1. For patients with symptomatic PAD and aspirin (75-100 mg/d) over aspirin alone for critical leg ischemia/rest pain who are not can- 1 year ( G r a d e 2 C ). For all other patients, we didates for vascular intervention, we suggest suggest single over dual antiplatelet therapy the use of prostanoids in addition to previously (Grade 2B). recommended antithrombotic therapies (aspirin 75-100 mg daily or clopidogrel 75 mg daily) 9.0 Patients With Carotid Artery Stenosis (Grade 2C). 9.1. For patients with asymptomatic carotid ste- Values and preferences: Patients who do not value nosis, we suggest aspirin 75 to 100 mg daily over uncertain relief of rest pain and ulcer healing greater no aspirin therapy (Grade 2B). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 37S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Remarks: Aspirin slightly reduces total mortality 3.0.3. For pregnant women, we suggest lim- regardless of cardiovascular risk profi le if taken over iting the use of fondaparinux and parenteral 10 years. In people at moderate to high risk of car- direct thrombin inhibitors to those with severe diovascular events, the reduction in MI is closely bal- allergic reactions to heparin (eg, HIT) who anced with an increase in major bleeds. Whatever cannot receive danaparoid (Grade 2C) . their risk status, people who are averse to taking medication over a prolonged time period for very 3.0.4. For pregnant women, we recommend small benefi ts will be disinclined to use aspirin for avoiding the use of oral direct thrombin (eg, primary prophylaxis. dabigatran) and anti-Xa (eg, rivaroxaban, apixaban) inhibitors (Grade 1C). 9.2-9.3. In patients with symptomatic carotid stenosis (including recent carotid endarterectomy), 4.0 Use of Antithrombotic Therapy in Nursing Women we recommend long-term antiplatelet therapy 4.0.1. For lactating women using warfarin, with clopidogrel (75 mg once daily) or aspirin- acenocoumarol, or UFH who wish to breast- extended-release dipyridamole (25 mg/200 mg bid) feed, we recommend continuing the use of war- or aspirin (75-100 mg once daily) over no anti- farin, acenocoumarol, or UFH (Grade 1A). platelet therapy (Grade 1A) . We suggest either clopidogrel (75 mg once daily) or aspirin-extended- 4.0.2. For lactating women using LMWH, dana- release dipyridamole (25 mg/200 mg bid) over paroid, or r-hirudin who wish to breast-feed, aspirin (75-100 mg) (Grade 2B) . we recommend continuing the use of LMWH, danaparoid, or r-hirudin (Grade 1B). 4.0.3. For breast-feeding women, we suggest VTE, Thrombophilia, Antithrombotic alternative anticoagulants rather than fonda- Therapy, and Pregnancy parinux (Grade 2C). For further details, see Bates et al.14 4.0.4. For breast-feeding women, we recom- 2.0 Maternal Consequences of Antithrombotic Therapy mend alternative anticoagulants rather than Use During Pregnancy oral direct thrombin (eg, dabigatran) and factor Xa inhibitors (eg, rivaroxaban, apixaban) 2.2.1. For pregnant patients, we recommend (Grade 1C) . LMWH for the prevention and treatment of VTE, instead of UFH (Grade 1B). 4.0.5. For lactating women using low-dose aspirin for vascular indications who wish to breast- 3.0 Fetal Consequence of Antithrombotic Therapy feed, we suggest continuing this medication Use in Pregnant Women (Grade 2C) . 3.0.1. For women receiving anticoagulation for 5.0 VTE in Patients Using Assisted Reproductive the treatment of VTE who become pregnant, Technology we recommend LMWH over VKAs during the fi rst trimester ( G r a d e 1 A ), in the second and 5.1.1. For women undergoing assisted repro- third trimesters ( G r a d e 1 B ), and during late preg- duction, we recommend against the use of rou- nancy when delivery is imminent (Grade 1A). tine thrombosis prophylaxis (Grade 1B).

3.0.2. For women requiring long-term VKAs who 5.1.2. For women undergoing assisted repro- are attempting pregnancy and are candidates duction who develop severe ovarian hyperstim- for LMWH substitution, we suggest perform- ulation syndrome, we suggest thrombosis ing frequent pregnancy tests and substituting prophylaxis (prophylactic LMWH) for 3 months LMWH for VKAs when pregnancy is achieved postresolution of clinical ovarian hyperstimu- rather than switching to LMWH while attempt- lation syndrome rather than no prophylaxis ing pregnancy (Grade 2C) . (Grade 2C) .

Remarks : Women who place little value on avoid- Remarks : Women who are averse to taking medica- ing the risks, inconvenience, and costs of LMWH tion for very small beneﬁ t and those who consider therapy of uncertain duration while awaiting preg- self-injecting a considerable burden will be disin- nancy and a high value on minimizing the risks of clined to use LMWH for extended thrombosis pro- early miscarriage associated with VKA therapy are phylaxis. Given that the absolute beneﬁ t decreases as likely to choose LMWH while attempting pregnancy. time from the hyperstimulation event increases, such

38S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians women will be very disinclined to continue prophy- LMWH at least 24 h prior to induction of labor laxis throughout the entire resultant pregnancy. or cesarean section (or expected time of neurax- ial anesthesia) rather than continuing LMWH 6.0 VTE Following Cesarean Section up until the time of delivery (Grade 1B) . 6.2.1. For women undergoing cesarean section without additional thrombosis risk factors, we 8.0 Prevention of Recurrent VTE in Pregnant Women recommend against the use of thrombosis pro- 8.2.1. For all pregnant women with prior VTE, phylaxis other than early mobilization ( G r a d e 1 B ). we suggest postpartum prophylaxis for 6 weeks with prophylactic- or intermediate-dose LMWH 6.2.2. For women at increased risk of VTE after or VKAs targeted at INR 2.0 to 3.0 rather than cesarean section because of the presence of no prophylaxis (Grade 2B) . one major or at least two minor risk factors, we suggest pharmacologic thromboprophylaxis 8.2.2. For pregnant women at low risk of recur- (prophylactic LMWH) or mechanical prophy- rent VTE (single episode of VTE associated with laxis (elastic stockings or intermittent pneumatic a transient risk factor not related to pregnancy compression) in those with contraindications or use of estrogen), we suggest clinical vigilance to anticoagulants while in hospital following antepartum rather than antepartum prophy- delivery rather than no prophylaxis (Grade 2B). laxis (Grade 2C) . Remarks: The reduced bleeding risk with mechanical 8.2.3. For pregnant women at moderate to high prophylaxis should be weighed against the inconve- risk of recurrent VTE (single unprovoked VTE, nience of elastic stockings and intermittent pneu- pregnancy- or estrogen-related VTE, or mul- matic compression. tiple prior unprovoked VTE not receiving long- 6.2.3. For women undergoing cesarean section term anticoagulation), we suggest antepartum who are considered to be at very high risk for VTE prophylaxis with prophylactic- or intermediate- and who have multiple additional risk factors for dose LMWH rather than clinical vigilance or thromboembolism that persist in the puerperium, routine care (Grade 2C). we suggest that prophylactic LMWH be combined with elastic stockings and/or intermittent pneu- 8.2.4. For pregnant women receiving long- matic compression over LMWH alone ( G r a d e 2 C ). term VKAs, we suggest adjusted-dose LMWH or 75% of a therapeutic dose of LMWH through- 6.2.4. For selected high-risk patients in whom out pregnancy followed by resumption of long- signiﬁ cant risk factors persist following delivery, term anticoagulants postpartum, rather than we suggest extended prophylaxis (up to 6 weeks prophylactic-dose LMWH (Grade 2C). after delivery) following discharge from the hospital (Grade 2C). 9.0 Prevention of VTE in Pregnant Women With Thrompophilia and No Prior VTE 7.0 Treatment of Patients With Proven Acute VTE During Pregnancy 9.2.1. For pregnant women with no prior history of VTE who are known to be homozygous 7.1.1. For pregnant women with acute VTE, for factor V Leiden or the prothrombin 20210A we recommend therapy with adjusted-dose mutation and have a positive family history for SC LMWH over adjusted-dose UFH (Grade 1B). VTE, we suggest antepartum prophylaxis with prophylactic- or intermediate-dose LMWH and 7.1.2. For pregnant women with acute VTE, we postpartum prophylaxis for 6 weeks with pro- recommend LMWH over VKA treatment ante- phylactic- or intermediate-dose LMWH or VKAs natally (Grade 1A). targeted at INR 2.0 to 3.0 rather than no pro- 7.1.3. For pregnant women with acute VTE, we phylaxis (Grade 2B). suggest that anticoagulants should be continued for at least 6 weeks postpartum (for a min- 9.2.2. For pregnant women with all other throm- imum total duration of therapy of 3 months) in bophilias and no prior VTE who have a positive comparison with shorter durations of treat- family history for VTE, we suggest antepartum ment ( G r a d e 2 C ). clinical vigilance and postpartum prophylaxis with prophylactic- or intermediate-dose LMWH 7.1.4. For pregnant women receiving adjusted- or, in women who are not protein C or S deﬁ - dose LMWH therapy and where delivery is cient, VKAs targeted at INR 2.0 to 3.0 rather planned, we recommend discontinuation of than routine care (Grade 2C). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 39S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 9.2.3. For pregnant women with no prior his- anticoagulant regimens in preference to no tory of VTE who are known to be homozygous anticoagulation (all Grade 1A): for factor V Leiden or the prothrombin 20210A mutation and who do not have a positive family (a) Adjusted-dose bid LMWH throughout preg- history for VTE, we suggest antepartum clin- nancy. We suggest that doses be adjusted to ical vigilance and postpartum prophylaxis for achieve the manufacturer’s peak anti-Xa LMWH 6 weeks with prophylactic- or intermediate- 4 h postsubcutaneous-injection or dose LMWH or VKAs targeted at INR 2.0 to (b) Adjusted-dose UFH throughout pregnancy 3.0 rather than routine care (Grade 2B). administered subcutaneously every 12 h in doses adjusted to keep the mid-interval acti- 9.2.4. For pregnant women with all other throm- vated partial thromboplastin time at least twice bophilias and no prior VTE who do not have a control or attain an anti-Xa heparin level of positive family history for VTE, we suggest 0.35 to 0.70 units/mL or antepartum and postpartum clinical vigilance rather than pharmacologic prophylaxis ( G r a d e 2 C ) . (c) UFH or LMWH (as above) until the 13th week, with substitution by VKAs until close to delivery 10.0 Prevention of Pregnancy Complications in when UFH or LMWH is resumed. Women With Thrombophilia Remarks: For pregnant women with mechanical heart 10.2.1. For women with recurrent early preg- valves, the decision regarding the choice of anti- nancy loss (three or more miscarriages before coagulant regimen is so value and preference depen- 10 weeks of gestation), we recommend screening dent (risk of thrombosis vs risk of fetal abnormalities) for antiphospolipid antibodies (APLAs) ( G r a d e 1 B ). that we consider the decision to be completely indi- 10.2.2. For women with a history of pregnancy vidualized. Women of childbearing age and pregnant complications, we suggest not to screen for women with mechanical valves, should be counseled inherited thrombophilia (Grade 2C). about potential maternal and fetal risks associated with various anticoagulant regimens, including con- 10.2.3. For women who fulﬁ ll the laboratory tinuation of VKAs with substitution by LMWH or criteria for APLA syndrome and meet the clin- UFH close to term, substitution of VKAs by LMWH ical APLA criteria based on a history of three or UFH until the 13th week and then close to term, or more pregnancy losses, we recommend and use of LMWH or UFH throughout pregnancy. antepartum administration of prophylactic- or Usual long-term anticoagulants should be resumed intermediate-dose UFH or prophylactic LMWH postpartum when adequate hemostasis is assured. combined with low-dose aspirin, 75 to 100 mg/d, over no treatment (Grade 1B). 12.1.2. In women judged to be at very high risk of thromboembolism in whom concerns exist 10.2.4. For women with inherited thrombo- about the efﬁ cacy and safety of UFH or LMWH philia and a history of pregnancy complications, as dosed above (eg, older generation prosthesis we suggest not to use antithrombotic prophy- in the mitral position or history of thrombo- laxis (Grade 2C). embolism), we suggest VKAs throughout pregnancy with replacement by UFH or LMWH 11.0 Prevention of Recurrent Preeclampsia or Pregnancy (as above) close to delivery rather than one of Loss in Women Without Known Thrombophilia the regimens above (Grade 2C). 11.1.1. For women considered at risk for pre- Remarks: Women who place a higher value on avoid- eclampsia, we recommend low-dose aspirin ing fetal risk than on avoiding maternal complica- throughout pregnancy, starting from the second tions (eg, catastrophic valve thrombosis) are likely to trimester, over no treatment (Grade 1B). choose LMWH or UFH over VKAs. 11.2.1. For women with two or more miscar- 12.1.3. For pregnant women with prosthetic riages but without APLA or thrombophilia, we valves at high risk of thromboembolism, we recommend against antithrombotic prophylaxis suggest the addition of low-dose aspirin, 75 to (Grade 1B). 100 mg/d (Grade 2C) . 12.0 Prevention of Thromboembolism in Pregnant Women With Mechanical Heart Valves Antithrombotic Therapy in 12.1.1. For pregnant women with mechanical Neonates and Children heart valves, we recommend one of the following For further details, see Monagle et al.15

40S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 1.0. We suggest that where possible, pediatric We suggest either initial anticoagulation or hematologists with experience in thromboem- supportive care with radiologic monitoring bolism manage pediatric patients with throm- for extension of thrombosis rather than no boembolism ( G r a d e 2 C ). When this is not possible, follow-up ( G r a d e 2 C ) ; however, in previously we suggest a combination of a neonatologist/ untreated patients, we recommend the start of pediatrician and adult hematologist supported anticoagulation if extension occurs (Grade 2C). by consultation with an experienced pediatric We suggest that anticoagulation should be with hematologist (Grade 2C) . either (1) LMWH or (2) UFH followed by LMWH. We suggest a total duration of anticoagulation 1.1 Heparin in Neonates and Children of between 6 weeks and 3 months rather than 1.1. We suggest that therapeutic unfraction- shorter or longer durations (Grade 2C). If either ated heparin (UFH) in children is titrated to a CVAD or a UVC is still in place on completion achieve a target range of anti-Xa activity of of therapeutic anticoagulation, we suggest a 0.35 to 0.7 units/mL or an activated partial prophylactic dose of anticoagulation until such thromboplastin time range that correlates to time as the CVAD or UVC is removed ( G r a d e 2 C ) . this anti-Xa range or to a protamine titration We suggest against thrombolytic therapy for range of 0.2 to 0.4 units/mL (Grade 2C). We sug- neonatal VTE unless major vessel occlusion is gest that when initiating UFH therapy, UFH causing critical compromise of organs or limbs boluses be no greater than 75 to 100 units/kg ( G r a d e 2 C ). We suggest if thrombolysis is required, and that boluses be withheld or reduced if there tissue plasminogen activator (tPA) is used rather are signiﬁ cant bleeding risks ( G r a d e 2 C ). We than other lytic agents ( G r a d e 2 C ), and we suggest avoiding long-term use of therapeutic suggest plasminogen (fresh frozen plasma) UFH in children (Grade 2C) . administration prior to commencing therapy (Grade 2C) . 1.2 LMWH in Neonates and Children 2.2-2.3 Renal Vein Thrombosis in Neonates 1.2. We suggest, for neonates and children receiving either once- or twice-daily therapeutic 2.2. For unilateral renal vein thrombosis (RVT) LMWH that the drug be monitored to a target in the absence of renal impairment or extension anti-Xa activity range of 0.5 to 1.0 units/mL in a into the inferior vena cava (IVC), we suggest sample taken 4 to 6 h after SC injection or 0.5 either (1) supportive care with radiologic moni- to 0.8 units/mL in a sample taken 2 to 6 h after toring for extension of thrombosis (if extension SC injection (Grade 2C) . occurs we suggest anticoagulation) or (2) anticoagulation with UFH/LMWH or LMWH in 1.3 VKAs in Neonates and Children therapeutic doses rather than no therapy. If 1.3. We suggest, for children receiving VKAs, anticoagulation is used, we suggest a total dura- that the drug be monitored to a target INR of tion of between 6 weeks and 3 months rather 2.5 (range, 2.0-3.0), except in the setting of pros- than shorter or longer durations of therapy thetic cardiac valves where we suggest adher- (Grade 2C) . For unilateral RVT that extends into ence to the adult recommendations outlined in the IVC, we suggest anticoagulation with UFH/ the article by Whitlock et al in this supplement LMWH or LMWH for a total duration of ( G r a d e 2 C ). We suggest that INR monitoring between 6 weeks and 3 months (Grade 2C). with point-of-care monitors be made available 2.3. For bilateral RVT with evidence of renal where resources make this possible (Grade 2C). impairment, we suggest anticoagulation with UFH/LMWH or initial thrombolytic therapy 1.5 Aspirin in Children with tPA followed by anticoagulation with UFH/ 1.5. We suggest that when aspirin is used for LMWH (Grade 2C) . antiplatelet therapy in children, it is used in doses of 1 to 5 mg/kg per day (Grade 2C). 2.4 CVAD Prophylaxis in Neonates 2.4. For neonates with CVADs, we recommend 2.1 VTE in Neonates to maintain CVAD patency with UFH contin- 2.1. We suggest that central venous access uous infusion at 0.5 units/kg per h over no devices (CVADs) or umbilical venous catheters prophylaxis ( G r a d e 1 A ) or intermittent local (UVCs) associated with conﬁ rmed thrombosis thrombolysis ( G r a d e 2 C ). For neonates with be removed after 3 to 5 days of therapeutic anti- blocked CVADs, we suggest local thrombolysis coagulation rather than left in situ (Grade 2C) . after appropriate clinical assessment ( G r a d e 2 C ). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 41S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 2.6 Thromboprophylaxis for Neonates and Children 2.13. For neonates with umbilical arterial cath- With Blalock-Taussig Shunts and Modifi ed Blalock- eters (UACs), we suggest UAC placement in a Taussig Shunts (MBTS) high rather than a low position (Grade 2B). 2.6. For neonates and children having modi- 2.14. For neonates with UAC, we suggest profi ed MBTS, we suggest intraoperative UFH phylaxis with a low-dose UFH infusion via the therapy ( G r a d e 2 C ). For neonates and children UAC (heparin concentration of 0.25-1 unit/mL, after MBTS surgery, we suggest either aspirin total heparin dose of 25-200 units/kg per day) to or no antithrombotic therapy as compared with maintain patency (Grade 2A). prolonged LMWH or VKAs (Grade 2C) . 2.16 Prophylaxis for Cardiac Catheterization in 2.9-2.10 Therapy for Femoral Artery Thrombosis in Neonates and Children Neonates and Children 2.16. For neonates and children requiring 2.9. For neonates and children with acute fem- cardiac catheterization via an artery, we recom- oral artery thrombosis, we recommend thera- mend administration of IV UFH as thrombo- peutic doses of IV UFH as initial therapy prophylaxis over no prophylaxis (Grade 1A) compared with aspirin or no therapy (Grade 1B) or aspirin ( G r a d e 1 B ). For neonates and chil- or LMWH (Grade 2C). We suggest subsequent dren requiring cardiac catheterization via an conversion to LMWH, or else continuation of artery, we recommend the use of UFH doses UFH, to complete 5 to 7 days of therapeutic of 100 units/kg as a bolus compared with a anticoagulation as compared with a shorter or 50-unit/kg bolus ( G r a d e 1 B ). In prolonged pro- longer duration (Grade 2C). cedures, we suggest further doses of UFH rather than no further therapy (Grade 2B). 2.10. For neonates and children with limb- threatening or organ-threatening (via proximal 2.17 Cerebral Sinovenous Thrombosis in Neonates extension) femoral artery thrombosis who fail to 2.17. For neonates with cerebral sinovenous respond to initial UFH therapy and who have thrombosis (CSVT) without signifi cant intracra- no known contraindications, we recommend nial hemorrhage, we suggest anticoagulation, thrombolysis (Grade 1C). For neonates and chil- initially with UFH or LMWH and subsequently dren with femoral artery thrombosis, we recom- with LMWH, for a total therapy duration mend surgical intervention compared with UFH between 6 weeks and 3 months rather than therapy alone when there is a contraindication shorter or longer treatment duration ( G r a d e 2 C ). to thrombolytic therapy and organ or limb death For neonates with CSVT with signifi cant hemor- is imminent (Grade 1C) . rhage, we suggest either (1) anticoagulation or 2.11 Prophylaxis for Peripheral Arterial Catheters in (2) supportive care with radiologic monitoring Neonates and Children of the thrombosis at 5 to 7 days and anticoagulation if thrombus extension is noted as compared 2.11. For neonates and children with periph- with no therapy (Grade 2C). eral arterial catheters in situ, we recommend UFH continuous infusion at 0.5 units/mL at 2.18-2.20 Arterial Ischemic Stroke in Neonates 1 mL/h compared with normal saline ( G r a d e 1 A ) . 2.18. For neonates with a fi rst arterial ischemic 2.12 Therapy for Peripheral Artery Thrombosis stroke (AIS), in the absence of a documented, Secondary to Peripheral Artery Catheters in Neonates ongoing cardioembolic source, we suggest sup- and Children portive care over anticoagulation or aspirin therapy (Grade 2C). 2.12. For neonates and children with a peripheral arterial catheter-related thromboembolism, 2.19. For neonates with a fi rst AIS and a docu- we suggest immediate removal of the catheter mented cardioembolic source, we suggest anti- (Grade 2B) . For neonates and children with a coagulation with UFH or LMWH (Grade 2C). symptomatic peripheral arterial catheter-related 2.20. For neonates with recurrent AIS, we sug- thromboembolism, we suggest UFH anticoag- gest anticoagulant or aspirin therapy ( G r a d e 2 C ) . ulation with or without thrombolysis or surgical thrombectomy and microvascular repair with 2.21 Neonates With Purpura Fulminans subsequent heparin therapy (Grade 2C) . 2.21. For neonates with clinical presentations of 2.13-2.14 Prophylaxis of Umbilical Arterial Catheters homozygous protein C defi ciency, we recom- in Neonates mend administration of either 10 to 20 mL/kg of

42S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians fresh frozen plasma every 12 h or protein C resolution of the precipitating factor but for a concentrate, when available, at 20 to 60 units/kg minimum of 3 months as compared with no fur- until the clinical lesions resolve (Grade 1A). For ther therapy (Grade 2C). neonates with homozygous protein C defi ciency, after initial stabilization, we recommend long- 2.22.6. In children with a CVAD in place who term treatment with VKA ( G r a d e 1 C ) , LMWH have a VTE, if a CVAD is no longer required or ( G r a d e 1 C ) , protein C replacement ( G r a d e 1 B ) , is nonfunctioning, we recommend it be removed or liver transplantation ( G r a d e 1 C ) compared with ( G r a d e 1 B ). We suggest at least 3 to 5 days of no therapy. anticoagulation therapy prior to its removal rather than no anticoagulation prior to removal 2.22 DVT and PE in Children (Grade 2C). If CVAD access is required and the 2.22.1. In children with fi rst VTE (CVAD and CVAD is still functioning, we suggest that the non-CVAD related) we recommend acute anti- CVAD remain in situ and the patient given coagulant therapy with either UFH or LMWH anticoagulants ( G r a d e 2 C ). For children with a ( G r a d e 1 B ). We recommend initial treatment fi rst CVAD-related VTE, we suggest initial man- with UFH or LMWH for at least 5 days (Grade agement as for secondary VTE as previously 1B) . For ongoing therapy, we recommend described. LMWH or UFH. For patients in whom clinicians 2.22.7. In children with CVAD in place who will subsequently prescribe VKAs, we recom- have a VTE and in whom the CVAD remains mend beginning oral therapy as early as day 1 necessary, we suggest, after the initial 3 months and discontinuing UFH/LMWH on day 6 or later of therapy, that prophylactic doses of VKAs than day 6 if the INR has not exceeded 2.0 com- (INR range, 1.5-1.9) or LMWH (anti-Xa level pared with no therapy (Grade 1B). range, 0.1-0.3 units/mL) be given until the CVAD 2.22.2. We suggest that children with idiopathic is removed ( G r a d e 2 C ). If recurrent thrombosis VTE receive anticoagulant therapy for 6 to occurs while the patient is receiving prophy- 12 months compared with no therapy ( G r a d e 2 C ) . lactic therapy, we suggest continuing therapeutic doses until the CVAD is removed and Values and preferences: Families who place a high for a minimum of 3 months following the VTE value on avoiding the unknown risk of recurrence in (Grade 2C) . the absence of an ongoing risk factor and a lower 2.23 Thrombolysis in Pediatric Patients With DVT value on avoiding the inconvenience of therapy or potential impact of therapy on growth and devel- 2.23. In children with VTE, we suggest that opment and bleeding risk associated with anti- thrombolysis therapy be used only for life- or thrombotic therapy are likely to choose to continue limb-threatening thrombosis ( G r a d e 2 C ). If throm- anticoagulant therapy beyond 6 to 12 months. bolysis is used in the presence of physiologically low levels or pathologic defi ciencies of plas- 2.22.3. In children with secondary VTE (ie, VTE minogen, we suggest supplementation with plas- that has occurred in association with a clinical minogen ( G r a d e 2 C ) . In children with VTE in risk factor) in whom the risk factor has resolved, whom thrombolysis is used, we suggest systemic we suggest anticoagulant therapy be admin- thrombolysis or catheter-directed thrombolysis, istered for 3 months ( G r a d e 2 C ) as compared depending on institutional experience and, in with no further therapy. In children who have the latter case, technical feasibility. ongoing, but potentially reversible risk factors, such as active nephrotic syndrome or ongoing 2.24 Thrombectomy and IVC Filter Use in Pediatric asparaginase therapy, we suggest continuing Patients With DVT anticoagulant therapy beyond 3 months in either 2.24. In children with life-threatening VTE, we therapeutic or prophylactic doses until the risk suggest thrombectomy (Grade 2C) . In children factor has resolved (Grade 2C) . who have had a thrombectomy, we suggest anticoagulant therapy as per recommendation (2.22) 2.22.4. In children with recurrent idiopathic (Grade 2C). In children . 10 kg body weight with VTE, we recommend indefi nite treatment with lower-extremity VTE and a contraindication VKAs (Grade 1A). to anticoagulation, we suggest placement of a 2.22.5. In children with recurrent secondary retrievable IVC fi lter ( G r a d e 2 C ) . In children VTEs with an existing reversible risk factor who receive a fi lter, we suggest that the fi lter be for thrombosis, we suggest anticoagulation until removed as soon as possible if thrombosis is not www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 43S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians present in the basket of the fi lter and when con- with large (. 2 cm) mobile right atrial thrombo- traindication to anticoagulation is resolved sis, we suggest anticoagulation, with appropri- (Grade 2C). In children who receive an IVC fi l- ately timed CVAD removal, and consid eration of ter, we recommend appropriate anticoagulation surgical intervention or thrombolysis based on for VTE (see 1.2) as soon as the contraindication individualized risk-benefi t assess ment compared to anticoagulation is resolved (Grade 1C). with no anticoagulation therapy (Grade 2C). 2.25 DVT in Children With Cancer 2.30-2.34 Children With CVADs 2.25. In children with cancer, we suggest that 2.30. For CVADs, we suggest fl ushing with nor- management of VTE follow the general recom- mal saline or heparin or intermittent recombi- mendations for management of VTE in children. nant urokinase to maintain patency as compared We suggest the use of LMWH in the treatment with no therapy ( G r a d e 2 C ). For blocked CVADs, we of VTE for a minimum of 3 months until the suggest tPA or recombinant urokinase to restore precipitating factor has resolved (eg, use of aspar- patency ( G r a d e 2 C ). If after at least 30 min aginase) (Grade 2C). following local thrombolytic instillation CVAD patency is not restored, we suggest a second Remarks : The presence of cancer, the need for surgery, dose be administered. If the CVAD remains chemotherapy, or other treatments may modify the blocked following two doses of local thrombo- risk-benefi t ratio for treatment of VTE, and clinicians lytic agent, we suggest radiologic imaging to should consider these factors on an individual basis. rule out a CVAD-related thrombosis ( G r a d e 2 C ). 2.26 Children With APLAs and DVT 2.31. For children with short- or medium-term 2.26. For children with VTE in the setting of CVADs, we recommend against the use of rou- APLAs, we suggest management as per general tine systemic thromboprophylaxis (Grade 1B) . recommendations for VTE management in 2.34. For children receiving long-term home children. total parenteral nutrition, we suggest thrombo- 2.27 Children With DVT and Positive Inherited prophylaxis with VKAs (Grade 2C). Thrombophilia Testing 2.35 Children Undergoing Glenn Procedure or 2.27. For children with VTE, independent of Bilateral Cavopulmonary Shunt the presence or absence of inherited thrombo- 2.35. For children who have bilateral cavopul- philic risk factors, we suggest that the duration monary shunt, we suggest postoperative UFH and intensity of anticoagulant therapy as per (Grade 2C) . 2.22. 2.36 Children Undergoing Fontan Surgery 2.28 Children With VTE and Structurally Abnormally Venous Systems 2.36. For children after Fontan surgery, we recommend aspirin or therapeutic UFH followed 2.28. For children with fi rst VTE secondary by VKAs over no therapy (Grade 1C). to structural venous abnormalities, we suggest anticoagulation as per other “spontaneous” VTE 2.37 Insertion of Endovascular Stents in Children (2.22) and consideration of subsequent percuta- 2.37. For children having endovascular stents neous or surgical interventions, depending on inserted, we suggest administration of UFH patient factors and institutional experience. perioperatively (Grade 2C). For children with recurrent VTE secondary to structural venous abnormalities, we suggest 2.38 Pediatric Patients With Dilated Cardiomyopathy indefi nite anticoagulation unless successful percutaneous or surgical interventions can be per- 2.38. For pediatric patients with cardiomyopathy, formed ( G r a d e 2 C ) . we suggest VKAs no later than their activation on a cardiac transplant waiting list (Grade 2C). 2.29 Children With Right Atrial Thrombosis Values and preferences: Parents who place a high 2.29. For children with right atrial thrombosis value on avoiding the inconvenience, discomfort, related to CVAD, we suggest removal of the and limitations of anticoagulant monitoring and a CVAD with or without anticoagulation, depend- lower value on the uncertain reduction in thrombotic ing on the individual risk factors, compared with complications are unlikely to choose VKA ther apy for leaving the CVAD in situ ( G r a d e 2 C ). For children their children who are eligible for transplant.

44S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 2.39 Children With Primary Pulmonary Hypertension 2.49. For children with moderate or giant coronary aneurysms following Kawasaki disease, we 2.39. For children with primary pulmonary suggest that warfarin in addition to low-dose hypertension, we suggest starting anticoagula- aspirin be given as primary thromboprophylaxis tion with VKAs at the same time as other med- (Grade 2C) . ical therapy (Grade 2C). 2.40-2.42 Children With Biologic and Mechanical 2.50. For children with Kawasaki disease who Prosthetic Heart Valves have giant aneurysms and acute coronary artery thrombosis, we suggest thrombolysis or acute 2.40-2.42. For children with biologic or mechan- surgical intervention (Grade 2C). ical prosthetic heart valves, we recommend that clinicians follow the relevant recommendations 2.51 CSVT in Children from the adult population. 2.51. For children with CSVT without signifi cant 2.44 Children With Ventricular Assist Devices (VADs) intracranial hemorrhage, we recommend anticoagulation initially with UFH or LMWH and sub- 2.44. For children with VADs we suggest admin- sequently with LMWH or VKA for a minimum of istration of UFH ( G r a d e 2 C ). We suggest start- 3 months relative to no anticoagulation ( G r a d e 1 B ). ing UFH between 8 and 48 h following In children who after 3 months of therapy still expe- implantation ( G r a d e 2 C ). In addition, we suggest rience occlusion of CSVT or ongoing symptoms, antiplatelet therapy (either aspirin or aspirin we suggest administration of a further 3 months dipyridamole) to commence within 72 h of of anticoagulation ( G r a d e 2 C ). For children with VAD placement (Grade 2C). For children with CSVT with signifi cant hemorrhage, we suggest ini- VAD, once clinically stable, we suggest switch- tial anticoagulation as for children without hemor- ing from UFH to either LMWH or VKA (target rhage or radiologic monitoring of the thrombosis INR 3.0 range, 2.5-3.5) until transplanted or at 5 to 7 days and anticoagulation if thrombus weaned from VAD (Grade 2C) . extension is noted at that time ( G r a d e 2 C ). In chil- 2.45-2.46 Primary Prophylaxis for Venous Access dren with CSVT and potentially recurrent risk Related to Hemodialysis factors (for example, nephrotic syndrome, asparaginase therapy), we suggest prophylactic anti- 2.45. For patients undergoing hemodialysis coagulation at times of risk factor recurrence via an arteriovenous fi stula, we suggest rou- (Grade 2C). We suggest thrombolysis, thrombec- tine use of VKAs or LMWH as fi stula throm- tomy, or surgical decompression only in children boprophylaxis as compared with no therapy with severe CSVT in whom there is no improve- ( G r a d e 2 C ). ment with initial UFH therapy (Grade 2C). 2.46. For patients undergoing hemodialysis via 2.52 AIS in Children CVAD, we suggest routine use of VKAs or LMWH for thromboprophylaxis as compared 2.52. For children with acute AIS, with or with- with no therapy (Grade 2C) . out thrombophilia, we recommend UFH or LMWH or aspirin as initial therapy until dissec- 2.47 Use of UFH or LMWH in Children Undergoing tion and embolic causes have been excluded Hemodialysis (Grade 1C). For children with acute AIS, we sug- 2.47. For children having hemodialysis, we suggest, once dissection and cardioembolic causes gest the use of UFH or LMWH during hemodi- are excluded, daily aspirin prophylaxis for a alysis to maintain circuit patency independent minimum of 2 years as compared with no anti- of type of vascular access (Grade 2C). thrombotic therapy ( G r a d e 2 C ). For children receiving aspirin who have recurrent AIS or 2.48-2.50 Children With Kawasaki Disease transient ischemic attacks (TIAs), we suggest 2.48. For children with Kawasaki disease, we changing to clopidogrel or anticoagulant therapy recommend aspirin in high doses (80-100 mg/kg with LMWH or VKA ( G r a d e 2 C ). For children with per day during the acute phase for up to 14 days) AIS, we recommend against the use of throm- as an antiinfl ammatory agent, then in lower bolysis (tPA) or mechanical thrombectomy out- doses (1-5 mg/kg per day for 6 to 8 weeks) as an side of specifi c research protocols (Grade 1C) . antiplatelet agent (Grade 1B). For children with 2.53 Embolic Stroke in Children Kawasaki disease, we recommend IV g -globulin (2 g/kg, single dose) within 10 days of the onset 2.53. For AIS secondary to cardioembolic causes, of symptoms (Grade 1A) . we suggest anticoagulant therapy with LMWH www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 45S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians or VKAs for at least 3 months (Grade 2C). For ical Pharmacy, the American Society of Health-System Pharma- cists, the American Society of Hematology, and the International AIS secondary to cardioembolic causes in chil- Society of Thrombosis and Hematosis. dren with demonstrated right-to-left shunts (eg, PFO), we suggest surgical closure of the shunt (Grade 2C). References 1. Holbrook A, Schulman S, Witt DM, et al. Evidence-based 2.54 Cerebral Arterial Dissection Underlying AIS management of anticoagulant therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American 2.54. For AIS secondary to dissection, we sug- College of Chest Physicians evidence-based clinical prac tice gest anticoagulant therapy with LMWH or VKAs guidelines. Chest. 2012;141(2)(suppl):e152S-e184S. for at least 6 weeks ( G r a d e 2 C ) . Ongoing treatment 2. Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in will depend on radiologic assessment of degree nonsurgical patients: antithrombotic therapy and preven- and extent of stenosis and evidence of recurrent tion of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. ischemic events. Chest. 2012;141(2)(suppl):e195S-226S. 3. Gould MK, Garcia DA, Wren SM, et al. Prevention of 2.55 Children With Cerebral Vasculopathies VTE in nonorthopedic surgical patients: antithrombotic 2.55. For children with acute AIS secondary to therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice non-Moyamoya vasculopathy, we recommend guidelines. Chest. 2012;141(2)(suppl):e227S-e277S. UFH or LMWH or aspirin for 3 months as initial 4. Falck-Ytter Y, Francis CW, Johanson NA, et al. Prevention therapy compared with no treatment (Grade of VTE in orthopedic surgery patients: antithrombotic 1C) . For children with AIS secondary to non- therapy and prevention of thrombosis, 9th ed: American Moyamoya vasculopathy, we suggest ongoing College of Chest Physicians evidence-based clinical prac tice guidelines. Chest. 2012;141(2)(suppl):e278S-e325S. antithrombotic therapy should be guided by 5. Douketis JD, Spyropoulos AC, Spencer FA, et al. Periop- repeat cerebrovascular imaging. erative management of antithrombotic therapy: antithrombotic therapy and prevention of thrombosis, 9th ed: American 2.56-2.57 Children With Moyamoya Disease College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2)(suppl):e326S-e350S. 2.56. For children with acute AIS secondary to 6. Bates SM, Jaeschke R, Stevens SM, et al. Diagnosis of Moyamoya, we suggest aspirin over no treat- DVT: antithrombotic therapy and prevention of throm- ment as initial therapy (Grade 2C) . bosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012; 2.57. For children with Moyamoya, we suggest 141(2)(suppl):e351S-e418S. they be referred to an appropriate center for 7. Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: antithrombotic therapy and pre- consideration of revascularization. vention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guide lines. Chest. 2012;141(2)(suppl):e419S-e494S. Acknowledgments 8. Linkins L-A, Dans AL, Moores LK, et al. Treatment and prevention of heparin-induced thrombocytopenia: antithrombotic Financial/nonfi nancial disclosures: In summary, the authors therapy and prevention of thrombosis, 9th ed: American have reported to CHEST the following confl icts of interest: Dr Crowther has served on various advisory boards, has assisted College of Chest Physicians evidence-based clinical practice in the preparation of educational materials, and has sat on data guidelines. Chest. 2012;141(2)(suppl):e495S-e530S. safety and monitoring boards. His institution has received research 9. You JJ, Singer DE, Howard PA, et al. Antithrombotic funds from the following companies: Leo Pharma A/S, Pfi zer Inc, therapy for atrial fi brillation: antithrombotic therapy Boerhinger Ingelheim GmbH, Bayer Healthcare Pharmaceuticals, and prevention of thrombosis, 9th ed: American College Octapharm AG, CSL Behring, and Artisan Pharma. Personal total of Chest Physicians evidence-based clinical practice guide- compensation for these activities over the past 3 years totals less lines. Chest. 2012;141(2)(suppl):e531S-e575S. than US $10,000. Dr Gutterman has had the following relationships 10. Whitlock RP, Sun JC, Fremes SE, Rubens FD, Teoh KH. that are entirely unrelated to the AT9 guidelines: ACCP President, Antithrombotic and thrombolytic therapy for valvular GlaxoSmithKline plc grant to study vasodilation in adipose tissue, National Institutes of Health grant to study human coronary dila- disease: antithrombotic therapy and prevention of throm- tion, and GE Healthcare consultation on a study for ECG evalua- bosis, 9th ed: American College of Chest Physicians evidence- tion of chronic heart disease. Drs Guyatt and Schü nemann are based clinical practice guidelines. Chest. 2012;141(2)(suppl): co-chairs of the GRADE Working Group, and Dr Akl is a member e576S-e600S. and prominent contributor to the GRADE Working Group. 11. Lansberg MG, O’Donnell MJ, Khatri P, et al. Antithrom botic Role of sponsors: The sponsors played no role in the develop- and thrombolytic therapy for ischemic stroke: antithrom- ment of these guidelines. Sponsoring organizations cannot botic therapy and prevention of thrombosis, 9th ed: American recommend panelists or topics, nor are they allowed prepubli- College of Chest Physicians evidence-based clinical prac- cation access to the manuscripts and recommendations. Guideline tice guidelines. Chest. 2012;141(2)(suppl):e601S-e636S. panel members, including the chair, and members of the Health & Science Policy Committee are blinded to the funding sources. 12. Vandvik PO, Lincoff AM, Gore JM, et al. Primary and sec- Further details on the Confl ict of Interest Policy are available ondary prevention of cardiovascular disease: antihrombotic online at http://chestnet.org. therapy and prevention of thrombosis, 9th ed: American Endorsements: This guideline is endorsed by the American College of Chest Physicians evidence-based clinical practice Association for Clinical Chemistry, the American College of Clin- guidelines. Chest. 2012;141(2)(suppl):e637S-e668S.

46S Executive Summary

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 13. Alonso-Coello P, Bellmunt S, McGorrian C, et al. Anti throm- and prevention of thrombosis, 9th ed: American College botic therapy in peripheral artery disease: anti thrombotic of Chest Physicians evidence-based clinical practice guide- therapy and prevention of thrombosis, 9th ed: American lines. Chest. 2012;141(2)(suppl):e691S-e736S. College of Chest Physicians evidence-based clinical prac- 15. Monagle P, Chan AKC, Goldenberg NA, et al. Anti- tice guidelines. Chest. 2012;141(2)(suppl):e669S-e690S. thrombotic therapy in neonates and children: anti thrombotic 14. Bates SM, Greer IA, Middeldorp S, Veenstra DL, therapy and prevention of thrombosis, 9th ed: American Prabulos A-M, Vandvik PO. VTE, thrombophilia, anti- College of Chest Physicians evidence-based clinical practice thrombotic therapy, and pregnancy: antithrombotic therapy guidelines. Chest. 2012;141(2)(suppl):e737S-e801S.

www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 47S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Executive Summary : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Gordon H. Guyatt, Elie A. Akl, Mark Crowther, David D. Gutterman, Holger J. Schuünemann and for the American College of Chest Physicians Antithrombotic Therapy and Prevention of Thrombosis Panel Chest 2012;141; 7S-47S DOI 10.1378/chest.1412S3

This information is current as of February 7, 2012

Supplementary Material View e-supplements related to this article at: http://chestjournal.chestpubs.org/content/suppl/2012/02/06/141.2_suppl.7S.DC1.html Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/141/2_suppl/7S.full.html References This article cites 15 articles, 15 of which can be accessed free at: http://chestjournal.chestpubs.org/content/141/2_suppl/7S.full.html#ref-list-1 Cited Bys This article has been cited by 1 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/141/2_suppl/7S.full.html#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Gordon H. Guyatt, Susan L. Norris, Sam Schulman, Jack Hirsh, Mark H. Eckman, Elie A. Akl, Mark Crowther, Per Olav Vandvik, John W. Eikelboom, Marian S. McDonagh, Sandra Zelman Lewis, David D. Gutterman, Deborah J. Cook and Holger J. Schünemann

Chest 2012;141;53S-70S DOI 10.1378/chest.11-2288

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/141/2_suppl/53S.full.html

Supplemental material related to this article is available at: http://chestjournal.chestpubs.org/content/suppl/2012/02/06/141.2_suppl. 53S.DC1.html

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians CHEST Supplement ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: ACCP GUIDELINES Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Gordon H. Guyatt , MD , FCCP ; Susan L. Norris , MD , MPH ; Sam Schulman , MD , PhD ; Jack Hirsh , MD , FCCP ; Mark H. Eckman , MD ; Elie A. Akl , MD , MPH , PhD ; Mark Crowther , MD ; Per Olav Vandvik , MD , PhD ; John W. Eikelboom , MBBS ; Marian S. McDonagh , PharmD ; Sandra Zelman Lewis , PhD ; David D. Gutterman , MD , FCCP ; Deborah J. Cook , MD ; Holger J. Schünemann , MD , PhD , FCCP

Background: To develop the Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: ACCP Evidence-Based Clinical Practice Guidelines (AT9), the American College of Chest Physi- cians (ACCP) assembled a panel of clinical experts, information scientists, decision scientists, and systematic review and guideline methodologists. Methods: Clinical areas were designated as articles, and a methodologist without important intellectual or fi nancial confl icts of interest led a panel for each article. Only panel members without signifi cant confl icts of interest participated in making recommendations. Panelists specifi ed the population, intervention and alternative, and outcomes for each clinical question and defi ned criteria for eligible studies. Panelists and an independent evidence-based practice center executed systematic searches for relevant studies and evaluated the evidence, and where resources and evidence permitted, they created standardized tables that present the quality of the evidence and key results in a transparent fashion. Results : One or more recommendations relate to each specifi c clinical question, and each recommendation is clearly linked to the underlying body of evidence. Judgments regarding the quality of evidence and strength of recommendations were based on approaches developed by the Grades of Recommendations, Assessment, Development, and Evaluation Working Group. Panel members constructed scenarios describing relevant health states and rated the disutility associated with these states based on an additional systematic review of evidence regarding patient values and preferences for antithrombotic therapy. These ratings guided value and preference decisions underlying the recommendations. Each topic panel identifi ed questions in which resource allocation issues were particularly important and, for these issues, experts in economic analysis provided additional searches and guidance. Conclusions: AT9 methodology refl ects the current science of evidence-based clinical practice guideline development, with reliance on high-quality systematic reviews, a standardized process for quality assessment of individual studies and the body of evidence, an explicit process for trans- lating the evidence into recommendations, disclosure of fi nancial as well as intellectual confl icts of interest followed by management of disclosed confl icts, and extensive peer review. CHEST 2012; 141(2)(Suppl):53S–70S

Abbreviations: A C C P 5 American College of Chest Physicians; A T 8 5 Antithrombotic and Thrombolytic Therapy: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition); AT9 5 Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines; GDP 5 gross domestic product; GRADE 5 Grades of Recommendation, Assessment, Development, and Evaluation; HSP 5 Health and Science Policy; PICO 5 population, intervention, comparator, and outcome; QALY 5 quality- adjusted life year; RCT 5 randomized controlled trial; WHO 5 World Health Organization

www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 53S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians his article describes the methodology used for Chest Physicians Evidence-Based Clinical Practice T the Antithrombotic Therapy and Prevention of Guidelines (8th Edition) (AT8), published in 20081 Thrombosis, 9th ed: American College of Chest Phy- (Dr Guyatt, Panel Chair, and Dr Schünemann, Vice sicians Evidence-Based Clinical Practice Guidelines Chair of Methodology); a third methodologist-clinician (AT9). This methodology incorporates current evidence- (Dr Akl); a leading thrombosis expert (Dr Crowther, based approaches to the appraisal and synthesis of Vice Chair for Thrombosis); and two liaisons with the evidence and to the formulation of clinical practice rec- ACCP Health and Science Policy (HSP) Committee ommendations. The process thus ensures explicit, trans- who had also served on the previous guideline Exec- parent, evidence-based clinical practice guidelines. utive Committee (Dr Gutterman, Vice Chair and The objective of AT9 is to optimize patient- HSP Liaison, and Dr Lewis, Project Manager). important health outcomes and the processes of care Articles within AT9 are defi ned by broad popula- for patients who have experienced or are at risk for tions (eg, pediatric, obstetric) or clinical conditions thrombotic events. The targeted users of these guide- (eg, prevention of VTE in medical patients or stroke lines are health-care providers in both primary and prophylaxis in atrial fi brillation). In addition, AT9 specialty care who assist patients in making treatment includes three articles addressing the basic science of choices that optimize benefi ts, minimize harms and oral and parenteral anticoagulants and platelet-active burdens, and are consistent with patient values and drugs and an article addressing new antithrombotic preferences. and thrombolytic drugs. Figure 1 summarizes the process for the development of the AT9 recommendations. The primary 1.0 Composition and Selection responsibility for AT9 rests with the American College of Topic Panel Members of Chest Physicians (ACCP) AT9 Executive Com- mittee. This committee includes two methodologist- The ACCP AT9 Executive Committee selected clinicians from the previous iteration, Antithrombotic panel members for each article. A topic editor and a and Thrombolytic Therapy: American College of deputy editor led each of the AT9 panels issuing recommendations. The topic editor was the person primarily responsible for each article and was required Revision accepted August 31, 2011. Affi liations: From the Department of Clinical Epidemiology to be a methodologist without serious fi nancial or and Biostatistics (Drs Guyatt, Akl, Cook, Schulman, and Schünemann) intellectual confl ict of interest for any of the article’s and Department of Medicine (Drs Guyatt, Schulman, Hirsh, recommendations. In all but one case, the topic edi- Crowther, Eikelboom, Cook, and Schünemann), McMaster University Faculty of Health Sciences, Hamilton, ON, Canada; tor also was a clinician. The Executive Committee Department of Medical Informatics and Clinical Epidemiology chose these individuals on the basis of their previous (Drs Norris and McDonagh), Oregon Health & Science University, experience with guideline development and, in par- Portland, OR; Division of General Internal Medicine and Center for Clinical Effectiveness (Dr Eckman), University of Cincinnati, ticular, their familiarity with methods developed by Cincinnati, OH; Departments of Medicine and Family Medicine the Grades of Recommendation, Assessment, Devel- (Dr Akl), State University of New York, Buffalo, NY; Department opment, and Evaluation (GRADE) Working Group.2 of Medicine Gjøvik-Innlandet Hospital Trust and Norwegian Knowledge Centre for the Health Services (Dr Vandvik), Oslo, These topic editors and all panel members were Norway; Evidence-Based Clinical Practice Guidelines and approved by the ACCP HSP Committee after review Clinical Standards (Dr Lewis), American College of Chest of their confl ict of interest disclosures (see section 7.0 Physicians, Northbrook, IL; and Cardiovascular Research Center (Dr Gutterman), Medical College of Wisconsin, Milwaukee, WI. “Disclosing and Managing Confl icts of Interest” ). Funding/Support: The Antithrombotic Therapy and Prevention Criteria for selection of the remainder of the panel of Thrombosis, 9th ed: American College of Chest Physicians members, including the deputy editor-thrombosis Evidence-Based Clinical Practice Guidelines received support from the National Heart, Lung, and Blood Institute [R13 HL104758] expert, were an established record in the relevant and Bayer Schering Pharma AG. Support in the form of educa- clinical or research area, international and gender tional grants was also provided by Bristol-Myers Squibb; Pfi zer, representation, and an absence of fi nancial confl icts Inc; Canyon Pharmaceuticals; and sanofi -aventis US . Disclaimer: American College of Chest Physician guidelines of interest that were judged unacceptable. Some of the are intended for general information only, are not medical advice, panelists had prior experience on ACCP guidelines in and do not replace professional medical care and physician this area and represented the thrombosis community, advice, which always should be sought for any medical condition. The complete disclaimer for this guideline can be accessed but there was substantial turnover from the previous at http://chestjournal.chestpubs.org/content/141/2_suppl/1S edition. After an international request for applications Correspondence to: Gordon H. Guyatt, MD, FCCP, Department broadcast through multiple medical societies, the of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, L8N 3Z5, Canada; e-mail: [email protected] Executive Committee nominated indivi dual topic edi- © 2012 American College of Chest Physicians. Reproduction tors and deputy editors and collaborated with them to of this article is prohibited without written permission from the identify and nominate other topic panel members. American College of Chest Physicians ( http://www.chestpubs.org/ site/misc/reprints.xhtml ). The ACCP HSP Committee reviewed all nominees DOI: 10.1378/chest.11-2288 and approved all panel members after review of their

54S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians proved as a result of the magnitude of fi nancial confl icts of interest. Articles associated with recommendations included from seven to 14 panel members. We did not include patients or representatives of specifi c stakeholder groups on topic panels. Each topic panel also included a frontline physician working in the relevant area who was neither an expert in thrombosis nor a methodologist or clinical investigator. These individuals were chosen in consultation with the topic editors and the ACCP HSP Committee. These clinicians were charged with the following: (1) proposing important real-world clinical questions on the prevention, diagnosis, and treatment of thrombosis that were not addressed in AT8 and (2) reviewing the draft manuscripts and recommendations to assess the usability of the guidelines and the feasibility of implementation of AT9 recom mendations. To address issues of economic effi ciency we included six health economist-physicians on the AT9 topic panels charged with making recommendations. These resource consultants were selected and approved through identical procedures to those for topic editors and panel members. We describe their roles more fully later in this discussion.

2.0 Ensuring Consistency Across Articles We used a number of strategies to ensure consistency across articles, and one of us (M. C.) participated extensively in the formulation of clinical questions for each article. To ensure consistency of judgments regarding bleeding, one of us (S. S.) was responsible for standardizing the approach to bleeding outcomes and participated in multiple topic panels (described in more detail later in this article). Additionally, to ensure consistency in the trade-offs between thrombotic and bleeding events, all articles used the same ratings of values and preferences (also described in more detail later). Because some of the same evidence summaries were relevant to several articles, fi ve individuals were chosen to participate in each of the articles addressing coronary artery disease, Figure 1. The steps in the process of development of the Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: stroke, and peripheral arterial disease. American College of Chest Physicians Evidence-Based Clinical In AT9, prevention of VTE is addressed in three Practice Guidelines. The Executive Committee, comprising system- articles as opposed to a single article as was done in atic review and guideline methodologists, clinical experts, and ACCP Health and Science Policy Committee liaisons, coordinated the AT8. The prevention topic editors and deputy editors process. Information within parentheses indicates who performed and those of the stroke article (which includes throm- each step in the process. ACCP 5 American College of Chest Physi- boprophylaxis recommendations) participated in mul- cians; EPC 5 Evidence-Based Practice Center; GRADE 5 Grades of Recommendations, Assessment, Development, and Evaluation; tiple conference calls to develop and harmonize the PICO 5 population, intervention, comparator, and outcome. approach to prevention and to ensure consistency among fi nal recommendations.3 Topic editors con- curricula vitae and confl ict of interest disclosures. sulted with one another when issues overlapped. For Of 150 nominees, 137 were approved, 18 were example, the decision regarding the use of a vitamin K approved with management of confl icts of interest antagonist, aspirin, and clopidogrel simultaneously in (ie, regular disclosures and review of ongoing con- patients with atrial fi brillation, valvular disease, and fl icts as the process progressed), and 13 were disap- intravascular stents is relevant for the atrial fi brillation, www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 55S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians coronary, and peripheral arterial disease articles. Topic primary responsibility for defi ning the scope of the panels deferred to the Evidence-Based Management clinical questions that each article would address. For of Anticoagulant Therapy AT9 topic panel for recom- each question, the topic editor and deputy editor mendations related to the dosing and monitoring of defi ned the relevant population, alternative manage- anticoagulation therapies. ment strategies (intervention and comparator), and The AT9 Executive Committee met at least once a the outcomes (ie, population, intervention, com- month and regularly issued statements of clarifi cation parator, and outcome [PICO] format). Each clinical of methods to topic editors and deputy editors (eg, question provided the framework for formulating use of fi xed- or random-effects models for meta- study inclusion and exclusion criteria and guided the analysis), confl ict of interest, preparation of tables, and search for relevant evidence (systematic reviews and issues of style and presentation. All these statements original studies). Panels typically restricted included were communicated directly to the topic editors and studies to randomized controlled trials (RCTs) for deputy editors and made available in a central repository intervention questions but included observational accessible to all AT9 panelists. The chair of the Exec- studies when there was a paucity of RCT data address- utive Committee (G. H. G.) was available for resolving ing an intervention and for questions of risk assess- any challenging issues related to the aforementioned ment. Readers can fi nd these PICO questions in the topics. Between September 2009 and September 2010, fi rst table of each article. One or more recommenda- two members of the Executive Committee (E. A. A. tions could be formulated for each clinical question. and S. Z. L.) held regular (every 3 months), separate The next subsections (3.2-3.5) deal with the approach conference calls with each topic editor and deputy to selection of outcomes. editor during which they addressed questions and concerns. Finally, the chair of the Executive Com- mittee reviewed every article to ensure consistency of 3.2 Patient-Important and Surrogate Outcomes evidence presentation, evaluation, and writing style. The outcomes for each clinical question were cho- In terms of writing style, we used consistent lan- sen by the topic editors and their panel members and guage to describe effects that did not reach statistical were generally consistent across articles. Outcomes signifi cance. The approach was as follows: were restricted to those of importance to patients.4 1. For adverse outcomes (such as thrombosis and Panels considered the burden of anticoagulation ther- bleeding), “A is associated with a trend toward apy as a patient-important outcome when its con- reduced thrombosis” if the lower boundary of sideration could tip the balance of benefi ts and harms. the CI around a relative effect is Յ 0.7 and the If we found no data for an outcome considered upper boundary of the CI is Յ 1.1 or if the lower at the outset as patient-important, we nevertheless boundary of the CI is Յ 0.8 and the upper included uncertainty about the effects of the inter- boundary of the CI is Յ 1.05. If the point esti- vention on that outcome when weighing its benefi ts mate is . 1.0, the language used was, “A i s a s s o - and harms. ciated with a trend toward increased bleeding” In the absence of data on patient-important out- if the lower boundary of the CI around a relative comes, surrogates could contribute to the estimation effect is . 0.9 and the upper boundary of the of the effect of an intervention on the outcomes that CI is . 1.3 or if the lower boundary of the CI are important. Examples of surrogate outcomes is . 0.95 and the upper boundary of the CI is . 1.2. include asymptomatic venous thrombosis detected 2. “ A appears to have little or no effect on throm- by venographic or ultrasound surveillance and the bosis” if the above conditions are not met, and the percentage of time that an international normal- boundaries of the CI lie between 0.80 and 1.2. ized ratio was in therapeutic range (used as a surro- 3. For all other results that fail to exclude a relative gate for bleeding and thrombosis in the assessment risk of 1.0, the language was, “Results failed of the effectiveness of centralized anticoagulation to demonstrate or exclude a benefi cial effect or services). detrimental effect of A on thrombosis.” Alterna- The issue of asymptomatic thrombosis detected tive wording with regard to an association is by venographic or ultrasound surveillance presented “failed to establish or refute.” particular challenges to the articles addressing VTE prevention in orthopedic and nonorthopedic surgery 3.0 Evidence Review populations, an article addressing nonsurgical prophylaxis, and an article addressing stroke preven- 3.1 Defi ning the Clinical Questions—Population, tion. We were explicit in considering the trade-offs Intervention, Comparator, and Outcome between VTE and bleeding events. An article by The thrombosis expert on the Executive Com- Guyatt et al 3 in this supplement addresses these issues mittee (M. C.) along with the deputy editors took in some detail.

56S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 3.3 Mortality cifi ed in footnotes the events attributable to each component. Options considered in summarizing mortal events We avoided any double-counting of events; for were (1) all-cause mortality and (2) mortality related example, a fatal hemorrhagic stroke would only to antithrombotic therapy (ie, deaths from pulmonary be reported under mortality. We also attempted to emboli and deaths from bleeding). Advantages of the report more specifi c and homogeneous bleeding out- former include its being the most patient-important comes than “major bleeding,” which includes events outcome and the diffi culty of ascertaining cause of with a wide range of patient importance and which death. Diffi culties ascertaining cause of death may investigators have defi ned in different ways. For be particularly problematic when adjudication is example, in Falck-Ytter et al8 i n t h i s s u p p l e m e n t , b l e e d - unblinded and therefore open to bias. ing outcomes of primary interest were (1) bleeding The disadvantage of all-cause mortality is that the requiring reoperation and (2) other major bleeding. signal from antithrombotic therapy-related deaths Because some authors, particularly in older studies, may be lost in noise from deaths due to other causes. failed to defi ne subsets of major bleeding, we were The decision about which mortal outcome to use not always able to achieve the desired specifi city. (all-cause mortality or antithrombotic therapy-related morality) was left to the authors of individual articles . 3.6 Identifying the Evidence Availability of data sometimes forced the choice of less satisfactory mortal outcomes (eg, if death related To identify the relevant evidence, a team of meth- to pulmonary embolus but not death related to bleed- odologists and medical librarians at the Oregon ing was reported). When mortality was one of the Health & Science University Evidence-based Prac- selected outcomes, we avoided double-counting by doc- tice Center conducted literature searches of Medline, umenting nonfatal events for the remaining outcomes the Cochrane Library, and the Database of Abstracts (eg, nonfatal thrombosis, nonfatal major bleeding) of Reviews of Effects. For each article, the team con- rather than all such events (fatal and nonfatal). ducted a search for systematic reviews and another for original studies encompassing the main popula- 3.4 Composite End Points tions and interventions for that article. These searches included studies indexed from week 1, January 2005, Many of the primary studies we reviewed, particu- forward because AT8 searches were carried out up to larly in the cardiovascular area, presented evidence that date (search strategies are available on request). in the form of composite end points.5 Particularly Many articles supplemented these searches with when the patient importance of the component end more-focused searches addressing specifi c clinical points and the magnitude of effect of the interven- questions. When clinical questions had not been cov- tion on the components differ, composite end points ered in AT8, searches commenced at a date relevant can be misleading.6,7 Therefore, we present results to each intervention. and base inferences on the effect of interventions on Titles and abstracts retrieved from bibliographic individual outcomes. database searches generally were screened in duplicate, and full-text articles were retrieved for further 3.5 Bleeding review. Consensus on whether individual studies ful- In view of the wide variation in how bleeding was fi lled inclusion criteria was achieved for each study assessed and reported in the included primary studies between two reviewers. If consensus could not be across chapters, one individual (S. S.) was responsible achieved, the topic editor and other topic panelists for standardizing the approach to bleeding outcomes. were brought into the discussion. Deputy editors He worked closely with the Executive Committee reviewed lists of included studies from the database and the topic editors and deputy editors to ensure the searches in order to identify any potentially missed uniform application of the approaches that were studies. Additional studies identifi ed were then developed. retrieved for further evaluation. We began by specifying the bleeding outcomes Topic panels also searched the same bibliographic that we believe patients consider important. We did databases for systematic reviews addressing each PICO not consider minor bleeding as incurring a burden question. The quality of reviews was assessed using that was important in comparison with symptomatic principles embodied in prior instruments addressing thromboembolic events. methodologic quality of systematic reviews,9,10 and We reported fatal hemorrhage as well as fatal stroke wherever possible, current high-quality systematic or pulmonary embolism in treatment-related or all- reviews were used as the source of summary estimates. cause mortality. Likewise, hemorrhagic stroke and Reviews were also used to identify additional studies ischemic stroke were reported as “stroke.” We spe- to complement the database searches. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 57S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 4.0 Assessing Studies and higher quality, individual prospective studies may Summarizing Evidence have a signifi cant risk of bias and specifi c retro- 4.1 Evaluating Risk of Bias in Individual Studies spective studies may not. For questions related to risk assessment, we evaluated the risk of bias of We developed and applied uniform criteria for individual studies using the following criteria: valid evaluating the risk of bias associated with individual outcome assessment, including blinding when appro- RCTs based on the criteria recommended by the priate; adjustment for between-group differences; Cochrane Collaboration11 ( Table 1 ). Although all and minimal loss to follow-up. authors assessed risk of bias for individual studies, because of resource limitations, we summarized the results of the risk of bias (eg, Table 112 ) for only a 4.2 Evaluating Quality of Bodies of Evidence minority of the recommendations. Readers can fi nd We assessed evidence across studies on an outcome- these assessments in the online data supplements. For by-outcome basis using criteria suggested by the most recommendations for which we did not develop GRADE Working Group.19 W e d e fi ned quality of such tables, we developed Evidence Profi les (see evidence as our confi dence in the estimate of the Table 213-15 ) that typically provide information on the effect to support a recommendation. 19 RCTs start risk of bias in footnotes. as high-quality evidence and observational studies as We also developed specifi c criteria for assessing low-quality evidence ( Fig 2 ). Additional factors that the risk of bias of observational studies (cohort studies affect this rating of quality include the risk of bias with concurrent controls, cohort studies with historical (as detailed earlier in this article); precision, con- controls, case-control studies, or case series). Again, sistency, and directness of results; likelihood of pub- these were based on the evidence-based domains lication bias; and presence of very large effects.19 The recommended by the Cochrane Collaboration for ACCP adaptation of the GRADE system ( Table 4 ) observational studies (eg, Table 316-18 ). differs only in that the quality of a body of evidence Studies without internal comparisons were termed can be high (A), moderate (B), or low (C) ( Fig 2 ); “cohort studies without internal controls” if they met GRADE also provides a category for very-low-quality the following criteria: evidence. 1. A protocol existed before the date of commence- Often, we found that the quality of the evidence ment of data collection. differed across outcomes. For example, in assessing 2. A defi nition of inclusion and exclusion criteria the quality of evidence for thienopyridines vs warfa- was available. rin in patients undergoing percutaneous coronary 3. The study reported the number of excluded interventions, we determined the evidence to be of patients. moderate quality for mortality, nonfatal myocardial 4. The study conducted a standardized follow-up, infarction, and revascularization but of low quality for including description of all of the following: major bleeding. schedule of follow-up, investigation of suspected We then made a rating of the quality of the entire outcomes, and criteria used to defi ne outcomes. body of evidence bearing on the effect of alternative 5. The study reported all losses to follow-up. management strategies for each clinical question. In other words, we assessed the quality across outcomes, We labeled studies that did not meet these crite- including both benefi ts and harms. Quality for each ria as “case series.” We did not make a distinction recommendation was the lowest quality rating of the between prospective and retrospective studies because outcomes judged as critical (as opposed to important, although prospective studies may on average be of but not critical).19

Table 1— [Section 4.1] Methodologic Quality of Randomized Trials: Fondaparinux vs No Fondaparinux for the Treatment of Superﬁ cial Vein Thrombosis

Randomization Stopping Early Author, Year Design Concealed Blinding Analysis for Benefi t Decousus RCT: randomization DY: “Through a Patients: PY ITT: DY for effi cacy outcomes No et al 12; CALISTO sequence generated central telephone Caregivers: PY (as-treated analysis for Study Group, “using a computer- system” Data collectors: PY safety outcomes) 2010 generated Adjudicators: DY Data for primary effi cacy randomization list” Data analysts: PN assessment available for 98.7% of randomized patients CALISTO 5 Comparison of ARIXTRA in Lower Limb Superfi cial Thrombophlebitis with placebo; DY 5 defi nitely yes; ITT 5 intent to treat; PN 5 probably no; PY 5 probably yes; RCT 5 randomized controlled trial.

58S Methodology for the Guidelines

b (Continued) fewer to 4 more) fewer to 16 more) fewer to 21 more) fewer to 16 fewer) fewer to 43 fewer) fewer to 108 fewer) LMWH (95% CI)

Risk Difference With With Difference Risk 4 fewer bleeds per 1,000 (from 9 15 fewer bleeds per 1,000 (from 36 7 fewer deaths per 1,000 (from 31 11 fewer VTE per 1,000 (from 5 30 fewer VTE per 1,000 (from 13 76 fewer VTE per 1,000 (from 32

a,13-15

d d d d f f Metastatic cancer Metastatic cancer Nonmetastatic cancer Anticipated Absolute Effects Anticipated Absolute Effects per 1,000 20 bleeds per 1,000 80 bleeds per 1,000 80 VTEs per 1,000 200 VTEs per 1,000 30 VTEs per 1,000 Relative (95% CI) Effect, RR Summary of Findings 0.96 (0.81-1.13) 164 deaths With With LMWH VKA Risk With 204/1,265 (16.1) 204/1,265 VKA With With Study Event Rates (%) 202/1,231 (16.4) 53/1,351 (3.9) 45/1,386 (3.2) 0.81 (0.55-1.2) No cancer or nonmetastatic 105/1,349 (7.8) 67/1,378 (4.9) 0.62 (0.46-0.84) No cancer imprecision imprecision bias of Evidence due to due to to risk of Overall Quality Overall Bias Publication Undetected Moderate Undetected Moderate Undetected Moderate due Major bleeding (critical outcome) Overall mortality (critical outcome) t and t and CI includes important beneﬁ harm CI includes important beneﬁ harm Recurrent symptomatic VTE (critical outcome): DVT and PE Serious imprecision Serious imprecision No serious imprecision No serious indirectness No serious indirectness No serious indirectness Quality Assessment No serious inconsistency No serious inconsistency No serious inconsistency

c — [Section 4.1] Evidence Proﬁ le: Question: Should LMWH Rather Than VKA be Used for Long-term Treatment of VTE? of VTE? le: Question: Should LMWH Rather Than VKA be Used for Long-term Treatment [Section 4.1] Evidence Proﬁ —

e Selective outcome reporting not serious Lack of blinding not serious Table 2 Table No serious risk of bias No serious risk of bias Serious risk of bias No studies were blinded 2,496 (7 RCTs), 2,496 (7 RCTs), 6 mo 2,737 (8 RCTs), 2,737 (8 RCTs), 6 mo Participants (Studies), Median Follow-up Risk of Bias Inconsistency Indirectness Imprecision 2,727 (8 RCTs), 2,727 (8 RCTs), 6 mo www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 59S

b to 46 fewer) LMWH (95% CI)

Risk Difference With With Difference Risk 30 fewer per 1,000 (from 12 fewer h

g Anticipated Absolute Effects Anticipated Absolute Effects per 1,000 Relative (95% CI) Effect, RR Summary of Findings With With LMWH VKA Risk With vitamin K agonist. (Kearon C, unpublished data). vitamin K agonist. (Kearon C, unpublished data). 5 VKA With With Study Event Rates (%) 31/44 (70.5) 34/56 (60.7) 0.85 (0.77- 0.94) 200 PTS risk ratio; VKA risk ratio; VKA 5 bias and indirectness of Evidence to risk of —Continued Overall Quality Overall Table 2 Table Bias Publication Undetected Low due randomized control trial; RR randomized control trial; RR 5 adjusted to 2-y time frame. All patients wore pressure stockings. 15 15 adjusted to 6-mo time frame. 14 14 PTS (important outcome): self-reported leg symptoms and signs No serious imprecision and Beth et al and Beth et al adjusted to 6-mo time frame . adjusted to 6-mo time frame 13 13 13 13 Predictive value from 3 mo to long term uncertain Serious indirectness postthrombotic syndrome; RCT postthrombotic syndrome; RCT Quality Assessment 5 50% of the acute treatment dose during extended phase treatment. Ն No serious inconsistency Patients and investigators not blinded Serious risk of bias low-molecular-weight heparin; PTS heparin; PTS low-molecular-weight 5 Time frame is 6 mo for all outcomes except PTS, which 2 y. Time Control event rates from cohort study by Prandoni et al, as referenced in the article text. The control event rate for mortality comes from this meta-analysis. Meta-analysis is based on RCTs Outcome less subjective: borderline decision. Control event rate comes from observational studies in review by Kahn et al Limited to LMWH regimens that used One study did not report deaths: borderline decision. Control event rates from cohort studies by Prandoni et al Control event rates from cohort studies by Prandoni et al

a b c d e f g h Participants (Studies), Median Follow-up100 (1 RCT), Risk of Bias 3 mo Inconsistency Indirectness Imprecision

LMWH LMWH

60S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 4.3 Estimating Relative and Absolute Effects Most patient-important outcomes in this guideline are binary or yes-no outcomes (death, stroke, VTE, myocardial infarction, bleeding). In general, relative effects are similar across subgroups of patients, including those with varying baseline risk. 20,21 The and cases of percent ve evidence summaries (Evidence Proﬁ les and Sum- HIT antibodies (remainder tested negative or were not tested). for HIT antibodies. controls tested positive for required positive testing were not provided. mary of Findings tables described later in this article), therefore, include a presentation of relative effects (where possible as relative risks because they are easier to understand than ORs) of intervention vs con- None Sixty-ﬁ Loss to

Follow-up Comments trol management strategies. Trading off desirable and undesirable consequences (eg, thrombosis vs bleeding) requires, however, estimates of absolute effect. For example, in patients with atrial ﬁ brillation, warfarin results in a 66% rela-

No tive risk reduction in nonfatal stroke. This comes at a cost of inconvenience, lifestyle restrictions, and risk

Effectively Blinded of bleeding. For patients with a CHADS (congestive Assessment of Outcomes of Assessment heart failure, hypertension, age Ն 75 years, diabetes mellitus, stroke) score of Ն 3, the 66% relative risk reduction translates into an absolute reduction of

None 6.3% (63 in 1,000) per year. Virtually all patients will consider this worthwhile. On the other hand, for patients with a CHADS score of 0, the 66% reduction translates into an absolute risk reduction of only 0.5% (5 in 1,000) per year. Many patients may consider this reduction not worth the undesirable consequences of

Not close warfarin use. We calculated absolute effects by applying relative Time Frame SimilarTime Adjustment Intervention/Control risks to estimates of control group risk. For instance, if control group risk of thrombosis is 4% and relative risk with an intervention is 50%, then the absolute difference between intervention and control is 4% of Very Not close None No None Cases and controls both Very 50% or 2%, and the number needed to treat to pre-

Somewhat Not close None Novent an episode None of HIT antibody data for cases thrombosis is 100/2 or 50. In many Setting Similar cases, the Summary of Findings tables present effects Intervention/Control as events prevented (or caused) per 1,000 patients. In this hypothetical example, the effect would be 20 events per 1,000 patients.

heparin-induced thrombocytopenia . heparin-induced thrombocytopenia Whenever valid prognostic data were available controls controls controls 5 Cohort with historical Cohort with historical Cohort with historical from observational studies, they were used to esti-

mate control group risks. When credible results

b ; b a from observational and prognostic studies were not available, risk estimates from control groups of RCTs 30)

5 were used. g/kg per min g/kg per min g/kg per min g/kg per min m m 4.4 Considering Subgroup-Speciﬁ c Relative [Section 4.1] Methodologic Quality of Observational Studies: Cohort Studies of the Treatment of Heparin-Induced Thrombocytopenia of Heparin-Induced Thrombocytopenia [Section 4.1] Methodologic Quality of Observational Studies: Cohort Studies the Treatment and Absolute Effects 24; phenoprocoumon, n 21; other, 5 5 and oral anticoagulation) n n and oral anticoagulation) Whenever we identiﬁ ed credible evidence that the Lepirudin: bolus 0.4 mg/kg Lepirudin: bolus 0.4 mg/kg 0.15 mg/kg Argatroban: 2 Argatroban: 2 (no bolus) (no bolus) Control: varied (typically discontinuation of heparin Control: varied (danaparoid, n Control: varied (typically heparin discontinuation relative effects vary across distinguishable subgroups Table 3— Table of patients (ie, interaction between the intervention activated partial thromboplastin time; HIT activated partial thromboplastin time; HIT /2005 /2003 /2001 16 16 17 18

5 and a patient characteristic), we considered the respec- tive relative effects separately. We then calculated et al et al et al et al et al et al APTT adjusted to 1.5 to 3.0 times baseline APTT . APTT adjusted to 1.5 3.0 times baseline APTT adjusted to 1.5 2.5 times baseline (or the mean laboratory normal range if was unavailable). Author/Year Intervention/Control Study Design a b

APTT APTT Lubenow Lewis Lewis the associated absolute effects. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 61S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Figure 2. GRADE approach to rating quality of evidence. See Figure 1 legend for expansion of abbreviation.

Even when the relative effect is the same, the abso- prevention for patients with atrial ﬁ brillation is likely lute magnitude of treatment effects may differ in close to 50% across risk groups, this translates into an patients with varying levels of risk. For instance, although absolute risk reduction of , 1% per year in the lowest- .per year in the highest-risk groups %5 ف the relative risk reduction of warfarin vs aspirin in stroke risk groups and

Table 4— Strength of the Recommendations Grading System

Benefi t vs Risk and Methodologic Strength of Supporting Grade of Recommendation Burdens Evidence Implications Strong recommendation, Benefi ts clearly outweigh Consistent evidence from randomized Recommendation can apply to most high-quality evidence (1A) risk and burdens or vice controlled trials without important patients in most circumstances. Further versa. limitations or exceptionally strong research is very unlikely to change our evidence from observational studies. confi dence in the estimate of effect. Strong recommendation, Benefi ts clearly outweigh Evidence from randomized controlled Recommendation can apply to most moderate-quality risk and burdens or vice trials with important limitations patients in most circumstances. evidence (1B) versa. (inconsistent results, methodologic Higher-quality research may well have fl aws, indirect or imprecise) or very an important impact on our confi dence strong evidence from observational in the estimate of effect and may studies. change the estimate. Strong recommendation, Benefi ts clearly outweigh Evidence for at least one critical Recommendation can apply to most low- or very-low-quality risk and burdens or vice outcome from observational studies, patients in many circumstances. evidence (1C) versa. case series, or randomized controlled Higher-quality research is likely to trials, with serious fl aws or indirect have an important impact on our evidence. confi dence in the estimate of effect and may well change the estimate. Weak recommendation, Benefi ts closely balanced Consistent evidence from randomized The best action may differ depending high-quality evidence (2A) with risks and burden. controlled trials without important on circumstances or patient or societal limitations or exceptionally strong values. Further research is very unlikely evidence from observational studies. to change our confi dence in the estimate of effect. Weak recommendation, Benefi ts closely balanced Evidence from randomized controlled Best action may differ depending on moderate-quality with risks and burden. trials with important limitations circumstances or patient or societal evidence (2B) (inconsistent results, methodologic values. Higher-quality research may fl aws, indirect or imprecise) or very well have an important impact on our strong evidence from observational confi dence in the estimate of effect and studies. may change the estimate. Weak recommendation, Uncertainty in the estimates Evidence for at least one critical Other alternatives may be equally low- or very-low-quality of benefi ts, risks, and outcome from observational studies, reasonable. Higher-quality research is evidence (2C) burden; benefi ts, risk, case series, or randomized controlled likely to have an important impact on and burden may be trials, with serious fl aws or indirect our confi dence in the estimate of effect closely balanced. evidence. and may well change the estimate.

62S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians We included control group risks and absolute effect estimates for different groups in the summaries of effect when (and only when) two conditions were present. First, we required validated prognostic outcomes: cacy cacy outcome: models or, at the very least, credible strategies for Results clinicians to easily identify higher- and lower-risk .05; major bleeding , patients. Second, we identifi ed varying risk groups RR, 0.15; 95% CI, 0.08-0.26 vs 22/1,500; RR, 0.18; 95% CI, 0.06-0.53 P by day 47, one event per group; RR, 1 only when recommendations differed in strength or Primary effi 13/1,502 vs 88/1,500; Composite of DVT and PE up to day 77: 4/1,502 effi other All direction between groups. Both conditions were met, for instance, in the atrial fi brillation recommendations in which strong recommendations in favor of anticoagulation were restricted to the higher-risk patients. cacy outcomes

4.5 Conducting Meta-analyses cial vein 1 thrombosis. for See expansion Table of other When pooled estimates of effects were not avail- Outcomes

able from existing high-quality systematic reviews, superfi 5 we performed meta-analyses if the data were suffi - cacy outcomes: ciently homogeneous. When pooling two studies, cacy outcome: we used a fi xed-effects model. When three or more

studies were available for generating a pooled esti- PE, or DVT symptomatic cause, any from death symptomatic extension to the saphenofemoral junction, recurrence of SVT -Composite of symptomatic events up to day 77 day to up events symptomatic of Primary effi 47: day to up events symptomatic of -Composite Secondary effi -Composite -Each component of primary effi -Composite of symptomatic PE or DVT -Surgery for SVT Safety outcomes (up to day 47 or up 77): -Major bleeding total bleeding minor, -Nonmajor, -Arterial thromboembolic event mate, we used a random-effects model as the primary risk ratio; SVT 5 analysis and a fi xed-effects model as a secondary analysis. If there were discrepancies between the two, we considered the following reasons: If there Placebo for 45 d was substantial heterogeneity leading to wider CIs with the random-effects model, we considered that model more trustworthy, and if the discrepancy was due to a single large dominant study with a result pulmonary embolism; RR pulmonary embolism; RR

substantially different from smaller studies, we con- 5 sidered the ﬁ xed-effects model more trustworthy. Intervention Control daily for 45 d Use of GCS encouraged for all patients

We also assessed statistical heterogeneity using both Fondaparinux, 2.5 mg subcutaneous once a x2 test and I 2 as well as assessed possible explana- tions of heterogeneity considering a priori-generated 22 72 h hypotheses. . ammatory drug; PE ammatory drug; PE 4.6 Summary Tables When resources permitted, we used a standardized approach for summarizing the evidence and methodology of individual studies (examples in Tables 1, 2, 5 ). 48 h or NSAID for . These summaries appear in the online data supple- nonsteroidal antiinﬂ 5 ments. Wherever possible, we report nonfatal events Patients (eg, nonfatal stroke) so that there is no overlap with

the number of fatal events reported. rmed SVT of the legs 3 wk 18 years with acute symptomatic, .

For a large number of recommendations, we sum- Ն marized the quality of the body of evidence ( Fig 2 ) and estimates of relative and absolute effect of alternative management strategies using the methods 23 for current episode of SVT -Ligation or stripping -Major surgery within 3 mo -Bleeding risk -Pregnant or childbearing age women not using reliable contraception of the GRADEmo 6 Workingpast the Group. within Evidence PE or ProfiDVT les mo, 3 past the Patients aged objectively confi Exclusion criteria: -Symptoms within -DVT or PE at presentation IV line -SVT following sclerotherapy, -SVT within 3 cm of the saphenofemoral junction -SVT for cancer within the past 6 mo -Treated -Antithrombotic therapy for summarize the quality of the body of evidence and ; [Section 4.1] Descriptive Table: Randomized Controlled Trials of Fondaparinux vs No Fondaparinux for the Treatment of Superfi cial Vein Thrombosis Thrombosis cial Vein of Superfi of Fondaparinux vs No for the Treatment Randomized Controlled Trials [Section 4.1] Descriptive Table: when evidence comes from randomized trials, gener- 12 ally include a presentation of reviewer assessment of

risk of bias, precision, consistency, directness, and graduated compression stockings; NSAID 5

publication bias associated with each outcome (eg, 5— Table CALISTO Study Group/2010 19 abbreviation. GCS GCS see Table 2 ). As speciﬁ ed in GRADE methodology, Author/Year Decousus et al www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 63S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians the overall quality of evidence represents the lowest quality of any critical outcome. Evidence Proﬁ les can be found in the online data (from supplement. The format for these tables was deter-

mined through a formal survey of panelists that evaluated the panelists’ preferences for alternative 24 presentations and the impact of these presentations on their understanding of the evidence. The text in the printed version of the AT9 recommendations includes more succinct Summary of Findings tables (eg, see 24

Table 6 ), which include the overall quality assess- more major bleeds per 1,000 to 3 more) to 14 fewer) ment as well as the relative and absolute effect sizes to 5 more 0 more to 12 more) for each outcome. Use of an associated computer program facilitated the production of the Evidence Proﬁ les and Summary of Findings tables.25 Anticipated Absolute Effects (1-y Time Frame) Anticipated Absolute Effects (1-y Time bleeds per 1,000a per bleeds 6 Grades of Recommendations, Assessment, Development, and

5.0 Values and Preferences 5 Making trade-offs between desirable and undesirable consequences of alternative management Aspirin in Patients With a Recent ACS and PCI a Recent ACS and PCI Aspirin in Patients With

strategies—the fundamental process of making 1 recommendations—requires making value and pref- dence in the estimate of effect and is likely to change the estimate; and very-low erence judgments. For antithrombotic therapy guide-

lines, this trade-off involves, in most instances, a CI)(95% RR Clopidogrel With Risk CI) (95% Prasugrel With Difference Risk Relative Effect, 0.76 (0.67-0.85) 95 MI per 1,000a 22 fewer MI per 1,000 (from 33 reduction in thrombotic events compared with an 1.32 (1.03-1.68) 22 major

increase in bleeding events. Ideally, the values and thrombolysis in myocardial infarction. preferences applied to this decision would be the 5 average values and preferences of the patient popula- coronary artery bypass graft; GRADE Aspirin vs Clopidogrel Aspirin vs Clopidogrel tion. We know, however, that patient values for health 5 1 High outcomes vary substantially from patient to patient. High (GRADE) Knowledge of the extent to which patient values and risk ratio; TIMI preferences vary is one factor in deciding on the 5 strength of a recommendation. The greater the vari- Quality of the Evidence ability in values and preferences, the more likely a imprecisionto due Moderate 0.89 (0.7-1.12) 24 deaths per 1,000aimprecisionto due Moderate 1.02 (0.71-1.45) (from 7 fewer 3 fewer deaths per 1,000 10 strokes per 1,000a (from 3 fewer 0 fewer strokes per 1,000 weak recommendation is appropriate.23

To inform these decisions, we conducted a system- Mean atic review of the literature bearing on patient values and preferences regarding antithrombotic therapy.26 acute coronary syndromes; CABG

The methodology of conducting such studies remains 5 to be fully developed, and the area remains under- investigated. Nevertheless, the results of the review Follow-up provided guidance for the values and preferences 26 percutaneous coronary intervention; RR that we adopted for these guidelines. 5 No. of Participants (Studies), 13,457 (1 study); median, 14.5 mo 13,608 (1 study), 14.5 mo 13,608 (1 study), 14.5 mo As an additional strategy for achieving meaningful 13,608 (1 study), 14.5 mo value and preference decisions by each topic panel and to facilitate consistency across articles, we con-

ducted a values rating exercise. Topic editors and Prasugrel [Section 4.6] Summary of Findings Table: deputy editors constructed patient scenarios for key outcomes of thrombosis and bleeding relevant to their articles. Then informed by the systematic review of myocardial infarction; PCI Table 6— Table values and preferences, panelists used these 5 scenarios to rate each outcome from 0 (death) to 100 (full health). The mean values of these rat ings guided the trade-offs between thrombotic and bleed- Outcomes Nonfatal MI Nonfatal stroke Vascular mortality Vascular ing events and, thus, the determination of strong Major extracranial bleed Non-CABG- related TIMI major bleeding estimate the in dence The anticipated absolute confi effect is our expressed on as risk impact difference (and important its 95% an CI) and have is to based on likely the baseline is risk in research the comparison further group and the relative effect of quality, the intervention moderate (and its effect; 95% of CI). High estimate the in dence confi our change to unlikely very is research further indicates quality of effect and may further change research the is estimate; very low likely quality, to have an important impact on our confi quality, we are very uncertain about the estimate. ACS Evaluation; MI

64S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians vs weak recommendations. The scenarios and results described in section 6.4 “Criteria for Resource Allo- of the rating exercise are available in the online data cation Issues to Affect Recommendations—Thresholds supplements. for Cost-Effectiveness.” The introductory section in each chapter includes a summary, quantitative wherever possible, of the 6.2 Identifying the Literature key values and preferences underlying the recom- The Oregon Health & Science University Evidence- mendations. Where value and preference judgments based Practice Center conducted thorough literature were particularly relevant or controversial, explicit searches for economic analyses relevant to the dif- statements of values and preferences accompany ferent AT9 articles. The resource use experts supple- individual recommendations. mented these by searches focused on the specifi c T h e l i t e r a t u r e r e v i e w 26 r e v e a l e d e x t e n s i v e h e t e r o - questions of interest for each article. The searches geneity of results across studies of patient values were conducted in Medline and the Cochrane Cen- and preferences—variability that often is diffi cult to tral Register of Clinical Trials. On the basis that explain. Both the variability between studies and data from studies appreciably more than a decade old the considerable variability in values and preferences would not refl ect the current situation, searches were among patients within studies, mandated circumspec- restricted to published studies from 1999 forward. tion in making strong recommendations. Therefore, Thus, bibliographic database searches encompassed we restricted strong recommendations to situations in publications from January 1999 forward: The end which the desirable consequences of an intervention date varied across articles and ranged between substantially and convincingly outweighed the unde- November 2009 and March 2010 when the searches sirable consequences (or the reverse) and to unusual were executed. situations in which there was reason to believe that values and preferences are relatively uniform. 6.3 Evaluating the Evidence A standardized data extraction form was used to 6.0 Resource Use Issues ensure uniform evaluation of the quality of relevant economic analyses. Quality assessment was based In addressing resource use (cost) issues in AT9, we on published criteria27-33 and included specifi cation of 27 followed previously developed principles. In partic- perspective of analysis (eg, societal, health system), ular, we restricted economic evaluation to recom- appropriateness of time horizon (preferably lifetime), mendations in which it was plausible that resource use of high-quality evidence for probabilities and use considerations might change the direction or rates, use of high-quality sources for costs (eg, primary strength of the recommendation and in which high- data, Medicare payments, claims data as proxies), quality economic evaluations were available. When use of appropriate methods for measurement of this was not the case, we did not consider resource preferences, and performance of sensitivity anal- use in the recommendations. yses to explore uncertainty (both deterministic and Six clinicians with the requisite expertise in decision probabilistic). and economic analyses participated in the guideline development process; each article had the benefi t of one of these experts as a full committee member. In 6.4 Criteria for Resource Allocation Issues the following subsections, we present key points in To Affect Recommendations—Thresholds the process of considering resource allocation issues for Cost-Effectiveness in the recommendations. The results of economic analyses may either increase the strength of an otherwise weak recommendation 6.1 Overview of the Process or weaken the strength of a strong recommenda- Panelists, in consultation with resource use consul- tion. If cost-effectiveness studies bolstered an already tants, determined questions for which resource use strong recommendation, no change to the recom- might change the direction or strength of recommen- mendation was necessary. We chose the following dations. For those questions, we sought high-quality thresholds for cost-effectiveness considerations affect- economic analyses. If such analyses were available, ing recommendations: we applied the evidence regarding resource use to 1. When the clinical evidence warrants a strong the relevant recommendation. If net costs or mar- recommendation for A over B : ginal cost-effectiveness ratios were very high, panelists considered rating down the quality of evidence a. Strong recommendation favoring A w h e n h i g h - for an intervention from high to low or possibly chang- quality evidence from economic evaluations ing the direction of the recommendation using guides shows that A c o s t s , 3 times the gross domes tic www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 65S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians product (GDP) per capita (approximately discussed in the previous section have been defi ned US $150,000)34 per quality-adjusted life year in terms of GDP per capita. Although referencing (QALY) gained relative to B thresholds for cost- effectiveness to average per-capita b. Weak recommendation favoring A w h e n h i g h - income in middle- and low-income countries can help quality evidence from economic evaluations to extend results of economic analyses performed in shows that A costs 3 to 5 times the GDP per high-income countries, such analyses may be less rel- per QALY gained evant in low-income countries because of signifi cantly (250,000$-150,000$ ف) capita relative to B different material and labor costs and, thus, may be c. Weak recommendation favoring B w h e n h i g h - diffi cult to extrapolate. Furthermore, the comparator quality evidence from economic evaluations strategies may not be feasible or customary in these shows that A c o s t s . 5 times the GDP per locales. per QALY gained relative (250,000$ ف) capita to B 7.0 Disclosing and Managing 2. When the clinical evidence warrants a weak Conflicts of Interest recommendation for A over B : All panelists were required to disclose both fi nan- a. Strong recommendation favoring A i f A results cial confl icts of interest, such as receipt of funds for in cost savings of . 10% to 20% of the GDP consulting with industry, and intellectual confl icts of relative to B interest, such as publication of original data bearing (10,000$-5,000$ ف ) per capita (Cost savings must represent all downstream directly on a recommendation. Financial and intel- costs and not just the actual cost of the inter- lectual confl icts of interest were classifi ed as primary vention, and analysis must demonstrate a high (more serious) or secondary (less serious). 38 T h e o p e r - level of confi dence that there is a cost savings.) ational defi nition of primary intellectual confl icts of b. Continued weak recommendation favoring A interest included authorship of original studies and when B is marginally more costly than A peer-reviewed grant funding (government, not-for- ( , 10% the GDP per capita) profi t organizations) directly bearing on a recommen- c. Continued weak recommendation favoring A dation. The operational defi nition of primary fi nancial when A costs 0 to 5 times the GDP per capita confl icts of interest included consultancies, advisory per QALY gained relative to B board membership, and the like from industry. Topic d. Weak recommendation favoring B i f A editors had no primary confl icts of interest, as noted. Some deputy editors, who were clinical experts in the (250,000$ ف) costs . 5 times the GDP per capita per QALY gained relative to B topic of the article, had relevant primary confl icts of interest. The ACCP HSP Committee deemed some of these confl icts serious enough to require “manage- 6.5 Extension of Economic Analyses to ment.” Management involved more frequent updates Low- and Middle-Income Countries of disclosures than required of the approved panelists Although certain interventions may be cost-effective without any confl icts and recusal from activities rele- in high-income countries (eg, , $20,000 per QALY vant to that confl ict. gained), in poor countries, $20,000 gained per QALY Topic panel members, including the deputy editor, may be prohibitive. The choice of a threshold will vary with primary confl icts related to a particular recom- depending on who is making resource allocation deci- mendation did not participate in the fi nal delibera- sions. To facilitate the use of already published cost- tions that led to the decision regarding the direction effectiveness analyses, the World Health Organization or strength of a recommendation, nor did they vote (WHO), through its WHO-CHOICE (Choosing Inter- on recommendations for which they were primarily ventions that are Cost Effective) program35 has used confl icted. Panelists with primary confl icts could, criteria suggested by the Commission on Macroeco- however, participate in discussions and offer their nomics and Health. 36 I n t e r v e n t i o n s t h a t c o s t , 1 times opinions on interpretations of the evidence. Readers the average per-capita income for a given country or will fi nd a record of panelist confl icts of interest on a region per QALY gained are considered very cost- recommendation-by-recommendation basis in the effective. Interventions that cost up to three times the online data supplement. average per-capita income per QALY gained are still considered cost-effective, whereas those that exceed 8.0 Finalizing the Recommendations this level are not considered to be cost-effective. To 8.1 Formulating Recommendations facilitate this process, WHO has developed tables of such threshold values for different regions and Following approaches recommended by the GRADE countries around the world.37 Thus, the thresholds Working Group,23 the topic editor, in some cases

66S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians aided by a panelist without confl icts, formulated the were recorded in written and audio formats and are draft recommendations. The formulation of recom- available on request to [email protected]. mendations considered the balance between the desirable and undesirable consequences of an intervention; the quality of evidence; the variability in 9.0 Review by ACCP and External Reviewers patient values and preferences; and, on occasion, resource use issues. The recommendations were The ACCP HSP Committee established a process graded as strong when desirable effects were much for the thorough review of all ACCP evidence-based greater than undesirable effects or vice versa. Strong clinical practice guidelines. After fi nal review by the recommendations were worded as “We recommend” AT9 Executive Committee, the guidelines underwent and labeled 1. Recommendations were graded as weak review by the Cardiovascular and Pulmonary Vascu- when desirable effects were not clearly greater or less lar NetWorks of the ACCP, the HSP Committee, and great than undesirable effects. Weak recommenda- the ACCP Board of Regents. The latter two groups tions were worded as “We suggest” and labeled 2. had the right of approval or disapproval but usually The rating of the quality of the evidence—high, A; worked with the topic panelists and editors to make moderate, B; or low, C—is provided with the strength necessary revisions prior to fi nal approval. Both the of each recommendation. HSP Committee and the Board of Regents identifi ed primary reviewers who read the full set of articles, and the remaining HSP Committee members were 8.2 Finalizing Recommendations responsible for reviewing several articles each. The After completing the steps described previously, reviewers considered both content and methodology the topic panel members without primary confl icts as well as whether there was balanced reporting discussed draft recommendations ( Fig 1 ). Initial dis- and adherence to HSP Committee processes. All cussions generally led to a consensus at the article reviewers were vetted through the same confl ict of level on the quality of evidence and the direction and interest disclosure and management process as strength of recommendations. At least two members described previously. Finally, the Editor in Chief of of the Executive Committee reviewed in detail drafts CHEST read and forwarded the manuscripts for of articles, including recommendations. Written cri- independent, external peer review prior to accep- tiques were prepared and returned to the authors for tance for publication. No recommendations or assess- revision. Articles were then made available to the ments of the quality of the evidence could be changed entire AT9 panel. without the express approval of the topic panel mem- Recommendations on which topic panels had diffi - bers, AT9 Executive Committee, HSP Committee, culty coming to a consensus were discussed at a fi nal and ACCP Board of Regents. conference in February 2011 attended by the topic editors and deputy editors and at least one other panel member from each article. Prior to the con- 10.0 Organization of Articles ference, all AT9 panelists updated their confl ict of interest disclosures. The ACCP invited a number of In order to provide a transparent, explicit link clinical organizations with interest in the guideline among PICO questions, evidence, tables, and recom- topic to attend the fi nal conference as observers. mendations, the section numbering in each article At this fi nal conference, a representative of each corresponds to numbers in Table 1 in each article, article presented potentially controversial issues in which specifi es the patients, interventions, and out- their article’s recommendations. Following discussion, comes for each question. The section numbering also which included those present and those attending by corresponds to the numbering of the recommenda- videoconference, all panelists without primary con- tions themselves. Evidence Profi les and other tables fl icts of interest voted on each recommendation. include these corresponding numbers in brackets The voting process used a GRADE grid and required in the title, as is true for the online data supplement that for a strong recommendation, Ն 80% of those tables. voting had to agree that a strong recommendation was appropriate.39 The AT9 Executive Committee members (G. H. G., 11.0 Revisions in the Process Since AT8 M. C., E. A. A., and D. D. G.) harmonized the articles and resolved remaining disagreements among them AT9 includes improvements from AT8 that refl ect through facilitated discussion with topic editors and the evolution of the science of systematic reviews and deputy editors without primary confl icts. All major clinical practice guidelines. In this supplement, 40 some correspondence and decisions at the fi nal conference of these improvements include augmented provisions www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 67S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians to decrease the likelihood of confl ict of interest infl u- in the PICO format along with study inclusion and ence, more stringent application of GRADE criteria exclusion criteria, an exhaustive search for relevant for evidence and recommendations (both facilitated literature, an evaluation of the risk of bias of included by methodologists without primary confl icts taking studies, and a rigorous and standardized assessment the role of topic editor), and a systematic review of of the quality of the body of evidence and its transla- values and preferences to guide the recommendations. tion into recommendations using GRADE Working Group methodologies. We incorporated specifi cation of values and preferences and resource consider- 12.0 Limitations of Methods ations into recommendations where particularly relevant and when such data were available. Finally, we Although encouraged to use Evidence Profi les and sought to minimize bias potentially introduced by Summary of Findings tables for all recommendations, intellectual and fi nancial confl icts of interest by com- there were some for which the authors were unable prehensive disclosure requirements and aggressive to produce such tables. However, those recommen- management of relevant confl icts. dations used an evidence-based systematic review and assessment of relevant studies. Some recommendations would have benefi ted from meta-analyses that would have clarifi ed aspects of the evidence. Acknowledgments Although panelists were instructed in completing Author contributions: Authors contributed to the AT9 guideline process in the roles described in the article. As Topic Editor the value and preference rating exercise to estimate and Chair of the guideline, Dr Guyatt oversaw the development patient values and preferences rather than to use of this article. their own, we cannot be assured that they succeeded Dr Guyatt: produced the fi rst draft and was responsible for the 41 fi nal article. in all instances. D r N o r r i s : undertook a major revision of a late draft of the article. Dr Schulman: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. 13.0 Plans for Updating AT9 Dr Hirsh: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. We plan to continue the tradition of the antithrom- Dr Eckman: took responsibility for sections relevant to their role botic guidelines to update recommendations when in AT9 and reviewed and approved the fi nal article. important new studies are published that might change Dr Akl: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. the current recommendations. In March 2011, the Dr Crowther: took responsibility for sections relevant to their role ACCP Board of Regents approved a proposal to in AT9 and reviewed and approved the fi nal article. revise the guideline development and updating pro- Dr Vandvik: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. cess to a “living guidelines” process, whereby the Dr Eikelbloom: took responsibility for sections relevant to their evidence-based guidelines will be periodically assessed role in AT9 and reviewed and approved the fi nal article. and updated as the literature warrants. From 1 year Dr McDonagh: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. after the publication of this ninth edition onward, all Dr Lewis: took responsibility for sections relevant to their role in clinical questions or sets of related questions will AT9 and reviewed and approved the fi nal article. become their own units. This process will be dis- Dr Gutterman: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. cussed in greater depth in future publications. Dr Cook: took responsibility for sections relevant to their role in In addition to the published guidelines, ACCP has AT9 and reviewed and approved the fi nal article. historically provided clinical resources, including a Dr Schünemann: took responsibility for sections relevant to their role in AT9 and reviewed and approved the fi nal article. quick reference to the recommendations, patient Financial/nonfi nancial disclosures: In summary, the authors education materials, and slide sets for presenta- have reported to CHEST the following confl icts of interest: tions. These resources will continue to be based Dr Eckman has received the following university grants: “Using Decision Analytic Modeling to Guide the ACCP Guideline Devel- on the guidelines, but they will be accessed online opment Process for Antithrombotic Therapy in Atrial Fibrillation” through the ACCP Web site. In addition, there will (Foundation for Informed Medical Deci sion Making; October 2011- be related resources, tools, and links to make the con- September 2013; $185,000); “Cost-Effectiveness of Screening for Chronic Hepatitis C Infection” (Merck/Schering-Plough; tent more easily useful and searchable. October 2011- September 2012; $58,000); “Greater Cincinnati BEACON Collaborative” (Offi ce of the National Coordinator for Health Information Technology [90BC0016/01]; September 2010- e f f o r t ) ; “Cincinnati Center for Clinical and % 5 1 ف ;March 2012 14.0 Conclusion Translational Science and Training (CTSA) ARRA Supplement for Development of Distance Learning Program in Medical Infor- For AT9, we used an explicit, transparent process matics” (National Institutes of Health [NIH]/National Center for that seeks to produce highly relevant and unbiased Research Resources [NCRR] [UL1 RR026314-01S1]; August 2009- effort); “Cincinnati Center for Clinical %20 ف ;August 2011 recommendations for clinical practice. This process and Translational Science and Training (CTSA)” (NIH/NCRR ;(effort %15 ف ;involved the a priori specifi cation of clinical questions [1U54 RR 025216]; January 2009-February 2014

68S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians “A Patient Specifi c Decision Support Tool for Bariatric Surgery” 4 . Guyatt G , Montori V , Devereaux PJ , Schünemann H , (National Institute of Diabetes and Digestive and Kidney Diseases Bhandari M . Patients at the center: in our practice, and in [K23 DK075599]; August 2007-June 2012 ; no fi nancial support); our use of language . ACP J Club . 2004 ; 140 ( 1 ): A11 - A12 . National Heart, Lung, and Blood Institute (K23 HL085387; 5 . Ferreira-González I , Busse JW , Heels-Ansdell D , et al . June 2008-March 2013; no fi nancial support); and “Cost- Problems with use of composite end points in cardiovascular Effectiveness of Screening for Chronic Hepatitis B Infection” . He trials: systematic review of randomised controlled trials . BMJ .(56,000$ ف ;Gilead Sciences Inc; March 2008-August 2010) has also served as consultant for Savient Pharmaceuticals (“Cost 2007 ; 334 ( 7597 ): 786 . . and as 6 . Ferreira-González I , Permanyer-Miralda G , Busse JW , et al (300$ ف ;Effectiveness Analysis of Gout Medication”; 2010 editorial consultant for the ACCP (“Physicians’ Information and Methodologic discussions for using and interpreting compos- Education Resource [PIER]: Module on Pre-Operative Assessment ite endpoints are limited, but still identify major concerns . .y). Dr Crowther J Clin Epidemiol . 2007 ; 60 ( 7 ): 651 - 662/250$ ف ;for Bleeding Disorders”; 2006-present has served on various advisory boards, has assisted in the prepara- 7 . Montori VM , Permanyer-Miralda G , Ferreira-González I , et al . tion of educational materials, and has sat on data safety and manage- Validity of composite end points in clinical trials . BMJ . ment boards. His institution has received research funds from 2005 ; 330 ( 7491 ): 594 - 596 . the following companies: Leo Pharma A/S, Pfi zer Inc, Boerhinger Ingelheim GmbH, Bayer Healthcare Pharmaceuticals, Octapharm 8 . Falck-Ytter Y, Francis CW, Johanson NA, et al. Prevention of AG, CSL Behring, and Artisan Pharma. Personal total compensa- VTE in orthopedic surgery patients: antithrombotic therapy tion for these activities over the past 3 years totals less than and prevention of thrombosis, 9th ed: American College of US $10,000. Dr Eikelboom has received consulting fees and hono- Chest Physicians evidence-based clinical practice guide lines. raria from AstraZeneca; Boehringer-Ingelheim GmbH; Bristol- Chest . 2012;141(2)(suppl):e278S-e325S. Myers Squibb; Corgenix; Daiichi-Sankyo, Inc; Eisai Co, Inc; 9 . Shea BJ , Hamel C , Wells GA , et al . AMSTAR is a reliable Eli Lilly and Company; GlaxoSmithKline plc; Haemonetics Corp; and valid measurement tool to assess the methodological McNeil Consumer Healthcare; and Sanofi -Aventis LLC and grants quality of systematic reviews . J Clin Epidemiol . 2009 ; 62 ( 10 ): and in-kind support from Accumetrics, Inc; AspirinWorks; Bayer 1013 - 1020 . Healthcare Pharmaceuticals; Boehringer Ingelheim GmbH; Bristol- Myers Squibb; Corgenix; Dade Behring Inc; GlaxoSmithKline 10 . Oxman AD , Guyatt GH , Singer J , et al . Agreement among plc; and Sanofi -Aventis LLC. Dr Gutterman has had the following reviewers of review articles . J Clin Epidemiol . 1991 ; 44 ( 1 ): relationships that are entirely unrelated to the AT9 guidelines: 91 - 98 . ACCP President, GlaxoSmithKline plc grant to study vasodila- 11 . Higgins JP , Altman D . Assessing the risk of bias in included tion in adipose tissue , National Institutes of Health grant to study studies. In: Higgins J , Green S , eds. Cochrane Handbook human coronary dilation, and GE Healthcare consultation on a for Systematic Reviews of Interventions 5.0.1 . Chichester, study for ECG evaluation of chronic heart disease. Drs Guyatt England: John Wiley & Sons ; 2008 : 17 - 21 . and Schünemann are co-chairs of the GRADE Working Group, 12 . Decousus H , Prandoni P , Mismetti P , et al; CALISTO Study and Drs Akl and Vandvik are members and prominent contrib- utors to the Grade Working Group. Dr Lewis is a full-time Group. Fondaparinux for the treatment of superfi cial-vein employee of the ACCP. Drs Norris, Schulman, Hirsh, McDonagh, thrombosis in the legs. N Engl J Med. 2010 ; 363 ( 13 ): 1222 - 1232 . and Cook have reported that no potential confl icts of interest exist 13 . Prandoni P , Lensing AW , Piccioli A , et al. Recurrent venous with any companies/organizations whose products or services may thromboembolism and bleeding complications during anti- be discussed in this article . coagulant treatment in patients with cancer and venous Role of sponsors: The sponsors played no role in the develop- thrombosis . Blood . 2002 ; 100 : 3484 - 3488 . ment of these guidelines. Sponsoring organizations cannot recom- 14 . Beyth RJ , Cohen AM , Landefeld CS . Long-term out- mend panelists or topics, nor are they allowed prepublication comes of deep-vein thrombosis . Arch Intern Med. 1995 ; 155 : access to the manuscripts and recommendations. Guideline panel 1031 - 1037 . members, including the chair, and members of the Health & Sci- ence Policy Committee are blinded to the funding sources. Fur- 15 . Kahn SR , Ginsberg JS . Relationship between deep venous ther details on the Confl ict of Interest Policy are available online thrombosis and the postthrombotic syndrome . Arch Intern at http://chestnet.org. Med. 2004 ; 164 : 17 - 26 . Endorsements: This guideline is endorsed by the American 16 . Lewis BE , Wallis DE , Berkowitz SD , et al. A rgatroban anti- Association for Clinical Chemistry, the American College of Clin- coagulant therapy in patients with heparin-induced thrombo- ical Pharmacy, the American Society of Health-System Pharmacists, cytopenia . Circulation . 2001 ; 103 : 1838 - 1843 . the American Society of Hematology, and the International Society 17 . Lewis BE , Wallis DE , Leya F , et al. Argatroban anticoagu- of Thrombosis and Hematosis. lation in patients with heparin-induced thrombocytopenia . Arch Intern Med. 2003 ; 163 : 1849 - 1856 . 18 . Lubenow N , Eichler P , Lietz T , et al. Lepirudin in patients with heparin-induced thrombocytopenia—results of the References third prospective study (HAT-3) and a combined analysis 1 . Hirsh J , Guyatt G , Albers GW , Harrington R , Schünemann HJ ; of HAT-1, HAT-2, and HAT-3 . J Thromb Haemost. 2005 ; 3 : American College of Chest Physicians . Antithrombotic and 2428 - 2436 . thrombolytic therapy: American College of Chest Physicians 19 . Guyatt GH , Oxman AD , Kunz R , Vist GE , Falck-Ytter Y , Evidence-Based Clinical Practice Guidelines (8th Edition) . Schünemann HJ ; GRADE Working Group . What is “quality Chest . 2008 ; 133 ( suppl 6 ): 110S - 112S . of evidence” and why is it important to clinicians? BMJ . 2 . Guyatt GH , Oxman AD , Vist GE , et al ; GRADE Working 2008 ; 336 ( 7651 ): 995 - 998 . Group . GRADE: an emerging consensus on rating quality 20 . Deeks JJ . Issues in the selection of a summary statistic for of evidence and strength of recommendations. BMJ . 2 0 0 8 ; meta-analysis of clinical trials with binary outcomes . Stat Med . 336 ( 7650 ): 924 - 926 . 2002 ; 21 ( 11 ): 1575 - 1600 . 3 . Guyatt GH , Eikelboom JW , Gould MK , et al . Approach to out- 21 . Furukawa TA , Guyatt GH , Griffi th LE . Can we individ- come measurement in the prevention of thrombosis in surgical ualize the ‘number needed to treat’? An empirical study of and medical patients: antithrombotic therapy and prevention summary effect measures in meta-analyses . Int J Epidemiol . of thrombosis, 9th ed: American College of Chest Physicians 2002 ; 31 ( 1 ): 72 - 76 . evidence-based clinical practice guidelines . Chest . 2 0 1 2 ; 1 4 1 ( 2 ) 22 . Higgins JP , Thompson SG , Deeks JJ , Altman DG . Measuring (suppl):e185S-e194S. inconsistency in meta-analyses . BMJ . 2003 ; 327 ( 7414 ): 557 - 560 . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT 69S

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians 23 . Guyatt GH , Oxman AD , Kunz R , et al ; GRADE Working 32 . Gold MR , Siegel JE , Russell LB , et al . Cost-Effectiveness Group. Going from evidence to recommendations. BMJ . in Health and Medicine . 1st ed. New York, NY : Oxford 2008 ; 336 ( 7652 ): 1049 - 1051 . University Press ; 1996 . 24 . Wiviott SD , Braunwald E , McCabe CH , et al. Prasugrel ver- 33 . Guyatt GH , Oxman AD , Kunz R , et al ; GRADE Working sus clopidogrel in patients with acute coronary syndromes . Group . Incorporating considerations of resources use into N Engl J Med. 2007 ; 357 ( 20 ): 2001 - 2015 . grading recommendations . BMJ . 2008 ; 336 ( 7654 ): 1170 - 1173 . 25 . Brozek J , Oxman A , Schünemann HS. GRADEpro [Computer 34. World Economic Outlook Database-April 2011 : International program]. Version 3.2 for Windows, 2008 . http://www.cc-ims. Monetary Fund. World Health Organization Web site. http:// net/gradepro. or http://mcmaster.ﬂ intbox.com/technology. www.who.int/macrohealth/documents/tough_choices/en/ asp?page5 3993 . index.html. Accessed June 27, 2011 . 26. MacLean S , Mulla S , Akl EA , et al. Patient values and pref- 35 . Hutubessy R , Chisholm D , Edejer TT . Generalized cost- erences in decision making for antithrombotic therapy: a effectiveness analysis for national-level priority-setting in the systematic review: antithrombotic therapy and prevention of health sector . Cost Eff Resour Alloc . 2003 ; 1 ( 1 ): 8 . thrombosis, 9th ed: American College of Chest Physicians 36 . WHO Commission on Macroeconomics and Health . evidence-based clinical practice guidelines. Chest . 2 0 1 2 ; Macroeconomics and Health: Investing in Health for Economic 141(2)(suppl):e1S-e23S. Development. Report of the Commission on Macroeconomics 27 . Guyatt G , Baumann M , Pauker S , et al . Addressing resource and Health: Executive Summary . Geneva, Switzerland : World allocation issues in recommendations from clinical practice Health Organization ; 2001 . guideline panels: suggestions from an American College of 37 . World Health Organization. Threshold Values for Intervention Chest Physicians task force . Chest . 2006 ; 129 ( 1 ): 182 - 187 . Cost-Effectiveness by Region . Geneva, Switzerland : World 28 . Detsky AS , Naglie IG . A clinician’s guide to cost-effectiveness Health Organization ; 2010 . analysis . Ann Intern Med . 1990 ; 113 ( 2 ): 147 - 154 . 38 . Guyatt G , Akl EA , Hirsh J , et al . The vexing problem of 29 . Drummond MF , Richardson WS , O’Brien BJ , Levine M , guidelines and conﬂ ict of interest: a potential solution . Ann Heyland D . Users’ guides to the medical literature. XIII. How Intern Med . 2010 ; 152 ( 11 ): 738 - 741 . to use an article on economic analysis of clinical practice. A. 39 . Jaeschke R , Guyatt GH , Dellinger P , et al ; GRADE Working Are the results of the study valid? Evidence-Based Medicine Group . Use of GRADE grid to reach decisions on clinical prac- Working Group . JAMA . 1997 ; 277 ( 19 ): 1552 - 1557 . tice guidelines when consensus is elusive . BMJ . 2008 ; 337 : a744 . 30 . O’Brien BJ , Heyland D , Richardson WS , Levine M , 40 . Guyatt GH, Akl EA, Crowther M, Schünemann HJ, Drummond MF. Users’ guides to the medical literature. Gutterman DD, Lewis SZ. Introduction to the ninth edition: XIII. How to use an article on economic analysis of clinical antithrombotic therapy and prevention of thrombosis, 9th ed: practice. B. What are the results and will they help me in American College of Chest Physicians evidence-based clinical caring for my patients? Evidence-Based Medicine Working practice guidelines. Chest . 2012;141(2)(suppl):48S-52S. Group . JAMA . 1997 ; 277 ( 22 ): 1802 - 1806 . 41 . Ubel PA , Angott AM , Zikmund-Fisher BJ . Physicians rec- 31 . Eisenberg JM . Clinical economics. A guide to the economic ommend different treatments for patients than they would analysis of clinical practices . JAMA . 1989 ; 262 ( 20 ): 2879 - 2886 . choose for themselves. Arch Intern Med . 2011 ; 171 ( 7 ): 630 - 634 .

70S Methodology for the Guidelines

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Gordon H. Guyatt, Susan L. Norris, Sam Schulman, Jack Hirsh, Mark H. Eckman, Elie A. Akl, Mark Crowther, Per Olav Vandvik, John W. Eikelboom, Marian S. McDonagh, Sandra Zelman Lewis, David D. Gutterman, Deborah J. Cook and Holger J. Schünemann Chest 2012;141; 53S-70S DOI 10.1378/chest.11-2288

This information is current as of February 7, 2012

Supplementary Material View e-supplements related to this article at: http://chestjournal.chestpubs.org/content/suppl/2012/02/06/141.2_suppl.53S.DC1.html Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/141/2_suppl/53S.full.html References This article cites 35 articles, 26 of which can be accessed free at: http://chestjournal.chestpubs.org/content/141/2_suppl/53S.full.html#ref-list-1 Cited Bys This article has been cited by 15 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/141/2_suppl/53S.full.html#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Downloaded from chestjournal.chestpubs.org by guest on February 7, 2012 © 2012 American College of Chest Physicians Antithrombotic Therapy for VTE Disease : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Clive Kearon, Elie A. Akl, Anthony J. Comerota, Paolo Prandoni, Henri Bounameaux, Samuel Z. Goldhaber, Michael E. Nelson, Philip S. Wells, Michael K. Gould, Francesco Dentali, Mark Crowther and Susan R. Kahn

Chest 2012;141;e419S-e494S DOI 10.1378/chest.11-2301

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/141/2_suppl/e419S.full.html

Supplemental material related to this article is available at: http://chestjournal.chestpubs.org/content/suppl/2012/02/03/141.2_suppl. e419S.DC1.html

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians CHEST Supplement ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: ACCP GUIDELINES Antithrombotic Therapy for VTE Disease Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Clive Kearon , MD , PhD ; Elie A. Akl , MD , MPH , PhD ; Anthony J. Comerota , MD ; Paolo Prandoni , MD , PhD ; Henri Bounameaux , MD ; Samuel Z. Goldhaber , MD , FCCP ; Michael E. Nelson , MD , FCCP ; Philip S. Wells , MD ; Michael K. Gould , MD , FCCP ; Francesco Dentali , MD ; Mark Crowther, MD ; and Susan R. Kahn , MD

Background: This article addresses the treatment of VTE disease. Methods: We generated strong (Grade 1) and weak (Grade 2) recommendations based on high- quality (Grade A), moderate-quality (Grade B), and low-quality (Grade C) evidence. Results: For acute DVT or pulmonary embolism (PE), we recommend initial parenteral anticoagulant therapy (Grade 1B) or anticoagulation with rivaroxaban. We suggest low-molecular-weight heparin (LMWH) or fondaparinux over IV unfractionated heparin (Grade 2C) or subcutaneous unfractionated heparin (Grade 2B). We suggest thrombolytic therapy for PE with hypotension (Grade 2C). For proximal DVT or PE, we recommend treatment of 3 months over shorter periods (Grade 1B). For a ﬁ rst proximal DVT or PE that is provoked by surgery or by a nonsurgical transient risk factor, we recommend 3 months of therapy (Grade 1B; Grade 2B if provoked by a nonsurgical risk factor and low or moderate bleeding risk); that is unprovoked, we suggest extended therapy if bleeding risk is low or moderate (Grade 2B) and recommend 3 months of therapy if bleeding risk is high (Grade 1B); and that is associated with active cancer, we recommend extended therapy (Grade 1B; Grade 2B if high bleeding risk) and suggest LMWH over vitamin K antagonists (Grade 2B). We suggest vitamin K antagonists or LMWH over dabigatran or rivaroxaban (Grade 2B). We suggest compression stockings to prevent the postthrombotic syndrome (Grade 2B). For extensive superﬁ cial vein thrombosis, we suggest prophylactic-dose fondaparinux or LMWH over no anticoagulation (Grade 2B), and suggest fondaparinux over LMWH (Grade 2C). Conclusion: Strong recommendations apply to most patients, whereas weak recommendations are sensitive to differences among patients, including their preferences. CHEST 2012; 141(2)(Suppl):e419S–e494S

Abbreviations: CALISTO 5 Comparison of ARIXTRA in Lower Limb Superﬁ cial Thrombophlebitis With Placebo; CDT 5 catheter-directed thrombolysis; CTPH 5 chronic thromboembolic pulmonary hypertension; HR 5 hazard ratio; INR 5 international normalized ratio; IVC 5 inferior vena cava; LMWH 5 low-molecular-weight heparin; PE 5 pulmonary embolism; PESI 5 Pulmonary Embolism Severity Index; PREPIC 5 Prevention du Risque d’Embolie Pulmonaire par Interruption Cave; PTS 5 postthrombotic (phlebitic) syndrome; RR 5 risk ratio; rt-PA 5 recombinant tissue plasminogen activator; SC 5 subcutaneous; SVT 5 superﬁ cial vein thrombosis; tPA 5 tissue plasminogen activator; UEDVT 5 upper-extremity DVT; UFH 5 unfractionated heparin; VKA 5 vitamin K antagonist

American College of Chest Physicians Evidence- Summary of Recommendations Based Clinical Practice Guidelines (8th Edition). Note on Shaded Text: Throughout this guideline, Recommendations that remain unchanged are not shading is used within the summary of recommenda- shaded. tions sections to indicate recommendations that are newly added or have been changed since the publica- 2.1. In patients with acute DVT of the leg treated tion of Antithrombotic and Thrombolytic Therapy: with vitamin K antagonist (VKA) therapy, we www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e419S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians recommend initial treatment with parenteral of diagnostic tests, provided test results are anticoagulation (low-molecular-weight heparin expected within 24 h (Grade 2C). [LMWH], fondaparinux, IV unfractionated heparin [UFH], or subcutaneous [SC] UFH) over no 2.3.1. In patients with acute isolated distal DVT such initial treatment (Grade 1B). of the leg and without severe symptoms or risk factors for extension, we suggest serial imaging 2.2.1. In patients with a high clinical suspicion of the deep veins for 2 weeks over initial antico- of acute VTE, we suggest treatment with paren- agulation (Grade 2C) . teral anticoagulants compared with no treatment while awaiting the results of diagnostic 2.3.2. In patients with acute isolated distal DVT tests (Grade 2C). of the leg and severe symptoms or risk factors for extension (see text), we suggest initial anti- 2.2.2. In patients with an intermediate clinical coagulation over serial imaging of the deep suspicion of acute VTE, we suggest treatment veins (Grade 2C) . with parenteral anticoagulants compared with no treatment if the results of diagnostic tests Remarks: Patients at high risk for bleeding are more are expected to be delayed for more than 4 h likely to beneﬁ t from serial imaging. Patients who (Grade 2C). place a high value on avoiding the inconvenience of repeat imaging and a low value on the inconvenience 2.2.3. In patients with a low clinical suspicion of of treatment and on the potential for bleeding are acute VTE, we suggest not treating with paren- likely to choose initial anticoagulation over serial teral anticoagulants while awaiting the results imaging.

2.3.3. In patients with acute isolated distal DVT Revision accepted August 31, 2011. Affi liations: From the Departments of Medicine and Clinical of the leg who are managed with initial anti- Epidemiology and Biostatistics (Dr Kearon), Michael De Groote coagulation, we recommend using the same School of Medicine, McMaster University, Hamilton, ON, Can- approach as for patients with acute proximal ada; Departments of Medicine, Family Medicine, and Social and Preventive Medicine (Dr Akl), State University of New York at DVT (Grade 1B). Buffalo, Buffalo, NY; Department of Surgery (Dr Comerota), Jobst Vascular Center, Toledo, OH; Department of Cardiotho- 2.3.4. In patients with acute isolated distal DVT racic and Vascular Sciences (Dr Prandoni), University of Padua, of the leg who are managed with serial imaging, Padua, Italy; Department of Medical Specialties (Dr Bounameaux), University Hospitals of Geneva, Geneva, Switzerland; Department we recommend no anticoagulation if the throm- of Medicine (Dr Goldhaber), Brigham and Women’s Hospital, bus does not extend (Grade 1B); we suggest anti- Harvard Medical School, Boston, MA; Department of Medicine coagulation if the thrombus extends but remains (Dr Nelson), Shawnee Mission Medical Center, Shawnee Mission, KS; Department of Medicine (Dr Wells), University of Ottawa, confi ned to the distal veins (Grade 2C); we rec- Ottawa, ON, Canada; Departments of Medicine and Preventive ommend anticoagulation if the thrombus Medicine (Dr Gould), Keck School of Medicine, University of extends into the proximal veins (Grade 1B). Southern California, Los Angeles, CA; Department of Medicine (Dr Dentali), University of Insubria, Varese, Italy; Department of Medicine (Dr Crowther), Michael De Groote School of Med- 2.4. In patients with acute DVT of the leg, we icine, McMaster University, Hamilton, ON, Canada; and Depart- recommend early initiation of VKA (eg, same day ments of Medicine and Clinical Epidemiology and Biostatistics as parenteral therapy is started) over delayed (Dr Kahn), McGill University, Montreal, QC, Canada. Funding/Support : The Antithrombotic Therapy and Prevention initiation, and continuation of parenteral anti- of Thrombosis, 9th ed: American College of Chest Physicians coagulation for a minimum of 5 days and until Evidence-Based Clinical Practice Guidelines received support from the international normalized ratio (INR) is 2.0 or the National Heart, Lung, and Blood Institute [R13 HL104758] and Bayer Schering Pharma AG. Support in the form of educational above for at least 24 h (Grade 1B). grants were also provided by Bristol-Myers Squibb; Pfi zer, Inc; Canyon Pharmaceuticals; and sanofi -aventis US. 2.5.1. In patients with acute DVT of the leg, we Disclaimer: American College of Chest Physician guidelines are suggest LMWH or fondaparinux over IV UFH intended for general information only, are not medical advice, and do not replace professional medical care and physician advice, (Grade 2C) and over SC UFH (Grade 2B for which always should be sought for any medical condition. The LMWH; Grade 2C for fondaparinux). complete disclaimer for this guideline can be accessed at http:// chestjournal.chestpubs.org/content/141/2_suppl/1S. Remarks: Local considerations such as cost, availability, Correspondence to: Elie A. Akl, MD, MPH, PhD, Department of Medicine, State University of New York at Buffalo, ECMC-CC 142, and familiarity of use dictate the choice between 462 Grider St, Buffalo, NY 14215 ; e-mail: [email protected] fondaparinux and LMWH. © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the LMWH and fondaparinux are retained in patients American College of Chest Physicians ( http://www.chestpubs.org/ site/misc/reprints.xhtml ). with renal impairment, whereas this is not a concern DOI: 10.1378/chest.11-2301 with UFH. e420S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 2.5.2. In patients with acute DVT of the leg 2.13.1. In patients with acute DVT of the leg, we treated with LMWH, we suggest once- over recommend against the use of an inferior vena twice-daily administration (Grade 2C). cava (IVC) fi lter in addition to anticoagulants (Grade 1B) . Remarks: This recommendation only applies when the approved once-daily regimen uses the same daily 2.13.2. In patients with acute proximal DVT of dose as the twice-daily regimen (ie, the once-daily the leg and contraindication to anticoagulation, injection contains double the dose of each twice- we recommend the use of an IVC fi lter (Grade daily injection). It also places value on avoiding an 1B) . extra injection per day. 2.13.3. In patients with acute proximal DVT of 2.7. In patients with acute DVT of the leg and the leg and an IVC fi lter inserted as an alterna- whose home circumstances are adequate, we tive to anticoagulation, we suggest a conven- recommend initial treatment at home over tional course of anticoagulant therapy if their treatment in hospital (Grade 1B). risk of bleeding resolves (Grade 2B).

Remarks: The recommendation is conditional on the Remarks: We do not consider that a permanent IVC adequacy of home circumstances: well-maintained fi lter, of itself, is an indication for extended anti- living conditions, strong support from family or coagulation. friends, phone access, and ability to quickly return to the hospital if there is deterioration. It is also condi- 2.14. In patients with acute DVT of the leg, we tional on the patient feeling well enough to be treated suggest early ambulation over initial bed rest at home (eg, does not have severe leg symptoms or (Grade 2C) . comorbidity). Remarks: If edema and pain are severe, ambula- 2.9. In patients with acute proximal DVT of tion may need to be deferred. As per section 4.1, we the leg, we suggest anticoagulant therapy suggest the use of compression therapy in these alone over catheter-directed thrombolysis (CDT) patients. (Grade 2C). 3.0. In patients with acute VTE who are treated Remarks: Patients who are most likely to benefi t from with anticoagulant therapy, we recommend CDT (see text), who attach a high value to prevention long-term therapy (see section 3.1 for recom- of postthrombotic syndrome (PTS), and a lower value mended duration of therapy) over stopping anti- to the initial complexity, cost, and risk of bleeding coagulant therapy after about 1 week of initial with CDT, are likely to choose CDT over anticoagu- therapy (Grade 1B). lation alone. 3.1.1. In patients with a proximal DVT of the leg provoked by surgery, we recommend treatment 2.10. In patients with acute proximal DVT of with anticoagulation for 3 months over (i) treat- the leg, we suggest anticoagulant therapy alone ment of a shorter period (Grade 1B) , (ii) treat- over systemic thrombolysis (Grade 2C). ment of a longer time-limited period (eg, 6 or Remarks: Patients who are most likely to benefi t 12 months) ( G r a d e 1 B ) , or (iii) extended therapy from systemic thrombolytic therapy (see text), who (Grade 1B regardless of bleeding risk). do not have access to CDT, and who attach a high value 3.1.2. In patients with a proximal DVT of the leg to prevention of PTS, and a lower value to the initial provoked by a nonsurgical transient risk factor, complexity, cost, and risk of bleeding with systemic we recommend treatment with anticoagulation thrombolytic therapy, are likely to choose systemic for 3 months over (i) treatment of a shorter thrombolytic therapy over anticoagulation alone. period ( G r a d e 1 B ), (ii) treatment of a longer time- 2.11. In patients with acute proximal DVT of limited period (eg, 6 or 12 months) (Grade 1B), the leg, we suggest anticoagulant therapy alone and (iii) extended therapy if there is a high over operative venous thrombectomy ( G r a d e 2 C ). bleeding risk (Grade 1B). We suggest treatment with anticoagulation for 3 months over extended 2.12. In patients with acute DVT of the leg who therapy if there is a low or moderate bleeding undergo thrombosis removal, we recommend risk (Grade 2B). the same intensity and duration of anticoagulant therapy as in comparable patients who do 3.1.3. In patients with an isolated distal DVT of not undergo thrombosis removal (Grade 1B). the leg provoked by surgery or by a nonsurgical www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e421S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians transient risk factor (see remark), we suggest Remarks (3.1.3, 3.1.4, 3.1.4.3): Duration of treatment treatment with anticoagulation for 3 months of patients with isolated distal DVT refers to patients over treatment of a shorter period (Grade 2C) in whom a decision has been made to treat with anti- and recommend treatment with anticoagulation coagulant therapy; however, it is anticipated that not for 3 months over treatment of a longer time- all patients who are diagnosed with isolated distal DVT limited period (eg, 6 or 12 months) (Grade 1B) will be prescribed anticoagulants (see section 2.3). or extended therapy (Grade 1B regardless of bleeding risk). In all patients who receive extended anticoagulant therapy, the continuing use of treatment 3.1.4. In patients with an unprovoked DVT of should be reassessed at periodic intervals (eg, the leg (isolated distal [see remark] or proximal), annually). we recommend treatment with anticoagulation for at least 3 months over treatment of a shorter 3.2. In patients with DVT of the leg who are duration ( G r a d e 1 B ). After 3 months of treatment, treated with VKA, we recommend a therapeutic patients with unprovoked DVT of the leg should INR range of 2.0 to 3.0 (target INR of 2.5) over be evaluated for the risk-beneﬁ t ratio of extended a lower (INR , 2) or higher (INR 3.0-5.0) range therapy. for all treatment durations (Grade 1B).

3.1.4.1. In patients with a fi rst VTE that is an 3.3.1. In patients with DVT of the leg and no unprovoked proximal DVT of the leg and who cancer, we suggest VKA therapy over LMWH have a low or moderate bleeding risk, we suggest for long-term therapy (Grade 2C). For patients extended anticoagulant therapy over 3 months with DVT and no cancer who are not treated of therapy (Grade 2B). with VKA therapy, we suggest LMWH over dabigatran or rivaroxaban for long-term therapy 3.1.4.2. In patients with a fi rst VTE that is an (Grade 2C). unprovoked proximal DVT of the leg and who have a high bleeding risk, we recommend 3.3.2. In patients with DVT of the leg and cancer, 3 months of anticoagulant therapy over extended we suggest LMWH over VKA therapy ( G r a d e 2 B ) . therapy (Grade 1B). In patients with DVT and cancer who are not treated with LMWH, we suggest VKA over dab- 3.1.4.3. In patients with a fi rst VTE that is igatran or rivaroxaban for long-term therapy an unprovoked isolated distal DVT of the leg ( G r a d e 2 B ). (see remark), we suggest 3 months of anticoagulant therapy over extended therapy in those Remarks (3.3.1-3.3.2): Choice of treatment in patients with a low or moderate bleeding risk (Grade 2B) with and without cancer is sensitive to the individual and recommend 3 months of anticoagulant patient’s tolerance for daily injections, need for labo- treatment in those with a high bleeding risk ratory monitoring, and treatment costs. (Grade 1B). LMWH, rivaroxaban, and dabigatran are retained in 3.1.4.4. In patients with a second unprovoked patients with renal impairment, whereas this is not a VTE, we recommend extended anticoagulant concern with VKA. therapy over 3 months of therapy in those who have a low bleeding risk (Grade 1B), and we sug- Treatment of VTE with dabigatran or rivaroxaban, in gest extended anticoagulant therapy in those addition to being less burdensome to patients, may with a moderate bleeding risk (Grade 2B). prove to be associated with better clinical outcomes than VKA and LMWH therapy. When these guide- 3.1.4.5. In patients with a second unprovoked lines were being prepared (October 2011), postmar- VTE who have a high bleeding risk, we sug- keting studies of safety were not available. Given the gest 3 months of anticoagulant therapy over paucity of currently available data and that new data extended therapy (Grade 2B). are rapidly emerging, we give a weak recommendation in favor of VKA and LMWH therapy over dab- 3.1.5. In patients with DVT of the leg and igatran and rivaroxaban, and we have not made any active cancer, if the risk of bleeding is not recommendations in favor of one of the new agents high, we recommend extended anticoagulant over the other. therapy over 3 months of therapy (Grade 1B), and if there is a high bleeding risk, we suggest 3.4. In patients with DVT of the leg who receive extended anticoagulant therapy (Grade 2B). extended therapy, we suggest treatment with the e422S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians same anticoagulant chosen for the ﬁ rst 3 months of diagnostic tests, provided test results are (Grade 2C). expected within 24 h (Grade 2C).

3.5. In patients who are incidentally found to 5.3. In patients with acute PE, we recommend have asymptomatic DVT of the leg, we suggest early initiation of VKA (eg, same day as paren- the same initial and long-term anticoagulation teral therapy is started) over delayed initiation, as for comparable patients with symptomatic and continuation of parenteral anticoagulation DVT (Grade 2B). for a minimum of 5 days and until the INR is 2.0 or above for at least 24 h (Grade 1B) . 4.1. In patients with acute symptomatic DVT of the leg, we suggest the use of compression 5.4.1. In patients with acute PE, we suggest stockings (Grade 2B) . LMWH or fondaparinux over IV UFH (Grade 2C for LMWH; Grade 2B for fondaparinux) and over Remarks: Compression stockings should be worn SC UFH (Grade 2B for LMWH; Grade 2C for for 2 years, and we suggest beyond that if patients fondaparinux). have developed PTS and fi nd the stockings help ful. Patients who place a low value on preventing PTS Remarks: Local considerations such as cost, availability, or a high value on avoiding the inconvenience and familiarity of use dictate the choice between and discomfort of stockings are likely to decline fondaparinux and LMWH. stockings. LMWH and fondaparinux are retained in patients 4.2.1. In patients with PTS of the leg, we sug- with renal impairment, whereas this is not a concern gest a trial of compression stockings ( G r a d e 2 C ). with UFH. 4.2.2. In patients with severe PTS of the leg that In patients with PE where there is concern about the is not adequately relieved by compression stock- adequacy of SC absorption or in patients in whom ings, we suggest a trial of an intermittent com- thrombolytic therapy is being considered or planned, pression device (Grade 2B). initial treatment with IV UFH is preferred to use of SC therapies. 4.3. In patients with PTS of the leg, we suggest that venoactive medications (eg, rutosides, defi - 5.4.2. In patients with acute PE treated with brotide, and hidrosmin) not be used ( G r a d e 2 C ) . LMWH, we suggest once- over twice-daily administration (Grade 2C). Remarks: Patients who value the possibility of response over the risk of side effects may choose to Remarks: This recommendation only applies when undertake a therapeutic trial. the approved once-daily regimen uses the same daily dose as the twice-daily regimen (ie, the once-daily 5.1. In patients with acute PE, we recommend injection contains double the dose of each twice- initial treatment with parenteral anticoagula- daily injection). It also places value on avoiding an tion (LMWH, fondaparinux, IV UFH, or SC UFH) extra injection per day. over no such initial treatment ( G r a d e 1 B ) . 5.5. In patients with low-risk PE and whose 5.2.1. In patients with a high clinical suspicion home circumstances are adequate, we suggest of acute PE, we suggest treatment with paren- early discharge over standard discharge (eg, teral anticoagulants compared with no treat- after fi rst 5 days of treatment) (Grade 2B). ment while awaiting the results of diagnostic tests (Grade 2C). Remarks: Patients who prefer the security of the hospital to the convenience and comfort of home are likely 5.2.2. In patients with an intermediate clinical to choose hospitalization over home treatment. suspicion of acute PE, we suggest treatment with parenteral anticoagulants compared with 5.6.1.1. In patients with acute PE associated no treatment if the results of diagnostic tests with hypotension (eg, systolic BP , 90 mm Hg) are expected to be delayed for more than 4 h who do not have a high bleeding risk, we (Grade 2C). suggest systemically administered thrombolytic therapy over no such therapy (Grade 2C). 5.2.3. In patients with a low clinical suspicion of acute PE, we suggest not treating with paren- 5.6.1.2. In most patients with acute PE not asso- teral anticoagulants while awaiting the results ciated with hypotension, we recommend against www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e423S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians systemically administered thrombolytic therapy 6.1. In patients with PE provoked by surgery, (Grade 1C). we recommend treatment with anticoagulation for 3 months over (i) treatment of a shorter 5.6.1.3. In selected patients with acute PE not period ( G r a d e 1 B ), (ii) treatment of a longer time- associated with hypotension and with a low limited period (eg, 6 or 12 months) (Grade 1B), bleeding risk whose initial clinical presentation, or (iii) extended therapy (Grade 1B regardless of or clinical course after starting anticoagulant bleeding risk). therapy, suggests a high risk of developing hypotension, we suggest administration of thrombo- 6.2. In patients with PE provoked by a nonsur- lytic therapy (Grade 2C) . gical transient risk factor, we recommend treatment with anticoagulation for 3 months over 5.6.2.1. In patients with acute PE, when a throm- (i) treatment of a shorter period ( G r a d e 1 B ) , bolytic agent is used, we suggest short infusion (ii) treatment of a longer time-limited period times (eg, a 2-h infusion) over prolonged infu- (eg, 6 or 12 months) ( G r a d e 1 B ), and (iii) extended sion times (eg, a 24-h infusion) (Grade 2C) . therapy if there is a high bleeding risk ( G r a d e 1 B ). We suggest treatment with anticoagulation for 5.6.2.2. In patients with acute PE when a throm- 3 months over extended therapy if there is a bolytic agent is used, we suggest administration low or moderate bleeding risk (Grade 2B). through a peripheral vein over a pulmonary artery catheter (Grade 2C) . 6.3. In patients with an unprovoked PE, we recommend treatment with anticoagulation for 5.7. In patients with acute PE associated with at least 3 months over treatment of a shorter dura- hypotension and who have (i) contraindication ( G r a d e 1 B ). After 3 months of treatment, tions to thrombolysis, (ii) failed thrombolysis, patients with unprovoked PE should be evaluated or (iii) shock that is likely to cause death for the risk-benefi t ratio of extended therapy. before systemic thrombolysis can take effect (eg, within hours), if appropriate expertise and 6.3.1. In patients with a fi rst VTE that is an resources are available, we suggest catheter- unprovoked PE and who have a low or moderate assisted thrombus removal over no such inter- bleeding risk, we suggest extended anticoagulant vention (Grade 2C) . therapy over 3 months of therapy (Grade 2B). 5.8. In patients with acute PE associated with 6.3.2. In patients with a fi rst VTE that is an hypotension, we suggest surgical pulmonary unprovoked PE and who have a high bleeding embolectomy over no such intervention if risk, we recommend 3 months of anticoagulant they have (i) contraindications to thrombolysis, therapy over extended therapy (Grade 1B). (ii) failed thrombolysis or catheter-assisted embolectomy, or (iii) shock that is likely to 6.3.3. In patients with a second unprovoked cause death before thrombolysis can take effect VTE, we recommend extended anticoagulant (eg, within hours), provided surgical expertise therapy over 3 months of therapy in those who and resources are available (Grade 2C). have a low bleeding risk (Grade 1B), and we suggest extended anticoagulant therapy in those 5.9.1. In patients with acute PE who are treated with a moderate bleeding risk (Grade 2B). with anticoagulants, we recommend against the use of an IVC fi lter (Grade 1B). 6.3.4. In patients with a second unprovoked VTE who have a high bleeding risk, we sug- 5.9.2. In patients with acute PE and contraindi- gest 3 months of therapy over extended therapy cation to anticoagulation, we recommend the (Grade 2B). use of an IVC fi lter (Grade 1B). 6.4. In patients with PE and active cancer, if 5.9.3. In patients with acute PE and an IVC fi l- there is a low or moderate bleeding risk, we rec- ter inserted as an alternative to anticoagulation, ommend extended anticoagulant therapy over we suggest a conventional course of anticoagu- 3 months of therapy ( G r a d e 1 B ), and if there is lant therapy if their risk of bleeding resolves a high bleeding risk, we suggest extended anti- (Grade 2B). coagulant therapy (Grade 2B). Remarks: We do not consider that a permanent Remarks: In all patients who receive extended anticoag- IVC fi lter, of itself, is an indication for extended ulant therapy, the continuing use of treatment should anticoagulation. be reassessed at periodic intervals (eg, annually). e424S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 6.5. In patients with PE who are treated with 8.1.1. In patients with superfi cial vein thrombo- VKA, we recommend a therapeutic INR range sis (SVT) of the lower limb of at least 5 cm in of 2.0 to 3.0 (target INR of 2.5) over a lower length, we suggest the use of a prophylactic (INR , 2) or higher (INR 3.0-5.0) range for all dose of fondaparinux or LMWH for 45 days treatment durations (Grade 1B). over no anticoagulation (Grade 2B). 6.6. In patients with PE and no cancer, we sug- Remarks: Patients who place a high value on avoiding gest VKA therapy over LMWH for long-term the inconvenience or cost of anticoagulation and a therapy (Grade 2C). For patients with PE and no low value on avoiding infrequent symptomatic VTE cancer who are not treated with VKA therapy, are likely to decline anticoagulation. we suggest LMWH over dabigatran or rivaroxaban for long-term therapy (Grade 2C) . 8.1.2. In patients with SVT who are treated with anticoagulation, we suggest fondaparinux 6.7. In patients with PE and cancer, we suggest 2.5 mg daily over a prophylactic dose of LMWH LMWH over VKA therapy (Grade 2B). In patients (Grade 2C) . with PE and cancer who are not treated with LMWH, we suggest VKA over dabigatran or 9.1.1. In patients with acute upper-extremity rivaroxaban for long-term therapy (Grade 2C). DVT (UEDVT) that involves the axillary or more proximal veins, we recommend acute treatment Remarks (6.6-6.7): Choice of treatment in patients with parenteral anticoagulation (LMWH, fonda- with and without cancer is sensitive to the individual parinux, IV UFH, or SC UFH) over no such acute patient’s tolerance for daily injections, need for labo- treatment (Grade 1B). ratory monitoring, and treatment costs. 9.1.2. In patients with acute UEDVT that Treatment of VTE with dabigatran or rivaroxaban, in involves the axillary or more proximal veins, we addition to being less burdensome to patients, may suggest LMWH or fondaparinux over IV UFH prove to be associated with better clinical outcomes (Grade 2C) and over SC UFH (Grade 2B). than VKA and LMWH therapy. When these guidelines were being prepared (October 2011), postmar- 9.2.1. In patients with acute UEDVT that involves keting studies of safety were not available. Given the the axillary or more proximal veins, we sug- paucity of currently available data and that new data gest anticoagulant therapy alone over throm- are rapidly emerging, we give a weak recommendation bolysis (Grade 2C) . in favor of VKA and LMWH therapy over dabigatran and rivaroxaban, and we have not made any recommen- Remarks: Patients who (i) are most likely to benefi t dation in favor of one of the new agents over the other. from thrombolysis (see text); (ii) have access to CDT; (iii) attach a high value to prevention of PTS; and (iv) 6.8. In patients with PE who receive extended attach a lower value to the initial complexity, cost, therapy, we suggest treatment with the same and risk of bleeding with thrombolytic therapy are anticoagulant chosen for the fi rst 3 months likely to choose thrombolytic therapy over anticoagu- (Grade 2C). lation alone. 6.9. In patients who are incidentally found to have asymptomatic PE, we suggest the same 9.2.2. In patients with UEDVT who undergo initial and long-term anticoagulation as for thrombolysis, we recommend the same inten- comparable patients with symptomatic PE sity and duration of anticoagulant therapy as in (Grade 2B). similar patients who do not undergo thrombolysis (Grade 1B). 7.1.1. In patients with chronic thromboembolic pulmonary hypertension (CTPH), we rec- 9.3.1. In most patients with UEDVT that is asso- ommend extended anticoagulation over stopping ciated with a central venous catheter, we sug- therapy (Grade 1B). gest that the catheter not be removed if it is functional and there is an ongoing need for the 7.1.2. In selected patients with CTPH, such catheter (Grade 2C). as those with central disease under the care of an experienced throm boendarterectomy team, 9.3.2. In patients with UEDVT that involves we suggest pulmonary thromboendarterectomy the axillary or more proximal veins, we sug- over no pulmonary thromboendarterectomy gest a minimum duration of anticoagulation of (Grade 2C). 3 months over a shorter period (Grade 2B).

www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e425S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Remarks: This recommendation also applies if the which are collectively referred to as VTE. We also UEDVT was associated with a central venous cath- provide recommendations for patients with (1) post- eter that was removed shortly after diagnosis. thrombotic syndrome (PTS), (2) chronic thromboembolic pulmonary hypertension (CTPH), (3) incidentally 9.3.3. In patients who have UEDVT that is asso- diagnosed (asymptomatic) DVT or PE, (4) acute ciated with a central venous catheter that is upper-extremity DVT (UEDVT), (5) superfi cial vein removed, we recommend 3 months of antico- thrombosis (SVT), (6) splanchnic vein thrombosis, agulation over a longer duration of therapy in and (7) hepatic vein thrombosis. patients with no cancer (Grade 1B), and we sug- Table 1 describes the populations, interventions, gest this in patients with cancer (Grade 2C). comparators, and outcomes (ie, PICO elements) for 9.3.4. In patients who have UEDVT that is asso- the questions addressed in this article and the design ciated with a central venous catheter that is not of the studies used to address them. Refer to 1 2 3 removed, we recommend that anticoagulation Garcia et al, Ageno et al, and Holbrook et al in is continued as long as the central venous cath- these guidelines for recommendations on the man- eter remains over stopping after 3 months of agement of parenteral anticoagulation (dosing and treatment in patients with cancer (Grade 1C), monitoring) and oral anticoagulation (dosing and 4 5 and we suggest this in patients with no cancer mon itoring). Refer to Bates et al and Monagle et al (Grade 2C). in these guidelines for recommendations for pregnancy and neonates and children. The current article 9.3.5. In patients who have UEDVT that is not builds on previous versions of these guidelines and, associated with a central venous catheter or with most recently, the eighth edition.6 cancer, we recommend 3 months of anticoagulation over a longer duration of therapy ( G r a d e 1 B ). 1.0 Methods 9.4. In patients with acute symptomatic UEDVT, we suggest against the use of compression 1.1 Presentation as DVT or PE sleeves or venoactive medications (Grade 2C) . In addressing DVT, we fi rst review studies that included (1) only patients who presented with symptomatic DVT or (2) patients 9.5.1. In patients who have PTS of the arm, we who presented with DVT or PE (ie, meeting the broader criterion suggest a trial of compression bandages or of VTE). For the PE components, we review studies (and sub- sleeves to reduce symptoms (Grade 2C). groups within studies) that required patients to have presented with symptomatic PE (who may also have had symptoms of DVT). 9.5.2. In patients with PTS of the arm, we sug- For this reason and because more patients with VTE present with gest against treatment with venoactive medica- symptoms of DVT alone than with symptoms of PE (including tions (Grade 2C). those who also have symptoms of DVT), the DVT section deals with a larger body of evidence than the PE section. 10.1. In patients with symptomatic splanch- In the evaluation of anticoagulant therapy, there are a number of justifi cations for inclusion of patients who present with DVT nic vein thrombosis (portal, mesenteric, and/or and PE in the same study, and for extrapolating evidence obtained splenic vein thromboses), we recommend antiin patients with one presentation of VTE (eg, DVT) to the other coagulation over no anticoagulation (Grade 1B). presentation (eg, PE). First, a majority of patients with symptomatic DVT also have PE (symptomatic or asymptomatic), and 10.2. In patients with incidentally detected a majority of those with symptomatic PE also have DVT (symp- splanchnic vein thrombosis (portal, mesenteric, tomatic or asymptomatic).7,8 Second, clinical trials of anticoagu- and/or splenic vein thromboses), we suggest no lant therapy have yielded similar estimates for effi cacy and safety in patients with DVT alone, in those with both DVT and PE, and anticoagulation over anticoagulation ( G r a d e 2 C ). in those with only PE. Third, the risk of recurrence appears to be similar after PE and after proximal DVT.7,9 Consequently, the 11.1. In patients with symptomatic hepatic vein results of all studies of VTE have been considered when formu- thrombosis, we suggest anticoagulation over no lating recommendations for short- and long-term anticoagula- anticoagulation (Grade 2C) . tion of proximal DVT and PE ( Fig 1 ), and these recommendations are essentially the same for proximal DVT or PE. 11.2. In patients with incidentally detected There are, however, some important differences between hepatic vein thrombosis, we suggest no antico- patients who present with PE and those who present with DVT agulation over anticoagulation (Grade 2C). that justify separate consideration of some aspects of the treatment of PE. First, the risk of early death (within 1 month) from VTE due to either the initial acute episode or recurrent VTE is much his article provides recommendations for the greater after presenting with PE than after DVT9 ; this difference may justify more aggressive initial treatment of PE (eg, thrombo- T use of antithrombotic agents as well as the use of lytic therapy, insertion of an inferior vena cava (IVC) fi lter, more devices or surgical techniques in the treatment of intensive anticoagulant therapy) compared with DVT. Second, patients with DVT and pulmonary embolism (PE), recurrent episodes of VTE are about three times as likely to be e426S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies (Continued) RCTs and RCTs RCTs RCTs and RCTs RCTs and RCTs RCTs and RCTs and RCTs and RCTs and RCTs RCTs RCTs and RCTs bleeding, mortality, QOL, PTS, PTS, QOL, mortality, bleeding, and complications of procedure bleeding, mortality, QOL, PTS, PTS, QOL, mortality, bleeding, and complications of procedure bleeding, mortality, QOL, PTS, PTS, QOL, mortality, bleeding, and complications of procedure bleeding, mortality, QOL, PTS, PTS, QOL, mortality, bleeding, and complications of procedure major bleeding, mortality, major bleeding, mortality, QOL, and PTS mortality bleeding, mortality, QOL, and PTS bleeding, mortality, QOL, and PTS bleeding, mortality, QOL, and PTS bleeding, mortality, QOL, and PTS QOL, and PTS mortality, major bleeding, mortality, QOL, PTS, shorter ICU and hospital stays, acute complications Recurrent DVT and PE, PE, major bleeding, and DVT extension, PE, major DVT extension, PE, major Recurrent DVT and PE, major Recurrent DVT and PE, major major PE, and DVT Recurrent Recurrent DVT and PE, major DVT extension, PE, major DVT extension, PE, major bleeding, major PE, extension, DVT Recurrent DVT and PE, lter lter of VKA rivaroxaban removal or another method of thrombus removal addition to IVC fi No IVC fi lter No IVC fi No anticoagulation No anticoagulation No anticoagulation Shorter duration No IVC fi LMWH, fondaparinux, No active thrombus No anticoagulation in lter Structured PICO Question —Structured Clinical Questions therapy thrombectomy thrombolysis addition to IVC fi IVC fi lter IVC fi Early ambulation Initial bed rest Recurrent DVT and PE, major Systemic thrombolytic Operative venous lter IVC fi Anticoagulation Anticoagulation Anticoagulation Early initiation of VKALonger duration Delayed initiation UFH IV or SQ In-hospital treatmentCatheter directed At-home treatment Anticoagulation in Patient with acute DVT of the leg (2.0-3.0) Table 1 Table lter, now contraindication lter, the leg and a contraindication to anticoagulation leg started on anticoagulant treatment of the leg started on anticoagulation the leg DVT of the leg awaiting results of diagnostic tests distal DVT of the leg of the leg of the leg of the leg of the leg DVT of the leg leg who initially received an IVC fi to anticoagulation resolved Patients with acute DVT of Patients with acute DVT of Patients with acute DVT Patients with suspected acute Patient with acute isolated Patients with acute DVT Patients with acute DVT Patients with acute DVT Patients with acute DVT Patients with acute proximal Patients with acute DVT of the lters when who initially received an IVC lter when contraindication to fi anticoagulation resolves (2.13.3) anticoagulation is contraindicated (2.13.2) anticoagulation (2.13.1) results of the diagnostic work-up (2.2.1- 2.2.3) thrombosis (2.3.1-2.3.4) to the initiation of parenteral anticoagulation (2.4) anticoagulation (2.4) anticoagulant (2.5.1, 2.5.2, 2.6) anticoagulation (2.7) interventions (2.9-2.12) Role of anticoagulation in patients Role of IVC fi Role of early ambulation (2.14) Patients with acute DVT of the Role of IVC fi lters in addition to Role of IVC fi Issue (Informal Question)Initial anticoagulant (2.1) Population Intervention Comparators Outcome Methodology Whether to treat while awaiting the Whether to treat isolated distal of initiation VKA relative Timing Duration of initial Choice and route of initial Setting of initial mechanical and thrombolytic of Role www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e427S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies (Continued) RCTs RCTs RCTs RCTs RCTs RCTs and RCTs and RCTs and RCTs RCTs and RCTs and RCTs RCTs and RCTs major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS DVT ulceration ulceration DVT major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, QOL, PTS, and recurrent QOL, symptomatic relief, QOL, symptomatic relief, QOL, PTS, and recurrent Recurrent DVT and PE, Recurrent DVT and PE, anticoagulation therapy stockings anticoagulation ranges stockings pneumatic compression medications of VKA Shorter duration No long-term VKA Recurrent DVT and PE, No intermittent Structured PICO Question —Continued Patient with acute PE Table 1 Table Patients with PTS of the leg therapy rivaroxaban compression Longer duration Long-term anticoagulation anticoagulation Long-term Longer duration Shorter duration Recurrent DVT and PE, Compression stockings No compression INR 2-3LMWH, dabigatran, Anticoagulation Higher or lower INR No anticoagulation Recurrent DVT and PE, Anticoagulation No anticoagulation Recurrent DVT and PE, the leg of the leg leg started on anticoagulant treatment of the leg of the leg. diagnosed asymptomatic DVT of the leg PE awaiting the results of the diagnostic tests Patients with acute VTE of Patients with an acute DVT Patients with acute DVT of the Patients with acute DVT Patients with incidentally Patients with PTS of the legPatients with PTS of the leg Compression stockingsPatients with PTS of the leg Intermittent pneumatic No compression medications Venoactive Patients with suspected acute No venoactive Patients with acute PE Early initiation of VKA Delayed initiation Patients with acute PE therapy (3.0) anticoagulation (3.1.1- 3.1.5) preventing PTS (4.1) (3.3.1, 3.3.2, 3.4) diagnosed asymptomatic acute DVT of the leg (3.5) in PTS (4.2.1) compression in PTS (4.2.2) PTS (4.3) results of the diagnostic work-up (5.2.1-5.2.3) to the initiation of parenteral anticoagulation (5.3) anticoagulation (5.3) Issue (Informal Question)Long-term anticoagulation Population Intervention Comparators Outcome Methodology Duration of long-term Role of compression stocking in Initial anticoagulant (5.1) Patients with acute PE Anticoagulation No initial Intensity of VKA (3.2)Choice of long-term anticoagulant Whether to treat an incidentally Patients with acute DVT Role of compression stocking Role of intermittent pneumatic Role of venoactive medications in Whether to treat while awaiting the of initiation VKA relative Timing Duration of initial e428S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohorts (Continued) RCTs and RCTs RCTs and RCTs RCTs and RCTs RCTs RCTs RCTs and RCTs and RCTs and RCTs and RCTs RCTs and RCTs RCTs and RCTs RCTs and RCTs major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS QOL, and PTS bleeding, mortality, QOL, and PTS bleeding, mortality, Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, major Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, major Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, Recurrent DVT and PE, lter addition to IVC fi and rivaroxaban therapy catheter-assisted catheter-assisted thrombus removal embolectomy INR range Higher or lower Shorter duration No IVC fi lter No IVC fi No anticoagulation in No IVC fi lter No IVC fi No anticoagulation No use of No surgical pulmonary VKA lter Structured PICO Question —Continued Table 1 Table addition to IVC fi thrombus removal embolectomy rivaroxaban INR 2-3 Anticoagulation in Longer duration LMWH, dabigatran, IVC fi lter IVC fi UFH IV or SQ LMWH, fondaparinux, Longer infusion timePeripheral vein Shorter infusion time Recurrent DVT and PE, Pulmonary catheter lter IVC fi Recurrent DVT and PE, Anticoagulation initially received an IVC now contraindication lter, fi to anticoagulation resolved a contraindication to anticoagulation of the leg requiring thrombolytic therapy requiring thrombolytic therapy on anticoagulation diagnosed asymptomatic PE Patients with acute PE who Patients with acute PE Patients with acute PE and Patients with acute DVT Patients with acute PEPatients with acute PE In-hospital treatment Thrombolytic therapy At-home treatment No thrombolytic Recurrent DVT and PE, Patients with acute PE Patients with acute PE Patients with acute PEPatients with acute PE Patients with acute PE started Use of catheter-assisted Surgical pulmonary Patients with acute PE Patients with incidentally Patients with acute PE lters when anticoagulation in patients with who initially received an IVC lter when contraindication to fi anticoagulation resolves (5.9.3) anticoagulation is contraindicated (5.9.2) anticoagulant (5.4.1, 5.4.2) anticoagulation (5.5) with acute PE (5.6.1.1, 5.6.1.2, 5.6.1.3) therapy (5.6.2.1) therapy (5.6.2.2) removal (5.7) embolectomy (5.8) anticoagulation in patients with acute PE (5.9.1) (6.6, 6.7, 6.8) diagnosed asymptomatic acute PE (6.9) acute PE (6.1-6.4) Duration of long-term Role of anticoagulation in patients Role of IVC fi Issue (Informal Question)Choice and route of initial Population Intervention Comparators Outcome Methodology Setting of initial Thrombolytic therapy in patients Infusion time for thrombolytic access for thrombolytic Venous thrombus Role of catheter-assisted Role of surgical pulmonary lter in addition to Role of IVC fi Intensity of VKA (6.5) Choice of long-term anticoagulant incidentally an treat to Whether www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e429S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians s and upper- 5 cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies cohort studies RCTs and RCTs and RCTs and RCTs and RCTs and RCTs and RCTs and RCTs RCTs and RCTs and RCTs and RCTs RCT pulmonary embolism; 5 cial vein thrombosis; UEDVT cial vein thrombosis; UEDVT superﬁ 5 major bleeding, mortality, major bleeding, mortality, QOL, and PTS QOL, bleeding, mortality, hospital and ICU shorter PTS, stays, and acute complications QOL, bleeding, mortality, hospital and ICU shorter PTS, stays, and acute complications QOL, bleeding, mortality, hospital and ICU shorter PTS, stays, and acute complications recurrent DVT and ulceration major bleeding, QOL, and symptomatic relief major bleeding, QOL, major bleeding, mortality, major bleeding, mortality, QOL, and PTS major bleeding, mortality, major bleeding, mortality, QOL, and PTS QOL, bleeding, mortality, symptomatic relief, and PTS and symptomatic relief Recurrent DVT and PE, major Recurrent DVT and PE, major QOL, PTS, and QOL, symptomatic relief, Recurrent DVT and PE, DVT and PE, major low-molecular-weight low-molecular-weight heparin, PE 5 randomized controlled trial; SVT randomized controlled trial; SVT anticoagulation Recurrent DVT and PE, 5 therapy indwelling central venous catheter or venoactive medications or venoactive medications thromboendarterectomy other anticoagulant No systemic thrombolytic thrombolytic systemic No No removal of No compression sleeves No compression sleeves No pulmonary inferior vena cava; LMWH 5 acute UEDVT Structured PICO Question quality of life; RCT quality of life; RCT —Continued 5 Patient with SVT Patient with CTPH Table 1 Table Patient with therapy central venous catheter or venoactive medications or venoactive medications thromboendarterectomy Parenteral anticoagulation Parenteral No Removal of indwelling Longer duration Shorter durationCompression sleeves Recurrent DVT and PE, major AnticoagulationAnticoagulation No anticoagulation No anticoagulation bowel ischemia, Mortality, liver failure, PE, Mortality, Patient with thrombosis in unusual sites international normalized ratio; IVC 5 postthrombotic syndrome, QOL postthrombotic syndrome, QOL 5 vitamin K antagonist. 5 indwelling central venous catheter indwelling central venous catheter the arm vein thrombosis vein thrombosis Patients with UEDVTPatients with UEDVT and Systemic thrombolytic Patients with UEDVT and Patients with PTS of Patients with splanchnic Patients with hepatic Patients with CTPHPatients with CTPH Oral anticoagulationPatients with SVT Pulmonary No oral anticoagulation Recurrent DVT and PE, Anticoagulation No anticoagulation or unfractionated heparin, VKA unfractionated heparin, VKA 5 chronic thromboembolic pulmonary hypertension; INR population, intervention, comparator, outcome; PTS population, outcome; intervention, PTS comparator, 5 5 vein thrombosis (10.1, 10.2) (9.2.1, 9.2.2) venous catheter should be removed (9.3.1) anticoagulation (9.3.2-9.3.5) (9.5.1, 9.5.2) vein thrombosis (11.1, 11.2) CTPH (7.1.1) thromboendarterectomy in CTPH (7.1.2) (8.1.1, 8.1.2) Role of anticoagulation in splanchnic Acute anticoagulation (9.1.1, 9.1.2) Patients with UEDVT Role of thrombolytic therapy Whether indwelling central Duration of long-term Prevention of PTS the arm (9.4) Patients with UEDVT of PTS the arm Treatment Compression sleeves Role of anticoagulation in hepatic Issue (Informal Question)Role of oral anticoagulation in Population Intervention Comparators Outcome Methodology Role of pulmonary Role of anticoagulation in SVT CTPH CTPH PICO PICO extremity DVT; UFH UFH extremity DVT; e430S Antithrombotic Therapy for VTE

PE after an initial PE than after an initial DVT (ie, about 60% and preferences for the most part were obtained from ratings after a PE vs 20% after a DVT) 7,9,10 ; this difference may justify that all panelists for these guidelines provided in response to more aggressive, or more prolonged, long-term therapy. Third, standardized descriptions of different outcomes and treatments, the long-term sequelae of PE are cardiorespiratory impairment, supplemented with the fi ndings of a systematic review of the especially due to pulmonary hypertension, rather than PTS of the literature on this topic.15 However, we also took into account that legs or arms. These differences are most important for recommen- values and preferences vary markedly among individual patients dations about the use of thrombus removal procedures (eg, thrombo- and that often there is appreciable uncertainty about the average lytic therapy) in patients who present with DVT and PE. patient values we used. On average, we assumed that patients attach equal value (or 1.2 Outcomes Assessed dislike [disutility]) to nonfatal thromboembolic and major bleeding events. Concern that the panelist rating exercise that attached a The outcomes important to patients we considered for most rec- similar disutility to vitamin K antagonist (VKA) therapy (frequent ommendations are recurrent VTE, major bleeding, and all-cause blood testing and telephone or clinic visits, attention to changes mortality. These outcomes are categorized in two different ways in other medications) and long-term low-molecular-weight- in the evidence profi les. Whenever data were available, fatal epi- heparin (LMWH) therapy (daily subcutaneous [SC] injection, sodes of recurrent VTE and bleeding were included in the mor- injection site bruising or nodules) may have been misguided led tality outcome, and nonfatal episodes of recurrent VTE and us to request a review of this issue at the fi nal meeting of all panelists. bleeding were reported separately in their own categories to avoid Our judgment that, on average, patients would prefer VKA therapy reporting an outcome more than once in an evidence profi le. to long-term LMWH therapy was confi rmed at that meeting. However, many original reports and published meta-analyses did not report fatal and nonfatal events separately. In this situa- 1.4 Infl uence of Bleeding Risk and Cost tion, we have reported the outcome categories of mortality, recurrent VTE, and major bleeding, with fatal episodes of VTE and Usually, we did not assess how an individual patient’s risk of bleeding included in both mortality and two specifi c outcomes bleeding would infl uence each recommendation because (1) we (ie, fatal episodes of VTE and bleeding are included in two out- considered that most recommendations would be unlikely to comes of the evidence profi le). change based on differences in risk of bleeding (eg, anticoagula- With both ways of reporting outcomes, we tried to specifi cally tion vs no anticoagulation for acute VTE, comparison of anticoag- identify deaths from recurrent VTEs and major bleeds. As part of ulant regimens), (2) there are few data assessing outcomes in the assessment of the benefi ts and harms of a therapy, we gener- patients with different risks of bleeding, and (3) there is a lack of of recurrent episodes of VTE are fatal11,12 well-validated tools for stratifying risk of bleeding in patients with %5 ف ally assume that of major bleeds are fatal,12-14 and if we deviated VTE. However, for a small number of the recommendations in %10 ف and that from these estimates, we noted the reasons for so doing. We did which the risk of bleeding is very infl uential (eg, use of extended- not consider surrogate outcomes (eg, vein patency) when there duration anticoagulation), we stratifi ed recommendations based were adequate data addressing the corresponding outcome of on this risk ( Table 2 ). Unless otherwise stated, the cost (eg, to the importance to patients (eg, PTS). patient, a third-party payer, or society) associated with different When developing evidence profi les, we tried to obtain the treatments did not infl uence our recommendations. In most situ- baseline risk of outcomes (eg, risk of recurrent VTE or major ations of uncertain benefi t of a treatment, particularly if it was bleeding) from observational studies because these estimates are potentially harmful, we took the position of primum non nocere most likely to refl ect real-life incidence. In many cases, however, (fi rst do no harm) and made a weak recommendation against the we used data from randomized trials because observational data treatment. were lacking or were of low quality. Methodologic issues specifi c to duration of anticoagulation are addressed in the section 3.1 under the subsection on general consideration in weighing the 2.0 Treatment of Acute DVT benefi ts and risks of different durations of anticoagulant therapy. 2.1 Initial Anticoagulation of Acute DVT of the Leg

1.3 Patient Values and Preferences T h e ﬁ rst and only randomized trial that compared anticoagulant therapy with no anticoagulant therapy In developing our recommendations, we took into account average patient values for each outcome and preferences for in patients with symptomatic DVT or PE was pub- different types of antithrombotic therapy. As described in lished in 1960 by Barritt and Jordan.50 Trial results MacLean et al15 and Guyatt et al16 in these guidelines, these values suggested that 1.5 days of heparin and 14 days of www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e431S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 2—[Section 2.3, 3] Risk Factors for Bleeding With Anticoagulant Therapy and Estimated Risk of Major Bleeding in Low-, Moderate-, and High-Risk Categories

Risk Factorsa

Age . 65 y17-25 Age . 75 y17-21,23,25-34 Previous bleeding18,24,25,30,33-36 Cancer20,24,30,37 Metastatic cancer36.38, Renal failure18,24,25,28,30,33 Liver failure19,21,27,28 Thrombocytopenia27,36 Previous stroke18,25,27,39 Diabetes18,19,28,32,34 Anemia18,21,27,30,34 Antiplatelet therapy19,27,28,34,40 Poor anticoagulant control22,28,35 Comorbidity and reduced functional capacity24,28,36 Recent surgery21,41,b Frequent falls27 Alcohol abuse24,25,27,34 Estimated Absolute Risk of Major Bleeding, % Categorization of Risk of Bleedingc Low Riskd (0 Risk Factors) Moderate Riskd (1 Risk Factor) High Riskd (Ն 2 Risk Factors) Anticoagulation 0-3 moe Baseline risk (%) 0.6 1.2 4.8 Increased risk (%) 1.0 2.0 8.0 Total risk (%) 1.6e 3.2 12.8f Anticoagulation after fi rst 3 mog Baseline risk (%/y) 0.3h 0.6 Ն 2.5 Increased risk (%/y) 0.5 1.0 Ն 4.0 Total risk (%/y) 0.8i 1.6i Ն 6.5 See Table 1 legend for expansion of abbreviations. aThe increase in bleeding associated with a risk factor will vary with (1) severity of the risk factor (eg, location and extent of metastatic disease, platelet count), (2) temporal relationships (eg, interval from surgery or a previous bleeding episode),29 and (3) how effectively a previous cause of bleeding was corrected (eg, upper-GI bleeding). bImportant for parenteral anticoagulation (eg, fi rst 10 d) but less important for long-term or extended anticoagulation. cAlthough there is evidence that risk of bleeding increases with the prevalence of risk factors,20,21,25,27,30,33,34,36,42,43 this categorization scheme has not been validated. Furthermore, a single risk factor, when severe, will result in a high risk of bleeding (eg, major surgery within the past 2 d, severe thrombocytopenia). dCompared with low-risk patients, moderate-risk patients are assumed to have a twofold risk and high-risk patients an eightfold risk of major bleed ing.18,20,21,27,28,30,36,44 eThe 1.6% corresponds to the average of major bleeding with initial UFH or LMWH therapy followed by VKA therapy (Table S6 Evidence Profi le: LMWH vs IV UFH for initial anticoagulation of acute VTE). We estimated baseline risk by assuming a 2.6 relative risk of major bleeding with anticoagulation (footnote g in this table). fConsistent with frequency of major bleeding observed by Hull et al41 in high-risk patients. gWe estimate that anticoagulation is associated with a 2.6-fold increase in major bleeding based on comparison of extended anticoagulation with no extended anticoagulation (Table S27 Evidence Profi le: extended anticoagulation vs no extended anticoagulation for different groups of patients with VTE and without cancer). The relative risk of major bleeding during the fi rst 3 mo of therapy may be greater that during extended VKA therapy because (1) the intensity of anticoagulation with initial parenteral therapy may be greater than with VKA therapy; (2) anticoagulant control will be less stable during the fi rst 3 mo; and (3) predispositions to anticoagulant-induced bleeding may be uncovered during the fi rst 3 mo of therapy.22,30,35 However, studies of patients with acute coronary syndromes do not suggest a Ն 2.6 relative risk of major bleeding with parenteral anticoagulation (eg, UFH or LMWH) compared with control.45,46 hOur estimated baseline risk of major bleeding for low-risk patients (and adjusted up for moderate- and high-risk groups as per footnote d in this table). iConsistent with frequency of major bleeding during prospective studies of extended anticoagulation for VTE22,44,47,48,49 (and Table S27 Evidence Profi le: extended anticoagulation vs no extended anticoagulation for different groups of patients with VTE and without cancer and Table S24).

VKA therapy markedly reduced recurrent PE (0/16 “S” before the number denote supplementary tables vs 10/19) and appeared to reduce mortality (1/16 not contained in the body of the article and avail- vs 5/19) in patients with acute PE. In the early 1990s, able instead in an online data supplement. See the a single randomized trial established the need for an “Acknowledgments” for more information.) The need initial course of heparin in addition to VKA as com- for an initial course of heparin is also supported by pared with starting treatment with VKA therapy the observation that there are high rates of recur- alone51 ( Table 3 , Table S1). (Tables that contain an rent VTE during 3 months of follow-up in patients e432S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 3—[Section 2.1] Summary of Findings: Parenteral Anticoagulation vs No Parenteral Anticoagulation in Acute VTEa, 51

Anticipated absolute effects

No. of Participants Quality of the Relative Effect Risk With No Parenteral Risk Difference With Parenteral Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Anticoagulation Anticoagulation (95% CI)

Mortality 120 (1 study), 6 mo Moderateb,c due to RR 0.5 33 per 1,000 16 fewer per 1,000 (from imprecision (0.05-5.37) 31 fewer to 144 more) VTE symptomatic 120 (1 study), 6 mo Moderateb,d due to RR 0.33 200 per 1,000 134 fewer per 1,000 (from extension or imprecision (0.11-0.98) 4 fewer to 178 fewer) recurrence Major bleeding 120 (1 study), 6 mo Moderateb,c due to RR 0.67 50 per 1,000 16 fewer per 1,000 (from imprecision (0.12-3.85) 44 fewer to 142 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. GRADE 5 Grades of Recommendations, Assessment, Development, and Evaluation; RR 5 risk ratio. aBoth groups treated with acenocoumarol. bStudy described as double blinded; outcome adjudicators blinded. None of the study participants were lost to follow-up. Intention-to-treat analysis. Study was stopped early for benefi t. cCI includes values suggesting no effect as well as values suggesting either appreciable benefi t or appreciable harm. dLow number of events caused by the early stoppage of the trial. with acute VTE treated with suboptimal heparin account that starting anticoagulant therapy in patients therapy.1,3,52,53 We discuss whether isolated distal (calf) who ultimately have DVT excluded is costly and is a DVT should be sought and if isolated distal DVT is burden to patients and the health-care system. Poor diagnosed, whether and how it should be treated in cardiopulmonary reserve may also encourage the use section 2.3. of anticoagulant therapy while awaiting diagnostic testing. If clinicians choose to administer anticoagu- Recommendation lant therapy and diagnostic testing will be completed within 12 h, we suggest using a 12-h over a 24-h dose 2.1. In patients with acute DVT of the leg treated of LMWH. VKA therapy usually should not be started with VKA therapy, we recommend initial treat- before VTE has been confi rmed. ment with parenteral anticoagulation (LMWH, fondaparinux, IV unfractionated heparin [UFH], Recommendations or SC UFH) over no such initial treatment (Grade 1B). 2.2.1. In patients with a high clinical suspicion of acute VTE, we suggest treatment with paren- 2.2 Whether to Treat With Parenteral teral anticoagulants compared with no treat- Anticoagulation While Awaiting the Results ment while awaiting the results of diagnostic of Diagnostic Work-up for VTE tests (Grade 2C) . W e i d e n t i fi ed no trial addressing this question. 2.2.2. In patients with an intermediate clinical The decision regarding treatment while awaiting test suspicion of acute VTE, we suggest treatment with results requires balancing (1) minimizing thrombotic parenteral anticoagulants compared with no treat- complications in patients with VTE and (2) avoiding ment if the results of diagnostic tests are expected bleeding in those without VTE. Our recommenda- to be delayed for more than 4 h (Grade 2C). tions are based on two principles. First, the higher the clinical suspicion for VTE (use of validated predic- 2.2.3. In patients with a low clinical suspicion of tion models for probability of having DVT54 or PE 55,56 acute VTE, we suggest not treating with paren- can usefully inform this assessment,57 the shorter the teral anticoagulants while awaiting the results acceptable interval without treatment until results of diagnostic tests, provided test results are of diagnostic testing become available. Second, the expected within 24 h (Grade 2C). higher the risk of bleeding, the longer the acceptable 2.3 Whether and How to Prescribe Anticoagulants interval without treatment until results are available. to Patients With Isolated Distal DVT Our recommendations assume that patients do not have major risk factors for bleeding, such as Whether to Look for Isolated Distal DVT and When recent surgery. The recommendations also take into to Prescribe Anticoagulants if Distal DVT Is Found: www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e433S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Whether patients with isolated distal DVT (DVT of the the risk for thrombus extension (the greater the calf [peroneal, posterior tibial, anterior tibial veins] risk, the stronger the indication for anticoagula- without involvement of the popliteal or more proximal tion), we suggest either (1) anticoagulation or (2) veins) are identifi ed depends on how suspected DVT withholding of anticoagulation while performing sur- is investigated.57 If all patients with suspected DVT veillance ultrasound examinations to detect thrombus have ultrasound examination of the calf veins (whole- extension. We consider the following to be risk fac- leg ultrasound), isolated distal DVT accounts for tors for extension: positive D -dimer, thrombosis that about one-half of all DVT diagnosed.58 If a diagnostic is extensive or close to the proximal veins (eg, . 5 cm approach is used that does not include ultrasound in length, involves multiple veins, . 7 mm in max- examination of the calf veins or that only performs imum diameter), no reversible provoking factor for ultrasound examination of the calf veins in selected DVT, active cancer, history of VTE, and inpatient patients, isolated distal DVT is rarely diagnosed.59 status.7,60,63,64 T h r o m b o s i s t h a t i s c o n fi ned to the mus- The primary goal of diagnostic testing for DVT is to cular veins has a lower risk of extension than true iso- identify patients who will benefi t from anticoagulant lated distal DVT.63,65 We anticipate that isolated distal therapy. This does not mean that all symptomatic DVT DVT detected using a selective approach to whole- need to be identifi ed. Isolated distal DVT do not need leg ultrasound often will satisfy criteria for initial anti- to be sought and treated provided that (1) there is coagulation, whereas distal DVT detected by routine strong evidence that the patient does not have a distal whole-leg ultrasound often will not. A high risk for DVT that will extend into the proximal veins (ie, the bleeding ( Table 2 ) favors ultrasound surveillance over patient is unlikely to have a distal DVT, and if a distal initial anticoagulation, and the decision to use sur- DVT is present, it is unlikely to extend); (2) if this crite- veillance or initial anticoagulation is expected to be rion is not satisfi ed, a follow-up proximal ultrasound sensitive to patient preferences. The evidence sup- is done after 1 week to detect distal DVT that has porting recommendations to prescribe anticoagu- extended into the proximal veins, in which case antico- lants for isolated calf DVT is low quality because it agulant therapy is started; and (3) the patient does not is not based on direct comparisons of the two man- have severe symptoms that would require anticoagu- agement strategies, and the ability to predict exten- lant therapy if the symptoms were due to a distal DVT. sion of distal DVT is limited. Diagnostic approaches to suspected DVT that do not examine the calf veins (eg, use of a combination How to Treat With Anticoagulants: A single con- of clinical assessment, D-dimer testing, single and trolled trial of 51 patients with symptomatic isolated serial proximal vein ultrasound examination to man- distal DVT, all of whom were initially treated with age patients) or only examine the calf veins in selected heparin, found that 3 months of VKA therapy pre- patients (eg, those who cannot have DVT excluded vented DVT extension and recurrent VTE (29% vs 0%, using the previously noted tests) have been proven P , .01). 66 The evidence in support of parenteral anti- safe and are presented in Bates et al 57 i n t h e s e g u i d e - coagulation and VKA ther apy for isolated distal DVT, lines. If the calf veins are imaged (usually with ultra- which includes indirect evidence from patients with sound) and isolated distal DVT is diagnosed, there acute proximal DVT and PE that is presented else- are two management options: (1) treat patients with where in this article, is of moderate quality (there is anticoagulant therapy or (2) do not treat patients with high-quality evidence that anticoagulation is effec- anticoagulant therapy unless extension of the DVT tive, but uncertainty that benefi ts outweigh risks). is detected on a follow-up ultrasound examination There have not been evaluations of alternatives to (eg, after 1 and 2 weeks or sooner if there is concern full-dose anticoagulation of symptomatic isolated dis- [there is no widely accepted protocol for surveil- tal DVT, and it is possible that less-aggressive anti- lance ultrasound testing]). 60 Natural history studies coagulant strategies may be adequate. Duration of of symptom- anticoagulation for isolated distal DVT is discussed in %15 ف ,suggest that when left untreated atic distal DVT will extend into the proximal veins section 3.1. and that if extension does not occur within 2 weeks, Recommendations it is unlikely to occur subsequently.7,60-62 The risk of extension of isolated distal DVT will vary among 2.3.1. In patients with acute isolated distal DVT patients (see later discussion). of the leg and without severe symptoms or risk A s n o t e d i n B a t e s e t a l ,57 these guidelines favor factors for extension (see text), we suggest serial diagnostic approaches to suspected DVT other imaging of the deep veins for 2 weeks over ini- than routine whole-leg ultrasound. If isolated dis- tial anticoagulation (Grade 2C) . tal DVT is diagnosed, depending on the severity of patient symptoms (the more severe the symptoms, 2.3.2. In patients with acute isolated distal DVT the stronger the indication for anticoagulation) and of the leg and severe symptoms or risk factors e434S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians for extension (see text), we suggest initial anti- 2.5 Choice of Initial Anticoagulant Regimen coagulation over serial imaging of the deep in Patients With Proximal DVT veins (Grade 2C). Initial anticoagulant regimens vary according to Remarks: Patients at high risk for bleeding are more the drug, the route of administration, and whether likely to benefi t from serial imaging. Patients who dose is adjusted in response to laboratory tests of place a high value on avoiding the inconvenience of coagulation. Six options are available for the initial repeat imaging and a low value on the inconve- treatment of DVT: (1) SC LMWH without moni- nience of treatment and on the potential for bleeding toring, (2) IV UFH with monitoring, (3) SC UFH given are likely to choose initial anticoagulation over serial based on weight initially, with monitoring, (4) SC UFH imaging. given based on weight initially, without monitoring, (5) SC fondaparinux given without monitoring, and Recommendations (6) rivaroxaban given orally. We considered the SC UFH options as a single category because results were 2.3.3. In patients with acute isolated distal DVT similar in studies that used SC UFH with and with- of the leg who are managed with initial anti- out laboratory monitoring ( Table 5 , Tables S3-S5). coagulation, we recommend using the same Rivaroxaban is used in the acute treatment of VTE approach as for patients with acute proximal without initial parenteral therapy; studies of its use DVT (Grade 1B). for the acute treatment of VTE are reviewed under long-term treatment of DVT (section 3.1) and PE 2.3.4. In patients with acute isolated distal (section 6) of this article. Recommendations for dosing DVT of the leg who are managed with serial and monitoring of IV UFH, SC UFH, and SC LMWH imaging, we recommend no anticoagulation are addressed in Garcia et al1 a n d H o l b r o o k e t a l3 if the thrombus does not extend ( G r a d e 1 B ) ; in these guidelines. Because LMWH, fondaparinux, we suggest anticoagulation if the thrombus and rivaroxaban have substantial renal excretion, these extends but remains confi ned to the distal veins agents should be avoided (eg, use UFH instead) or (Grade 2C); we recommend anticoagulation if should be used with coagulation monitoring (test the thrombus extends into the proximal veins selection is specifi c to each agent and requires expert (Grade 1B). interpretation) in patients with marked renal impairment (eg, estimated creatinine clearance , 30 mL/min 2.4 Timing of Initiation of VKA and Associated [in a 70-year-old weighing 70 kg, a creatinine clear- Duration of Parenteral Anticoagulant Therapy ance of 30 mL/min corresponds to a serum creati- years ago, initiation of VKA therapy was nine of about 200 m mol/L (2.3 mg/dL) in a man and 20 ف U n t i l delayed until patients had received about 5 days of 175 m mol/L (2.0 mg/dL) in a woman] http://www. heparin therapy, which resulted in patients remain- nephron.com/cgi-bin/CGSIdefault.cgi). days 10 ف ing in the hospital until they had received of heparin. Three randomized trials41,67,68 provided LMWH Compared With IV UFH for the Initial moderate-quality evidence that early initiation of Treatment of DVT: A n u m b e r o f m e t a - a n a l y s e s 72-75 .days, have summarized the trials addressing this question 5 ف VKA, with shortening of heparin therapy to is as effective as delayed initiation of VKA with about The evidence suggests that LMWH is associated with a 10-day course of heparin ( Table 4 , Table S2). Short- decreased mortality, lower recurrence of VTE, and ening the duration of initial heparin therapy from decreased incidence of major bleeding compared about 10 to 5 days is expected to have the added advan- with IV UFH ( Table 6 , Table S6). However, the qual- tage of reducing the risk of heparin-induced throm- ity of supporting evidence is low due to a high risk of bocytopenia.69 If the international normalized ratio bias in the primary studies, and evidence of publica- (INR) exceeds the therapeutic range (ie, INR . 3.0) tion bias in favor of LMWH. LMWH has the advan- prematurely, it is acceptable to stop parenteral therapy tage over IV UFH that it is much easier to administer before the patient has received 5 days of treatment. (which makes outpatient treatment feasible) and that it has a lower potential for heparin-induced Recommendation thrombocytopenia,69 but the disadvantage is that it accumulates in patients with renal failure. 2.4. In patients with acute DVT of the leg, we recommend early initiation of VKA (eg, same day SC UFH Compared With LMWH for the Initial as parenteral therapy is started) over delayed Treatment of DVT: Four randomized trials have initiation, and continuation of parenteral antico- compared SC UFH with SC LMWH ( Table 6 , agulation for a minimum of 5 days and until the Tables S3-S5).70,71,76,77 This evidence suggests that INR is 2.0 or above for at least 24 h (Grade 1B). SC UFH is associated with a similar frequency of www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e435S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 4—[Recommendation 2.4] Summary of Findings: Early Warfarin (and Shorter Duration Heparin) vs Delayed Warfarin (and Longer Duration Heparin) for Acute VTEa-d,41,67,68

Anticipated Absolute Effects

Quality of the Relative Risk With Delayed Warfarin Risk Difference With Early Warfarin No. of Participants Evidence Effect (95% Initiation (and Longer Initiation (and Shorter Duration Outcomes (Studies), Follow-up (GRADE) CI) Duration Heparin) Heparin) (95% CI)

Mortality 688 (3 studies), 3 moe Moderatef,g due to RR 0.9 24 per 1,000h 2 fewer per 1,000 (from 14 fewer imprecision (0.41-1.95) to 23 more) Recurrent VTE 688 (3 studies), 3 moe Moderatef,g due to RR 0.83 47 per 1,000h 8 fewer per 1,000 (from 28 fewer imprecision (0.4-1.74) to 35 more) Major bleeding 688 (3 studies), 3 moi Highf,j,k RR 1.48 16 per 1,000h 14 more per 1,000 (from 9 fewer (0.68-3.23) to 66 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aMost patients had proximal DVT, some had isolated distal DVT, most DVT were symptomatic (asymptomatic DVT included in Hull et al41), and few had PE (only included in Gallus et al67). bThe early initiation of VKA was associated with a fewer number of days of heparin therapy (4.1 vs 9.5 in Gallus et al67; 5 vs 10 in Hull et al41) and a fewer number of days of hospital stay (9.1 vs 13.0 in Gallus et al; 11.7 vs 14.7 in Hull et al; 11.9 vs 16.0 in Leroyer et al68). cVKA therapy started within 1 day of starting heparin therapy (UFH in two studies and LMWH in one study). dVKA therapy delayed for 4 to 10 d. eOutcome assessment was at hospital discharge in the study by Gallus et al67 (although there was also extended follow-up) and 3 mo in the studies by Hull et al41 and Leroyer et al.68 fPatients and investigators were not blinded in two studies (Gallus et al67 and Leroyer et al68) and were blinded in one study (Hull et al41). Concealment was not clearly described but was probable in the three studies. Primary outcome appears to have been assessed after a shorter duration of follow- up in the shorter treatment arm of one study because of earlier discharge from hospital, and 20% of subjects in this study were excluded from the fi nal analysis postrandomization (Gallus et al). gThe 95% CI on relative effect includes both clinically important benefi t and clinically important harm. hEvent rate corresponds to the median event rate in the included studies. iBleeding was assessed early (in hospital or in the fi rst 10 d) in two studies (Gallus et al67 and Hull et al41) and at 3 mo in one study (Leroyer et al68). jIt is unclear whether bleeding was assessed at 10 d in all subjects or just while heparin was being administered, which could yield a biased estimate in favor of short-duration therapy in one study (Hull et al41). kBecause the shorter duration of heparin therapy is very unlikely to increase bleeding, the wide 95% CIs around the relative effect of shorter therapy on risk of bleeding is not a major concern. mortality, recurrent VTE, and major bleeding as equivalence of fondaparinux to LMWH for the treat- LMWH. However, the quality of the evidence is ment of acute VTE. mod erate because of imprecision. LMWH has the disadvantage of a higher cost but is more conve- Fondaparinux Compared With IV UFH for the Ini- nient to use (LMWH can be administered once daily tial Treatment of DVT: In the absence of direct evi- [see later discussion]), is more widely available for dence in patients with DVT, indirect evidence in use in outpatients, has a lower potential for heparin- patients with acute PE (section 5.4) suggests that induced thrombocytopenia,69 and there is much fondaparinux is equivalent to IV UFH.79 As noted more experience with its use than with SC UFH. previously, we judge that fondaparinux and LMWH are equivalent; fondaparinux also shares the advan- Fondaparinux Compared With LMWH for the tages that LMWH has over IV UFH and the disad- Initial Treatment of DVT: T h e M a t i s s e - D V T t r i a l78 vantage that it is renally excreted (section 2.5). The compared fondaparinux with LMWH for short-term quality of the evidence regarding the comparison of treatment of DVT ( Table 7 , Table S7). This study sug- fondaparinux and UFH is moderate as, although gests that fondaparinux is associated with a similar there is some indirectness, it is minor. frequency of mortality, recurrent VTE, and major bleeding as LMWH. However, the quality of the Fondaparinux Compared With SC UFH for the evidence from this study was moderate because of Initial Treatment of DVT: There is no direct evidence imprecision. Evidence that fondaparinux is effective for this comparison in any patient population. Our for the treatment of PE79 (section 5.4) supports the recommendation is based on our assessment that e436S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 5—[Section 2.5.1] Summary of Findings: LMWH vs SC UFH for Initial Anticoagulation of Acute VTE70,71,76,77

Anticipated Absolute Effects

No. of Participants (Studies), Quality of the Evidence Relative Effect Risk Difference With Outcomes Follow-up (GRADE) (95% CI) Risk With SC UFH LMWH (95% CI)

All-cause 1,566 (3 studies), 3 mo Moderatea,b due to RR 1.1 (0.68-1.76) 33 per 1,000c 3 more per 1,000 (from mortality imprecision 11 fewer to 25 more) Recurrent VTE 1,563 (3 studies), 3 mo Moderatea,b due to RR 0.87 (0.52-1.45) 42 per 1,000c 5 fewer per 1,000 (from imprecision 20 fewer to 19 more) Major bleeding 1,634 (4 studies), 3 mo Moderatea,b due to RR 1.27 (0.56-2.9) 16 per 1,000c 4 more per 1,000 (from imprecision 7 fewer to 30 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. SC 5 subcutaneous. See Table 1 and 3 legends for expansion of abbreviations. aIn the two largest trials (Prandoni et al,70 Kearon et al;71 87% of patients), allocation was concealed, outcome adjudicators and data analysts were concealed, analysis was intention to treat, and there were no losses to follow-up. bPrecision judged from the perspective of whether SC heparin is noninferior to LMWH. The total number of events and the total number of participants were relatively low. cEvent rate corresponds to the median event rate in the included studies. fondaparinux and LMWH are equivalent and that fi ve of these studies83-87 that had unconfounded com- fondaparinux shares the advantages that LMWH has parisons. This evidence suggests that LMWH once over SC UFH (section 2.5). This recommendation daily and twice daily are associated with similar mor- does not take into account difference in purchase tality, recurrent VTE, and major bleeding. However, cost between SC UFH and fondaparinux and is based the quality of the evidence is low because of impre- on low-quality evidence. cision and inconsistency. The sixth study that used a lower total daily dose of LMWH with once-daily Once- vs Twice-Daily Administration of LMWH compared with twice-daily administration (enox- for Initial Treatment of DVT: Two meta-analyses80,81 aparin 1.5 mg/kg once daily vs 1.0 mg/kg bid; enox- summarized six studies comparing once-daily and aparin 2 mg/kg once daily is not used) suggested twice-daily administrations of the same LMWH.82-87 that outcomes might be inferior with this once-daily T a b l e 8 a n d T a b l e S 8 s u m m a r i z e t h e fi ndings of regimen.85

Table 6—[Section 2.5.1] Summary of Findings: LMWH vs IV UFH for Initial Anticoagulation of Acute VTE74

Anticipated Absolute Effects

No. of Participants (Studies), Quality of the Evidence Relative Effect Risk Difference With Outcomes Follow-up (GRADE) (95% CI) Risk With IV UFH LMWH (95% CI)

All-cause 7,908 (17 studies), 3 mo Lowa,b due to risk of bias, RR 0.79 (0.66-0.95) 46 per 1,000c 10 fewer per 1,000 (from mortality publication bias 2 fewer to 16 fewer) Recurrent VTE 7,976 (17 studies), 3 mo Lowa,b due to risk of bias, RR 0.72 (0.58-0.89) 55 per 1,000c 15 fewer per ,1000 (from publication bias 6 fewer to 23 fewer) Major bleeding 6,910 (20 studies), 3 mo Lowa,b,d due to risk of bias, RR 0.67 (0.45-1) 15 per 1,000c 5 fewer per 1,000 (from publication bias 8 fewer to 0 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aOf the 20 trials, allocation was concealed in nine and was unclear whether concealed in the remaining 11. In 18 trials, outcome assessors were blinded. Seven trials did not have any postrandomization exclusions or losses to follow-up. Ten trials reported the number of participants lost to follow-up, which ranged from 1.0% to 12.7%. One trial did not report the drop-outs. bInverted funnel plot very suggestive of publication bias. Many of the included studies are of small size, and all were funded by industry. cEvent rate corresponds to the median event rate in the included studies. dCI includes values suggesting signifi cant benefi t and no effect. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e437S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 7—[Section 2.5.1] Summary of Findings: Fondaparinux vs LMWH for Initial Anticoagulation of Acute DVTa-c,78

Anticipated Absolute Effects

No. of Participants (Studies), Quality of the Evidence Relative Effect Risk Difference With Outcomes Follow-up (GRADE) (95% CI) Risk With LMWH Fondaparinux (95% CI)

Mortality 2,205 (1 study), 3 mo Moderated,e due to RR 1.25 (0.8-1.97) 30 per 1,000 7 more per 1,000 (from imprecision 6 fewer to 29 more) Recurrent VTE 2,205 (1 study), 3 mo Moderated,e due to RR 0.96 (0.64-1.45) 41 per 1,000f 2 fewer per 1,000 (from imprecision 15 fewer to 18 more) Major bleeding 2,205 (1 study), 3 mo Moderated,e due to RR 0.93 (0.43-2.03) 12 per 1,000g 1 fewer per 1,000 (from imprecision 7 fewer to 12 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aAll patients had acute symptomatic DVT. bFondaparinux 7.5 mg (5.0 mg in patients weighing , 50 kg and 10.0 mg in patients weighing . 100 kg) SC once daily for at least 5 d and until VKAs induced an INR . 2.0. cEnoxaparin 1 mg/kg of body weight SC bid for at least 5 d and until VKAs induced an INR . 2.0. dAllocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Analysis excluded 0.6% of randomized patients. Not stopped early for benefi t. eCI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. fFive fatal VTE in fondaparinux group and fi ve fatal VTE in LMWH group. g Twelve patients in the fondaparinux group and 13 in the LMWH group had a major bleeding event during the initial period (7 d). Of these, two in the fondaparinux group and none in the LMWH group were fatal.

Recommendations Remarks: Local considerations such as cost, avail- 2.5.1. In patients with acute DVT of the leg, we ability, and familiarity of use dictate the choice suggest LMWH or fondaparinux over IV UFH between fondaparinux and LMWH. LMWH and (Grade 2C) and over SC UFH (Grade 2B for LMWH; fondaparinux are retained in patients with renal Grade 2C for fondaparinux). impairment, whereas this is not a concern with UFH.

Table 8—[Section 2.5.2] Summary of Findings: LMWH Once vs Twice Daily for Initial Anticoagulation of Acute VTEa,b,81

Anticipated Absolute Effects

No. of Participants Quality of the Evidence Relative Effect Risk With Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) Twice Daily LMWH Once Daily (95% CI)

Mortality 1,261 (3 studies), 3 mo Lowc-e due to inconsistency RR 1.05 (0.57-1.94) 31 per 1,000 2 more per 1,000 (from and imprecision 13 fewer to 29 more) VTE recurrence 1,261 (3 studies), 3 mo Lowc,e,f due to inconsistency RR 0.86 (0.52-1.42) 49 per 1,000 7 fewer per 1,000 (from and imprecision 24 fewer to 21 more) Major bleeding 1,522 (5 studies), 10 d Moderatec,e due to RR 1.13 (0.48-2.66) 12 per 1,000 2 more per 1,000 (from imprecision 6 fewer to 20 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aOf the fi ve included studies, one included patients with PE and DVT and four included only patients with DVT. All studies addressed the initial management of VTE. bThe fi ve included studies used four brands of LMWH (enoxaparin, tinzaparin, dalteparin, and nadroparin). In Merli et al,85 enoxaparin 1 mg/kg bid was compared with 1.5 mg/kg once daily. Holmström et al84 adjusted the dose to anti-Xa levels, which resulted in different daily doses after a number of days. In the remaining studies, the dose of the once-daily administration was double the dose of the twice-daily administration (equal total daily dose). cAll included studies concealed allocation. Two studies had a double-blind design, and two others were single blind. One study did not mention blinding. Intention to treat likely used in all studies. Participants were lost to follow-up in only two studies (0.3% and 2.2%). dI2 5 37%; point effect estimate in favor of twice-daily dose in Merli et al85 and in favor of once-daily dose in Charbonnier et al.83 eImprecision judged relative to no difference. fI2 5 65%; point effect estimate in favor of twice-daily dose in Merli et al85 and in favor of once-daily dose in Charbonnier.83 e438S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 2.5.2. In patients with acute DVT of the leg conditional on the patient feeling well enough to be treated with LMWH, we suggest once- over treated at home (eg, does not have severe leg symp- twice-daily administration (Grade 2C). toms or comorbidity).

Remarks: This recommendation only applies when 2.8 Treatment Strategies of Thrombus the approved once-daily regimen uses the same daily Removal for Acute DVT dose as the twice-daily regimen (ie, the once-daily injection contains double the dose of each twice- Treatments that actively remove thrombus in patients daily injection). It also places value on avoiding an with acute DVT have the potential to reduce acute extra injection per day. symptoms and the risk of developing PTS. Patients with DVT that involves the iliac and common femoral 2.6 Initial Treatment With Rivaroxaban veins are at highest risk for PTS and, therefore, are vs Parenteral Therapy the subset with the greatest potential to benefi t from 102 One trial directly compared short- and long-term thrombus removal strategies. Thrombus removal rivaroxaban (without initial parenteral anticoagula- strategies are indicated in patients with the very rare tion) with parenteral anticoagulation (LMWH) and complication of impending venous gangrene despite VKA in patients with acute DVT.88 The fi ndings of optimal anticoagulant therapy; such patients are not this study and associated recommendations are pre- the focus of the following sections. A recent trial that sented in section 3.3. randomized 183 patients with proximal DVT to percutaneous endovascular intervention or to anticoagu- 2.7 At-Home vs In-Hospital Initial Treatment of DVT lant therapy alone reported reduced acute symptoms, hospital stay, recurrent VTE, and PTS at 6 months in One trial of 201 patients directly compared outpa- the thrombus removal group. 103 This trial, which had a tient and inpatient administration of the same initial high potential for bias (randomization not described, anticoagulant regimen (three LMWH preparations were no blinding), is not considered further because it was used); there were few recurrent VTE and major bleeds not possible to determine outcomes in patients treated in each group.89 A n u m b e r o f t r i a l s90-94 have compared with mechanical thrombectomy alone and in those LMWH administered at home (without hospital admis- treated with thrombolytic therapy. sion or after early discharge) in a substantial proportion of patients, with IV UFH administered in the hospital 2.9 Catheter-Directed Thrombolysis for Acute DVT ( Table 9 , Table S9). This evidence suggests that home treatment is not associated with an increase in mortality, The rationale for catheter-directed thrombolysis recurrent VTE, or major bleeding and may be asso- (CDT) is that compared with systemic thromboly- ciated with improved outcomes. However, the quality sis, it will achieve lysis of thrombus more rapidly and of the evidence is moderate because of indirectness with lower doses of thrombolytic therapy, thereby (patients were not explicitly randomized to home reducing serious bleeding. The addition of mechan- therapy in most studies) and imprecision. ical thrombus fragmentation (collectively referred to Health economic evaluations that have assessed as pharmacomechanical thrombolysis) with or with- initial treatment of DVT at home, although they have out aspiration can further reduce the dose of throm- weaknesses (eg, industry funded, not derived from bolytic therapy and shorten the procedure.104 trials in which LMWH was used both in the hospital One randomized trial of CDT has been com- and at home, short time horizon (ie, Յ 3 months), and pleted,105 and a second has reported short-term limited use of sensitivity analyses), all conclude that outcomes (but not the development of PTS).106,107 home treatment is cost-saving (about US $500-$2,500 Table 10 and Table S10 present the combined fi nd- per patient).95-101 ings from these studies (see also Tables S11 and S12). This evidence suggests that CDT may reduce PTS Recommendation and improve quality of life without being associated 2.7. In patients with acute DVT of the leg and with an unacceptable increase in bleeding. However, whose home circumstances are adequate, we rec- the quality of evidence is low for mortality, recur- ommend initial treatment at home over treat- rent VTE, and major bleeding because of very serious ment in hospital (Grade 1B). imprecision, and is low for PTS because of indirectness (ie, use of surrogate outcome [PTS has yet to be Remarks: The recommendation is conditional on the measured directly during follow-up]). adequacy of the following home circumstances: well- In addition to the two randomized trials,105,106 fi ndings maintained living conditions, strong support from of observational studies suggest that CDT improves family or friends, phone access, and ability to quickly venous patency and preserves venous valve function return to hospital if there is deterioration. It is also (Tables S11 and S12). Use of CDT, however, requires www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e439S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 9—[Section 2.7] Summary of Findings: Home Treatment vs Hospital Treatment of Acute DVTa-d,442

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With Hospital Risk Difference With Outcomes (Studies), Follow-up Evidence(GRADE) (95% CI) Treatment Home Treatment (95% CI)

Mortality 1,708 (6 studies), Lowc-f due to indirectness and RR 0.72 (0.45-1.15) 46 per 1,000 13 fewer per 1,000 (from 3 mo imprecision 25 fewer to 7 more) Recurrent VTE 1,708 (6 studies), Moderatec-e due to indirectness RR 0.61 (0.42-0.9) 74 per 1,000 29 fewer per 1,000 (from 3 mo 7 fewer to 43 fewer) Major bleeding 1,708 (6 studies), Moderatec-e,g due to indirectness RR 0.67 (0.33-1.36) 21 per 1,000 7 fewer per 1,000 (from 3 mo 14 fewer to 8 more) QOL 0 (3 studiesh), Lowi-k due to indirectness and Not estimable … h 3 mo imprecision The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aStudies included in the systematic review should have recruited patients whose home circumstances were adequate. bAll studies included patients with lower-extremity DVT and excluded patients with suspected or confi rmed PE. Studies also excluded patients who were pregnant. cFour studies had partial hospital treatment of many in the home arm: Koopman et al92 (mean hospital stay, 2.7 in home arm vs 8.1 d in hospital arm), Levine et al93 (2.1 vs 6.5 d), Boccalon et al89 (1 vs 9.6 d), and Ramacciotti et al94 (3 vs 7 d). In Daskalopoulos et al,91 there was no hospital stay at all in the home group. Chong et al90 did not report duration of hospital stay. dOnly one study (Boccalon et al89) used LMWH in both treatment arms. Remaining studies used UFH in the inpatient arm and LMWH in the outpatient arm. eOut of six studies, allocation was clearly concealed in three (unclear in remaining three), outcome adjudicators were blinded in the two largest studies (unclear in remaining four), loss to follow-up was signifi cant in only one small study, intention-to-treat analysis was conducted in four (unclear in remaining two), and no study was stopped early for benefi t. Overall, the judgment was that these limitations would not warrant downgrading of quality; it has already been downgraded by at least one level based on other factors. fThe CI includes values suggesting benefi t and harm. gJudged as precise based on the narrow CI around absolute effect. hBäckman et al95 reported evaluation of health-related QOL using the EQ-5D. They found no differences in mean QOL scores or in the proportion of patients showing improvement in self-rated health state. Koopman et al92 evaluated health-related QOL using the Medical Outcome Study Short Form-20 and an adapted version of the Rotterdam Symptom Checklist. The changes over time were similar in both groups, except that the patients receiving LMWH had better scores for physical activity (P 5 .002) and social functioning (P 5 .001) at the end of the initial treatment. The authors did not report enough data to assess precision and clinical signifi cance of results. O’Brien et al96 assessed changes in QOL using the Medical Outcome Study Short Form-36 in 300 patients participating in Levine et al.93 They found that the change in scores from baseline to day 7 was not signifi cantly different between the treatment groups for seven of the eight domains. The one exception was the domain of social functioning, where a greater improvement was observed for the outpatient group. iPotential inconsistency as Bäckman et al95 showed no effect, whereas Koopman et al92 and O’Brien et al96 showed potential benefi t. jTwo of the three studies had partial hospital treatment of many in the home arm: Koopman et al92 (mean hospital stay, 2.7 in home arm vs 8.1 d in hospital arm) and Levine et al93 (2.1 vs 6.5 d). kNot able to evaluate, but imprecision is possible. Taken together with the potential inconsistency, we downgraded the quality of evidence by one level. substantial resources and expertise. Patients who are assess the benefi t or risk of inserting an IVC fi lter in most likely to benefi t from CDT have iliofemoral DVT, patients who have CDT performed (recommended symptoms for , 14 days, good functional status, life by manufacturer with some endovascular devices and expectancy of Ն 1 year, and a low risk of bleeding techniques, whereas not with others). Percutaneous ( Table 11 ). Because the balance of risks and benefi ts mechanical venous thrombectomy without concomi- with CDT is uncertain, anticoagulant therapy alone is an tant thrombolysis has not been evaluated in random- acceptable alternative to CDT in all patients with acute ized trials, and its use is discouraged because small DVT who do not have impending venous gangrene. retrospective studies suggest that it often fails to There is no single standardized approach to per- remove much of the thrombus115,116 and is associated forming CDT or pharmacomechanical thrombolysis. with a high risk of PE.117,118 If these interventions are performed, the technique used will vary with local resources and expertise. Recommendation If CDT has been successful but there are residual lesions in the common femoral or more proximal 2.9. In patients with acute proximal DVT of veins, balloon angioplasty and stenting often are used the leg, we suggest anticoagulant therapy alone to relieve obstruction. There are inadequate data to over CDT (Grade 2C) . e440S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 10—[Section 2.9] Summary of Findings: CDT vs No CDT for Extensive Acute DVT of the Lega,b,105,106

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With Risk Difference With Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) No CDT CDT (95% CI)

Mortality 153 (2 studies), 3 mo Lowc,d due to RR 0.14 (0.01-2.71) 39 per 1,000e 34 fewer per 1,000 (from imprecision 39 fewer to 67 more) Nonfatal recurrent VTE 153 (1 study), 3 mo Lowc,d due to RR 0.35 (0-8.09) 48 per 1,000f 31 fewer per 1,000 (from imprecision 48 fewer to 340 more) Nonfatal major bleeding 153 (2 studies), 7 d Lowc,d due to RR 2.00 (0.19-19.46) 29 per 1,0006,7 29 more per 1,000 (from imprecision 23 fewer to 535 more) PTS 138 (2 studies), 2 y Moderatec,g due RR 0.46 (0-0.79) 588 per 1,000h 318 fewer per 1,000 (from (complete lysis on to indirectness 123 fewer to 588 fewer)i venography [Elsharawy et al73]; patency on ultrasound and air plethysmography [Enden et al74]) QOL 98 (1 studyj), 16 mo Lowk,l See footnotem (SF-12, HUI Mark version 2/3 questionnaires) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. HUI 5 Health Utilities Index; SF-12 5 Medical Outcomes Survey Short Form-12; VETO 5 Venous Thrombosis Outcomes. See Table 1 and 3 legends for expansion of other abbreviations. aIn selected patients with extensive acute proximal DVT (eg, iliofemoral DVT, symptoms for , 14 d, good functional status, life expectancy Ն 1 y) who have a low risk of bleeding. bAll patients prescribed anticoagulants per protocol, but the intervention group receives CDT in addition to anticoagulation. cAllocation was concealed in Enden et al106 but unclear in Elsharawy et al.105 Outcome assessor blinded in both studies. Follow-up rates were 87% in Enden et al and 100% in Elsharawy et al. Neither of the studies was stopped early for benefi t. dCI includes values suggesting both benefi t and harm. eThree control patients died of cancer. fBaseline risks for nonfatal recurrent VTE and for major bleeding derived from Douketis et al.108 gSurrogate outcome: absence of patency at 6 mo in Enden et al106 study; absence of complete lysis at 6 mo in Elsharawy et al105 study. hThis estimate is based on the fi ndings of the VETO study.102 This probably underestimates PTS baseline risk given that overall, 52% of patients reported the current use of compression stockings during study follow-up. iSevere PTS: assuming the same RR of 0.46 and a baseline risk of 13.8%,102 the absolute reduction is 75 fewer severe PTS per 1,000 (from 29 fewer to 138 fewer) over 2 y. jCamerota et al.109 kParticipation rate was 65%. lRecall was used to measure QOL prior to the thrombotic event; we did not consider these measurements. mAt the initial follow-up (mean, 16 mo), patients treated with CDT reported a trend toward a higher mental summary scale (P 5 .087) and improved HUI (P 5 .078). They reported better overall role physical functioning (P 5 .046), less stigma (P 5 .033), less health distress (P 5 .022), and fewer overall symptoms (P 5 .006) compared with patients who were treated with anticoagulation alone.

Remarks: Patients who are most likely to beneﬁ t bleeding, and PTS (Table 12, Table S15). This evi- from CDT (see text) and attach a high value to dence suggests that systemic thrombolysis has the prevention of PTS and a lower value to the initial potential to reduce PTS at the expense of an increase complexity, cost, and risk of bleeding with CDT in major bleeding. However, the overall quality of are likely to choose CDT over anticoagulation alone. this evidence is low because of imprecision and risk of bias. There have been no direct comparisons of different 2.10 Systemic Thrombolytic Therapy for Acute DVT thrombolytic agents; however, prolonged infusions of Many trials of systemic thrombolysis for the treat- streptokinase that were used predominantly in the ment of DVT assessed early lysis, often reported earlier studies appear to be associated with higher bleeding, but rarely reported recurrent VTE or bleeding rates than other regimens. No random- development of PTS (Table S13 and S14). 119-137 ized trial has compared systemic thrombolysis with A meta-analysis138 summarized the ﬁ ndings of trials CDT, but a single-center, retrospective study127 sug- that assessed mortality, recurrent VTE, major gested that systemic thrombolysis achieves less www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e441S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 11—[Section 2.9, 2.10, 5.6, 9.2] Risk Factors systemic thrombolysis are expected to be associ- for Bleeding With and Contraindications to Use of ated with a higher risk of nonprocedure-related Thrombolytic Therapy (Both Systemic and Locally Administered) bleeding. We believe that systemic thrombolysis should Major contraindicationsa be considered only in patients who meet all of Structural intracranial disease the following criteria: iliofemoral DVT, symptoms Previous intracranial hemorrhage for , 14 days, good functional status, life expec- Ischemic stroke within 3 mo tancy of Ն 1 year, and low risk of bleeding ( Table 11 ). Active bleeding Based on low-quality evidence of greater effective- Recent brain or spinal surgery ness and less bleeding, if resources and expertise Recent head trauma with fracture or brain injury Bleeding diathesis are available to perform CDT, we consider it the Relative contraindicationsb preferable approach. Because the balance of risks Systolic BP . 180 mm Hg and benefi ts with systemic thrombolysis is uncertain, Diastolic BP . 110 mm Hg and particularly because of concerns about major Recent bleeding (nonintracranial) bleeding, anticoagulant therapy alone is an accept- Recent surgery Recent invasive procedure able alternative to systemic thrombolysis in all patients Ischemic stroke more that 3 mo previously with acute DVT who do not have impending venous Anticoagulation (eg, VKA therapy) gangrene. Traumatic cardiopulmonary resuscitation Pericarditis or pericardial fl uid Recommendation Diabetic retinopathy Pregnancy 2.10. In patients with acute proximal DVT of Age . 75 y the leg, we suggest anticoagulant therapy alone Low body weight (eg, , 60 kg) Female sex over systemic thrombolysis (Grade 2C). Black race Remarks: Patients who are most likely to benefi t from Among 32,000 Medicare patients (Ն 65 y) with myocardial infarc- systemic thrombolytic therapy (see text), who do not tion who were treated with thrombolytic therapy, the following factors were independently associated with intracranial hemorrhage: have access to CDT, and who attach a high value age Ն 75 y (OR, 1.6), black race (OR, 1.6), female sex (OR, 1.4), to prevention of PTS and a lower value to the initial previous stroke (OR, 1.5), systolic BP Ն 160 mm Hg (OR, 1.8), complexity, cost, and risk of bleeding with systemic women weighing Յ 65 kg or men weighing Յ 80 kg (OR, 1.5), and thrombolytic therapy are likely to choose systemic 110 INR . 4 (OR, 2.2). The rate of intracranial hemorrhage increased thrombolytic therapy over anticoagulation alone. from 0.7% with none or one of these risk factors to 4.1% with fi ve or more of these risk factors. Among 32,000 patients with myocardial infarction who were treated with thrombolytic therapy in fi ve clini- 2.11 Operative Venous Thrombectomy cal trials, the following factors were independently associated with for Acute DVT moderate or severe bleeding: older age (OR, 1.04 per year), black race (OR, 1.4), female sex (OR, 1.5), hypertension (OR, 1.2), and lower Operative venous thrombectomy, with contempo- weight (OR, 0.99/kg).111 We estimated that systemic thrombolytic rary operative techniques140 and more effective anti- therapy is associated with a relative risk of major bleeding of 3.5 for intracranial bleeding); about three-fourths coagulant regimens, appears to achieve improved 7ف ,within 35 d (RR 141,142 of the excess of major bleeds with thrombolytic therapy occur in the outcomes compared with earlier reports. A single fi rst 24 h.112 See Table 1 legend for expansion of abbreviations. small randomized trial with extended follow-up com- aThe presence of major contraindications usually precludes use of pared iliofemoral venous thrombectomy with a tem- thrombolytic therapy and, consequently, these factors have not been porary arteriovenous fi stula plus anticoagulation with well studied as risk factors for bleeding associated with thrombolytic 143-145 therapy. The factors listed in this table are consistent with other anticoagulation alone. Results at 6 months, 5 years, recommendations for the use of thrombolytic therapy in patients with and 10 years suggested improved iliac vein patency, PE.72,259,260,454 104,111,113,114 less leg swelling, and fewer leg ulcers with throm- bRisk factors for bleeding during anticoagulant therapy noted in bectomy (Table 13, Tables S16-S18).143-145 Evidence Table 10 that are not included in this table are also likely to be from this trial is of low quality because of impreci- relative contraindications to thrombolytic therapy. The increase in bleeding associated with a risk factor will vary with (1) severity of the sion and risk of bias. risk factor (eg, extent of trauma or recent surgery) and (2) temporal We believe that operative venous thrombec- relationships (eg, interval from surgery or a previous bleeding episode tomy should be considered only if all of the -wk). Risk factors for bleeding following criteria are met: iliofemoral DVT, symp 2ف believed to decrease markedly after at critical sites (eg, intracranial, intraocular) or noncompressible sites toms for , 7 days (criterion used in the single are stronger contraindications for thrombolytic therapy. randomized trial), good functional status, life expec- lysis (31% vs 50%) and less preservation of valve tancy of Ն 1 year, and both resources and expertise function (13% vs 44%) (Tables S13 and S14), and are available. Based on low-quality evidence of the higher doses of thrombolytic agent used with greater effectiveness and less bleeding, we consider e442S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 12—[Section 2.10] Summary of Findings: Systemic Lysis vs No Systemic Lysis for Extensive Acute DVT of the Leg138

Anticipated Absolute Effects

No. of Participants Quality of the Evidence Relative Effect Risk With No Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) Systemic Lysis Systemic Lysis (95% CI)

Mortality 688 (5 studies), 3 moa Lowb-e due to imprecision RR 0.86 (0.27-2.68) 21 per 1,000 3 fewer per 1,000 (from 16 fewer to 36 more) Nonfatal 687 (3 studies), 3 mof Lowd,e,g due to imprecision RR 1.28 (0.25-6.68) 48 per 1,000h 13 more per 1,000 (from recurrent VTE 36 fewer to 273 more) Nonfatal major 688 (10 studies), 3 mof Moderatec,d,i due to RR 1.84 (0.94-3.59) 29 per 1,000h,j 24 more per 1,000 (from bleeding imprecision 2 fewer to 75 more) PTS 678 (2 studies), 2 yk Lowd,e,l,m due to risk of bias RR 0.71 (0.49-1.04) 588 per 1,000n 171 fewer per 1,000 (from and imprecision 300 fewer to 24 more)o QOL not …………… measured The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1, 3, and 10 legends for expansion of abbreviations. aRange of follow-up in included studies, 1 to 72 mo. bAllocation was concealed in three of fi ve studies. Follow-up inadequate in one of fi ve studies (Common et al139). Excluding this study from the analysis does not change the effect estimate. All studies had blinded outcome assessors. None of the studies used a placebo control. cThe population of one study (Schulman et al131) comprised patients with calf vein thrombosis. dInterventions varied across studies with regard to agent (eg, tissue plasminogen activator, streptokinase, urokinase), dose, use of the pedal vein administration, duration of treatment, and concomitant drugs (eg, steroids). However, we did not downgrade for indirectness given that there was no standard regimen and all analyses showed no heterogeneity in results. eCI included both no effect and a potentially signifi cant effect. fRange of follow-up in included studies, 1 to 30 d. gAllocation was concealed in two of three studies. Follow-up adequate in all studies. All studies had blinded outcome assessors. None of the studies used a placebo control. hBaseline risks for nonfatal recurrent VTE and for major bleeding derived from Douketis et al.108 iAllocation was concealed in seven of 10 studies. Follow-up inadequate in one of 10 studies (Common et al139). Excluding this study from the analysis does not affect the effect estimate. All studies had blinded outcome assessors. Two studies used placebo (Turpie et al135 and Verhaeghe et al136). jOnly 4% of all major bleeding events were intracranial bleeds. kRange of follow-up in included studies, 1 to 6 y. lAllocation was concealed in two of two studies. Follow-up adequate in both studies. Both studies had blinded outcome assessors. Neither study used placebo control. mNo use of a standardized validated tool reported. nThis estimate is based on the fi ndings of the VETO study.102 This probably underestimates PTS baseline risk, given that overall, 52% of patients reported the current use of compression stockings during study follow-up. oSevere PTS: Assuming the same RR of 0.71 and a baseline risk of 13.8%,102 the absolute reduction is 40 fewer severe PTS per 1,000 (from 70 fewer to 6 more) over 2 y.

CDT preferable to the operative venous thrombec- method of thrombus removal (including systemic tomy approach. thrombolysis). Mechanical components of these procedures are associated with a high early risk of Recommendation early recurrent thrombosis, and thrombus removal is not known to alter the long-term risk of recurrent 2.11. In patients with acute proximal DVT of the VTE. We used evidence from patients with DVT leg, we suggest anticoagulant therapy alone over who did not have thrombus removal to guide anti- operative venous thrombectomy ( G r a d e 2 C ). coagulant decisions in those who had thrombus removal. This evidence is rated down to moderate 2.12 Anticoagulation in Patients Who Have Had quality because of its indirectness in this patient Any Method of Thrombus Removal Performed population. There are no randomized trials or observational Recommendation studies that have compared different anticoagulant regimens or durations of therapy in patients with 2.12. In patients with acute DVT of the leg who acute proximal DVT of the leg who have had any undergo thrombosis removal, we recommend www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e443S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 13—[Section 2.11] Summary of Findings: Surgical Thrombectomy vs No Surgical Thrombectomy for Extensive Acute DVT of the Lega,144

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With No Surgical Risk Difference With Surgical Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Thrombectomy Thrombectomy (95% CI) Mortality not ………… … reported Nonfatal 51 (1 study), 3 mo Lowb,c due to risk of bias RR 0.37 48 per 1,000d,e 30 fewer per 1,000 (from 47 fewer recurrent and imprecision (0.02-8.75) to 372 more) VTE Nonfatal major 51 (1 study), 3 mo Lowb,c due to risk of bias Not estimable …- bleeding and imprecision (no events) PTS 51 (1 study), 2 y Lowf,g due to risk of bias RR 0.63 588 per 1,000i 218 fewer per 1,000 (from 59 and imprecision (0.44-0.9)h fewer to 329 fewer)j QOL not ………… … measured The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1, 3 and 10 legends for expansion of abbreviations. aThe study included patients with DVT with symptoms of leg swelling not exceeding 7 d and a proximal extension of the thrombus above the inguinal ligament but not into the vena cava. bNot clear whether allocation was concealed. No blinding reported. Not clear whether analysis was intention to treat. Follow-up rate was 88% at 6 mo. Study not stopped early for benefi t. cCI includes values suggesting either harm or benefi t. dBaseline risks for nonfatal recurrent VTE derived from Douketis et al.108 eOne event was a symptomatic PE. fIn addition to other study limitations, this outcome was assessed by those who did the surgery and anticoagulation. No standardized tool was used. One surgical patient had an amputation secondary to venous gangrene and was not counted in the PTS assessment. gFew number of events. This warrants rating down the quality of evidence by a second level when considered along with study limitations. hThe RR is based on the 6-mo data. iThis estimate is based on the fi ndings of the VETO study.102 This probably underestimates PTS baseline risk, given that overall, 52% of patients reported the current use of compression stockings during study follow-up. jSevere PTS: assuming the same RR of 0.63 and a baseline risk of 13.8% over 2 y,102 the absolute reduction is 51 fewer severe PTS per 1,000 (from 14 fewer to 77 fewer) over 2 y. the same intensity and duration of anticoagu- bleeding), we assume that the relative risk of out- lant therapy as in similar patients who do not comes will be the same as in patients who received undergo thrombosis removal (Grade 1B). anticoagulant therapy in the Prevention du Risque d’Embolie Pulmonaire par Interruption Cave (PREPIC) 2.13 Vena Caval Filters for the Initial study. However, their absolute rate of symptomatic Treatment of DVT PE and recurrent DVT will be higher compared with No randomized trial or prospective observational the PREPIC participants who were prescribed antico- study has evaluated IVC fi lters as sole therapy (ie, agulants. A comprehensive review of mostly retrospec- without concurrent anticoagulation) in patients with tive case series of IVC fi lter insertions (6,500 patients DVT. A single, large, randomized controlled trial in 89 reports) suggested that venous thrombosis at of patients %10 ف evaluated permanent IVC fi lter insertion as an the site of fi lter insertion occurs in adjunct to anticoagulant therapy in patients with and that fi lters can be placed above the renal veins and acute DVT who were considered to be at high risk in the superior vena cava if necessary.150 A prospec- for PE ( Table 14 , Table S19). The fi ndings at 2 years 146 tive observational study also suggested that symp- and 8 years 149 of follow-up, suggest that IVC fi lters tomatic VTE and asymptomatic fi lter thrombosis are increase the risk of recurrent DVT, reduce the risk common,151 and a systematic review suggested that of PE, do not alter the combined frequency of DVT the prevalence of PTS may be increased152 in patients and PE (ie, recurrent VTE), do not increase the risk with permanent IVC fi lters. A small single-center of PTS, and do not alter mortality. randomized trial suggested a higher complication In assessing the role of an IVC fi lter in patients rate with the Trapease compared with the Green- who cannot receive anticoagulant therapy (eg, actively fi eld permanent fi lter.153 e444S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 14—[Section 2.13] Summary of Findings: Vena Cava Filter vs No Vena Cava Filter for Acute Proximal DVT of the Leg Treated With Anticoagulationa,b,146,443

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With No Risk Difference With Vena Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) vena cava Filters cava Filters (95% CI)

Mortality 400 (1 study), 8 y Moderatec,d due to RR 0.95 515 per 1,000 26 fewer per 1,000 imprecision (0.78-1.16)e (from 113 fewer to 82 more) Symptomatic PE 304 (1 study), 8 y Moderatec,f due to RR 0.41 151 per 1,000 89 fewer per 1,000 imprecision (0.2-0.86)g (from 21 fewer to 121 fewer) Recurrent DVT 310 (1 study), 8 y Moderatec,f due to RR 1.3 273 per 1,000 82 more per 1,000 imprecision (0.93-1.82)h (from 19 fewer to 224 more) Major bleeding 337 (1 study), 8 y Moderatec,d due to RR 0.83 185 per 1,000 31 fewer per 1,000 imprecision (0.52-1.34)i (from 89 fewer to 63 more) PTS 308 (1 study), 8 y Lowd,j due to risk of RR 0.87 699 per 1,000 91 fewer per 1,000 bias and imprecision (0.66-1.13) (from 238 fewer to 91 more) Complications 379 (1 study), 2 y Moderatef due to …… k imprecision QOL not reported … … … … … The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. PREPIC 5 Prevention du Risque d’Embolie Pulmonaire par Interruption Cave. See Table 1 and 3 legends for expansion of other abbreviations. aAnticoagulation consisted of LMWH or UFH initially (according to a 2 3 2 factorial design) followed by oral anticoagulation for at least 3 mo. bFour types of permanent vena cava fi lters were used: Vena Tech LGM (B. Braun Melsugen AG), titanium Greenfi eld (Boston Scientifi c Corporation), Cardial (C.R. Bard, Inc), and Bird’s Nest (Cook Group Incorporated). cAllocation was concealed. Data collectors and outcome adjudicators were blinded. Intention-to-treat analysis. Data missing for 4% at 2 y and 1% at 8 y. Enrollment was stopped at 400 instead of targeted 800 because of slow recruitment. dCI includes both negligible effect and appreciable benefi t or appreciable harm. eRR, 1.0 (95% CI, 0.29-3.4) at 12 d; RR, 1.08 (95% CI, 0.73-1.58) at 2 y. fSmall number of events. gRR. 0.23 (95% CI, 0.05-1.05) at 12 d (both symptomatic and asymptomatic PE). RR, 0.54 (0.21-1.41) at 2 y (symptomatic PE). hRR, 1.78 (95% CI, 1.09-2.94) at 2 y. iRR, 1.5 (95% CI, 0.54-4.14) at 12 d. RR, 0.74 (95% CI, 0.41-1.36) at 2 y. jNo standardized validated tool used to measure PTS. kNo complications directly related to the fi lter or its insertion reported in the PREPIC trial.146 Mismetti et al147 (prospective study) reported an incidence of 3.2% (excluding fi lter tilting and puncture site hematoma) among 220 patients receiving a retrievable vena cava fi lter for secondary prevention of VTE, whereas Athanasoulis et al148 (retrospective study) reported an incidence of 0.3% for major complications among 1,731 patients receiving vena cava fi lters predominantly for secondary prevention of VTE.

If an IVC fi lter is indicated in a patient with be treated for the same length of time as if the same acute DVT or PE because anticoagulant therapy is patient had not had an IVC fi lter inserted (see section temporarily contraindicated (eg, active bleeding), 3.1). The duration of anticoagulation, therefore, will there is the option of inserting a retrievable fi lter vary according to whether the DVT was provoked by and removing it when it is safe to start anticoagu- a temporary risk factor, was unprovoked, or was asso- lant therapy. However, most retrievable fi lters are ciated with cancer, and may be infl uenced by the not removed; retrievable fi lters that are not removed patient’s ongoing risk of bleeding and preferences. may have a higher long-term complication rate than Our recommendation to treat patients with an IVC permanent fi lters, and there currently is no good fi lter with anticoagulants when contraindications to evidence that retrievable IVC fi lters improve patient anticoagulation resolve is weaker than for anticoagu- outcomes.104,147,154,155 lation of most patients with VTE because the risks of Insertion of an IVC fi lter does not eliminate the bleeding may remain elevated, and the patient’s risk risk of PE and increases the risk of DVT ( Table 14 , of recurrence is expected to be lower if the acute Table S19). Consequently, we suggest that patients episode of thrombosis occurred remotely. The evidence who have an IVC fi lter inserted should receive a con- for IVC fi lter use in patients with acute proximal DVT ventional course of anticoagulation (eg, parenteral who cannot be treated with anticoagulation is mod- and long-term anticoagulation) if the contraindica- erate because of serious imprecision and indirectness tion to anticoagulation resolves. Such patients should (ie, extrapolated from the PREPIC study in which www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e445S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians patients were routinely treated with anticoagulants; 2.14 Early Ambulation of Patients With Acute DVT this indirectness, however, is minor). Treatment of acute DVT with bed rest and antico- Recommendations agulation (originally IV UFH) has given way to early mobilization with anticoagulation (often administered 2.13.1. In patients with acute DVT of the leg, we SC). Two meta-analyses156,157 summarized evidence recommend against the use of an IVC fi lter in from four relevant trials ( Table 15 , Tables S20-S22). addition to anticoagulants (Grade 1B). This evidence is of low quality because of risk of bias and imprecision. We suggest early ambulation (eg, with- 2.13.2. In patients with acute proximal DVT of the out a period of bed rest) when feasible because of its leg and contraindication to anticoagulation, we potential to decrease PTS and improve quality of life. recommend the use of an IVC fi lter (Grade 1B). Recommendation 2.13.3. In patients with acute proximal DVT of the leg and an IVC fi lter inserted as an alterna- 2.14. In patients with acute DVT of the leg, we tive to anticoagulation, we suggest a conventional suggest early ambulation over initial bed rest course of anticoagulant therapy if their risk of (Grade 2C) . bleeding resolves ( G r a d e 2 B ). Remarks: If edema and pain are severe, ambulation Remarks: We do not consider that a permanent may need to be deferred. As per section 4.1, we rec- IVC fi lter, of itself, is an indication for extended ommend the use of compression therapy in these anticoagulation. patients.

Table 15—[Section 2.14] Summary of Findings: Early Ambulation vs Delayed Ambulation for Acute DVT of the Lega,b,273,309,310,314,315

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With Delayed Risk Difference With Early Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Ambulation Ambulation (95% CI)

Mortality 385 (4 studies), 3 moc Lowd,e due to risk of RR 1.3 11 per 1,000 3 more per 1,000 bias, imprecision (0.23-7.55) (from 8 fewer to 70 more) PE (symptomatic or 385 (4 studies), 4-12 d Lowd-g due to risk of RR 1.16 118 per 1,000 19 more per 1,000 asymptomatic) bias, imprecision (0.66-2.05) (from 40 fewer to 124 more) QOL questionnaire in 53 (1 study), 2 y Lowh,i due to risk of See footnotej chronic limb venous bias, indirectness insuffi ciency (CIVIQ) PTS Villata-Prandoni 37 (1 study), 2 y Lowe,h due to risk of RR 0.66 400 per 1,000 136 fewer per 1,000 scores (value, . 5) bias, imprecision (0.42-1.03) (from 232 fewer to 12 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. CIVIQ 5 ChronIc Venous Insuffi ciency Questionnaire. See Table 1 and 3 legends for expansion of other abbreviations. aTwo of four eligible studies excluded patients with symptomatic PE; in the third study, 24% of participants had symptomatic PE at baseline. It was not clear whether the fourth study excluded patients with symptomatic PE. bIn two of four eligible trials, all patients received early compression therapy (bandages or stockings). In the two other trials, only patients randomized to early ambulation received early compression therapy. cThree studies reporting acute-phase mortality reported no deaths. dConcealment of allocation was reported in one of four studies; blinding of outcome assessors was reported in two of four studies; intention-to-treat analysis reported in two of four studies. Follow-up was 97% to 100%. In two of four trials, only patients randomized to early ambulation received early compression therapy (bandages or stockings). In the other two trials, all patients received early compression therapy. eCI includes both values of clinically signifi cant benefi t and values of clinically signifi cant harms. fPE assessed as both symptomatic and asymptomatic PE. gFunnel plot reported as not asymmetrical by Aissaoui et al.156 hConcealment of allocation was not reported; outcome assessors were not blinded for this outcome. Seventy percent follow-up rate; compression stockings used on patients with early mobilization but not in patients with delayed mobilization. iNo explanation was provided. jPsychologic and overall somatic QOL did not differ signifi cantly between the treatment groups, whereas DVT-related items, especially those refl ecting the ease of locomotion, showed signifi cantly greater improvement with compression than with bed rest (P , .001 for bandages, P , .05 for stockings). e446S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 3.0 Long-term Anticoagulation continued until (1) the reduction of recurrent VTE no of Acute DVT of the Leg longer clearly outweighs the increase in bleeding or (2) it is patient preference (which may be infl uenced by In this review, the term long-term treatment r e f e r s fi nancial burden) to stop treatment, even if the reduc- to treatments (eg, VKA therapy, LMWH, dabigatran) tion in VTE would outweigh the increase in bleeding. that are continued after initial therapy (eg, parenteral Increase in Risk of Recurrent VTE After Stopping anticoagulation, thrombolytic therapy) ( Fig 1 ). In Therapy— Current evidence suggests that the risk of addition, we consider treatment with rivaroxaban, recurrence after stopping therapy is largely deter- which is used without initial parenteral therapy. Long- mined by two factors: (1) whether the acute episode term therapy has two goals: (1) to complete treat- of VTE has been effectively treated (duration of ment of the acute episode of VTE and (2) to prevent therapy) and (2) the patient’s intrinsic risk of having a new episodes of VTE that are not directly related to new episode of VTE (individual risk of recurrence). the acute event. During the early phase of long-term Duration of therapy: The primary goal of trials that treatment (ie, fi rst 3 months), treatment of the acute compare different time-limited durations of antico- episode of VTE predominates. During the late phase agulation is to identify the shortest duration of therapy of long-term treatment (ie, after the fi rst 3 months), that results in a posttreatment risk of recurrence that prevention of new episodes of VTE predominates. is as low as can be achieved. The fi ndings of these We use the term extended anticoagulation to refer to trials generally are not sensitive to differences in indi- anticoagulation that is continued beyond 3 months vidual patient risk of bleeding. without a scheduled stop date. However, regular (eg, Individual risk of recurrence: Primary factors yearly) reassessments are needed to assess whether for estimating risk of recurrence: Presence of a a patient’s risk of bleeding increased or the patient’s reversible provoking risk factor,47,159,160,162-168,169,170 preferences changed. unprovoked VTE, 47,159,160,162-168,169,170 and presence of Three lines of evidence from randomized trials sup- active cancer9,47,165,166 are the most important factors port the need for long-term anticoagulant treatment that infl uence risk of recurrent VTE after stopping of DVT (ie, after 5-10 days of initial heparin therapy): VKA. Among patients with VTE provoked by a (1) a randomized controlled trial of long-term anti- reversible factor, the risk of recurrence is much lower coagulant therapy in 51 patients with symptomatic if the provoking factor was recent surgery com- calf-vein thrombosis that documented a 25% rate of pared with a nonsurgical trigger (eg, estrogen therapy, symptomatic extension of thrombosis within 3 months pregnancy, leg injury, fl ight of . 8 h).162,171 In patients 66 in the control group, ( 2 ) a r a n d o m i z e d t r i a l c o m - with proximal DVT and PE, the estimated cumula- paring long-term SC low-dose UFH (5,000 units bid) tive risk of recurrent VTE after stopping anticoagu- with VKA therapy in patients with proximal DVT lant therapy of each of these categories is as follows: that found that low-dose UFH was ineffective and VTE provoked by surgery, 1% after 1 year and 3% resulted in a high rate of recurrent VTE (47% within after 5 years; VTE provoked by a nonsurgical revers- 158 3 months), and (3) randomized trials in which ible risk factor, 5% after 1 year and 15% after 5 years; reduced durations of treatment of 4 or 6 weeks and unprovoked VTE, 10% recurrence after 1 year resulted in important increases in recurrent VTE and 30% after 5 years. compared with conventional durations of treat- There are sparse data addressing the risk of recur- 159-161 ment of 3 or 6 months (section 3.1). This evi- rent VTE after stopping therapy in patients with dence is of moderate quality because of imprecision cancer because treatment is rarely stopped in these and indirectness. patients because of a high risk for recurrence. 38,47,165,166,172 A reasonable estimate for this risk, expressed as an Recommendation annualized rate, may be 15%. However, the risk of 3.0. In patients with acute VTE who are treated recurrence is expected to vary according to whether with anticoagulant therapy, we recommend long- the cancer is metastatic, being treated with chemo- 38,165 term therapy (see section 3.1 for recommended therapy, or rapidly progressing. The high mortality duration of therapy) over stopping anticoag- in patients with VTE and cancer (40% at 6 months 173 ulant therapy after 1 week of initial therapy in one large study ) precludes estimating the cumu- (Grade 1B). lative risk of recurrence after long-term follow-up. We categorize patients with VTE according to these primary individual risk factors for recurrence when 3.1 Duration of Long-term Anticoagulant Therapy we make recommendations for duration of anticoag- Weighing the Benefi ts and Risks of Different Dura- ulant therapy. tions of Anticoagulant Therapy: General Consider- Secondary factors for estimating risk of recurrence: ations: Anticoagulant therapy for VTE should be Additional factors that infl uence the risk of recurrence www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e447S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians strongly enough to modify some recommendations VKA therapy targeted to an INR of 2.5 was the anti- about duration of therapy include (1) whether DVT coagulant regimen in all comparisons. was confi ned to the distal veins (isolated distal [or Short vs Intermediate Durations of Therapy—Five calf] DVT), which is estimated to be associated with trials have evaluated shortening the duration of oral about one-half of the risk of recurrence of proximal anticoagulant therapy from 3 or 6 months to 4 or DVT and PE,10,164,167,169,170 and (2) whether the VTE 6 weeks in patients with mostly fi rst episodes of VTE was a second or subsequent episode of VTE, which is ( Table 16 , Tables S23-S25).159,160,167,169 This evidence, estimated to be associated with about a 50% higher which is high quality, indicates that with the shorter risk of recurrence compared with a fi rst VTE.9,173,175 duration of therapy, the absolute decrease in bleeding Additional factors for estimating risk of recur- was small compared with the absolute increase in rence: Other factors predict risk of recurrence, but recurrent VTE. Patients with isolated distal DVT not strongly or consistently enough to infl uence rec- provoked by a major transient risk factor have a very ommendations on duration of therapy once the pri- low risk of recurrence after anticoagulant therapy is per year170 ). It is uncertain whether %1 ف) mary and secondary factors noted previously have stopped been considered. These factors, which have mostly this risk is lowered by treating for 3 months compared been evaluated in patients with unprovoked VTE, with 4 or 6 weeks (hazard ratio [HR] for 4 or 6 weeks include negative D -dimer testing 1 month after vs Ն 3 months at 2 years after stopping therapy, 0.36; CI, 0.09-1.54170 ). For this reason, we make a %95 181-48,176, ( 4 . 0 ف ,[withdrawal of VKA (risk ratio [RR h e r e d i - weaker recommendation for 3 months compared 185-182, ( 2 ف ,antiphospholipid antibody (RR with a shorter duration of therapy in patients with 190-182,185-162,163,174,177,180 , ( 5 . 1 ف ,tary thrombophilia (RR Asian ethnicity isolated distal DVT that was provoked by a reversible 191,192 , ( 6 . 1 ف ,male vs female sex (RR -and residual thrombosis in the prox- risk factor. The evidence supporting this weaker rec 193,(0.8 ف ,RR) Combinations ommendation is rated down to low quality because of 198-149,182,185,194 . ( 5 . 1 ف ,imal veins (RR of factors have the potential to be more important serious imprecision and because it is a post hoc predictions of recurrence risk than single factors observation. (eg, low risk of recurrence in women with unpro- Different Intermediate Durations of Therapy (6 or voked proximal DVT or PE who have a negative 12 months vs 3 months)—We considered trials that D - d i m e r t e s t b e f o r e 185 o r 1 m o n t h a f t e r 180,199 s t o p - randomized patients with VTE to 3 months vs to 6 or ping anticoagulant therapy). PTS may be a risk 12 months of treatment to determine, when using a factor for recurrent VTE, 183,185,200 and recurrent time-limited duration of therapy, whether there was ipsilateral DVT is a risk factor for development of any benefi t to treating for . 3 months. Five reports, PTS. 201,202 Both associations may contribute to a which included six randomized comparisons, con- decision to use extended therapy in a patient with tributed to this analysis (Table 17, Tables S24- established PTS. S26).167,194,203-205 These studies found that 6 or Increase in Risk of Bleeding While Remaining on 12 months of therapy did not convincingly lower risk Anticoagulant Therapy— Although the decision to of recurrence but increased major bleeding about treat patients with different time-limited durations of 2.5-fold. In a meta-analysis of individual patient data anticoagulant therapy generally are insensitive to from randomized trials, during the 2 years after stop- an individual’s risk of bleeding, the decision to use ping anticoagulant therapy, treatment of 3 months extended anticoagulation, particularly in patients compared with Ն 6 months was associated with an with an unprovoked proximal DVT or PE, is sensitive HR of 1.19 (95% CI, 0.86-1.65) in all patients and an to risk of bleeding. There is no validated prediction HR of 1.39 (95% CI, 0.96-2.01) in patients with tool to stratify the risk of major bleeding during unprovoked DVT or PE.170 T h e r e f o r e , a l t h o u g h a n t i - extended anticoagulant therapy specifi cally in patients coagulants are very effective at preventing recurrence with VTE, but this risk appears to increase with the while patients are receiving therapy, when anticoagu- prevalence of the factors noted in Table 2 . This table lants are stopped, there is a similar risk of recurrence also provides our estimate of the absolute risk of whether patients have been treated for 3 months or bleeding without anticoagulation (baseline risk), the longer.170 This evidence is of moderate quality because increase with anticoagulation, and the sum of these of serious imprecision. two risks (ie, risk of bleeding on therapy). As an alternative to comparing two time-limited durations of anticoagulant therapy, the AESOPUS Comparisons of Time-Limited Durations of (Ultrasound Findings to Adjust the Duration of Anti- Therapy: Randomized trials have compared either a coagulation) trial compared a predefi ned duration short (eg, 4 or 6 weeks) with an intermediate (eg, 3 or of therapy with a fl exible duration of therapy that 6 months) duration of therapy, or two intermediate depended on whether there was residual thrombo- durations of therapy, (eg, 3 months vs 6 or 12 months). sis during follow-up in patients with a fi rst proximal e448S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 16—[Section 3.1.1-3.1.4] Summary of Findings: Four or Six Weeks vs Three or Six Months as Minimum Duration of Anticoagulation for VTEa,b,16,159,167,169,195

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With Risk Difference With 4 or 6 wk vs 3 or 6 mo Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Control Months of Anticoagulation (95% CI)

Recurrent VTE 2,185 (5 studiesc), 1-2 yd Highe-g RR 1.83 64 per 1,000 53 more per 1,000 (1.39-2.42) (from 25 more to 91 more) Major bleeding 2,185 (5 studies), 1-2 y Highf RR 0.54 12 per 1,000 5 fewer per 1,000 (from 9 fewer to 4 more) (0.22-1.32) Mortality 2,098 (5 studies), 1-2 y Highe,f,h RR 0.97 55 per 1,000 2 fewer per 1,000 (0.68-1.38) (from 18 fewer to 21 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confidence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aPopulations varied among studies: fi rst provoked isolated distal DVT, proximal DVT, or PE provoked in Kearon et al195; fi rst isolated distal DVT in Pinede et al167; fi rst isolated distal DVT, proximal DVT, or PE in Schulman et al169; proximal DVT (21% had cancer) in Levine et al159; DVT or PE (29% not objectively confi rmed) in British Thoracic Society.160 bShort vs longer duration of anticoagulation was 6 wk vs 6 mo for Schulman et al169, 6 wk vs 3 mo for Pinede et al,167 and 4 wk vs 3 mo for the other three studies. cTiming of randomization relative to the start of treatment varied across studies: Pinede et al,167 Schulman et al,169 and British Thoracic Society160 randomized at diagnosis, and Kearon et al195 and Levine et al159 randomized to stop or to continue treatment for 2 mo more after the initial 4 wk of treatment. .y in all studied except for Schulman et al169 in which it was 2 y 1ف dFollow-up was for eGenerally, study design was strong. No study stopped early for benefi t; two stopped early because of slow recruitment (Kearon et al,195 Pinede et al167). In one study (British Thoracic Society160), 44 randomized patients were excluded centrally as they did not satisfy eligibility criteria. Patients and caregivers were blinded in two studies (Kearon et al, Levine et al159). Adjudicators of outcomes were blinded in all but one study (British Thoracic Society). All studies appear to have used effective randomization concealment, intention-to-treat analysis, and a low unexplained drop-out frequency. fNo heterogeneity with I2 5 0%. gNo imprecision for overall estimates. However, for the subgroup of patients with isolated distal DVT, who are known to have a very low risk of recurrence, there is imprecision and the possibility that the shorter duration of anticoagulation is adequate and not associated with a clinically important higher risk of recurrence. hDifferences in mortality are expected to be mediated by differences in recurrent VTE and bleeding.

DVT206 (Tables S24 and S25). Its fi ndings suggest bleeding. The quality of evidence for the reduction in that the latter approach may be helpful for tailoring recurrent VTE with extended therapy is high but is the duration of therapy. rated down to moderate for bleeding and mortality because of imprecision. Extended vs Time-Limited Anticoagulant Therapy: Weighing the Benefi ts and Risks of Extended VKA Five trials compared extended anticoagulation with Therapy— The decision to extend anticoagulation VKA therapy (target INR 2.0-2.85,161 2.0-3.0,48,182,207 therapy beyond 3 months is sensitive to both baseline and 1.5-2.0 174 ) with stopping VKA therapy at 3 or risks of recurrent VTE and major bleeding. We did 6 months in patients who were judged to have a high not identify a validated prediction tool for either risk of recurrence ( Table 18 , Table S24, S25, and S27). outcome that takes into account all relevant risk fac- The results indicate that randomization to indefi nite tors. As an alternative, for patients without cancer, treatment with conventional-intensity VKA (target we chose to stratify our recommendations according INR 2.5) reduces recurrent VTE by about 90% (RR to four primary risk groups for recurrent VTE (sec- for the four studies, 0.12; 95% CI, 0.05-0.2548,161,182,207 ), tion 3.1) and three risk groups for major bleeding and randomization to low-intensity therapy (target (section 3.1) ( Table 2 ). This approach resulted in a INR 1.75) reduces VTE by 64% (95% CI for HR, total of 12 combinations of risk profi les. Table 19 23%-81%),174 with about one-half of recurrent VTE shows the estimated total (and fatal) number of in the active treatment groups in these studies recurrent episodes of VTE prevented and the number occurring in patients who had prematurely stopped of major bleeds caused by 5 years of extended therapy VKA therapy. Extended anticoagulant therapy was for each of the 12 combinations. In the absence of associated with about a 2.6-fold increase in major robust trial data for mortality for recurrent VTE and www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e449S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 17—[Section 3.1.1-3.1.4] Summary of Findings: Six or Twelve Months vs Three Months as Minimum Duration of Anticoagulation for VTEa,b,167,203,204

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk Difference With 6 or Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Risk With 3 mo 12 mo (95% CI)

Recurrent VTE 2,061 (6 studies), 1-3 y Moderatec-e due to RR 0.89 115 per 1,000 13 fewer per 1,000 imprecision (0.69-1.14) (from 36 fewer to 16 more) Major bleeding 2,061 (6 studies), 1-3 y Highf RR 2.49 9 per 1,000 13 more per 1,000 (1.2-5.16) (from 2 more to 37 more) Mortalityg 1,331 (5 studies), 1-3 y Moderated due to RR 1.3 44 per 1,000 13 more per 1,000 (from 8 imprecision (0.81-2.08) fewer to 47 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of other abbreviations. aStudy populations varied across studies: Pinede et al167 enrolled provoked and unprovoked proximal DVT and PE; Campbell et al203 enrolled provoked and unprovoked isolated distal DVT, proximal DVT, and PE; Agnelli et al194 had separate randomizations for provoked PE (3 vs 6 mo) and unprovoked (3 vs 12 mo); and Agnelli et al204 enrolled unprovoked proximal DVT. bTiming of randomization relative to the start of treatment and length of treatment in the non-3 mo group varied across studies: Pinede et al167 and Campbell et al203 randomized at diagnosis; and Agnelli et al194,204 randomized after the initial 3 mo of treatment to stop or continued treatment. The longer duration of treatment was 6 mo in Agnelli et al194 (provoked PE) and 12 mo in Agnelli et al204 and Agnelli et al194 (unprovoked PE). cGenerally, study design was strong. No study stopped early for benefi t; two stopped early because of slow recruitment (Campbell et al,203 Pinede et al167) and one because of lack of benefi t (Agnelli et al204). In one study (Campbell et al), 20% of VTE outcomes were not objectively confi rmed. Patients and caregivers were not blinded in any study. Adjudicators of outcomes were blinded in all but one study (Campbell et al). All studies used effective randomization concealment, intention-to-treat analysis, and a low unexplained drop-out frequency. dCIs include both values suggesting no effect and values suggesting either benefi t or harm. eLow number of events and a total number of participants , 2,000. fOne study may have confi ned the assessment of bleeding to when subjects were receiving anticoagulant therapy, which could have infl ated the increase in bleeding associated with the longer duration of therapy (Campbell et al203). gDifferences in mortality are expected to be mediated by differences in recurrent VTE and bleeding. bleeding, we assumed that 3.6% of recurrent VTE others fi nd it a major burden that greatly erodes their and 11.3% of major bleeds will be fatal.12 sense of well-being.210 We make (1) a strong recommendation for extended The presence of additional risk factors for VTE therapy when it is associated with a reduction in recurrence (section 3.1), the patient’s relative value VTE that is substantially more frequent than the for the different outcome of interest (recurrence of increase in major bleeding and with a mortality VTE, major bleeding, PTS), and the patient’s per- advantage, (2) a weak recommendation for extended ceived burden of anticoagulant therapy may infl uence therapy when it is associated with a reduction in VTE decisions about the use of extended anticoagulant that is more frequent than the increase in major therapy in patient groups for which we provide a bleeding but the magnitude of this difference and weak recommendation but are unlikely to infl uence the suggested mortality advantage are more modest, this decision in patients groups for which we provide (3) a weak recommendation against extended therapy a strong recommendation.16 Similarly, the costs of when extended therapy is associated with a reduc- therapy and how those costs are paid (eg, patient, tion in VTE that is less frequent than the increase in third party) are more likely to infl uence treatment major bleeding and no mortality advantage exists, decisions when there is a weak recommendation and (4) a strong recommendation against extended (Grade 2) in favor of extended therapy. therapy when extended therapy is associated with a Patients With VTE and Cancer: As previously noted reduction in VTE that is less frequent than the (section 3.1), because they have a high risk of recur- increase in major bleeding and a mortality disadvan- rence, patients with active cancer (eg, treated within tage may exist. We assume that on average, extended the past 6 months, persistent or progressive) should anticoagulation with VKA therapy is a modest burden benefi t from extended anticoagulant therapy unless to patients.15 However, this differs markedly among they have a very high risk of bleeding. The quality patients; some do not fi nd anticoagulant therapy a of the evidence supporting this recommendation is burden and have an enhanced feeling of well-being moderate because of indirectness (the relative effects because they feel protected from recurrence, whereas of anticoagulation are based, in part, on evidence e450S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 18—[Section 3.1.1-3.1.4] Summary of Findings: Extended Anticoagulation vs No Extended Anticoagulation for Different Groups of Patients with VTE and Without Cancera,b,48,161,182,207

Anticipated Absolute Effects

Mortality 1,184 (4 studies), Moderatec-e RR 0.57 (0.31-1.03) 63 per 1,000 27 fewer per 1,000 (from 10-36 mo due to 44 fewer to 2 more) imprecision Recurrent VTE at 1,184 (4 studies), High RR 0.12 (0.09-0.38) First VTE provoked by surgeryf-j 1 y 10-36 mo 10 per 1,000 10 fewer per 1,000 (from 6 fewer to 9 fewer) First proximal DVT or PE provoked nonsurgical/ ﬁ rst unprovoked distal DVTf-j 50 per 1,000 44 fewer per 1,000 (from 31 fewer to 45 fewer) First unprovoked VTEf-j 100 per 1,000 88 fewer per 1,000 (from 62 fewer to 91 fewer) Second unprovoked VTEf-j 150 per 1,000 132 fewer per 1,000 (from 93 fewer to 137 fewer) Major bleeding 1,184 (4 studies), Moderate RR 2.63 (1.02-6.76) Lowk,l (see Table 2) at 1 y 10-36 mo due to 3 per 1,000 5 more per 1,000 (from imprecision 0 more to 15 more) Moderatek,l (see Table 2) 6 per 1,000 10 more per 1,000 (from 1 more to 29 more) Highk,l (see Table 2) 25 per 1,000 40 more per 1,000 (from 3 more to 122 more) Recurrent VTE 1,184 (4 studies), High RR 0.12 (0.09-0.38) First VTE provoked by surgeryf-j at 5 y 10-36 mo 30 per 1,000 26 fewer per 1,000 (from 19 fewer to 27 fewer) First proximal DVT or PE provoked nonsurgical/ ﬁ rst unprovoked distal DVTf-j 150 per 1,000 132 fewer per 1,000 (from 93 fewer to 137 fewer) First unprovoked VTEf-j 300 per 1,000 264 fewer per 1,000 (from 186 fewer to 273 fewer) Second unprovoked VTEf-j 450 per 1,000 396 fewer per 1,000 (from 279 fewer to 409 fewer) Major bleeding 1,184 (4 studies), Moderate RR 2.63 (1.02-6.76) Lowk,l (see Table 2) at 5 y 10-36 mo due to 15 per 1,000 24 more per 1,000 (from imprecision 2 more to 73 more) Moderatek,l (see Table 2) 30 per 1,000 48 more per 1,000 (from 4 more to 146 more) Highk,l (see Table 2) 125 per 1,000 199 more per 1,000 (from 17 more to 609 more) (Continued)

www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e451S

Anticipated Absolute Effects

Quality of Risk With No Risk Difference With No. of Participants the Evidence Relative Effect Extended Duration Extended Duration Oral Outcomes (Studies), Follow-up (GRADE) (95% CI) Oral Anticoagulation Anticoagulation (95% CI) Burden of … … … Warfarin: daily … anticoagulation medication, not reported dietary and activity restrictions, frequent blood testing/monitoring, increased hospital/ clinic visitsm PTS not reported … … … n … The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. PREVENT 5 Prevention of Recurrent Venous Thromboembolism. See Table 1 and 3 legends for expansion of other abbreviations. aStudies vary in follow-up duration (10 mo to 3 y) and in duration of time-limited VKA (3-6 mo). bWe excluded PREVENT trial because target INR was 1.75 (low intensity), which has been shown in an RCT44 to be less effective than a target of 2.5. cI2 5 0%. dCI includes both values suggesting no effect and values suggesting either appreciable harms or appreciable benefi t. eSmall number of events. Decision to rate down also takes into account that two studies were stopped early for benefi t. fAnnual risk of VTE recurrence after discontinuing oral anticoagulation therapy in patients with fi rst VTE provoked by surgery: 1% (Iorio et al171); we assumed a 0.5% yearly risk thereafter (3% over 5 y). the fi rst year (Iorio et al171); we assumed a 2.5% yearly thereafter (15% %5ف :gAnnual risk in patients with fi rst VTE provoked by nonsurgical factor over 5 y). hAnnual risk in patients with fi rst episode of unprovoked VTE: 9.3% over 1 y in Rodger et al185; 11.0% over 1 y, 19.6% over 3 y, and 29.1% over 5 y in Prandoni et al.208 We assumed a risk of 10% the fi rst year after discontinuation and 5% yearly thereafter (30% over 5 y). iAnnual risk in patients with second episode of unprovoked VTE: we assumed an RR of 1.5 compared with a fi rst episode of unprovoked VTE: 15% the fi rst year after discontinuation, 7.5% yearly thereafter (45% over 5 y). jCase fatality rate of recurrent VTE after discontinuing oral anticoagulation therapy: 3.6% (Carrier et al12). kAnnual risk of major bleeding is based on three risk levels: low, intermediate, and high. The corresponding 0.3%, 0.6%, and 1.2% risks are estimates based on control arms of included studies (see Table 2). lCase fatality rate of major bleeding during initial oral anticoagulation therapy: 11.3% (Carrier et al12) (no data available for after discontinuing oral anticoagulation therapy). mBurden of anticoagulation: endured by all patients who continue extended-duration anticoagulation (100%) and applies to patients who stop anticoagulation (no extended duration anticoagulation) who subsequently experienced a recurrent VTE (5%/10%/15% at 1 y; 15%/30%/45% at 5 y). nPTS: baseline risk over 2 y of 58.8% for PTS and 13.8% for severe PTS (Kahn et al102). There was a threefold (Prandoni et al202) to 10-fold (van Dongen et al209) increase in PTS with recurrent VTE in the ipsilateral leg. from patients without cancer). Presence of factors of selecting a patient for extended therapy if these associated with a lower risk of recurrence that may have become available, and (4) they are being treated support stopping anticoagulant therapy, particularly with the anticoagulant regimen that best suits them. if the risk of bleeding was high, include the following: LMWH for Extended Therapy: We identifi ed no (1) VTE was associated with a superimposed revers- direct evidence for LMWH compared with (1) VKA, ible risk factor (eg, recent surgery, chemotherapy), (2) other anticoagulant strategies, or (3) control for (2) the cancer has responded to treatment, (3) the the extended phase of anticoagulation in patients who cancer has not metastasized, and (4) the VTE was an were treated with a standardized initial long-term isolated distal DVT. anticoagulant regimen. Based on indirect evidence Follow-up of Patients on Extended Therapy: Patients from comparisons of LMWH with VKA therapy during who are treated with extended anticoagulant therapy the initial 3 or 6 months of long-term therapy, we should be reviewed regularly (eg, annually) to ensure judged LMWH to be at least as effective in terms of that (1) they have not developed contraindications recurrent VTE and as safe in terms of major bleeding to extended therapy, (2) their preferences have not ( Table 20 , Table S28). The potential for drug-induced changed (eg, anticoagulation has become an exces- osteoporosis, however, may be greater with extended sive burden), (3) they can benefi t from improved ways therapy LMWH than with VKA therapy.1 e452S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 19—[Section 3.1.1-3.1.4] Estimated Absolute Difference in Recurrent VTE and Major Bleeding Events (Including Fatal Events) With 5 Years of vs No Extended Anticoagulation

Risk of Bleeding Outcomes After 5 y of Treatment Low Intermediate Higha

First VTE provoked by Recurrent VTE reduction ↓ 26 (19-27) (1 fatal)b ↓ 26 (19-27) (1 fatal)b ↓ 26 (19-27) (1 fatal)b surgery per 1,000 Major bleeding increase ↑ 24 (2-73) (3 fatal)b ↑ 49 (1-173) (5 fatal)b ↑ 98 (1-346) (11 fatal)b per 1,000 First VTE provoked by a Recurrent VTE reduction ↓ 132 (93-137) (5 fatal)c ↓ 132 (93-137) (5 fatal)c ↓ 132 (93-137) (5 fatal)b nonsurgical factor/fi rst per 1,000 unprovoked distal DVT Major bleeding increase ↑ 24 (2-73) (3 fatal)c ↑ 49 (1-173) (5 fatal)c ↑ 98 (1-346) (11 fatal)b per 1,000 First unprovoked proximal Recurrent VTE reduction ↓ 264 (186-273) (10 fatal)d ↓ 264 (186-273) (10 fatal)d ↓ 264(186-273) (10 fatal)b DVT or PE per 1,000 Major bleeding increase ↑ 24 (2-73) (3 fatal)d ↑ 49 (1-173) (5 fatal)d ↑ 98 (1-346) (11 fatal)b per 1,000 second unprovoked VTE Recurrent VTE reduction ↓ 396 (279-409) (14 fatal)e ↓ 396 (279-409) (14 fatal)d ↓ 396 (279-409) (14 fatal)c per 1,000 Major bleeding increase ↑ 24 (2-73) (3 fatal)e ↑ 49 (1-173) (5 fatal)d ↑ 98 (1-346) (11 fatal)c per 1,000 Recommendations: Risk of dying in patients with a recurrent VTE or a major bleed: • Case fatality rate of recurrent VTE after discontinuing oral anticoagulation therapy: 3.6% (Carrier et al12). • Case fatality rate of major bleeding during initial oral anticoagulation therapy: 11.3% (Carrier et al12) (no data available for after discontinuing oral anticoagulation therapy). Annual risks of recurrent VTE after discontinuation of anticoagulation: • First VTE provoked by surgery: 1% (Iorio et al171); we assumed a 0.5% yearly risk thereafter (3% over 5 y). (the fi rst year (Iorio et al171); we assumed a 2.5% yearly thereafter (15% over 5 y %5ف :First episode of VTE provoked by nonsurgical factor • • First episode of unprovoked VTE: 9.3% over 1 y (Rodger et al185); 11.0% over 1 y, 19.6% over 3 y, 29.1% over 5 y (Prandoni et al208). We assumed a risk of 10% the fi rst year after discontinuation and 5% yearly thereafter (30% over 5 y). • Second episode of unprovoked VTE: we assumed that this infl icts 1.5 the risk of recurrent VTE relative to fi rst episode of unprovoked VTE: 15% the fi rst year after discontinuation, 7.5% yearly thereafter (45% over 5 y). Relative risk reduction with extended anticoagulant therapy: • 82% based on Table 18 Annual risks of major bleeding in patients not on anticoagulant therapy: • Low risk, 0.3%/y; intermediate risk 0.6%/y; high risk, 2.4%/y (Table 2). Relative risk of major bleeding with extended anticoagulant therapy: • 2.6 based on Table 18. Criteria used to decide on direction and strengths of recommendations: • Criterion for a strong recommendation against whenever the estimated number of fatal bleeding events exceeded the estimated number of fatal recurrent VTE prevented. • Criterion to go from a strong recommendation against to weak recommendation against: difference between the lower boundary of increased major bleeding and upper boundary of reduction in recurrent VTE , 2% (risk over 5 y averaged per year). • Criterion to go from a weak recommendation against to a weak recommendation in favor of: difference between point estimate of reduction of recurrent VTE and point estimate for increase in major bleeding is . 2% (risk over 5 y averaged per year) (2% to account for the burden and cost of VKA). • Criterion to go from a weak recommendation for to strong recommendation for: difference between the lower boundary of reduction in VTE and upper boundary of increased major bleeding . 4% (risk over 5 y averaged per year). Another way of interpreting the direction and strength of recommendation based on the number of deaths (related to either bleeding or recurrent VTE) is as follows: • A strong recommendation against: extended anticoagulation is estimated to be associated with an increase in deaths. • A weak recommendation against: extended anticoagulation is estimated to be associated with from no effect on deaths to only a very small reduction in deaths (0-4/1,000 prevented over 5 y or , 0.5%/patient-y). • A weak recommendation for: extended anticoagulation is estimated to be associated with a small reduction in deaths (5 to 9/1,000 prevented over 5 y or 0.5%-0.9%/patient-y). • A strong recommendation for: extended anticoagulation is estimated to be associated with a large reduction in deaths (. 10/1,000 prevented over 5 y or . 1%/patient-y). aWith an eightfold risk of bleeding in the high-risk group compared with the low-risk group, a strong recommendation against extended anticoagulation for a second unprovoked VTE is justifi ed. The high-risk group, however, includes patients who have a risk of bleeding that is less than this estimate (eg, patients aged . 75 y without additional risk factors for bleeding [Table 2]) and, therefore, may benefi t from extended anticoagulant therapy. For this reason, we provide a weak rather than a strong recommendation against extended anticoagulation for patients with a second unprovoked VTE in the high-bleeding-risk group. bStrong against cWeak against dWeak in favor eStrong in favor www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e453S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Rivaroxaban for Extended Therapy: Use of rivar- lation for at least 3 months over treatment of a oxaban compared with initial parenteral therapy shorter duration ( G r a d e 1 B ) . After 3 months of followed by VKA therapy for the short- and long- treatment, patients with unprovoked DVT of the term treatment of DVT is reviewed in section 3.3 leg should be evaluated for the risk-benefi t ( Table 21 , Table S29). In the current section, we con- ratio of extended therapy. sider rivaroxaban compared with no anticoagulation in patients with proximal DVT or PE who have com- 3.1.4.1. In patients with a fi rst VTE that is an pleted 6 or 12 months of anticoagulant therapy, unprovoked proximal DVT of the leg and who which has been evaluated in a single study ( Table 22 , have a low or moderate bleeding risk ( Table 2 ), Table S30).127 This study found that rivaroxaban we suggest extended anticoagulant therapy over markedly reduced recurrent VTE at the expense of a 3 months of therapy (Grade 2B). modest absolute increase in major bleeding. The evidence from this one study is of moderate quality 3.1.4.2. In patients with a fi rst VTE that is an because of serious imprecision. unprovoked proximal DVT of the leg and who Dabigatran for Extended Therapy: There are no have a high bleeding risk ( Table 2 ), we recom- completed studies that have compared dabigatran mend 3 months of anticoagulant therapy over with no anticoagulation for extended treatment of extended therapy (Grade 1B). VTE. Choice of Anticoagulant Regimen for Extended 3.1.4.3. In patients with a fi rst VTE that is an Therapy: This question is addressed in section 3.3. unprovoked isolated distal DVT of the leg (see remark), we suggest 3 months of anticoagulant Recommendations therapy over extended therapy in those with a low bleeding risk ( Table 2 ) ( G r a d e 2 B ), and 3.1.1. In patients with a proximal DVT of the leg recommend 3 months of anticoagulant treat- provoked by surgery, we recommend treatment ment in those with a moderate or high bleeding with anticoagulation for 3 months over: (i) treat- risk ( Table 2 ) (Grade 1B). ment of a shorter period (Grade 1B) , (ii) treatment of a longer time-limited period (eg, 6 or 3.1.4.4. In patients with a second unprovoked 12 months) ( G r a d e 1 B ) , or (iii) extended therapy VTE, we recommend extended anticoagulant (Grade 1B regardless of risk of bleeding). therapy over 3 months of therapy in those who have a low bleeding risk ( Table 2 ) (Grade 1B), 3.1.2. In patients with a proximal DVT of the leg and we suggest extended anticoagulant therapy provoked by a nonsurgical transient risk factor, in those with a moderate bleeding risk ( Table 2 ) we recommend treatment with anticoagulation (Grade 2B). for 3 months over (i) treatment of a shorter period ( G r a d e 1 B ), (ii) treatment of a longer time- 3.1.4.5. In patients with a second unprovoked limited period (eg, 6 or 12 months) (Grade 1B), VTE who have a high bleeding risk ( Table 2 ), and (iii) extended therapy if there is a high we suggest 3 months of anticoagulant therapy bleeding risk ( Table 2 ) (Grade 1B). We suggest over extended therapy (Grade 2B). treatment with anticoagulation for 3 months over extended therapy if there is a low or mod- 3.1.5. In patients with DVT of the leg and erate bleeding risk ( Table 2 ) (Grade 2B). active cancer, if the risk of bleeding is not high ( Table 2 ), we recommend extended anti- 3.1.3. In patients with an isolated distal DVT of coagulant therapy over 3 months of therapy the leg provoked by surgery or by a nonsurgical (Grade 1B), and if there is a high bleeding risk transient risk factor (see remark), we suggest ( Table 2 ), we suggest extended anticoagulant treatment with anticoagulation for 3 months therapy (Grade 2B). over treatment of a shorter period (Grade 2C) and recommend treatment with anticoagulation Remarks (3.1.3, 3.1.4, 3.1.4.3): Duration of treatment for 3 months over treatment of a longer time- of patients with isolated distal DVT refers to patients limited period (eg, 6 or 12 months) (Grade 1B) or in whom a decision has been made to treat with anti- extended therapy (Grade 1B regardless of bleeding coagulant therapy; however, it is anticipated that not risk) . all patients who are given a diagnosis of isolated distal DVT will be prescribed anticoagulants (see section 2.3). 3.1.4. In patients with an unprovoked DVT of In all patients who receive extended anticoagulant the leg (isolated distal [see remark] or prox- therapy, the continuing use of treatment should be imal), we recommend treatment with anticoagu- reassessed at periodic intervals (eg, annually). e454S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 20—[Section 3.3] Summary of Findings: LMWH vs VKA for Long-term Treatment of VTEa-c,173,211,213,223,224,226-228

Anticipated Absolute Effects No. of Quality of the Participants Evidence Relative Effect Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) Risk With VKA LMWH (95% CI)

Death 2,496 (7 studies), Moderated,e due RR 0.96 164 per 1,000 7 fewer per 1,000 6 mo to imprecision (0.81-1.13) (from 31 fewer to 21 more) Recurrent VTE 2,727 (8 studies), Moderatef due RR 0.62 No cancerg 6 mo to risk of bias (0.46-0.84) 30 per 1,000 11 fewer per 1,000 (from 5 fewer to 16 fewer) Nonmetastatic cancerg 80 per 1,000 30 fewer per 1,000 (from 13 fewer to 43 fewer) Metastatic cancerg 200 per 1,000 76 fewer per 1,000 (from 32 fewer to 108 fewer) Major bleeding 2,737 (8 studies), Moderateh,i due RR 0.81 No cancer or nonmetastatic cancerj 6 mo to imprecision (0.55-1.2) 20 per 1,000 4 fewer per 1,000 (from 9 fewer to 4 more) Metastatic cancerj 80 per 1,000 15 fewer per 1,000 (from 36 fewer to 16 more) Burden of … Highk … Warfarin: daily medication, dietary LMWH: daily injection, anticoagulation restrictions, frequent blood no dietary restrictions, testing/monitoring, increased no frequent blood hospital/clinic visits testing/monitoring PTS (self-reported 100 (1 study), 2 y Lowl,m due to RR 0.85 200 per 1,000n 30 fewer per 1,000 leg symptoms risk of bias, (0.77-0.94) (from 12 fewer to 46 fewer) and signs) indirectness The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aTwo of these studies enrolled only patients without cancer, three enrolled only patients with cancer, and three enrolled both patients with and without cancer (separate data provided for patients with cancer and without cancer in one study). bLimited to LMWH regimens that used Ն 50% of the acute treatment dose during the extended phase of treatment. cThe initial parenteral anticoagulation was similar in both arms for all except one study (Hull et al211) in which patients randomized to LMWH initially received the same LWMH, whereas patients randomized to VKA received initially UFH. dOne study did not report deaths, which is unusual and could refl ect selective reporting of outcomes. eCI includes both no effect and harm with LMWH. fNone of the studies were blinded, whereas the diagnosis of recurrent VTE has a subjective component and there could be a lower threshold for diagnosis of recurrent VTE in VKA-treated patients because switching the treatment of such patients to LMWH is widely practiced. At the same time, there is reluctance to diagnose recurrent VTE in patients who are already taking LMWH because there is no attractive alternative treatment option. gRisk of recurrent VTE: low corresponds to patients without cancer (3% estimate taken from recent large RCTs of acute treatment), intermediate corresponds to patients with local or recently resected cancer (based on average rate across the six studies in this analysis and appears to be consistent with Prandoni et al38 [particularly if low risk is increased to 4%]), and high corresponds to patients with locally advanced or distant metastatic cancer (Prandoni et al). hNo study was blinded; diagnosis of major bleeding has a subjective component. iThe 95% CIs for the RR for major bleeding includes a potentially clinically important increase or decrease with LMWH and may vary with the dose of LMWH used during the extended phase of therapy. jRisk of bleeding: low corresponds to patients without risk factor for bleeding (ie, age . 75 y, cancer, metastatic disease; chronic renal or hepatic failure; platelet count , 80,0000; requirement for antiplatelet therapy; history of bleeding without a reversible cause) (Table 2). Based on Prandoni et al38 and Beyth et al212 and adjusted to a 6-mo time frame. kHull et al213 reported no signifi cant difference in QOL but suggested greater satisfaction with LMWH over VKA (questionnaire did not directly assess the burden of injections). lPatients and investigators not blinded. Self-reported leg symptoms and signs after 3 mo of treatment. mThe association between leg symptoms and signs at 3 mo and long-term PTS is uncertain. nBaseline risk assumes that patients all wear pressure stockings. Control event rate comes from observational studies in a review by Kahn et al,214 adjusted to a 2-y time frame.

www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e455S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 21—[Section 3.3] Summary of Findings: Rivaroxaban vs LMWH and VKA Therapy for Acute and Long-term Treatment of VTEa-c,88

Anticipated Absolute Effects No. of Quality of Participants the Evidence Relative Effect Risk With LMWH and Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) VKA Therapy Rivaroxaban (95% CI)

Mortality 3,449 (1 study), Moderatee,f HR 0.67 29 per 1,000 9 fewer per 1,000 6-12 mod due to (0.44-1.02) (from 16 fewer to 1 more) imprecision Recurrent VTE 3,449 (1 study), Moderatee,g HR 0.68 30 per 1,000h 9 fewer per 1,000 6-12 mod due to (0.44-1.04) (from 17 fewer to 1 more) imprecision Major bleeding 3,429 (1 study), Moderatee,g HR 0.68 11 per 1,000j 4 fewer per 1,000 6-12 mod due to (0.34-1.38)i (from 7 fewer to 4 more) imprecision Burden of … … … Warfarin: daily medication, Rivaroxaban: daily medication, no anticoagulation dietary restrictions, frequent dietary restrictions, no frequent blood not reported blood testing/monitoring, testing/monitoring increased hospital/clinic visits The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. HR 5 hazard ratio. See Table 1 and 3 legends for expansion of other abbreviations. aIncluded patients had acute, symptomatic, and objectively verifi ed proximal DVT of the legs (unprovoked, 62%; cancer, 6%; previous, VTE 19%). bRivaroxaban 15 mg bid for 3 wk and then 20 mg/d for a total of 3 (12%), 6 (63%), or 12 (25%) mo. .d and then VKA therapy targeted to an INR of 2.5 for 3, 6, or 12 mo 8ف cEnoxaparin 1 mg/kg bid for dFollow-up was prespecifi ed to be 3 (12%), 6 (63%), or 12 (25%) mo. eAllocation was concealed. Patients, providers, and data collectors were not blinded, but outcome adjudicators were blinded. Intention-to-treat analysis; 1.0% were loss to follow-up. Not stopped early for benefi t. fCI includes values suggesting benefi t or no effect; relatively low number of events. gCI includes values suggesting benefi t and harm. hOne defi nite or possible fatal VTE in the rivaroxaban group and one in the LMWH/VKA group. iCalculated from reported data. jBleeds contributing to death: one in the rivaroxaban group and fi ve in the warfarin group.

3.2 Intensity of Anticoagulant Effect laboratory monitoring and twice-daily injection and is associated with osteoporosis.215,216 SC LMWH Ageno et al2 and Holbrook et al3 in these guidelines present evidence for the optimal intensity of VKA has advantages over SC UFH in that it does not therapy (ie, target INR) during the long-term (eg, require laboratory monitoring, is less likely to cause 216 ﬁ rst 3 months of treatment) and extended phases of osteoporosis, and can be given once a day. For treatment of VTE. these reasons, LMWH has been used for the long- term treatment of VTE. The synthetic pentasaccha- Recommendation ride, fondaparinux, has not been evaluated or widely used for long-term treatment of VTE. Idraparinux, 3.2. In patients with DVT of the leg who are a long-acting pentasaccharide, is effective for the treated with VKA, we recommend a therapeutic long-term treatment of VTE,217 but it is not being INR range of 2.0 to 3.0 (target INR of 2.5) over marketed because of concerns about bleeding given , a lower (INR 2) or higher (INR 3.0-5.0) range its prolonged duration of action (once weekly injec- for all treatment durations (Grade 1B). tion) and lack of reversibility. The direct antithrombin dabigatran and the direct factor Xa inhibitors apixa- 3.3 Choice of Anticoagulant Regimen ban and rivaroxaban have been evaluated for treat- for Long-term Therapy ment of VTE and are now available in many countries. VKA therapy has been the standard method of In this section, we compare VKA therapy (target anticoagulant therapy for VTE. Adjusted-dose SC INR 2.5), LMWH, dabigatran, and rivaroxaban for UFH is an effective alternative to VKA therapy, but the long-term treatment of VTE (ie, ﬁ rst 3 months it has never been popular because it requires initial and extended therapy). e456S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 22—[Section 3.3] Summary of Findings: Rivaroxaban vs Placebo for Extended Anticoagulation of VTEa,b,88

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk Difference With Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Risk With Placebo Rivaroxaban (95% CI)

Mortality 1,196 (1 study), Moderated,e due to RR 0.49 3 per 1,000 2 fewer per 1,000 6 or 12 moc imprecision (0.04-5.4)f (from 3 fewer to 15 more) Recurrent VTE 1,196 (1 study), Highd HR 0.18 71 per 1,000g 58 fewer per 1,000 6 or 12 moc (0.09-0.39) (from 43 fewer to 64 fewer) Major bleeding 1,188 (1 study), Moderated,h due to RR 4.9 i 7 more per 1,000 6 or 12 mo imprecision (0.58-42)f (from 3 more to 16 more) Burden of … Rivaroxaban: daily medication anticoagulation not reported PTS not reported … … … … j The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1, 3, and 21 legends for expansion of abbreviations. aIncluded patients had acute, symptomatic, and objectively verifi ed proximal DVT of the legs or PE (unprovoked, 73%; cancer, 5%; previous VTE, 19%). bRivaroxaban 20 mg/d for 6 or 12 mo after initial long-term therapy. cFollow-up was prespecifi ed to be 6 (60%) or 12 (40%) mo. dAllocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Intention-to-treat analysis; 0.2% were loss to follow-up. Not stopped early for benefi t. eCI includes values suggesting benefi t or no effect; relatively low number of events. fCalculated from reported data with addition of one event to each event rate because event rate was 0 in the control group. gOne defi nite or possible fatal VTE in the rivaroxaban group and one in the LMWH/VKA group. hCI includes values suggesting benefi t and harm. iBleeds contributing to death: none in the rivaroxaban group and none in the warfarin group. jPTS: baseline risk over 2 y of 58.8% for PTS and 13.8% for severe PTS (Kahn et al102). There is threefold (Prandoni et al202) to 10-fold (van Dongen et al209) increase in PTS with recurrent VTE in the ipsilateral leg.

LMWH vs VKA Therapy for the Long-term Treat- C a n c e r v s N o C a n c e r — There are differences ment of DVT: Two meta-analyses compared LMWH between patients with and without cancer with VTE in widely differing doses with VKAs. 218,219 In an that may infl uence response to anticoagulant ther- analysis by Iorio and colleagues,218 which included apies. These include about a 10-fold higher risk of seven studies168,221-225 and a total of 1,379 patients, dying and a threefold higher risk of recurrent VTE among study differences of mean daily dose of LMWH, and major bleeding during the fi rst 3 or 6 months of little effect on the effi cacy of LMWH compared with treatment38 ; different mechanisms of thrombosis that VKA therapy was found, but the dose of LMWH may be associated with a poor response to VKA ther- appeared to infl uence the risk of major bleeding apy229 ; use of cancer chemotherapy that is associated with thrombocytopenia, vomiting, and anorexia and 0.7 ف International Units/d to 4,000 ف w i t h 2 . 0 ف ,OR) with 12,000 International Units/d, relative to the may have other interactions with VKA therapy; and VKA groups; P 5 .03 for dose-dependent interac- the need for invasive therapeutic interventions (eg, tion). Because prophylactic doses of LMWH are drainage procedures) that require reversal of antico- rarely used as an alternative to VKA therapy in agulation. Many of these factors make LMWH more patients with VTE, we restricted our analysis to attractive and VKA therapy less attractive in patients eight studies that used Ն 50% of the full therapeutic with VTE and cancer and suggest that cancer may dose of LMWH for long-term treatment of VTE alter the response of VTE to LMWH vs VKA therapy ( Table 20 , Table S28).173,211,213,223,224,226-228 This evidence (ie, presence of an interaction). suggests that compared with VKA therapy, LMWH Among the eight studies included in Table 20 is associated with a reduction of recurrent VTE and Table S28, separate data are provided for and a similar frequency of major bleeding and mor- 1,114 patients with cancer and 660 patients with- tality. The quality of this evidence is moderate because out cancer. Subgroup analyses suggest the possi- of potential for bias in the assessment of recurrent bility that the response to LMWH vs VKA therapy VTE (nonblinded outcome assessment) and serious may differ between patients with cancer and with- imprecision for major bleeding and mortality. out cancer (recurrent VTE: RR, 0.52 with cancer www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e457S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians [95% CI, 0.36-0.76] vs 0.99 without cancer [95% CI, patients treated with VKA therapy, patients treated 0.46-2.13]); major bleeding: RR, 0.92 with cancer with dabigatran initially received parenteral therapy [95% CI, 0.59-1.44] vs 0.43 without cancer [95% CI, (usually IV UFH or LMWH). This study suggests 0.16-1.17]); mortality: RR, 0.93 with cancer [95% CI, that treatment with dabigatran or VKA therapy is 0.79-1.09] vs 1.85 without cancer [95% CI, 0.59- associated with a similar frequency of recurrent VTE, 5.77]). However, none of these differences is sta- major bleeding, and death. This evidence is of mod- tistically signifi cant, making it less likely that there is erate quality because of serious imprecision for each a true difference in response to LMWH vs VKA in outcome and lack of long-term safety data for dab- the two patient populations. For this reason, we have igatran in this patient population. Because the study applied the same relative effects for LMWH vs VKA included few patients with cancer, we were unable to in patients with and without cancer. The baseline risks assess whether its fi ndings apply equally to patients of events, however, are clearly different in the two with and without cancer. In the absence of evidence populations. of such an interaction, however, we have not further Patient Preferences— As discussed in the Methods rated down the quality of evidence for patients with section, the ultimate judgment of the entire Anti- VTE and cancer. thrombotic Therapy and Prevention of Thrombo- sis, 9th ed: American College of Chest Physicians Rivaroxaban vs VKA Therapy for the Long-term Evidence-Based Clinical Practice Guidelines (AT9) Treatment of DVT: A single study has directly com- panel is that most patients prefer VKA therapy over pared rivaroxaban (without initial parenteral antico- LMWH therapy. The higher purchase cost of LMWH agulation) with parenteral anticoagulation and VKA 88 compared with VKA therapy is an additional barrier in patients with acute DVT ( Table 21 , Table S29). to the long-term use of LMWH. Results suggested that treatment with rivaroxaban Quality of Evidence and Strength of Recommen- and VKA therapy are associated with a similar fre- dations— Evidence for the comparison of long-term quency of recurrent VTE, major bleeding, and death. LMWH vs VKA therapy in patients without cancer is This evidence is of moderate quality because of of low quality. The subgroup effect discussed previ- serious imprecision for each outcome. Because the ously was not suffi ciently convincing to allow us to study included few patients with cancer, we were generate an effect estimate specifi cally for patients unable to assess whether its fi ndings apply equally to without cancer, but still reduced our confi dence that patients with and without cancer. In the absence of the overall effect estimate applies to the noncancer evidence of such an interaction, however, we have subgroup that contributed a minority of data. Consid- not further rated down the quality of evidence for erations favoring use of VKA over LMWH in patients patients with VTE and cancer. without cancer include (1) the evidence of benefi t Comparisons Among LMWH, Dabigatran, and with LMWH is of low quality, (2) the estimated abso- Rivaroxaban for the Long-term Treatment of DVT: lute reductions in recurrent VTE events with LMWH There are no direct comparisons of these three agents compared with VKA therapy is small, (3) the high for the long-term treatment of VTE. Recommendations cost of LMWH, and (4) our assessment that LMWH about the use of one of these agents over the other is a greater burden to patients than VKA therapy. are based on indirect comparisons, and the evidence Considerations favoring use of LMWH over warfarin is low quality. in patients with cancer include (1) a large absolute reduction in recurrent VTE with LMWH over VKA Recommendations therapy and (2) that LMWH is better suited to the care of patients with cancer than is VKA therapy. 3.3.1. In patients with DVT of the leg and no Among patients with VTE and cancer, the advan- cancer, we suggest VKA therapy over LMWH tages of LMWH over VKA therapy are expected to for long-term therapy (Grade 2C). For patients be greatest in patients (1) with metastatic disease, with DVT and no cancer who are not treated (2) being treated with aggressive chemotherapy, with VKA therapy, we suggest LMWH over dab- (3) presenting with extensive VTE, (4) with liver igatran or rivaroxaban for long-term therapy dysfunction, (5) with poor or unstable nutritional (Grade 2C) . status, and (6) who wish to avoid laboratory monitoring of coagulation. 3.3.2. In patients with DVT of the leg and cancer, we suggest LMWH over VKA therapy Dabigatran vs VKA Therapy for the Long-term (Grade 2B). In patients with DVT and cancer Treatment of DVT: One completed study has directly who are not treated with LMWH, we suggest compared dabigatran and VKA for the fi rst 6 months VKA over dabigatran or rivaroxaban for long- of treatment of VTE ( Table 23 , Table S31 ).343 Like term therapy (Grade 2B). e458S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 23—[Section 3.3] Summary of Findings: Dabigatran vs VKA Therapy for Long-term Treatment of VTEa-c,343

Anticipated Absolute Effects Quality of No. of Participants the Evidence Relative Effect Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) Risk With Warfarin Dabigatran (95% CI)

Mortality 2,539 (1 study), Moderated,e due HR 0.98 17 per 1,000 0 fewer per 1,000 6 mo to imprecision (0.53-1.79) (from 8 fewer to 13 more) Recurrent VTE 2,539 (1 study), Moderated,e due HR 1.01 19 per 1,000f 0 more per 1,000 6 mo to imprecision (0.65-1.84) (from 7 fewer to 16 more) Major bleeding 2,539 (1 study), Moderated,e due HR 0.82 19 per 1,000g 3 fewer per 1,000 6 mo to imprecision (0.45-1.48) (from 10 fewer to 9 more)

Burden of … … … Warfarin: daily medication, Dabigatran: daily medication, no anticoagulation dietary restrictions, frequent dietary restrictions, no frequent blood not reported blood testing/monitoring, testing/monitoring increased hospital/clinic visits The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1, 3, and 21 legends for expansion of abbreviations. aIncluded patients had acute, symptomatic, and objectively verifi ed proximal DVT of the legs or PE. bDabigatran 150 mg bid taken orally for 6 mo after an initial treatment with LMWH or IV UFH. cWarfarin adjusted to achieve an INR of 2.0 to 3.0 for 6 mo after an initial treatment with LMWH or IV UFH. dAllocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Modifi ed intention-to-treat analysis; 1.1% loss to follow-up. Not stopped early for benefi t. eCI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. fOne fatal VTE in dabigatran group and three fatal VTE in warfarin group. gOne fatal major bleeding in dabigatran group and one fatal major bleeding in warfarin group.

Remarks (3.3.1-3.3.2): Choice of treatment in patients decision about using extended therapy occurs after with and without cancer is sensitive to individual an initial period of anticoagulation (eg, 3 months) and patient’s tolerance for daily injections, need for labo- because the relative effi cacy and safety of anticoagulant ratory monitoring, and treatment costs. LMWH, rivar- regimens are expected to be similar during the early oxaban, and dabigatran are retained in patients with and extended phases of therapy, we anticipate that renal impairment, whereas this is not a concern with most patients will continue to use their initial antico- VKA. Treatment of VTE with dabigatran or rivaroxa- agulant regimen for extended therapy. Possible rea- ban, in addition to being less burdensome to patients, sons for switching from LMWH to VKA therapy may prove to be associated with better clinical out- include the following: cancer becomes less active, comes than VKA and LMWH therapy. When these chemotherapy is completed, patient tires of SC injec- guidelines were being prepared (October 2011), tions, development of renal impairment causes con- postmarketing studies of safety were not available. cern about accumulation of LMWH (also applies Given the paucity of currently available data and that to rivaroxaban and dabigatran), and LMWH costs new data are rapidly emerging, we give a weak rec- become prohibitive . Reasons for switching from ommendation in favor of VKA and LMWH therapy VKA therapy to LMWH could include diffi culty with over dabigatran and rivaroxaban, and we have not INR control and need for repeated invasive proce- made any recommendations in favor of one of the dures. New anticoagulant therapies may expand indi- new agents over the other. cations for extended anticoagulant therapy because they are less burdensome than VKA or LMWH 3.4 Choice of Anticoagulant Regimen therapy and because they may be associated with for Extended Therapy improved clinical outcomes (ie, more effective or Other than a comparison of low-intensity (target safer). INR 1.75) with conventional-intensity VKA therapy44 Recommendation (Table S24 and S25), there are no completed stud ies that have compared different anticoagulant agents 3.4. In patients with DVT of the leg who choose or regimens for extended therapy after a standard- extended therapy, we suggest treatment with ized initial period (eg, Ն 3 months) of anticoagulation the same anticoagulant chosen for the fi rst in either patients without or with cancer. Because a 3 months (Grade 2C) . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e459S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 3.5 Treatment of Asymptomatic DVT of the Leg 4.0 PTS of the Leg Rather than screen postoperative patients for the PTS is a cluster of leg symptoms and signs attribut- presence of asymptomatic DVT, clinicians should able to previous DVT. PTS occurs in about one-third prescribe primary prophylaxis for VTE to surgical of patients after acute DVT and up to two-thirds who patients.230,231 If imaging studies performed for other have had an iliofemoral DVT.102,232 The initial treat- reasons (eg, CT scanning for staging of cancer) inci- ment of acute DVT, particularly with the use of dentally detect asymptomatic proximal DVT, the high thrombus removal strategies, may infl uence the risk frequency of false-positive results in patients without of developing PTS (section 2.8). The most prominent a prior suspicion of DVT dictates caution in assuming symptoms are chronic dependent swelling and pain, that a DVT is truly present. Reasons for a high rate discomfort on walking, and skin discoloration. The of false-positive results include (1) the imaging tech- severity of symptoms may vary over time, and the most nique may not have been optimal for the diagnosis of severe manifestation is a venous ulcer of the lower DVT, (2) incidentally diagnosed DVT often is seen in leg. In this section, we address prevention of PTS the pelvis where DVT is harder to image (eg, unable fi rst and then its treatment. Unlike the last edition to be assessed with compression ultrasound), and of these guidelines,6 we do not address treatment of (3) the prevalence of DVT in asymptomatic patients leg ulcers associated with venous insuffi ciency. is much lower than in symptomatic patients. Conse- quently, when there is evidence of incidental DVT, 4.1 Compression Stockings and Bandages additional diagnostic testing (eg, ultrasound) to con- to Prevent PTS fi rm the presence of DVT may be necessary. Because Five randomized trials evaluated compression stock- many cases of asymptomatic VTE are detected as PE, ings used at various times after diagnosis of acute see also section 6.9 of this article for recommenda- DVT for the prevention of PTS (Tables S32 and tions on the management of this condition. S33). 201,202,233-235 Two of these trials 163,164 randomized No randomized trials have evaluated anticoagu- patients soon after diagnosis of a fi rst episode of lant therapy in patients with incidental VTE; there- symptomatic proximal DVT to prolonged use of fore, evidence is of moderate quality because of stockings (30-40 mm Hg pressure at the ankles) or indirectness. Moreover, benefi ts of anticoagulant no stockings and otherwise treated patients the same therapy may be less than in symptomatic patients way ( Table 24 , Table S34). These trials suggest that because asymptomatic DVT may be chronic or less compression stockings started within 2 weeks of DVT extensive and because the prevalence of false-positive and continued for 2 years reduce PTS by about 50% results will be higher than in patients who were sus- and do not alter the frequency of recurrent VTE. pected of having DVT. Patients with proximal DVT and a previous DVT in Factors that justify a more-aggressive approach to the same leg and who have marked symptoms are anticoagulation in patients with incidentally diagnosed expected to gain the most benefi t from compression DVT include certainty of diagnosis, extensive throm- stockings.102,201,202 The evidence is of moderate quality bosis that appears to be acute (eg, not present on a because the assessment of PTS, which includes a previous imaging study), progression of thrombosis large subjective component, was not blinded, and the on a follow-up imaging study, ongoing risk factors for estimate for recurrent VTE was imprecise. VTE (eg, cancer), and a low risk of bleeding. A less- Compression stockings applying an ankle pressure aggressive approach to anticoagulation could include of 30 to 40 mm Hg and a lower pressure higher up (1) withholding of anticoagulation with surveillance the leg (ie, graduated pressure) should be started as to detect DVT extension or (2) limiting anticoagulant soon as feasible after starting anticoagulant therapy. therapy to 3 months in patients with continuing risk Bandages may be used to provide initial compressive factors for VTE (eg, cancer). Many patients have left therapy because it may not be possible to fi t compres- the hospital by the time incidental DVT is reported. sion stockings immediately and if stockings can be If it would be diffi cult for patients to return the same worn, a rapid decrease in leg swelling will require them day, it is often reasonable to defer further assessment to be refi tted. The fi ndings from a randomized trial of and anticoagulant therapy until the next day. 69 patients suggest that immediate compression ban- Recommendation daging improves acute symptoms but does not reduce PTS at 1 year (RR, 0.9; 95% CI, 0.4-1.8). 236 Patients or 3.5. In patients who are incidentally found to their caregivers need to be able to apply and remove have asymptomatic DVT of the leg, we suggest stockings for their use to be feasible. Alternative the same initial and long-term anticoagulation approaches to the use of stockings, such as routinely as for comparable patients with symptomatic wearing stockings after acute DVT but stopping them DVT (Grade 2B). if there are no symptoms of PTS after 6 months 235,237 or e460S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians only wearing stockings if there is persistent leg swell- quality of the evidence is low because of imprecision ing, have not been adequately evaluated. and risk of bias. There is anecdotal evidence that compression therapy is of benefi t in many patients Recommendation with PTS, and the potential benefi t of a trial of compression stockings in individual patients is likely to 4.1. In patients with acute symptomatic DVT outweigh its harm and cost. We suggest below-knee of the leg, we suggest the use of compression stockings in most patients, but thigh-length stockings stockings (Grade 2B) . may be preferable in those with marked thigh swelling. Two small crossover randomized trials have evalu- Remarks : Compression stockings should be worn ated the treatment of severe PTS with intermittent for 2 years, and we suggest beyond that if patients compression devices239,240 ( Table 26 , Table S36). Both have developed PTS and fi nd the stockings helpful. studies suggested benefi t from intermittent compres- Patients who place a low value on preventing PTS or sion therapy. The quality of the evidence, however, is a high value on avoiding the inconvenience and dis- moderate because of imprecision. The goal of inter- comfort of stockings are likely to decline stockings. mittent compression therapy is to reduce PTS symptoms rather than to alter the natural history of its 4.2 Physical Treatment of PTS development. These devices can be used with or Treatment of PTS with compression stockings without compression stockings, depending on patient has only been evaluated in two small trials 233, 238 preference. Leg swelling and associated symptoms ( Table 25 , Tables S35 and S37) (all patients received (eg, heaviness, tightness) are more likely to respond rutosides in one study238 ). These studies did not fi nd to compression stockings or intermittent compres- compression stockings to be of benefi t. However, the sion devices than are other symptoms.

Table 24—[Section 4.1] Summary of Findings: Elastic Compression Stockings vs No Elastic Compression Stockings to Prevent PTS of the Lega,b,451

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With No Elastic Risk Difference With Elastic Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Compression Stockings Compression Stockings (95% CI)

PTS 421 (2 studies), 2 y Moderatec due to risk RR 0.46 479 per 1,000e,f 259 fewer per 1,000 of bias (0.34-0.63)d (from 177 fewer to 316 fewer)g Recurrent 374 (2 studies), 5 y Moderateh,i due to RR 1.01 210 per 1,000j 2 more per 1,000 VTE imprecision (0.61-1.67)d (from 82 fewer to 141 more) QOL not …… k … reported The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1, 3 and 10 legends for expansion of abbreviations. aPrandoni et al202 excluded patients with recurrent ipsilateral DVT, preexisting leg ulcers, or signs of chronic venous insuffi ciency, bilateral thrombosis, a short life expectancy or a contraindication for use of stockings (eg, advanced-stage peripheral arterial insuffi ciency). Brandjes et al201 excluded patients with short life expectancy, paralysis of the leg, bilateral thrombosis, leg ulcers, or extensive varicosis. bBrandjes201 used graded elastic compression stockings (40 mm Hg of pressure at the ankle, 36 mm Hg at the lower calf, and 21 mm Hg at the upper calf); stockings were applied 2 to 3 wk after the fi rst episode of proximal DVT. Prandoni et al202 used fl at-knitted stockings (30 to 40 mm Hg of pressure at the ankle); stockings were started at hospital discharge, an average of 1 wk after admission. In both studies, stockings were used for 2 y. cPatients were not blinded to the treatment assignment, and outcomes were partly based on subjective report of symptoms. dThe effect estimate shown here results from a meta-analysis (Mantel-Haenszel fi xed-effects model) of the two relevant trials. A fi xed-effects model was chosen because of the small number of studies available. eThis estimate is based on the fi ndings of the VETO study.70 This probably underestimates PTS baseline risk given that overall, 52% of patients reported the current use of compression stockings during study follow-up. fIn Prandoni et al,202 most events occurred during the fi rst 6 mo. The cumulative incidence of the PTS in the control group was 40% after 6 mo, 47% after 1 y, and 49% after 2 y. gSevere PTS: assuming the same RR of 0.46 and a baseline risk of 8.1% over 2 y, the absolute reduction is 44 fewer severe PTS per 1,000 (from 30 fewer to 53 fewer) over 2 y. hWe did not rate down the quality of evidence for recurrent VTE for the lack of blinding because this a more objective outcome than PTS. iCI includes both negligible effect and appreciable benefi t or appreciable harm. jThis estimate is the mean of two estimates derived from two studies: 12.4% probable/defi nite VTE (Heit et al165) and 29.1% confi rmed VTE (Prandoni et al208). kThis is an important outcome that should be considered in future studies. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e461S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Recommendations sion. Furthermore, rutosides may be associated with important side effects. 4.2.1. In patients with PTS of the leg, we suggest a trial of compression stockings ( G r a d e 2 C ). Recommendation 4.2.2. In patients with severe PTS of the leg 4.3. In patients with PTS of the leg, we suggest that is not adequately relieved by compres- that venoactive medications (eg, rutosides, defi - sion stockings, we suggest a trial of an inter- brotide, and hidrosmin) not be used ( G r a d e 2 C ). mittent compression device ( G r a d e 2 B ). Remarks: Patients who value the possibility of response 4.3 Pharmacologic Treatment of PTS over the risk of side effects may choose to undertake Hydroxyrutosides, a class of fl avonoid drug pro- a therapeutic trial duced from plant glycosides, may reduce capillary permeability, reduce infl ammation, improve lym- 5.0 Initial Treatment of Acute PE phatic function, and promote ulcer healing in patients with chronic venous insuffi ciency.6,241,242 Two studies As we noted in Methods (section 1.1), recommenda- compared treatment of PTS (without ulceration) with tions for management of patients with PE, particu- rutosides vs control238,243 (all patients wore compres- larly those addressing anticoagulant therapy and IVC sion stockings in one study238 ), and one study com- fi lter insertion, are based on studies that enrolled pared rutosides with hidrosmina244 (Table S37). The patients with only DVT , patients with both DVT and two controlled studies suggest that rutosides do not PE, and patients with only symptoms of PE. The reduce most symptoms of PTS, although they may following sections emphasize studies that enrolled reduce ankle swelling ( Table 27 ). This evidence is only patients with symptoms of PE (who could also of low quality because of inconsistency and impreci- have symptoms of DVT), emphasize differences in

Table 25—[Section 4.2.1] Summary of Findings: Compression Stockings vs No Compression Stockings for Patients With PTSa-c,233,238

Anticipated Absolute Effects

Risk Difference With No. of Participants Quality of the Relative Effect Risk With No Compression Stockings Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Compression Stockings (95% CI)

Symptomatic relief 115 (2 studies), Lowd-f due to risk of bias RR 0.96 579 per 1,000 23 fewer per 1,000 (from 174 treatment successg 12 to 26 mo and imprecision (0.70-1.31) fewer to 179 more) QOL not reported … Not estimable Recurrent VTE not … Not estimable reported Ulcerationh not … Not estimableh reported The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. CLOTS1 5 Clots in Legs or Stocking After Stroke. See Table 1 and 3 legends for expansion of other abbreviations. aGinsberg et al233 included patients with PTS 1 y after chronic, typical proximal DVT. Frulla et al238 included patients with clinical symptoms and signs suggestive of PTS. bGinsberg et al233: Graduated compression stockings (30-40 mm Hg, calf or thigh length, depending on symptoms). Patients were encouraged to wear stockings as much as possible during waking hours. Frulla et al238: below-knee graded elastic compression stockings (ECS) (30-40 mm Hg at the ankle). Patients in both study arms received hydroxyethylrutosides (HR) (we considered the ECS 1 HR vs HR comparison). cGinsberg et al233: placebo stockings (calf or thigh length, depending on symptoms). dGinsberg et al233: Adequacy of sequence generation and allocation concealment were unclear; patients and outcome assessors were adequately blinded; unclear whether analysis followed the intention-to-treat principle; unclear whether follow-up was complete. Frulla et al238: outcome assessors were blinded; follow-up was complete; intention-to-treat principle was adhered to, but sequence generation and allocation concealment were unclear, and patients were not blinded. eVery small number of patients fPublication bias was not detected but not ruled out given that we identifi ed only one small study partially supported by industry (provision of graduated compression stockings). gGinsberg et al233 reported treatment failure (defi ned a priori based on any of fi ve clinical criteria, including symptoms and ulcer development). Treatment success refers to the absence of treatment failure. Frulla et al238 used the Villalta scale. hIndirect evidence from the CLOTS1 trial suggests that compression stockings is associated with an RR of 4 for skin complications. e462S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 26—[Section 4.2.2] Summary of Findings: Intermittent Compression Device vs No Intermittent Compression Device for Patients With Severe PTSa-c,233,240

Anticipated Absolute Effects No. of Participants Quality of Relative Risk With No Risk Difference With (Studies), the Evidence Effect (95% Intermittent Intermittent Compression Outcomes Follow-up (GRADE) CI) Compression Device Device (95% CI)

Symptomatic relief: symptom 82 (2 studiesd), Moderatee-i due The mean symptomatic The mean symptomatic relief in score includes scoring of pain, 8 wk to imprecision relief in the control the intervention groups was swelling, and limitation of groups was 0 0.41 SDs higher (0.02 lower activity on a scale of 10-70 to 0.85 higher) QOL: VEINES-QOL scale of 0 (1 studyd,j), Moderateg-i,k,l due The mean QOL in the The mean QOL in the 0-100. 8 wk to imprecision control groups was intervention groups was 2.3 50.2 higher (1.04 lower to 5.64 higher) Recurrent VTEm not reported … Not estimablem Ulcerationn not reported … Not estimablen The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. VEINES 5 Venous Insuffi ciency Epidemiological and Economic Study. See Table 1, 3, and 25 legends for expansion of other abbreviations. aPatients with previous DVT with symptoms of severe PTS. bIntervention group: Ginsberg et al239: Extremity pump used bid for 20 min each session; 50 mm Hg (therapeutic pressure) for 1 mo. O’Donnell et al240: Venowave lower-limb venous return assist device to wear for most of the day for 8 wk. cControl group: Ginsberg et al239: Extremity pump used bid for 20 min each session; 15 mm Hg (placebo pressure) for 1 mo. O’Donnell et al240: Venowave lower-limb venous return assist device with no connection between motor and planar sheet for 8 wk. dCrossover RCTs. eIn both studies, sequence generation was adequate; patients were blinded, analysis adhered to intention-to-treat principle, and there were no missing outcome data. In Ginsberg et al239 (but not O’Donnell et al240), outcome assessors were not blinded, and it was not clear whether allocation was concealed. fI2 5 0%. gSome concerns with indirectness, given relatively short follow-up (8 wk). hVery small number of patients. CI includes both values suggesting no effect and values suggesting a benefi cial effect. iPublication bias not detected but not ruled out given that we identifi ed only two small studies, with one (Ginsberg et al239) partially supported by industry (provision of devices). jO’Donnell et al et al.240 kSequence generation was adequate; patients were blinded, analysis adhered to intention-to-treat principle, and there were no missing outcome data. However, outcome assessors were not blinded, and it was not clear whether allocation was concealed. lPublication bias was not detected but not ruled out given that we identifi ed only a small study. mO’Donnell et al240 indicated no cases of recurrent VTE by the end of this study but judged the follow-up period to be short. nO’Donnell et al240 indicated that one patient in the control group developed a venous ulceration. Three other participants developed nonserious skin-related side effects. Indirect evidence from the CLOTS1 suggests that compression stockings are associated with an RR of 4 for skin complications. Common side effects attributed to Venowave were heat sensation, skin irritation, and increased sweating. the management of patients who present with PE Recommendation compared with DVT, and make recommendations for the management of patients with PE. We do not 5.1. In patients with acute PE, we recommend repeat evidence that was presented in the corre- initial treatment with parenteral anticoagula- sponding section that addresses treatment of DVT; tion (LMWH, fondaparinux, IV UFH, or SC instead, the reader is directed to those sections of the UFH) over no such initial treatment (Grade 1B). article and to the related tables. We do not comment in the text on the quality of the evidence that under- 5.2 Whether to Treat With Parenteral lies treatment recommendations for PE unless the Anticoagulation While Awaiting the quality of this evidence differs from that for patients Results of Diagnostic Work-up for PE who present with DVT. See section 2.2. For the purpose of implement- ing this recommendation, validated prediction rules 5.1 Initial Anticoagulation for Acute PE help with estimation of clinical probability of having See section 2.1, Table 3 , and Table S1. PE.55,56 www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e463S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 27—[Section 4.3] Summary of Findings: Venoactive Medication vs No Venoactive Medication for Patients With PTSa,b,202,238

Anticipated Absolute Effects No. of Participants Quality of Risk With No Risk Difference (Studies), the Evidence Relative Effect Venoactive With Venoactive Outcomes Follow-up (GRADE) (95% CI) Medication Medication (95% CI)

Symptomatic relief: PTS score 163 (2 studies) Lowd-g due to RR 1.14 (0.85-1.52) 476 per 1,000 67 more per 1,000 (Villalta scale) , 5 or decreased by 30% inconsistency, (from 71 fewer to at 12 mo compared with baseline in imprecision 247 more) Frulla et al238; improved tiredness of the leg at 8 wk in de Jongste et al243,c QOL not reported … Not estimable Recurrent VTE not reported … Not estimable Ulceration not reported … Not estimable Side effects 203 (2 studies) Moderated,f-h due RR 2.04 (0.76-5.51) 61 per 1,000 63 more per 1,000 to imprecision (from 15 fewer to 275 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aPatients with PTS and history of DVT in PTS leg. bIncluded studies that assessed rutosides. cInvestigators assessed other symptoms (pain, heaviness, swelling feeling, restless legs, and cramps) but did not report a composite score. The symptom we chose to report showed the most benefi t; the effect estimates for the other symptoms ranged from 0.8 to 1.4, and none were statistically signifi cant. dIn both studies, sequence generation and allocation concealment were unclear. Both studies blinded outcome assessors and had complete follow- up. Although de Jongste et al243 blinded patients, the authors did not adhere to the intention-to-treat principle and did not use a validated scale to measure symptomatic relief. Although Frulla et al238 adhered to the intention-to-treat principle, the author did not blind patients. eI2 5 77%. fSmall number of patients. CI included both values suggesting harms and values suggesting benefi ts. gPublication bias was not detected but not ruled out given that we identifi ed only two small studies, and it unclear whether they were funded by industry. hI2 5 7%.

Recommendations Recommendation 5.2.1. In patients with a high clinical suspicion 5.3. In patients with acute PE, we recommend of acute PE, we suggest treatment with paren- early initiation of VKA (eg, same day as parenteral anticoagulants compared with no treat- teral therapy is started) over delayed initiation, ment while awaiting the results of diagnostic and continuation of parenteral anticoagulation tests (Grade 2C). for a minimum of 5 days and until the INR is 2.0 or above for at least 24 h (Grade 1B) . 5.2.2. In patients with an intermediate clinical suspicion of acute PE, we suggest treatment with parenteral anticoagulants compared with 5.4 Choice of Initial Parenteral Anticoagulant no treatment if the results of diagnostic tests Regimen in Patients With PE are expected to be delayed for more than 4 h See section 2.5. (Grade 2C). 5.2.3. In patients with a low clinical suspicion of LMWH Compared With IV UFH for the Initial acute PE, we suggest not treating with parenteral Treatment of PE: S e e s e c t i o n 2 . 5 a n d T a b l e 6 . C o n - anticoagulants while awaiting the results of sistent with ﬁ ndings in patients with DVT, LMWH diagnostic tests provided that test results are has been found to be as effective and safe as IV UFH expected within 24 h (Grade 2C). in studies that included both patients with PE and DVT or only in patients with PE ( Table 6 ). A meta-analysis of 12 studies85,146,245-254 that included 5.3 Timing of Initiation of VKA and Associated a total of 1,951 patients with either submassive symp- Duration of Parenteral Anticoagulant Therapy tomatic PE or asymptomatic PE in conjunction See section 2.4, Table 4 , and Table S2. with symptomatic DVT failed to demonstrate or e464S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians exclude a beneﬁ cial or detrimental effect of LMWH for this comparison. In making recommendations, we on recurrent VTE (OR, 0.63; 95% CI, 0.33-1.18), considered that fondaparinux and LMWH are equiv- major bleeding (OR, 0.67; 95% CI, 0.36-1.27), and alent and that fondaparinux shares the advantages all-cause mortality (OR, 1.20; 95% CI, 0.59-2.45).255 that LMWH has over IV UFH. We did not take into account the lower purchase cost of SC UFH com- SC UFH Compared With SC LMWH for the Ini- pared with fondaparinux. tial Treatment of PE: See section 2.5, Table 5 , and Tables S3 through S5. Once- vs Twice-Daily Administration of LMWH for Initial Treatment of PE: See section 2.5, Table 8 , Fondaparinux Compared With IV UFH for the Ini- and Table S8. Patients who presented with PE were tial Treatment of PE: The Matisse-PE trial compared included in only one of the ﬁ ve studies with an fondaparinux with IV UFH for acute treatment of unconfounded comparison of once- and twice-daily 79 PE ( Table 28 , Table S38). This study suggested that LMWH85 and in one additional large study that com- fondaparinux is associated with a similar frequency of pared once-daily LMWH therapy with IV UFH in mortality, recurrent VTE, and major bleeding as patients who presented with PE.252 LMWH. The quality of this evidence is moderate because of imprecision. In making recommendations, Recommendations we also considered evidence that fondaparinux is equivalent to LMWH for the treatment of DVT (see sec- 5.4.1. In patients with acute PE, we suggest tion 2.5, Table 7 , and Table S38) and that fondaparinux LMWH or fondaparinux over IV UFH (Grade 2C shares the advantages that LMWH has over IV UFH. for LMWH; Grade 2B for fondaparinux), and over SC UFH (Grade 2B for LMWH; Grade 2C for Fondaparinux Compared With LMWH for the Ini- fondaparinux). tial Treatment of PE: In the absence of direct evidence in patients with PE, indirect evidence in Remarks: Local considerations such as cost, avail- patients with acute DVT (see section 2.5, Table 7 , ability, and familiarity of use dictate the choice and Table S7) suggests that fondaparinux is equiva- between fondaparinux and LMWH. LMWH and lent to LMWH. fondaparinux are retained in patients with renal impairment, whereas this is not a concern with UFH. Fondaparinux Compared With SC UFH for the In patients with PE where there is concern about Initial Treatment of PE: There is no direct evidence the adequacy of SC absorption or in patients in whom

Table 28—[Section 5.4] Summary of Findings: Fondaparinux vs IV UFH for Initial Anticoagulation of Acute PEa-c,79

Anticipated Absolute Effects Quality of the No. of Participants Evidence Relative Effect Risk Difference With Outcomes (Studies), Follow-up (GRADE) (95% CI) Risk With UFH Fondaparinux (95% CI)

Mortality 2,213 (1 study), 3 mo Moderated,e due RR 1.20 (0.82-1.74) 43 per 1,000 9 more per 1,000 (from 8 fewer to imprecision to 32 more) Recurrent VTE 2,213 (1 study), 3 mo Moderated,e due RR 0.75 (0.51-1.12) 50 per 1,000f 13 fewer per 1,000 (from 25 fewer to imprecision to 6 more) Major bleeding 2,213 (1 study), 3 mo Moderated,e due RR 0.85 (0.49-1.49) 23 per 1,000g 4 fewer per 1,000 (from 12 fewer to imprecision to 11 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aAll patients had acute, symptomatic, hemodynamically stable PE. bFondaparinux (5.0, 7.5, or 10.0 mg in patients weighing , 50, 50 to 100, or . 100 kg, respectively) SC once daily given for at least 5 days and until the use of VKAs resulted in an INR . 2.0. cUFH continuous IV infusion (ratio of the activated partial thromboplastin time to a control value of 1.5-2.5) given for at least 5 days and until the use of VKAs resulted in an INR . 2.0. dAllocation was concealed. Patients, providers, and data collectors were not blinded. Outcome adjudicators were blinded; 0.6% of randomized patients were lost to follow-up. Not stopped early for benefi t. eCI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. fSixteen fatal VTE in fondaparinux group and 15 fatal VTE in UFH group. gFourteen patients in the fondaparinux group and 12 patients in the LMWH group had a major bleeding event during the initial period (6-7 d). Of these, one in the fondaparinux group and one in the UFH group were fatal. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e465S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians thrombolytic therapy is being considered or planned, to be treated at home (eg, absence of severe symp- initial treatment with IV UFH is preferred to use of toms or comorbidity). SC therapies. Consistent with the fi ndings of these two trials, a systematic review of 11 observational studies (seven 5.4.2. In patients with acute PE treated with prospective, four retrospective; 928 patients)267 and LMWH, we suggest once- over twice-daily admin- four more recent observational studies (two retro- istration (Grade 2C). spective with 584 patients268,269 ; two prospective with 449 patients270,271 ) reported a very low frequency of Remarks: This recommendation only applies when the complications in low-risk patients with acute PE who approved once-daily regimen uses the same daily dose were initially treated partially or entirely at home. as the twice-daily regimen (ie, the once-daily injec- About one-third to one-half of outpatients with acute tion contains double the dose of each twice-daily PE appear to be in this low-risk group. 272 The evidence injection). It also places value on avoiding an extra from the randomized trials is of moderate quality injection per day. (rated down for serious imprecision), with additional supportive fi ndings from the observational studies. 5.5 Early vs Standard Discharge of Patients With Acute PE Recommendation Consistent with our discussion of outpatient treat- 5.5. In patients with low-risk PE and whose ment of acute DVT (section 2.7), LMWH has made home circumstances are adequate, we suggest it feasible to treat acute PE at home either without early discharge over standard discharge (eg, admission to the hospital (ie, discharge from the after the fi rst 5 days of treatment) (Grade 2 B). emergency department) or with admission and early discharge. However, because acute PE is associated Remarks: Patients who prefer the security of the hos- with much higher short-term mortality than acute pital to the convenience and comfort of home are DVT, the safety of treating PE at home is uncertain. likely to choose hospitalization over home treatment. Consequently, PE is treated at home much less often than DVT, and the proportion of outpatients with PE 5.6 Systemic Thrombolytic Therapy for PE that clinical centers treat at home varies from almost none to about 50%. 5.6.1 Systemic Thrombolytic Therapy vs Anticoag- Two studies randomized patients with acute PE ulation Alone for PE: Randomized trials have estab- and a low risk of complications to receive LMWH lished that, at 24 h, thrombolytic therapy improves either (1) in the hospital for only 3 days vs entirely in (1) pulmonary artery hemodynamic measurements the hospital256 or (2) entirely out of the hospital (dis- (eg, mean pulmonary artery pressure improvement, charged within 24 h) vs at least partly in hospital257 4.4 mm Hg; 95% CI, 2 4.6-4.2 mm Hg), (2) arterio- ( Table 29 , Table S39). This evidence suggests that venous oxygen (difference of 2 0.3 [2 0.4 to 2 0.2] ), treating appropriately selected patients with acute (3) pulmonary perfusion (50% early improvement in PE at home does not increase recurrent VTE, perfusion scan, OR, 3.8; 95% CI, 0.9-15.7), and (4) echo- bleeding, or mortality. cardiographic assessment (OR for improved right There are a number of prediction rules for identi- ventricular wall movement, 3.1; 95% CI, 1.5-6.3). 273 fying patients with acute PE who have a low risk of Thrombolytic therapy, however, does not appear to serious complications and may be suitable for treat- reduce the extent of residual thrombosis. It is uncer- ment at home.258-263 Of these, the PE Severity Index tain whether the benefi ts of more-rapid resolution of (PESI) is best validated261,262,264-266 and was used to PE outweigh the risk of increased bleeding associ- select patients for home treatment in the larger of the ated with thrombolytic therapy. In patients with PE, previously noted clinical trials ( Table 29 , Table S39). severity of presentation is expected to depend on the Patients with acute PE who meet the following extent of embolism (ie, degree of pulmonary artery criteria appear to be suitable for treatment out of obstruction) and the presence and severity of chronic the hospital: (1) Clinically stable with good car- cardiopulmonary impairment. 104,274,275 Patients with diopulmonary reserve (eg, PESI score of , 85257 or the most severe presentations who have the highest simplifi ed PESI score of 0, 262 including none of risk of dying from an acute PE have the most to gain hypoxia, systolic BP , 100, recent bleeding, severe from thrombolysis. chest pain, platelet count , 7 0 , 0 0 0 / m m3 , PE while Prognosis in Patients With Acute PE— Of patients %5 ف ,on anticoagulant therapy, and severe liver or renal who are diagnosed with PE and start treatment disease)257 , (2) good social support with ready access die of the initial PE or another PE within the next to medical care, and (3) expected to be compliant 7 days.9,104,276-279 However, although the risk of dying with follow-up. Patients also need to feel well enough of PE differs markedly among patients, no validated e466S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 29—[Section 5.5] Summary of Findings: Early Discharge vs Standard Discharge in the Treatment of Acute PEa,b,256,257

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With Standard Risk Difference With Early Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Discharge Discharge (95% CI)

Mortality 471 (2 studies), 3 mo Moderatec,d due to RR 0.58 (0.17-1.97) 26 per 1,000 11 fewer per 1,000 (from 22 imprecision fewer to 26 more) Nonfatal recurrent PE 471 (2 studies), 3 mo Moderatec,d due to RR 1.23 (0.25-6.03) 9 per 1,000 2 more per 1,000 (from 7 imprecision fewer to 44 more) Major bleeding 471 (2 studies), 3 mo Moderatec,d due to RR 2.74 (0.45-16.71) 4 per 1,000 8 more per 1,000 (from 2 imprecision fewer to 69 more) QOL not reported … … … … … PTS not reported … … … … … The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aThe two RCTs included only patients with low risk: risk classes I or II on the Pulmonary Embolism Severity Index in Aujesky et al216; low risk on clinical prediction rule by Uresandi et al.258 bMean length of hospital stay: 3.4 (SD 1.1) vs 9.3 (SD 5.7) d in Otero et al256 and 0.5 (SD 1.0) vs 3.9 (SD 3.1) d in Aujesky et al257; low risk on clinical prediction rule by Uresandi et al258 in Otero et al. cOtero et al256: allocation concealed; no patients lost to follow-up; intention-to-treat analysis; no blinding of outcome assessors reported; study stopped early because the rate of short-term mortality was unexpectedly high in the early discharge group (2 [2.8%] vs 0 [0%]). Aujesky et al257: unclear whether allocation was concealed; three (1%) patients had missing outcome data; intention-to-treat analysis; blinding of outcome adjudicators; no early stoppage. dCI includes both values suggesting no effect and values suggesting appreciable harm or appreciable benefi t. risk prediction tool is available. Risk of dying of PE tion tool for bleeding with thrombolytic therapy if cardiopulmonary arrest in patients with PE. However, we assume that the %70 ف is estimated to be -of patients at presentation), 30% if assessment of bleeding risk with thrombolytic ther %1 ف) occurs apy is similar in patients with PE and with acute %5 ف) there is shock requiring inotropic support in patients who are not ST-segment elevation myocardial infarction.104,110-113,298,299 %2 ف of patients), and hypotensive.104,276-278,280,281 In the presence of normal Table 11 lists risk factors for bleeding with thrombolytic systemic arterial pressure, prognosis can also differ, therapy, categorized as major and relative contra- depending on (1) clinical evaluation, 276 (2) cardiac indications. biomarkers such as troponin or brain natriuretic Trials Evaluating Thrombolytic Therapy in Patients peptide,279,282-291 and (3) assessment of right ventric- With Acute PE— T h e fi ndings of 13 randomized ular size and function.279,280,283,285,290-295 trials that compared thrombolytic therapy to antico- Clinical evaluation involves assessment of general agulant therapy alone in patients with acute PE are appearance, BP, heart rate, respiratory rate, temper- summarized in Table 30 and Tables S40 through ature, pulse oximetry, and signs of right ventricular S42.300-313 A number of meta-analyses of these studies dysfunction (eg, distended jugular veins, tricuspid have been performed.104,273,315,315 This evidence sug- 104 regurgitation, accentuated P2 ). Clues on the ECG gests that thrombolysis may be associated with a reduc- include right bundle branch block, S1 Q 3 T 3 , and tion in mortality and recurrent PE and is associated 296 T-wave inversion in leads V1 through V4 . Elevation with an increase in major bleeding, as has been estab- of cardiac troponins indicates right ventricular micro- lished in patients with myocardial infarction.112 The infarction, and echocardiography may show right ven- quality of evidence regarding mortality and recurrent tricular hypokinesis; both are risk factors for early PE is low because of risk of bias, serious imprecision, mortality and are associated with a worse outcome and suspected publication bias. A previous meta- when they occur together. 238,241-245,250 Right ventricular analysis that categorized studies as either including, enlargement on the CT pulmonary angiogram, defi ned or not including, patients with cardiopulmonary com- as a right ventricular diameter Ն 90% than the left promise, suggested that thrombolytic therapy reduced ventricular diameter may also be an independent risk the composite outcome of death and recurrent PE in factor for death and nonfatal complications.279,291,293,295,297 studies that included the sickest patients.315 However, Risk of Bleeding With Thrombolytic Therapy — we found that the data available from these studies We have not identifi ed any validated risk predic- are not suffi ciently detailed to enable a subgroup www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e467S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians analysis evaluating outcomes in patients with hemo- tension, we suggest administration of thrombo- dynamic compromise or other markers of heightened lytic therapy (Grade 2C). risk of death (eg, right ventricular dysfunction). Balancing Benefi ts and Harms of Thrombolytic 5.6.2 Systemic Thrombolytic Therapy Regimen for Therapy— In patients who present with PE and PE: Twelve randomized trials (total of 938 patients) hypotension (eg, systolic pressure BP , 90 mm Hg or have compared the rate of thrombus resolution a documented drop in systolic BP . 40 mm Hg with achieved with various IV thrombolytic regimens.316-327 evidence of poor perfusion), especially if they have a These regimens included urokinase given over 2 h319,326 low risk of bleeding, even modest effi cacy of throm- or 12 h 316,321,326 ; streptokinase given over 2 h, 12 h,320 bolytic therapy is likely to reduce deaths from PE or 24 h316 ; and recombinant tPA (rt-PA) given over more than it would increase fatal bleeds and nonfatal 15 min317,322,325 o r 2 h ,317-325,327 r e t e p l a s e i n t w o b o l u s e s intracranial bleeds ( Table 30 , Tables S40-S42). The 30 min apart,325 and desmoteplase in three different ultimate judgment of the entire AT9 panel was to doses as a bolus323 . issue a weak recommendation for patients with PE An additional study compared IV with pulmonary and hypotension given the uncertainty of the benefi t. artery catheter administration of rt-PA (50 mg over In most patients with PE, given the certain risks of 2 h).328 The results of studies that compared different bleeding and less-certain benefi ts, thrombolysis is approaches to thrombolysis in patients with PE likely to be harmful. Selected patients without hypo- (noted previously) suggest that (1) prolonged infusions tension may benefi t from thrombolysis because their of thrombolytic agents (eg, Ն 12 h) are associated initial clinical presentation or clinical course after with higher rates of bleeding316,318 ; ( 2 ) 2 - h i n f u - starting anticoagulant therapy suggest that they are at sions achieve more rapid clot lysis than 12- or 24-h high risk of dying. infusions318,320,321 ; (3) when a high-concentration 2-h There is no explicit clinical prediction rule that infusion of thrombolysis is administered, there is no identifi es this subgroup of patients. We suggest that clear difference in the effi cacy or safety of rt-PA mg 50 ف ,such patients are identifi ed predominantly by clinical v s s t r e p t o k i n a s e327 ; (4) bolus rt-PA regimens (eg evidence of instability (eg, a decrease in systolic BP in Յ 15 min) appears to be as effective and safe as that still remains . 90 mm Hg, tachycardia, elevated a 2-h infusion of 100 mg of rt-PA313,317,322,325 ; and (5) jugular venous pressure, clinical evidence of poor infusion of rt-PA directly into a pulmonary artery as tissue perfusion, hypoxemia) and failure to improve opposed to a peripheral vein does not accelerate on anticoagulant therapy. As noted previously, labo- thrombolysis but does cause more frequent bleeding ratory (eg, troponin, brain natriuretic peptide), ECG, at the catheter insertion site (there was no attempt echocardiography, and CT evidence of right ventric- to infuse rt-PA directly into or to mechanically dis- ular dysfunction or enlargement, can supplement the rupt the thrombus in this study from 1988). 328 When clinical evaluation of instability; however, they are a lytic agent is appropriate for PE, current evidence not suffi ciently predictive to serve as indications for supports that thrombolytic therapy should be infused thrombolytic therapy on their own,289 and we do not into a peripheral vein over Յ 2 h. At a dose of 100 mg recommend that they are routinely measured. over 2 h, rt-PA is currently the most widely used and evaluated regimen. In patients with imminent or Recommendations actual cardiac arrest, bolus infusion of thrombolytic therapy is indicated. 5.6.1.1. In patients with acute PE associated The quality of evidence for comparisons of systemic with hypotension (eg, systolic BP , 90 mm Hg) thrombolytic agents and regimens (eg, different who do not have a high risk of bleeding, we sug- doses or durations of infusion) is low based on very gest systemically administered thrombolytic serious imprecision and risk of bias. In addition, therapy over no such therapy (Grade 2C). there is substantial potential for publication bias. Based on this evidence, we provide only weak rec- 5.6.1.2. In most patients with acute PE not asso- ommendations for all comparisons of thrombolytic ciated with hypotension, we recommend against agents and regimens in the short-term treatment systemically administered thrombolytic therapy of PE. (Grade 1C). Recommendations 5.6.1.3. In selected patients with acute PE not associated with hypotension and with a low risk 5.6.2.1. In patients with acute PE, when a throm- of bleeding whose initial clinical presentation bolytic agent is used, we suggest short infusion or clinical course after starting anticoagulant times (eg, a 2-h infusion) over prolonged infu- therapy suggests a high risk of developing hypo- sion times (eg, a 24-h infusion) (Grade 2C). e468S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 30—[Section 5.6.1] Summary of Findings: Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With Acute PEa-d,273,309,310,314,315

Anticipated Absolute Effects

Risk With No Risk Difference With No. of Participants Quality of the Relative Effect Systemically Administered Systemically Administered Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Thrombolytic Therapy Thrombolytic Therapy (95% CI)

Mortality 847 (12 studies), Lowe-h due to risk of RR 0.7 Lowi,j 30 d bias and imprecision (0.37-1.31) 11 per 1,000 3 fewer per 1,000 (from 7 fewer to 3 more) Highi,j 89 per 1,000 27 fewer per 1,000 (from 56 fewer to 28 more) Recurrent PE 801 (9 studies), Lowe-h due to risk of bias RR 0.7 57 per 1,000 17 fewer per 1,000 (from 34 30 d and imprecision (0.4-1.21) fewer to 12 more) Major bleeding 847 (12 studies), Moderatee,f,h,k due to RR 1.63 Lowm 10 d risk of bias and (1-2.68)l 1 per 1,000 1 more per 1000 (from 0 more imprecision to 2 more) Highm 62 per 1,000 39 more per 1,000 (from 0 more to 104 more) The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aOne study included exclusively patients with hemodynamic compromise (shock), six excluded them, whereas the rest either included a number of such patients or did not specify related eligibility criteria. Of studies not restricted to patients with hemodynamic compromise (n 5 11), only three were clearly restricted to patients with right ventricular dysfunction; the rest either did not specify related eligibility criteria or included both patients with and without right ventricular dysfunction. As a result, it was not possible to perform reliable categorization of studies to conduct subgroup analyses based on the presence or absence of right ventricular dysfunction or hemodynamic compromise. bStudies included patients at low risk for bleeding. cIncluded studies that used different thrombolytic agents with varying doses and durations of administration; no statistical heterogeneity was noted. dThrombolysis was in addition to anticoagulation (most of the studies used heparin followed by warfarin; three studies used warfarin only). eReport of methodologic quality was poor in most studies. Of the 12 eligible studies, allocation was concealed in fi ve, three were single blind (outcome assessor), six were double blind, and three were not blinded. Most studies did not report on missing outcome data. None of the studies were stopped early for benefi t. For the increase in bleeding with thrombolytic therapy, quality of evidence is increased from low to moderate because there is high quality evidence of this association in patients with myocardial infarction and the indirectness of this evidence to patients with PE is minor. fI2 5 0%. gCI includes values suggesting both benefi t and no effect or harm; small number of events. hInverted funnel plots suggestive of possible publication bias in favor of thrombolytics. iRecurrent PE stratifi cation based on the simplifi ed Pulmonary Embolism Severity Index validated in the RIETE (Registro Informatizado de la Enfermedad Tromboembólica) cohort.262 jSome studies suggested that the baseline risk of mortality in patients with hemodynamic instability is as high as 30%.274 In that case, the absolute number of deaths associated with thrombolytics would be 90 fewer per 1,000 (from 189 fewer to 93 more). kCI includes values suggesting both harm and no effect; small number of events. lMajor bleeding risk stratifi cation derived from the RIETE cohort.30 The median risk of bleeding over the fi rst 10 d reported in the eligible trials was 3.1%. In that case, the absolute number of major bleeding events with thrombolysis would be 20 more per 1,000 (from 0 more to 52 more). mIndirect evidence from studies of thrombolysis for myocardial infarction and acute stroke provide more precise estimates of increase major bleeding with thrombolytics use.

5.6.2.2. In patients with acute PE, when a throm- conjunction with thrombolytic therapy ( Table 30 , bolytic agent is used, we suggest administration Tables S40-S42), and no randomized trials have com- through a peripheral vein over a pulmonary pared different regimens of IV UFH in this setting. artery catheter (Grade 2C) . IV UFH should be given in full therapeutic doses1,3 before thrombolytic therapy is administered, and it 5.6.3 Initial Anticoagulant Therapy in Patients is acceptable to either continue or suspend the Treated With Thrombolytic Therapy: Trials that UFH infusion during administration of thrombo- evaluated thrombolysis for PE used IV UFH in lytic therapy (these two practices have never been www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e469S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians compared). During a 2-h infusion of 100 mg of tPA, whom thrombolysis has been unsuccessful. The US regulatory bodies recommend suspension of IV procedure is best performed on a warm, beating UFH, whereas IV UFH is continued during the tPA heart, without aortic cross-clamping, cardioplegia, infusion in many other countries. US authorities rec- or fi brillatory arrest. ommend checking the activated partial thrombo- No randomized trials or prospective observational plastin time immediately after completion of the tPA studies have evaluated surgical embolectomy in patients infusion and, provided that the activated antithrombin with acute PE. Consequently, evidence related to time is not . 80 s, restarting IV UFH without a bolus surgical embolectomy in patients with acute PE is of at the same infusion rate as before tPA was started. low quality, and our recommendations are weak. Recommendation 5.7 Catheter-Based Thrombus Removal for the Initial Treatment of PE 5.8. In patients with acute PE associated with Interventional catheterization techniques for mas- hypotension, we suggest surgical pulmonary sive PE include mechanical fragmentation of throm- embolectomy over no such intervention if bus with a standard pulmonary artery catheter, clot they have (i) contraindications to thrombolysis, pulverization with a rotating basket catheter, percu- (ii) failed thrombolysis or catheter-assisted taneous rheolytic thrombectomy, or pigtail rotational embolectomy, or (iii) shock that is likely to cause catheter embolectomy.329-336 Pharmacologic throm- death before thrombolysis can take effect (eg, bolysis and mechanical interventions are usually within hours), provided surgical expertise and combined unless bleeding risk is high. Catheter resources are available (Grade 2C). embolectomy does not result in extraction of intact pulmonary arterial thrombus; instead, clot fragments 5.9. Vena Caval Filters for the Initial are suctioned through the catheter or displaced dis- Treatment of PE tally with modest angiographic improvement. As previously noted in section 2.13, IVC fi lters No randomized trials have evaluated interventional can be used instead of initial anticoagulant therapy catheterization techniques for PE. Most observation in patients with acute PE if there is an unacceptable studies are retrospective series of , 30 patients.334 risk of bleeding or as an adjunct to anticoagulation. Consequently, evidence for the use of interventional As in DVT, no randomized trials or prospective catheter techniques in patients with acute PE is of cohort studies have evaluated IVC fi lters as sole low quality, and our recommendations are weak. therapy for acute PE (ie, without concurrent antico- Catheter selection, catheter deployment, and adjunc- agulation). As described in section 2.13, the PREPIC tive thrombolytic regimen should be based on local study, which evaluated IVC fi lters as an adjunct to expertise and resources. anticoagulation in 400 high-risk patients with proximal DVT, showed that fi lters reduced PE, increased Recommendation DVT, and did not change overall frequency of VTE 5.7. In patients with acute PE associated with (DVT and PE combined) or mortality146,149 ( Table 14 ; hypotension and who have (i) contraindications Table S19). to thrombolysis, (ii) failed thrombolysis, or (iii) The PREPIC study included 145 (36%) patients shock that is likely to cause death before systemic with symptomatic PE and 52 (13%) patients with thrombolysis can take effect (eg, within hours), asymptomatic PE at enrollment. If a patient has if appropriate expertise and resources are avail- an acute PE and a short-term contraindication to able, we suggest catheter-assisted thrombus anticoagulation, provided there is no proximal DVT removal over no such intervention (Grade 2C). on ultrasound, it is reasonable not to insert an IVC fi lter immediately; serial ultrasound examinations can be performed to ensure that the patient 5.8 Surgical Embolectomy for the Initial remains free of proximal DVT while anticoagulation Treatment of PE is withheld. Emergency surgical embolectomy with cardio- There is uncertainty about the risk and benefi ts of pulmonary bypass is another management strategy inserting IVC fi lters as an adjunct to anticoagulant for acute PE associated with hypotension. 337-340 and thrombolytic therapy in patients with PE and This operation is also suited for patients with acute hypotension. Among patients with hemodynamic PE who require surgical excision of a right atrial compromise in the International Cooperative Pulmo- thrombus, paradoxical arterial embolism, or clo- nary Embolism Registry, insertion of an IVC fi lter sure of a patent foramen ovale. Surgical embolec- was associated with a reduction of early recurrent tomy also can be performed to rescue patients in PE and death.280 Consequently, our recommendation e470S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians against insertion of an IVC fi lter in patients with acute with long-term VKA exclusively in patients who pre- PE who are treated with anticoagulants may not sented with PE.341,342 The two studies combined found apply to this select subgroup of patients. a similar frequency of recurrent VTE (enoxaparin, 4/60; VKA, 1/40) and major bleeding (enoxaparin, Recommendations 1/60; VKA, 2/40) with the two treatments.341 Of the 12 other studies that compared LMWH with VKA 5.9.1. In patients with acute PE who are treated therapy for long-term treatment of VTE (see section with anticoagulants, we recommend against the 3.3), only two173,227 included patients with PE. In use of an IVC fi lter (Grade 1B). these two studies, all patients had cancer, and 295 5.9.2. In patients with acute PE and contraindi- had PE (36% of all enrolled patients; some PE may cation to anticoagulation, we recommend the have been asymptomatic in one study227 ); subgroup use of an IVC fi lter (Grade 1B). analyses were not reported for the patients with PE. 5.9.3. In patients with acute PE and an IVC fi l- Dabigatran for the Long-term Treatment of PE: I n ter inserted as an alternative to anticoagulation, the one completed study that compared dabigatran we suggest a conventional course of anticoagu- with VKA therapy after initial parenteral therapy lant therapy if their risk of bleeding resolves ( Table 23 , Table S31), 786 (31%) patients had symp- (Grade 2B). tomatic PE at enrollment.343 Subgroup analysis did not suggest that patients with symptomatic PE have a Remarks: We do not consider that a permanent different response to dabigatran vs VKA therapy in IVC fi lter of itself is an indication for extended terms of either recurrent VTE or bleeding. anticoagulation. Rivaroxaban for the Long-term Treatment of PE: In the Einstein Extension study that compared rivar- 6.0 Long-term Treatment of PE oxaban with placebo after an initial period of long- term anticoagulation ( Table 22 , Table S30), 454 (38%) In the following sections, we emphasize studies patients had symptomatic PE at enrollment.88 Sub- that were performed exclusively in patients with PE group analysis did not suggest that patients with and patients with PE who were enrolled in other symptomatic PE had a different response to rivaroxa- studies. For the reasons noted in section 1.1, we make ban vs VKA therapy in terms of either recurrent VTE the same recommendations for long-term treatment or bleeding. of PE as for DVT and rate the quality of the underlying evidence as the same (see corresponding sec- Recommendations tions for treatment of DVT). 6.1. In patients with PE provoked by surgery, VKA for the Long-term Treatment of PE: There we recommend treatment with anticoagulation has been only one evaluation of duration of VKA for 3 months over (i) treatment of a shorter therapy exclusively in patients with PE. After period ( G r a d e 1 B ), (ii) treatment of a longer time- 3 months of initial treatment, patients with PE pro- limited period (eg, 6 or 12 months) ( G r a d e 1 B ), voked by a temporary risk factor were randomized to or (iii) extended therapy (Grade 1B regardless of stop or to receive 3 more months of therapy, and bleeding risk). those with unprovoked PE were randomized to stop or to receive 6 more months of therapy (WODIT PE 6.2. In patients with PE provoked by a nonsur- [Warfarin Optimal Duration Italian Trial in patients gical transient risk factor, we recommend treat- with Pulmonary Embolism]) (Table S24 and S25).194 ment with anticoagulation for 3 months over (i) Consistent with studies that included patients who treatment of a shorter period (Grade 1B), (ii) presented with DVT, extended VKA therapy was treatment of a longer time-limited period (eg, 6 effective while treatment was being received. How- or 12 months) (Grade 1B), and (iii) extended ever, extending the duration of treatment beyond therapy if there is a high bleeding risk ( Table 2 ) 3 months did not lower the rates of recurrence that (Grade 1B) . We suggest treatment with antico- were observed when anticoagulants were subsequently agulation for 3 months over extended therapy stopped. if there is a low or moderate bleeding risk ( Table 2 ) (Grade 2B) . LMWH for the Long-term Treatment of PE: Two small studies from the same investigator group have 6.3. In patients with an unprovoked PE, we compared long-term LMWH (enoxaparin 1 mg/kg recommend treatment with anticoagulation for days followed by 1.5 mg/kg SC daily) at least 3 months over treatment of a shorter 14 ف SC bid for www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e471S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians duration (Grade 1B). After 3 months of treat- LMWH, we suggest VKA over dabigatran or ment, patients with unprovoked PE should be rivaroxaban for long-term therapy (Grade 2C). evaluated for the risk-benefi t ratio of extended therapy. Remarks (6.6-6.7): Choice of treatment in patients with and without cancer is sensitive to individual 6.3.1. In patients with a fi rst VTE that is an patient tolerance for daily injections, need for labora- unprovoked PE and who have a low or moderate tory monitoring, and treatment costs. Treatment of bleeding risk ( Table 2 ), we suggest extended VTE with dabigatran or rivaroxaban, in addition to anticoagulant therapy over 3 months of therapy being less burdensome to patients, may prove to be (Grade 2B). associated with better clinical outcomes that VKA and LMWH therapy. When these guidelines were 6.3.2. In patients with a fi rst VTE that is an being prepared (October 2011), postmarketing stud- unprovoked PE and who have a high bleeding ies of safety were not available. Given the paucity of risk, we recommend 3 months of anticoagulant currently available data and that new data are rapidly therapy over extended therapy (Grade 1B). emerging, we give a weak recommendation in favor 6.3.3. In patients with a second unprovoked of VKA and LMWH therapy over dabigatran and VTE, we recommend extended anticoagulant rivaroxaban, and we have not made any recommen- therapy over 3 months of therapy in those who dations in favor of one of the new agents over the have a low bleeding risk ( Table 2 ) (Grade 1B), other. and we suggest extended anticoagulant therapy in those with a moderate bleeding risk ( Table 2 ) 6.8. In patients with PE who receive extended (Grade 2B). therapy, we suggest treatment with the same anticoagulant chosen for the fi rst 3 months 6.3.4. In patients with a second unprovoked (Grade 2C) . VTE who have a high bleeding risk ( Table 2 ), we suggest 3 months of therapy over extended 6.9 Treatment of Asymptomatic PE therapy (Grade 2B). of %1 ف Diagnosis of asymptomatic PE occurs in -of inpatients who have contrast %4 ف In patients with PE and active cancer, if outpatients and .6.4 the risk of bleeding is not high ( Table 2 ), we rec- enhanced CT scans, with a majority being in patients ommend extended anticoagulant therapy over with known malignancy.40,344-355 When PE is diagnosed 3 months of therapy ( G r a d e 1 B ) , and if there is a unexpectedly in patients with cancer, in retrospect, high bleeding risk ( Table 2 ), we suggest extended the clinical history may reveal symptoms that were anticoagulant therapy (Grade 2B). aggravated by the PE (eg, an increase in fatigue).345 About one-half of such incidental PE involve the Remarks: In all patients who receive extended lobar or more central pulmonary arteries, whereas the anticoagulant therapy, the continuing use of treat- other one-half are more distal.40,353 ment should be reassessed at periodic intervals When there is evidence of an asymptomatic PE, (eg, annually). the fi rst priority is to review the CT scans to determine whether the fi ndings are convincing for acute 6.5. In patients with PE who are treated with PE. Other recent CT scans may be available for com- VKA, we recommend a therapeutic INR range parison, or the current scan may also reveal DVT in of 2.0 to 3.0 (target INR of 2.5) over a lower the central deep veins (eg, subclavian vein, IVC, iliac (INR , 2) or higher (INR 3.0-5.0) range for all vein). If there is any uncertainty about the pres- treatment durations ( G r a d e 1B) . ence of an acute PE, additional diagnostic testing is required (eg, ultrasonography of the deep veins, 6.6. In patients with PE and no cancer, we suggest dedicated CT pulmonary angiography, D -dimer). VKA therapy over LMWH for long-term therapy Consistent with recommendations for the treat- (Grade 2C). For patients with PE and no cancer ment of asymptomatic DVT (section 3.5) in patients who are not treated with VKA therapy, we suggest in whom clinicians are convinced that an asymptom- LMWH over dabigatran or rivaroxaban for long- atic PE has occurred, based on moderate-quality term therapy ( G r a d e 2 C ). evidence, we suggest the same initial and long-term anticoagulation as for similar patients with symp- 6.7. In patients with PE and cancer, we suggest tomatic PE. The indication for anticoagulation is LMWH over VKA therapy (Grade 2B). In patients most compelling when the presence of PE is with PE and cancer who are not treated with unequivocal, PE involves the lobar and more central e472S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians pulmonary arteries, PE is a new fi nding on CT, There are no randomized trials of CTPH therapy ultrasound reveals proximal DVT, there are ongoing and, overall, evidence is low quality. There is, how- risk factors for VTE such as active cancer, and the ever, high-quality indirect evidence that anticoagu- patient is not at high risk for bleeding ( Table 2 ). lant therapy is very effective at preventing recurrent Many patients have left the hospital by the time an VTE in other patient populations (see section 3.1; incidental PE is reported. If PE is less extensive and Table 18 ; and Tables S24, S25, and S27). Conse- it would be diffi cult for patients to return the same quently, the evidence supporting long-term antico- day, it often is reasonable to defer further assessment agulation in patients with CTPH is of moderate and anticoagulant therapy until the next day. quality (rated down for indirectness). Features that are expected to be associated with greater benefi t with Recommendation pulmonary thromboendarterectomy include younger 6.9. In patients who are incidentally found to age, central disease, progressive clinical deteriora- have asymptomatic PE, we suggest the same ini- tion, and access to an expert multidisciplinary throm- tial and long-term anticoagulation as for similar boendarterectomy team. patients with symptomatic PE (Grade 2B). Recommendations 7.1.1. In patients with CTPH, we recommend 7.0 Chronic Thromboembolic extended anticoagulation over stopping therapy Pulmonary Hypertension (Grade 1B). Prospective studies suggest that CTPH occurs 7.1.2. In selected patients with CTPH, such as of patients who are treated for PE. 104,354-358 those with central disease under the care of an %3 ف in About one-third of patients have a history of VTE, experienced thromboendarterectomy team, we whereas two-thirds have had single or recurrent epi- suggest pulmonary thromboendarterectomy sodes of PE that were not diagnosed and may have over no pulmonary thromboendarterectomy been asymptomatic.359 Patients with CTPH are likely (Grade 2C) . to have a high risk of recurrent VTE because they have had previous VTE and have cardiopulmonary impairment. Recurrent VTE may be fatal more often 8.0 Superficial Vein Thrombosis in patients with severe cardiopulmonary impairment SVT has been less well studied than DVT but is than in those without such impairment . After PE ini- estimated to occur more often.371,372 It usually affects tiates CTPH, pulmonary vascular remodeling may the lower limbs; often involves a varicose vein; is cause severe pulmonary hypertension out of propor- associated with chronic venous insuffi ciency, malig- tion with pulmonary vascular thrombosis.104,359,360 nancy, thrombophilia, pregnancy or estrogen therapy, obesity, sclerotherapy, long-distance travel, and a his- 7.1 Pulmonary Thromboendarterectomy, tory of VTE; or may be unprovoked.371-373 The long Anticoagulant Therapy, and Vena Caval saphenous vein is involved in about two-thirds of Filter for the Treatment of CTPH lower-limb SVT.374 Primary therapy for CTPH is pulmonary throm- Although traditionally considered a benign disease, boendarterectomy, which, if successful, can reduce or a number of studies indicate that the consequences cure pulmonary hypertension and associated symp- of SVT may be more serious.371,372 A prospective study toms.104,359-367 The operation is lengthy and complex, of 844 patients with acute SVT of Ն 5 cm found that -of patients had symp %4 ف ,requiring a median sternotomy, cardio pulmonary at initial presentation bypass, deep hypothermia with periods of circu- tomatic PE, and routine ultrasound detected prox- latory arrest, and exploration of both pulmonary imal DVT in 10% and distal DVT in an additional arteries. At the most experienced centers, mortality 13% of patients. 374 In patients without VTE at presen- Management often includes inser- tation, despite 90% being treated with anticoagulant 367-104,360,363,365. % 5 ف is tion of a permanent IVC fi lter before or during pul- therapy (therapeutic doses in two-thirds, prophylac- monary endarterectomy and indefi nite anticoagulant tic doses in one-third, median duration of 11 days), therapy.359,361,363,365 Patients with CTPH who are not 3.1% developed symptomatic VTE (0.5% PE, 1.2% candidates for pulmonary endarterectomy because of proximal DVT, 1.4% distal DVT), 1.9% had recur- comorbid disease or surgically inaccessible lesions rent SVT (different location), and 3.3% had an exten- may be candidates for vasodilator therapy, balloon sion of SVT (same location) at 3 months. Male sex, pulmonary angioplasty, or lung transplantation and history of VTE, cancer, and absence of varicose veins may benefi t from referral to a center with expertise in each was associated with about a doubling of the risk pulmonary hypertension.359,363-365,368-370 of VTE during follow-up. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e473S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Given the high prevalence of concomitant prox- is low because there is no direct comparison in imal DVT in patients with SVT and the need to treat patients with SVT. Factors that favor the use of such patients with higher doses of anticoagulant anticoagulant therapy in patients with SVT (see therapy (ie, therapeutic doses), patients with SVT Recommendation 8.1.1) include: extensive SVT; above the knee should have ultrasonography to involvement above the knee, particularly if close exclude proximal DVT. Ultrasound can also help with to the saphenofemoral junction; severe symptoms; the diagnosis of SVT if the clinical presentation is involvement of the greater saphenous vein; history uncertain. With greater appreciation of the serious- of VTE or SVT; active cancer; and recent surgery. ness of SVT, investigators have evaluated anticoagu- An economic evaluation found that treatment of lant therapy, often in prophylactic or intermediate SVT with fondaparinux was not cost-effective; it cost doses, as a way to reduce acute symptoms, extension, $500,000 per quality-adjusted life year gained com- recurrence, and progression to VTE ( Table 31 , Tables pared with no treatment.383 S43-S45).375 Graduated compression stockings often are used in patients with SVT (eg, 83% of patients in the CALI- STO study). Oral nonsteroidal antiinfl ammatory 8.1 Treatment of SVT agents may be used to alleviate symptoms if patients Most studies that have evaluated anticoagulant are not treated with anticoagulants. Topical nonste- therapy for SVT have been small (eg, Յ 100 patients roidal antiinfl ammatory agents may reduce symptoms per treatment group), with additional methodologic and can be used with anticoagulant therapy. Sur- weaknesses376-381 (Tables S44, S45). Although these gical therapy, with ligation of the saphenofemoral studies suggest that prophylactic-dose LMWH, inter- junction or stripping of thrombosed superfi cial veins mediate-dose UFH or LMWH, warfarin therapy, and appears to be associated with higher rates of VTE oral nonsteroidal antiinfl ammatory agents are benefi - than treatment with anticoagulants.380,384,385 Anticoag- cial in patients with SVT, the supporting evidence is ulant therapy generally is not used to treat SVT that of low quality. The recently published Comparison of occurs in association with an IV infusion (ie, infusion ARIXTRA™ in lower LImb Superfi cial Thrombo- thrombophlebitis). phlebitis with placebo (CALISTO) study, which compared fondaparinux (2.5 mg/d for 45 days) with Recommendations Ն placebo in 3,000 patients with SVT ( 5 cm in length), 8.1.1. In patients with SVT of the lower limb has helped to clarify the role of anticoagulants for the of at least 5 cm in length, we suggest the use treatment of SVT ( Table 31 , Table S43), and the nat- of a prophylactic dose of fondaparinux or 382 ural history of this condition. LMWH for 45 days over no anticoagulation CALISTO found that fondaparinux is very effec- (Grade 2B). tive at reducing VTE, recurrent SVT, extension of SVT, and the need for venous surgery, and is associ- Remarks: Patients who place a high value on avoid- ated with little bleeding. In the placebo group, throm- ing the inconvenience or cost of anticoagulation and botic complications occurred more often if SVT a low value on avoiding infrequent symptomatic VTE involved the greater saphenous vein (92% of patients are likely to decline anticoagulation. in the control group), extended to within 10 cm from the saphenofemoral junction (9% of patients), and 8.1.2. In patients with SVT who are treated involved veins above the knee (46% of patients) and with anticoagulation, we suggest fondaparinux if VTE (7% of patients) or SVT (12% of patients) had 2.5 mg daily over a prophylactic dose of LMWH occurred previously.382 Age, sex, and presence of var- (Grade 2C). icose veins were not convincingly associated with the frequency of thrombotic complications, and there 9.0 Acute Upper-Extremity DVT were too few patients with cancer in CALISTO to assess that association. About 5% to 10% of VTE involve the upper ex tre m- The evidence is moderate quality. We have inter- ities.386-390 UEDVT includes two etiologic groups: preted the fi ndings of the CALISTO study as evi- primary (unprovoked with or without thrombophilia, dence for anticoagulation in general and assume effort-related and thoracic outlet syndrome) and sec- that prophylactic doses of LMWH and fondaparinux ondary (provoked by central venous catheters, pace- have similar antithrombotic effi cacy and safety. makers, or cancer). Secondary UEDVT accounts of cases.386,388,389,391-394 %75 ف Because it is direct and more extensive, the evidence for in support of fondaparinux is higher quality than the UEDVT involves the subclavian, axillary, or bra- evidence in support of LMWH. Quality of the evi- chial veins and may include extension to the bra- dence for comparison of fondaparinux with LMWH chiocephalic vein, superior vena cava, or the internal e474S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians Table 31—[Section 8.1] Summary of Findings: Fondaparinux vs Placebo for Acute SVTa-c,382

Anticipated Absolute Effects

No. of Participants Quality of the Relative Effect Risk With No Risk Difference With Outcomes (Studies), Follow-up Evidence (GRADE) (95% CI) Fondaparinux Fondaparinux (95% CI)

Mortality 3,002 (1 study), 3 mo Moderated-g due to RR 1.99 (0.18-21.87) 4 per 1,000h 4 more per 1,000 (from 3 imprecision fewer to 83 more) VTE 3,002 (1 study), 3 mo Highd RR 0.18 (0.06-0.53) 33 per 1,000h 27 fewer per 1,000 (from 16 fewer to 31 fewer) SVT recurrence 3,002 (1 study), 3 mo Highd RR 0.31 (0.14-0.68) 19 per 1,000h 13 fewer per 1,000 (from 6 fewer to 16 fewer) Major bleeding 2,987(1 study), 47 d Moderated,e,i due to RR 0.99 (0.06-15.86)e 1 per 1,000 0 fewer per 1,000 (from 1 imprecision fewer to 10 more) QOL not measured … … … … … The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confi dence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confi dence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confi dence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of abbreviations. aPatients with infusion-related SVT were excluded from CALISTO (Comparison of ARIXTRA in lower Limb Superfi cial Thrombophlebitis with Placebo). bFondaparinux 2.5 mg for 45 d. cPatients in the two treatment groups benefi ted from close clinical monitoring with adequate diagnostic procedures in the event of new or persistent symptoms. dAllocation concealed. Outcome adjudicators, steering committee, patients, providers, and data collectors blinded. Follow-up rate was 98%. Intention-to-treat analysis for effi cacy outcomes. Not stopped early for benefi t. eCI includes values suggesting large benefi t and values suggesting large harm. fWe rated down by only one level because of the low event rate and large sample size. gSmall number of events. hBaseline risk derived from a large prospective cohort study.374 iThe upper limit of the CI for absolute effect (10 more bleeds) is not low enough to suggest a clear balance of benefi ts vs harms. jugular vein. Clinical manifestations include acute not be removed. If the catheter is not functioning and and chronic arm pain, swelling, discoloration, and cannot be made to function (even after a period of dilated collateral veins over the arm, neck, or chest. systemic anticoagulation), it should be removed . UEDVT may lead to complications, including symp- As with treatment of leg DVT and PE, treatment of of patients387-389,393,395 ), recurrent UEDVT may be divided into acute (eg, parenteral %5 ف) tomatic PE at 5 years of follow-up389,390,395 ), and anticoagulants, thrombolytic therapy) and long-term %8 ف) UEDVT -of patients).386,387,395,396 In the phases (eg, anticoagulation, treatment of upper %20 ف ) PTS of the arm absence of a central venous catheter, the dominant extremity PTS). Because no randomized trials have arm is more often affected.387 Complications of evaluated treatment of UEDVT, recommendations are UEDVT are expected to occur much more often and based on indirect evidence from studies performed to be more severe when UEDVT involves the axillary in patients with leg DVT, observational studies (gen- or more-proximal veins than if thrombosis is confi ned erally small), and understanding of the natural his- to the brachial vein. In general, when we refer to tory of UEDVT. Therefore, quality of evidence is, at UEDVT, we are referring to thrombosis that involves best, moderate. the axillary or more-proximal veins. The most frequent risk factor for UEDVT is a cen- 9.1 Acute Anticoagulation for UEDVT tral venous catheter.386,388,397,398 If UEDVT occurs in association with a central venous catheter and the No randomized controlled studies have evaluated catheter is no longer required, it should be removed. acute anticoagulation for initial treatment of UEDVT. There are no data to guide whether catheter removal Several small prospective cohort studies have reported should be preceded by an initial period of anticoagu- low rates of recurrent DVT, PE, and major bleeding lation, and we do not have a preference for imme- when UEDVT was treated similarly to leg DVT diate or deferred removal. If UEDVT occurs in (Tables S46 and S47).395,399-401 Anticoagulant therapy association with a central venous catheter and there is is used to treat UEDVT because (1) UEDVT causes a continuing need for the catheter, the catheter need acute symptoms, can cause PE (including fatal www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e475S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians episodes), and is associated with PTS; (2) observa- the axillary vein, symptoms for , 14 days, good functional studies support its use; and (3) there is strong tional status, life expectancy of Ն 1 year, and low risk evidence for benefi t in patients with leg DVT. for bleeding ( Table 11 ). If thrombolysis is used, in Uncertainty exists about the need to prescribe antico- order to reduce the dose of thrombolytic therapy agulants to patients with thrombosis confi ned to the and the associated risk of bleeding, we encourage cath- brachial vein. Acceptable alternatives to full-dose anti- eter-based therapy over systemic thrombolysis. In coagulation in such patients include clinical or ultra- addition, because the balance of risks and benefi ts sound surveillance to detect extension of UEDVT with all forms of thrombolytic therapy is uncertain, while withholding anticoagulation, or treatment with anticoagulant therapy alone is acceptable initial ther- prophylactic-dose anticoagulation, or treatment with apy in all patients with UEDVT. There is no evidence therapeutic doses of anticoagulation for , 3 months. to suggest that thrombolysis reduces the risk of recur- We favor anticoagulation if isolated brachial vein rent VTE. thrombosis is symptomatic, associated with a central Resection of the fi rst rib has been advocated when venous catheter that will remain in place, or associ- UEDVT is believed to have been due to entrapment ated with cancer in the absence of a central venous of the subclavian vein as it passes between the clav- catheter. A high risk of bleeding argues against full- icle and the fi rst rib. 6,386,411-421 I n s e r t i o n o f a fi lter in dose anticoagulation. the superior vena cava has also been used in patients with acute UEDVT who cannot be given anticoagu- Recommendations lants. Complications, however, may be more than with IVC fi lters. 389,413,422 The evidence in support of 9.1.1. In patients with acute UEDVT that these procedures is of low quality, and because there involves the axillary or more proximal veins, is the potential to cause harm, their use should be we recommend acute treatment with paren- confi ned to exceptional circumstances in specialized teral anticoagulation (LMWH, fondaparinux, centers. IV UFH, or SC UFH) over no such acute treatment (Grade 1B). Recommendations 9.1.2. In patients with acute UEDVT that involves 9.2.1. In patients with acute UEDVT that the axillary or more proximal veins, we suggest involves the axillary or more proximal veins, we LMWH or fondaparinux over IV UFH (Grade 2C) suggest anticoagulant therapy alone over throm- and over SC UFH ( G r a d e 2 B ). bolysis (Grade 2C) .

9.2 Thrombolytic Therapy for the Remarks: Patients who (i) are most likely to beneﬁ t Initial Treatment of UEDVT from thrombolysis (see text); (ii) have access to CDT; (iii) attach a high value to prevention of PTS; and (iv) No randomized controlled studies have evaluated attach a lower value to the initial complexity, cost, thrombolytic therapy compared with anticoagula- and risk of bleeding with thrombolytic therapy are tion alone in patients with UEDVT. A number of likely to choose thrombolytic therapy over anticoagu- retrospective and small prospective observational lation alone. studies have evaluated streptokinase, urokinase, or rt-PA administered with varying doses, methods 9.2.2. In patients with UEDVT who undergo of administration (IV, catheter directed), and infu- thrombolysis, we recommend the same inten- sion durations.402-409 Three of these studies included sity and duration of anticoagulant therapy as in nonrandomized control groups who received antico- similar patients who do not undergo thromboly- agulation alone.402,407,408 In some studies, a few patients sis (Grade 1B). also had venous angioplasty405 o r s u r g i c a l d e c o m - pression402,405,408 (Tables S48 and S49). 9.3 Long-term Anticoagulation for UEDVT These studies suggest that thrombolysis can improve early and late venous patency but is associ- No randomized studies have evaluated duration or ated with increased bleeding. However, it is not intensity of long-term anticoagulation in patients known whether thrombolytic therapy reduces PTS of with UEDVT. In prospective observational studies, the arm or recurrent VTE. PTS of the arm appears to patients with UEDVT generally were treated with be a less common complication of thrombosis than VKA (target INR 2.5) for periods of 3 to 6 months.393,399- PTS of the leg.386-388,393,395,410 We believe that throm- 401,423 Rates of recurrent VTE and PTS varied (Tables bolysis should be considered only in patients who S50 and S51), but as previously noted, these rates meet all of the following criteria: severe symptoms, generally were lower than those observed in patients thrombus involving most of the subclavian vein and with leg DVT. e476S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians The factors that infl uence long-term anticoagu- eter remains over stopping after 3 months of lation in patients with leg DVT (section 3.1) are treatment in patients with cancer ( G r a d e 1 C ), relevant to long-term treatment of UEDVT. Some and we suggest this in patients with no cancer differences between UEDVT and leg DVT are wor- (Grade 2C) . thy of emphasis, especially that UEDVT often is associated with a central venous catheter that may or may 9.3.5. In patients who have UEDVT that is not not be removed (section 9.0). The most important associated with a central venous catheter or continuing risk factors for UEDVT are (1) the pres- with cancer, we recommend 3 months of anti- ence of a central venous catheter in the same arm and coagulation over a longer duration of therapy (2) active cancer in patients with UEDVT not associ- (Grade 1B). ated with a central venous catheter. Another important distinction between UEDVT 9.4 Prevention of PTS of the Arm and leg DVT relates to long-term anticoagulation in of patients after %20 ف patients with unprovoked thrombosis. Because the PTS of the arm occurs in risk of recurrent VTE is substantially lower in patients treatment for UEDVT38,395,396 a n d c a n b e a d i s a b l - with UEDVT compared with in those with proximal ing condition that adversely affects quality of life, leg DVT,387,388,393,395 we discourage extended anticoag- particularly if the dominant arm is involved. 410,424 ulant therapy (ie, beyond 3 months) in patients with No randomized trials have evaluated compression an unprovoked UEDVT. bandages, compression sleeves, or venoactive drugs No data are available for the long-term use of to prevent PTS after UEDVT. We have not con- LMWH monotherapy or newer anticoagulants for sidered indirect evidence from the legs for use of the long-term treatment of UEDVT. We make the compression therapy to prevent PTS of the arms same recommendations for choice of initial, long-term, because (1) the pathophysiology of PTS is believed and extended anticoagulant regimens for UEDVT to be different in the arms than in the legs (less as for leg DVT (recommendations 3.1.1, 3.1.2, and dependent venous hypertension), (2) arm sleeves 3.1.4) and note that the supporting evidence for these are more diffi cult to fi t than stockings, and (3) PTS weak recommendations is further weakened in this occurs less often after UEDVT than after leg DVT. population because of indirectness. Recommendation Recommendations 9.4. In patients with acute symptomatic UEDVT, 9.3.1. In most patients with UEDVT that is asso- we suggest against the use of compression sleeves ciated with a central venous catheter, we sug- or venoactive medications (Grade 2C). gest that the catheter not be removed if it is functional and there is an ongoing need for the 9.5 Treatment of PTS of the Arm catheter (Grade 2C). Symptoms of PTS of the arm include swelling, 9.3.2. In patients with UEDVT that involves the heaviness, and limb fatigue with exertion.395,410 No axillary or more proximal veins, we suggest a randomized trials have evaluated compression ban- minimum duration of anticoagulation of 3 months dages, compression sleeves (as are used for lym- over a shorter period (Grade 2B). phedema), or venoactive drugs to treat PTS after UEDVT. We considered anecdotal evidence that Remarks: This recommendation also applies if the compression therapy benefi ts some patients with PTS UEDVT is associated with a central venous catheter of the arm and that the benefi ts of a trial of compres- that was removed shortly after diagnosis. sion therapy will outweigh its harms and costs. There is no evidence that venoactive drugs are of benefi t in 9.3.3. In patients who have UEDVT that is asso- PTS of the arm. ciated with a central venous catheter that is removed, we recommend 3 months of antico- Recommendations agulation over a longer duration of therapy in patients with no cancer (Grade 1B) , and we sug- 9.5.1. In patients who have PTS of the arm, we gest this in patients with cancer (Grade 2C). suggest a trial of compression bandages or sleeves to reduce symptoms (Grade 2C). 9.3.4. In patients who have UEDVT that is associated with a central venous catheter that is not 9.5.2. In patients with PTS of the arm, we sug- removed, we recommend that anticoagulation gest against treatment with venoactive medica- is continued as long as the central venous cath- tions (Grade 2C). www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e477S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 10.0 Splanchnic Vein Thrombosis ernous transformation), progression of thrombosis on a follow-up imaging study, and ongoing cancer che- Thrombosis in the portal venous system, which motherapy. Esophageal varices secondary to acute includes the superior mesenteric, inferior mesen- portal vein thrombosis are not necessarily a contrain- teric, splenic, and portal veins, is collectively termed dication to anticoagulant therapy because such treat- splanchnic vein thrombosis . Depending on the loca tion ment may improve the portal hypertension. LMWH and extent of thrombosis, how rapidly thrombosis may be preferred over VKA if there is active malig- develops, speed and extent of thrombus recannula- nancy, liver disease, or thrombocytopenia.431 T h e p r e s - tion, presence of collateral portal venous drainage, ence of a reversible provoking factor for splanchnic and adequacy of arterial inflow, splanchnic vein vein thrombosis, such as intraabdominal sepsis or thrombosis may result in bowel or splenic infarction recent surgery, supports stopping anticoagulant therapy and chronic portal hypertension.425-430 Acute and after 3 months. Absence of a reversible risk factor chronic splanchnic vein thrombosis may be symp- (eg, “unprovoked” thrombosis or presence of a per- tomatic, but many episodes are detected incidentally sistent risk factor, such as myeloproliferative disease) in imaging studies performed for other indications, and a low risk of bleeding support extended anticoag- such as assessing response to surgical or medical ulant therapy. therapy in patients with cancer.348,429 Limited understanding of the natural history of both symptomatic Recommendations and incidentally detected splanchnic vein thrombosis in patients who are not treated with anticoagulants 10.1. In patients with symptomatic splanchnic (ie, frequency of bowel infarction, development of vein thrombosis (portal, mesenteric, and/or portal hypertension, recurrence), a paucity of data splenic vein thromboses), we recommend anti- from prospective cohort studies,428,431 and a lack of coagulation over no anticoagulation (Grade 1B). randomized trials of standardized anticoagulant therapy 10.2. In patients with incidentally detected for splanchnic vein thrombosis result in uncertainty splanchnic vein thrombosis (portal, mesenteric, about the role of anticoagulation for this condition. and/or splenic vein thromboses), we suggest no Increased risk of bleeding associated with esophageal anticoagulation over anticoagulation ( G r a d e 2 C ). varices (secondary to portal hypertension), thrombocytopenia (secondary to hypersplenism), and the pres- 11.0 Hepatic Vein Thrombosis ence of cirrhosis and malignancy (which predispose to splanchnic vein thrombosis) add to this uncertainty. Hepatic vein thrombosis, particularly Budd-Chiari A number of retrospective,425,429,430 and two pro- syndrome with occlusion of the main hepatic vein, spective,428,431 studies suggested that bowel ischemia can result in impairment of liver function and an is uncommon in patients with symptomatic splanchnic associated coagulopathy.432,433 Because there is limited vein thrombosis who are treated with anticoagulants understanding of the natural history of this condition -that recurrent venous thrombosis (both and a paucity of prospective studies that have evalu ,( 428% 2 ف ) involving the splanchnic and nonsplanchnic veins) is ated anticoagulant therapy, the role of anticoagulant common without anticoagulation or after stopping therapy is uncertain.432 -p e r y e a r425,429,430 ) , a n d t h a t a n t i - In a prospective registry of 163 patients with Budd % 5 ف) anticoagulation coagulation is effective at preventing progression Chiari syndrome of variable extent, of whom 86% and recurrent thrombosis,425,428-431 although it is asso- were treated with anticoagulation, one-half of patients ciated with an increased (particularly GI), but usually did not require invasive interventions (ie, transjugu- acceptable, frequency of bleeding.428-431 E f fi cacy of lar intrahepatic portosystemic shunting in 34% of all anticoagulant therapy in other forms of symptom- patients, surgical portosystemic shunting in 2% of atic venous thrombosis also provides indirect evi- patients, liver transplantation in 12% of patients), and dence for anticoagulation of patients with symptomatic survival was 82% after 2 years. 432 In an earlier retro- splanchnic vein thrombosis and, supported by the spective study of 237 patients with Budd-Chiari syn- previously noted observational studies, this evidence drome performed when use of surgical portosystemic is of moderate quality. We are not aware of studies of shunting was common (49% of patients), use of anti- treated or untreated asymptomatic splanchnic vein coagulant therapy (72% of patients) had no apparent thrombosis. effect on survival (RR, 1.05; 95% CI, 0.62-1.76). Factors that may encourage anticoagulant therapy Factors that encourage anticoagulant therapy in in patients with incidental splanchnic vein thrombo- patients with incidental hepatic vein thrombosis sis include extensive thrombosis that appears to be include extensive thrombosis that appears to be acute acute (eg, not present on a previous imaging study, (eg, not present on a previous imaging study, pres- presence of an intraluminal fi lling defect, lack of cav- ence of an intraluminal fi lling defect), progression e478S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians of thrombosis on a follow-up imaging study, and • Which patients with unprovoked VTE or cancer- ongoing cancer chemotherapy. Coagulopathy due to associated VTE have an unacceptable risk of liver dysfunction caused by hepatic vein thrombosis bleeding if they remain on extended anticoagu- is not a contraindication to anticoagulant therapy lant therapy? because anticoagulants may improve hepatic func- • How should risk of recurrent VTE if anticoagu- tion. LMWH usually will be preferred to VKA therapy lant therapy is stopped be balanced against risk when there is hepatic dysfunction and if there is of bleeding is anticoagulant therapy is continued? active malignancy. Presence of a reversible provoking • What is the preferred anticoagulant regimen for factor for hepatic vein thrombosis, such as oral con- the short- and long-term treatment of VTE in traceptive therapy, encourages a time-limited course patients with and without cancer? of therapy. Absence of a reversible risk factor encour- • Should patients receiving an incidental diagno- ages the use of extended therapy. Treatment of sis of asymptomatic VTE routinely be treated hepatic vein obstruction is complex and best under- with anticoagulant therapy, or should they have taken by a multidisciplinary team. serial testing to determine whether they have evolving DVT and only be treated if this is Recommendations detected? 11.1. In patients with symptomatic hepatic vein • Should patients with symptomatic DVT rou- thrombosis, we suggest anticoagulation over no tinely wear graduated compression stockings anticoagulation (Grade 2C) . from the time of diagnosis, or should stockings be used selectively (eg, in selected patients, in 11.2. In patients with incidentally detected patients whose symptoms do not rapidly resolve)? hepatic vein thrombosis, we suggest no antico- • Should patients with PE that causes right ven- agulation over anticoagulation (Grade 2C). tricular dysfunction be treated with anticoagulant therapy alone, or should they be treated with thrombolytic therapy? 12.0 Future Research • If patients have catheter-associated UEDVT Several questions in the treatment of VTE need to and the catheter is removed, should they be be answered. Current evidence relating to these treated with anticoagulant therapy or can they questions is of moderate or low quality. We list the be treated without anticoagulant therapy? questions roughly as they arise in this article rather • Can UEDVT be treated with less-intense or a than in order of importance. We do not present the shorter duration of anticoagulant therapy than rationale for each question because this is addressed leg DVT? in the corresponding sections of the article. We have confi ned ourselves to the primary question (eg, Should Acknowledgments patients with proximal DVT be treated with antico- Author contributions: As Topic Editor, Dr Akl oversaw the agulant therapy alone, or should they be treated development of this article, including the data analysis and subse- with pharmacomechanical CDT?); however, once quent development of the recommendations contained herein. the primary question is answered, we anticipate that Dr Kearon: served as Deputy Editor . Dr Akl: served as Topic Editor. secondary questions will need to be addressed (eg, Dr Comerota: served as a panelist. Which patients with proximal DVT should, or should Dr Prandoni: served as a panelist. not, be treated with CDT?). We are pleased to note Dr Bounameaux: served as a panelist. Dr Goldhaber: served as a panelist. that many of the these questions are being addressed Dr Nelson: served as a frontline clinician. in ongoing trials. Dr Wells: served as a panelist. Dr Gould: served as a resource consultant. • Should patients with an isolated distal DVT rou- Dr Dentali: served as a panelist. Dr Crowther: served as a panelist. tinely be treated with anticoagulant therapy, or Dr Kahn: served as a panelist. should they have serial testing to determine Financial/nonfi nancial disclosures: The authors of this guide- whether the DVT is extending and only be line provided detailed confl ict of interest information related to each individual recommendation made in this article. A grid of these treated if extension is detected? disclosures is available online at http://chestjournal.chestpubs. • Should patients with proximal DVT be treated org/content/141/2_suppl/e419S/suppl/DC1. In summary, the authors with anticoagulant therapy alone, or should they have reported to CHEST the following confl icts of interest: Dr Kearon is a consultant to Boehringer Ingelheim and has received peer-reviewed be treated with pharmacomechanical CDT? funding for studies in the treatment of VTE. Dr Bounameaux • Which patients with unprovoked proximal DVT has received grant monies and honoraria for consultancy or lec- or PE or cancer-associated VTE should stop tures from Bayer Schering Pharma AG; Daiichi-Sankyo Co, Ltd; Sanofi -Aventis LLC; Pfi zer Inc; Bristol-Myers Squibb; Servier; anticoagulant therapy at 3 months, and which and Canonpharma Production Ltd. Dr Goldhaber has received should remain on extended anticoagulant therapy? funds from Eisai Co, Ltd; EKOS Corporation; Sanofi -Aventis LLC; www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e479S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians and Johnson & Johnson. He serves as a consultant for Boeh- 5 . Monagle P , Chan AKC , Goldenberg NA , et al . Antithrombotic ringer Ingelheim GmbH; Bristol-Myers Squibb; Daiichi San- therapy in neonates and children: antithrombotic therapy kyo Co, Ltd; Eisai Co, Ltd; EKOS Corporation; Medscape; Merck and prevention of thrombosis, 9th ed: American College & Co, Inc; Portola Pharmaceuticals, Inc; and Sanofi -Aventis LLC. of Chest Physicians evidence-based clinical practice guide- Dr Wells has received peer-reviewed and investigator-initiated lines. Chest . 2 0 1 2 ; 1 4 1 ( 2 ) ( s u p p l ) : e737S-e801S. industry research funding for projects related to venous thrombosis treatment. He has received honoraria for industry-sponsored 6 . Kearon C , Kahn SR , Agnelli G , Goldhaber S , Raskob GE , (Bayer, Boehringer-Ingelheim, Pfi zer, BioMerieux, sanofi -aventis) Comerota AJ ; American College of Chest Physicians . Anti- talks pertaining to venous thrombosis and has attended advisory thrombotic therapy for venous thromboembolic disease: boards for Bayer, Boehringer-Ingelheim, Pfi zer and Bristol-Myers- American College of Chest Physicians evidence-based clinical Squibb. Dr Crowther has served on various advisory boards, has practice guidelines (8th Edition) . Chest . 2008 ; 133 ( 6 suppl ): assisted in the preparation of educational materials, has sat on 454S - 545S . data safety management boards, and his institution has received 7. Kearon C. Natural history of venous thromboembolism. research funds from the following companies: Leo Pharma A/S, Circulation . 2003 ;107(23)(suppl 1):I-22-I30. Pfi zer Inc, Boerhinger Ingelheim GmbH, Bayer Healthcare 8 . Stein PD , Matta F , Musani MH , Diaczok B . Silent pulmo- Pharmaceuticals, Octapharm AG, CSL Behring, and Artisan Pharma. Personal total compensation for these activities over the past 3 years nary embolism in patients with deep venous thrombosis: a totals less than US $10,000. Dr Kahn has received peer-reviewed systematic review . Am J Med . 2010 ; 123 ( 5 ): 426 - 431 . and investigator-initiated industry research funding for projects 9 . Murin S , Romano PS , White RH . Comparison of outcomes related to venous thrombosis and postthrombotic syndrome pre- after hospitalization for deep venous thrombosis or pulmo- vention and treatment. She has received honoraria for industry- nary embolism . Thromb Haemost . 2002 ; 88 ( 3 ): 407 - 414 . sponsored talks pertaining to venous thrombosis. Dr Akl is a 10 . Baglin T , Douketis J , Tosetto A , et al . Does the clinical member of and prominent contributor to the GRADE Working presentation and extent of venous thrombosis predict likeli- Group. Dr Comerota discloses that he is a consultant to and hood and type of recurrence? A patient-level meta-analysis . speaker for Covidien, Inc. Drs Prandoni, Nelson, Gould, and Dentali J Thromb Haemost . 2010 ; 8 ( 11 ): 2436 - 2442 . have reported that no potential confl icts of interest exist with any companies/organizations whose products or services may be dis- 11 . Douketis JD , Kearon C , Bates S , Duku EK , Ginsberg JS . cussed in this article . Risk of fatal pulmonary embolism in patients with treated Role of sponsors: The sponsors played no role in the develop- venous thromboembolism . JAMA . 1998 ; 279 ( 6 ): 458 - 462 . ment of these guidelines. Sponsoring organizations cannot recom- 12 . Carrier M , Le Gal G , Wells PS , Rodger MA . Systematic mend panelists or topics, nor are they allowed prepublication review: case-fatality rates of recurrent venous thromboem- access to the manuscripts and recommendations. Guideline panel bolism and major bleeding events among patients treated for members, including the chair, and members of the Health & Sci- venous thromboembolism . Ann Intern Med . 2010 ; 152 ( 9 ): ence Policy Committee are blinded to the funding sources. Fur- 578 - 589 . ther details on the Confl ict of Interest Policy are available online at http://chestnet.org . 13 . Linkins L , O’Donnell M , Julian JA , Kearon C . Intracranial Other contributions: We acknowledge the contributions of the and fatal bleeding according to indication for long-term writing groups of previous editions of this guideline and thank oral anticoagulant therapy . J Thromb Haemost . 2010 ; 8 ( 10 ): Gordon H. Guyatt , MD, FCCP, for his comments and suggestions 2201 - 2207 . and Aman Rajpal, MD, for his help with reviewing the proofs. 14 . Linkins LA, Choi PT , Douketis JD . Clinical impact of Endorsements: This guideline is endorsed by the American bleeding in patients taking oral anticoagulant therapy for Association for Clinical Chemistry, the American College of Clin- venous thromboembolism: a meta-analysis . Ann Intern Med . ical Pharmacy, the American Society of Health-System Pharma- 2003 ; 139 ( 11 ): 893 - 900 . cists, the American Society of Hematology, and the International 15 . MacLean S , Mulla S , Akl EA , et al . Patient values and pref- Society of Thrombosis and Hematosis. Additional information: The supplement Tables can be found erences for decision making in antithrombotic therapy: in the Online Data Supplement at http://chestjournal.chestpubs. antithrombotic therapy and prevention of thrombosis, org/content/141/2_suppl/e419S/suppl/DC1. 9th ed: American College of Chest Physicians evidence- based clinical practice guidelines. Chest . 2012;141(2)(suppl): e1S-e23S . References 16 . Guyatt GH , Norris SL , Schulman S , et al . Methodology for 1 . Garcia DA , Baglin TP , Weitz JI , Samama MM . Parenteral the development of antithrombotic therapy and prevention anticoagulants: antithrombotic therapy and prevention of of thrombosis guidelines: antithrombotic therapy and pre- thrombosis, 9th ed: American College of Chest Physicians vention of thrombosis, 9th ed: American College of Chest evidence-based clinical practice guidelines . Chest . 2012 ; Physicians evidence-based clinical practice guidelines. Chest . 141(2)(suppl):e24S-e43S. 2012;141(2)(suppl):53S-70S. 2 . Ageno W , Gallus AS , Wittkowsky A , Crowther M , Hylek EM , 17 . van der Meer FJ , Rosendaal FR , Vandenbroucke JP , Palareti G . Oral anticoagulant therapy: antithrombotic Briët E . Bleeding complications in oral anticoagulant ther- therapy and prevention of thrombosis, 9th ed: American apy. An analysis of risk factors. Arch Intern Med . 1993 ; College of Chest Physicians evidence-based clinical practice 153 ( 13 ): 1557 - 1562 . guidelines . Chest . 2012 ;141(2)(suppl):e44S-e88S. 18 . Beyth RJ , Quinn LM , Landefeld CS . Prospective evaluation 3 . Holbrook A , Schulman S , Witt DM , et al . Evidence-based of an index for predicting the risk of major bleeding in out- management of anticoagulant therapy: antithrombotic ther- patients treated with warfarin . Am J Med . 1998 ; 105 ( 2 ): 91 - 99 . apy and prevention of thrombosis, 9th ed: American College 19 . Douketis JD , Arneklev K , Goldhaber SZ , Spandorfer J , of Chest Physicians evidence-based clinical practice guide- Halperin F , Horrow J . Comparison of bleeding in patients lines . Chest . 2012 ;141(2)(suppl):e152S-e184S. with nonvalvular atrial fi brillation treated with ximelagatran 4 . Bates SM , Greer IA , Middeldorp S , Veenstra DL , Prabulos A-M , or warfarin: assessment of incidence, case-fatality rate, time Vandvik PO . VTE, thrombophilia, antithrombotic therapy, course and sites of bleeding, and risk factors for bleeding . and pregnancy: anti thrombotic therapy and prevention of Arch Intern Med . 2006 ; 166 ( 8 ): 853 - 859 . thrombosis, 9th ed: American College of Chest Physicians 20 . Kuijer PMM , Hutten BA , Prins MH , Büller HR . Prediction of evidence-based clinical practice guidelines. Chest . 2 0 1 2 ; the risk of bleeding during anticoagulant treatment for venous 141(2)(suppl):e691S-e736S. thromboembolism . Arch Intern Med . 1999 ; 159 ( 5 ): 457 - 460 . e480S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 21 . Landefeld CS , McGuire E III , Rosenblatt MW . A bleeding 36 . Nieto JA , Solano R , Ruiz-Ribó MD , et al ; Riete Investigators . risk index for estimating the probability of major bleeding Fatal bleeding in patients receiving anticoagulant therapy in hospitalized patients starting anticoagulant therapy . Am J for venous thromboembolism: fi ndings from the RIETE Med . 1990 ; 89 ( 5 ): 569 - 578 . registry . J Thromb Haemost . 2010 ; 8 ( 6 ): 1216 - 1222 . 22 . Palareti G , Leali N , Coccheri S , et al ; Italian Study on 37 . Hutten BA , Prins MH , Gent M , Ginsberg J , Tijssen JG , Complications of Oral Anticoagulant Therapy . Bleeding Büller HR . Incidence of recurrent thromboembolic and complications of oral anticoagulant treatment: an inception- bleeding complications among patients with venous throm- cohort, prospective collaborative study (ISCOAT) . Lancet . boembolism in relation to both malignancy and achieved 1996 ; 348 ( 9025 ): 423 - 428 . international normalized ratio: a retrospective analysis . 23 . Torn M , Bollen WL , van der Meer FJ , van der Wall EE , J Clin Oncol . 2000 ; 18 ( 17 ): 3078 - 3083 . Rosendaal FR . Risks of oral anticoagulant therapy with 38 . Prandoni P , Lensing AW , Piccioli A , et al . Recurrent venous increasing age . Arch Intern Med . 2005 ; 165 ( 13 ): 1527 - 1532 . thromboembolism and bleeding complications during 24 . White RH , Beyth RJ , Zhou H , Romano PS . Major bleeding anticoagulant treatment in patients with cancer and venous after hospitalization for deep-venous thrombosis . Am J Med . thrombosis . Blood . 2002 ; 100 ( 10 ): 3484 - 3488 . 1999 ; 107 ( 5 ): 414 - 424 . 39 . Hylek EM , Singer DE . Risk factors for intracranial hem- 25 . Olesen JB , Lip GY , Hansen PR , et al . Bleeding risk in ‘real orrhage in outpatients taking warfarin . Ann Intern Med . world’ patients with atrial fi brillation: comparison of two 1994 ; 120 ( 11 ): 897 - 902 . established bleeding prediction schemes in a nationwide 40 . Dentali F , Ageno W , Becattini C , et al . Prevalence and clin- cohort . J Thromb Haemost . 2011 ; 9 ( 8 ): 1460 - 1467 . ical history of incidental, asymptomatic pulmonary embo- 26 . Fihn SD , Callahan CM , Martin DC , McDonell MB , Henikoff lism: a meta-analysis . Thromb Res . 2010 ; 125 ( 6 ): 518 - 522 . JG , White RH ; The National Consortium of Anticoagulation 41 . Hull RD , Raskob GE , Rosenbloom D , et al . Heparin for Clinics . The risk for and severity of bleeding complications 5 days as compared with 10 days in the initial treatment of in elderly patients treated with warfarin . Ann Intern Med . proximal venous thrombosis . N Engl J Med . 1990 ; 322 ( 18 ): 1996 ; 124 ( 11 ): 970 - 979 . 1260 - 1264 . 27 . Gage BF , Yan Y , Milligan PE , et al . Clinical classifi cation 42 . Dahri K , Loewen P . The risk of bleeding with warfarin: a schemes for predicting hemorrhage: results from the systematic review and performance analysis of clinical pre- National Registry of Atrial Fibrillation (NRAF) . Am Heart J . diction rules . Thromb Haemost . 2007 ; 98 ( 5 ): 980 - 987 . 2006 ; 151 ( 3 ): 713 - 719 . 43 . P a l a r e t i G , C o s m i B . B l e e d i n g with anticoagulation therapy- 28 . Lip GY , Frison L , Halperin JL , Lane DA . Comparative who is at risk, and how best to identify such patients . Thromb validation of a novel risk score for predicting bleeding risk Haemost . 2009 ; 102 ( 2 ): 268 - 278 . in anticoagulated patients with atrial fi brillation: the HAS- 44 . Kearon C , Ginsberg JS , Kovacs MJ , et al ; Extended BLED (Hypertension, Abnormal Renal/Liver Function, Low-Intensity Anticoagulation for Thrombo-Embolism Stroke, Bleeding History or Predisposition, Labile INR, Investigators . Comparison of low-intensity warfarin therapy Elderly, Drugs/Alcohol Concomitantly) score . J Am Coll with conventional-intensity warfarin therapy for long-term Cardiol . 2011 ; 57 ( 2 ): 173 - 180 . prevention of recurrent venous thromboembolism . N Engl J 29 . Nieto JA , Bruscas MJ , Ruiz-Ribo D , et al ; RIETE Med . 2003 ; 349 ( 7 ): 631 - 639 . Investigators . Acute venous thromboembolism in patients 45 . Collins R , MacMahon S , Flather M , et al . Clinical effects with recent major bleeding. The infl uence of the site of of anticoagulant therapy in suspected acute myocardial bleeding and the time elapsed on outcome. J Thromb infarction: systematic overview of randomised trials . BMJ . Haemost . 2006 ; 4 ( 11 ): 2367 - 2372 . 1996 ; 313 ( 7058 ): 652 - 659 . 30 . Ruíz-Giménez N , Suárez C , González R , et al ; RIETE 46 . Yusuf S , Mehta SR , Xie C , et al ; CREATE Trial Group Investigators . Predictive variables for major bleeding events Investigators . Effects of reviparin, a low-molecular-weight in patients presenting with documented acute venous heparin, on mortality, reinfarction, and strokes in patients thromboembolism. Findings from the RIETE Registry . with acute myocardial infarction presenting with ST-segment Thromb Haemost . 2008 ; 100 ( 1 ): 26 - 31 . elevation . JAMA . 2005 ; 293 ( 4 ): 427 - 435 . 31 . van der Meer FJ , Rosendaal FR , Vandenbroucke JP , Briët E . 47 . Prandoni P , Lensing AWA , Cogo A , et al . The long-term Assessment of a bleeding risk index in two cohorts of clinical course of acute deep venous thrombosis . Ann Intern patients treated with oral anticoagulants . Thromb Haemost . Med . 1996 ; 125 ( 1 ): 1 - 7 . 1996 ; 76 ( 1 ): 12 - 16 . 48 . Palareti G , Cosmi B , Legnani C , et al ; PROLONG 32 . Pengo V , Legnani C , Noventa F , Palareti G ; ISCOAT Investigators . D-dimer testing to determine the duration Study Group.(Italian Study on Complications of Oral Anti- of anticoagulation therapy . N Engl J Med . 2006 ; 355 ( 17 ): coagulant Therapy) . Oral anticoagulant therapy in patients 1780 - 1789 . with nonrheumatic atrial fi brillation and risk of bleeding. 49 . Wells PS , Forgie MA , Simms M , et al . The outpatient bleeding A Multicenter Inception Cohort Study . Thromb Haemost . risk index: validation of a tool for predicting bleeding rates 2001 ; 85 ( 3 ): 418 - 422 . in patients treated for deep venous thrombosis and pulmo- 33 . Fang MC , Go AS , Chang Y , et al . A new risk scheme to pre- nary embolism . Arch Intern Med . 2003 ; 163 ( 8 ): 917 - 920 . dict warfarin-associated hemorrhage: The ATRIA (Anti- 50 . Barritt DW , Jordan SC . Anticoagulant drugs in the treat- coagulation and Risk Factors in Atrial Fibrillation) Study . ment of pulmonary embolism. A controlled trial . Lancet . J Am Coll Cardiol . 2011 ; 58 ( 4 ): 395 - 401 . 1960 ; 1 ( 7138 ): 1309 - 1312 . 34 . Shireman TI , Mahnken JD , Howard PA , Kresowik TF , 51 . Brandjes DPM , Heijboer H , Büller HR , de Rijk M , Jagt H , Hou Q , Ellerbeck EF . Development of a contemporary ten Cate JW . Acenocoumarol and heparin compared with bleeding risk model for elderly warfarin recipients . Chest . acenocoumarol alone in the initial treatment of proximal- 2006 ; 130 ( 5 ): 1390 - 1396 . vein thrombosis . N Engl J Med . 1992 ; 327 ( 21 ): 1485 - 1489 . 35 . Fihn SD , McDonell M , Martin D , et al ; Warfarin Optimized 52 . Hull RD , Raskob GE , Hirsh J , et al . Continuous intravenous Outpatient Follow-up Study Group . Risk factors for compli- heparin compared with intermittent subcutaneous heparin cations of chronic anticoagulation. A multicenter study . Ann in the initial treatment of proximal-vein thrombosis . N Engl Intern Med . 1993 ; 118 ( 7 ): 511 - 520 . J Med . 1986 ; 315 ( 18 ): 1109 - 1114 . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e481S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 53 . Raschke RA , Reilly BM , Guidry JR , Fontana JR , Srinivas S . College of Chest Physicians evidence-based clinical prac- The weight-based heparin dosing nomogram compared tice guidelines. Chest . 2012;141(2)(suppl):e495S-e530S. with a “standard care” nomogram. A randomized controlled 70 . Prandoni P , Carnovali M , Marchiori A ; Galilei Investigators . trial . Ann Intern Med . 1993 ; 119 ( 9 ): 874 - 881 . Subcutaneous adjusted-dose unfractionated heparin vs fi xed- 54 . Wells PS , Hirsh J , Anderson DR , et al . A simple clinical dose low-molecular-weight heparin in the initial treatment model for the diagnosis of deep-vein thrombosis combined of venous thromboembolism . Arch Intern Med . 2 0 0 4 ; with impedance plethysmography: potential for an improve- 164 ( 10 ): 1077 - 1083 . ment in the diagnostic process . J Intern Med . 1998 ; 243 ( 1 ): 71 . Kearon C , Ginsberg JS , Julian JA , et al ; Fixed-Dose Heparin 15 - 23 . (FIDO) Investigators . Comparison of fi xed-dose weight- 55 . Le Gal G , Righini M , Roy PM , et al . Prediction of pulmo- adjusted unfractionated heparin and low-molecular-weight nary embolism in the emergency department: the revised heparin for acute treatment of venous thromboembolism . Geneva score . Ann Intern Med . 2006 ; 144 ( 3 ): 165 - 171 . JAMA . 2006 ; 296 ( 8 ): 935 - 942 . 56 . Wells PS , Anderson DR , Rodger M , et al . Derivation of a 72 . Dolovich LR , Ginsberg JS , Douketis JD , Holbrook AM , simple clinical model to categorize patients probability of Cheah G . A meta-analysis comparing low-molecular-weight pulmonary embolism: increasing the models utility with the heparins with unfractionated heparin in the treatment of SimpliRED D-dimer . Thromb Haemost . 2000 ; 83 ( 3 ): 416 - 420 . venous thromboembolism: examining some unanswered 57 . Bates SM , Jaeschke R , Stevens SM , et al . Diagnosis of DVT: questions regarding location of treatment, product type, and antithrombotic therapy and prevention of thrombosis, 9th ed: dosing frequency . Arch Intern Med . 2000 ; 160 ( 2 ): 181 - 188 . American College of Chest Physicians evidence-based clin- 73 . Gould MK , Dembitzer AD , Doyle RL , Hastie TJ , Garber AM . ical practice guidelines. Chest . 2012;141(2)(suppl):e351S-e418S. Low-molecular-weight heparins compared with unfraction- 58 . Johnson SA , Stevens SM , Woller SC , et al . Risk of deep ated heparin for treatment of acute deep venous thrombosis. vein thrombosis following a single negative whole-leg com- A meta-analysis of randomized, controlled trials . Ann Intern pression ultrasound: a systematic review and meta-analysis . Med . 1999 ; 130 ( 10 ): 800 - 809 . JAMA . 2010 ; 303 ( 5 ): 438 - 445 . 74. van Dongen CJJ , van den Belt AG , Prins MH , Lensing AW. 59 . Bernardi E , Camporese G , Büller HR , et al ; Erasmus Study Fixed dose subcutaneous low molecular weight heparins Group . Serial 2-point ultrasonography plus D-dimer vs whole- versus adjusted dose unfractionated heparin for venous leg color-coded Doppler ultrasonography for diagnosing thromboembolism. Cochrane Database Syst Rev . 2004;(4): suspected symptomatic deep vein thrombosis: a randomized CD001100. controlled trial . JAMA . 2008 ; 300 ( 14 ): 1653 - 1659 . 75 . Erkens PM , Prins MH . Fixed dose subcutaneous low molec- 60 . Masuda EM , Kistner RL . The case for managing calf vein ular weight heparins versus adjusted dose unfractionated thrombi with duplex surveillance and selective anticoagula- heparin for venous thromboembolism . Cochrane Database tion . Dis Mon . 2010 ; 56 ( 10 ): 601 - 613 . Syst Rev . 2010 ;( 9 ): CD001100 . 61 . Palareti G , Cosmi B , Lessiani G , et al . Evolution of 76 . Faivre R , Neuhart Y , Kieffer Y , et al . [A new treatment of untreated calf deep-vein thrombosis in high risk symptom- deep venous thrombosis: low molecular weight heparin frac- atic outpatients: the blind, prospective CALTHRO study . tions. Randomized study] . Presse Med . 1988 ; 17 ( 5 ): 197 - 200 . Thromb Haemost . 2010 ; 104 ( 5 ): 1063 - 1070 . 77 . Lopaciuk S , Meissner AJ , Filipecki S , et al . Subcutaneous 62 . Righini M , Paris S , Le Gal G , Laroche JP , Perrier A , low molecular weight heparin versus subcutaneous unfrac- Bounameaux H . Clinical relevance of distal deep vein tionated heparin in the treatment of deep vein thrombosis: thrombosis. Review of literature data . Thromb Haemost . a Polish multicenter trial . Thromb Haemost . 1992 ; 68 ( 1 ): 2006 ; 95 ( 1 ): 56 - 64 . 14 - 18 . 63 . Macdonald PS , Kahn SR , Miller N , Obrand D . Short-term 78 . Büller HR , Davidson BL , Decousus H , et al ; Matisse natural history of isolated gastrocnemius and soleal vein Investigators . Fondaparinux or enoxaparin for the initial thrombosis . J Vasc Surg . 2003 ; 37 ( 3 ): 523 - 527 . treatment of symptomatic deep venous thrombosis: a ran- 64 . Parisi R , Visonà A , Camporese G , et al . Isolated distal deep domized trial . Ann Intern Med . 2004 ; 140 ( 11 ): 867 - 873 . vein thrombosis: effi cacy and safety of a protocol of treat- 79 . Büller HR , Davidson BL , Decousus H , et al ; Matisse ment. Treatment of Isolated Calf Thrombosis (TICT) Study . Investigators . Subcutaneous fondaparinux versus intrave- Int Angiol . 2009 ; 28 ( 1 ): 68 - 72 . nous unfractionated heparin in the initial treatment of pul- 65 . Schwarz T , Buschmann L , Beyer J , Halbritter K , Rastan A , monary embolism . N Engl J Med . 2003 ; 349 ( 18 ): 1695 - 1702 . Schellong S . Therapy of isolated calf muscle vein throm- 80 . Couturaud F , Julian JA , Kearon C . Low molecular weight bosis: a randomized, controlled study . J Vasc Surg . 2010 ; heparin administered once versus twice daily in patients 52 ( 5 ): 1246 - 1250 . with venous thromboembolism: a meta-analysis . Thromb 66 . Lagerstedt CI , Olsson CG , Fagher BO , Oqvist BW , Haemost . 2001 ; 86 ( 4 ): 980 - 984 . Albrechtsson U . Need for long-term anticoagulant treat- 81 . van Dongen CJ , MacGillavry MR , Prins MH . Once ver- ment in symptomatic calf-vein thrombosis . Lancet . 1985 ; sus twice daily LMWH for the initial treatment of venous 2 ( 8454 ): 515 - 518 . thromboembolism . Cochrane Database Syst Rev . 2005 ;( 3 ): 67 . Gallus AS , Jackaman J , Tillett J , Mills W , Wycherley A . CD003074 . Safety and effi cacy of warfarin started early after submas- 82 . Breddin HK , Hach-Wunderle V , Nakov R , Kakkar VV ; sive venous thrombosis or pulmonary embolism . Lancet . CORTES Investigators. Clivarin: Assessment of Regression 1986 ; 2 ( 8519 ): 1293 - 1296 . of Thrombosis, Efficacy, and Safety . Effects of a low- 68 . Leroyer C , Bressollette L , Oger E , et al ; The ANTENOX molecular-weight heparin on thrombus regression and recur- Study Group . Early versus delayed introduction of oral rent thromboembolism in patients with deep-vein thrombosis . vitamin K antagonists in combination with low-molecular- N Engl J Med . 2001 ; 344 ( 9 ): 626 - 631 . weight heparin in the treatment of deep vein thrombosis. a 83 . Charbonnier BA , Fiessinger JN , Banga JD , Wenzel E , randomized clinical trial . Haemostasis . 1998 ; 28 ( 2 ): 70 - 77 . d’Azemar P , Sagnard L . Comparison of a once daily with a 69 . Linkins L-A , Dans A L, Moores LK, et al. Treatment and pre- twice daily subcutaneous low molecular weight heparin reg- vention of heparin-induced thrombocytopenia: antithrombotic imen in the treatment of deep vein thrombosis. FRAXODI therapy and prevention of thrombosis, 9th ed: American group . Thromb Haemost . 1998 ; 79 ( 5 ): 897 - 901 . e482S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 84 . Holmoström M , Berglund MC , Granquist S , Bratt G , enoxaparin vs inpatient treatment with unfractionated hep- Törnebohm E , Lockner D . Fragmin once or twice daily arin . Chest . 2002 ; 122 ( 1 ): 108 - 114 . subcutaneously in the treatment of deep venous thrombosis 99 . Tillman DJ , Charland SL , Witt DM . Effectiveness and of the leg . Thromb Res . 1992 ; 67 ( 1 ): 49 - 55 . economic impact associated with a program for outpatient 85 . Merli G , Spiro TE , Olsson CG , et al ; Enoxaparin Clinical management of acute deep vein thrombosis in a group Trial Group . Subcutaneous enoxaparin once or twice daily model health maintenance organization . Arch Intern Med . compared with intravenous unfractionated heparin for 2000 ; 160 ( 19 ): 2926 - 2932 . treatment of venous thromboembolic disease . Ann Intern 100 . Rodger M , Bredeson C , Wells PS , Beck J , Kearns B , Med . 2001 ; 134 ( 3 ): 191 - 202 . Huebsch LB . Cost-effectiveness of low-molecular-weight 86 . Partsch H , Kechavarz B , Mostbeck A , Köhn H , Lipp C . heparin and unfractionated heparin in treatment of deep Frequency of pulmonary embolism in patients who have vein thrombosis . CMAJ . 1998 ; 159 ( 8 ): 931 - 938 . iliofemoral deep vein thrombosis and are treated with once- 101 . van den Belt AG , Bossuyt PM , Prins MH , Gallus AS , or twice-daily low-molecular-weight heparin . J Vasc Surg . Büller HR ; TASMAN Study Group . Replacing inpatient 1996 ; 24 ( 5 ): 774 - 782 . care by outpatient care in the treatment of deep venous 87 . Siegbahn A , Y-Hassan S , Boberg J , et al . Subcutaneous thrombosis—an economic evaluation . Thromb Haemost . treatment of deep venous thrombosis with low molecular 1998 ; 79 ( 2 ): 259 - 263 . weight heparin. A dose fi nding study with LMWH-Novo . 102 . Kahn SR , Shrier I , Julian JA , et al . Determinants and Thromb Res . 1989 ; 55 ( 6 ): 767 - 778 . time course of the postthrombotic syndrome after acute 88 . Bauersachs R , Berkowitz SD , Brenner B , et al ; EINSTEIN deep venous thrombosis . Ann Intern Med . 2008 ; 149 ( 10 ): Investigators . Oral rivaroxaban for symptomatic venous 698 - 707 . thromboembolism . N Engl J Med . 2010 ; 363 ( 26 ): 2499 - 2510 . 103 . Sharifi M , Mehdipour M , Bay C , Smith G , Sharifi J . 89 . Boccalon H , Elias A , Chalé JJ , Cadène A , Gabriel S . Clinical Endovenous therapy for deep venous thrombosis: the outcome and cost of hospital vs home treatment of proximal TORPEDO trial . Catheter Cardiovasc Interv . 2010 ; 76 ( 3 ): deep vein thrombosis with a low-molecular-weight hep- 316 - 325 . arin: the Vascular Midi-Pyrenees study . Arch Intern Med . 104 . Jaff MR , McMurtry MS , Archer SL , et al. ; American Heart 2000 ; 160 ( 12 ): 1769 - 1773 . Association Council on Cardiopulmonary, Critical Care, 90 . Chong BH , Brighton TA , Baker RI , Thurlow P , Lee CH ; Perioperative and Resuscitation; American Heart Association ASTH DVT Study Group . Once-daily enoxaparin in the Council on Peripheral Vascular Disease; American Heart outpatient setting versus unfractionated heparin in hospital Association Council on Arteriocsclerosis, Thrombosis and for the treatment of symptomatic deep-vein thrombosis . J Vascular Biology. Management of massive and submassive Thromb Thrombolysis . 2005 ; 19 ( 3 ): 173 - 181 . pulmonary embolism, iliofemoral deep vein thrombosis, 91 . Daskalopoulos ME , Daskalopoulou SS , Tzortzis E , et al . and chronic thromboembolic pulmonary hypertension: a Long-term treatment of deep venous thrombosis with a low scientifi c statement from the American Heart Association. molecular weight heparin (tinzaparin): a prospective ran- Circulation . 2011;123(16):1788-1830. domized trial . Eur J Vasc Endovasc Surg . 2005 ; 29 ( 6 ): 638 - 650 . 105 . Elsharawy M , Elzayat E . Early results of thrombolysis 92 . Koopman MMW , Prandoni P , Piovella F , et al ; The Tasman vs anticoagulation in iliofemoral venous thrombosis. A ran- Study Group . Treatment of venous thrombosis with intrave- domised clinical trial . Eur J Vasc Endovasc Surg . 2 0 0 2 ; 2 4 ( 3 ) : nous unfractionated heparin administered in the hospital as 209 - 214 . compared with subcutaneous low-molecular-weight hep- 106 . Enden T , Kløw NE , Sandvik L , et al ; CaVenT study group . arin administered at home. N Engl J Med . 1 9 9 6 ; 3 3 4 ( 1 1 ) : Catheter-directed thrombolysis vs. anticoagulant therapy 682 - 687 . alone in deep vein thrombosis: results of an open random- 93 . Levine M , Gent M , Hirsh J , et al . A comparison of low- ized, controlled trial reporting on short-term patency . J molecular-weight heparin administered primarily at home Thromb Haemost . 2009 ; 7 ( 8 ): 1268 - 1275 . with unfractionated heparin administered in the hospi- 107 . Enden T , Sandvik L , Kløw NE , et al . Catheter-directed tal for proximal deep-vein thrombosis . N Engl J Med . 1996 ; venous thrombolysis in acute iliofemoral vein thrombosis— 334 ( 11 ): 677 - 681 . the CaVenT study: rationale and design of a multicenter, 94 . Ramacciotti E , Araújo GR , Lastoria S , et al ; CLETRAT randomized, controlled, clinical trial (NCT00251771) . Am Investigators . An open-label, comparative study of the Heart J . 2007 ; 154 ( 5 ): 808 - 814 . effi cacy and safety of once-daily dose of enoxaparin versus 108 . Douketis JD , Foster GA , Crowther MA , Prins MH , Ginsberg unfractionated heparin in the treatment of proximal lower JS . Clinical risk factors and timing of recurrent venous limb deep-vein thrombosis . Thromb Res . 2004 ; 114 ( 3 ): 149 - 153 . thromboembolism during the initial 3 months of anticoagu- 95 . Bäckman K , Carlsson P , Kentson M , Hansen S , Engquist L , lant therapy . Arch Intern Med . 2000 ; 160 ( 22 ): 3431 - 3436 . Hallert C . Deep venous thrombosis: a new task for pri- 109 . Comerota AJ , Throm RC , Mathias SD , Haughton S , mary health care. A randomised economic study of outpa- Mewissen M . Catheter-directed thrombolysis for iliofemo- tient and inpatient treatment . Scand J Prim Health Care . ral deep venous thrombosis improves health-related quality 2004 ; 22 ( 1 ): 44 - 49 . of life . J Vasc Surg . 2000 ; 32 ( 1 ): 130 - 137 . 96 . O’Brien B , Levine M, Willan A , et al . Economic evalua- 110 . Brass LM , Lichtman JH , Wang Y , Gurwitz JH , Radford MJ , tion of outpatient treatment with low-molecular-weight Krumholz HM . Intracranial hemorrhage associated with heparin for proximal vein thrombosis . Arch Intern Med . thrombolytic therapy for elderly patients with acute myocar- 1999 ; 159 ( 19 ): 2298 - 2304 . dial infarction: results from the Cooperative Cardiovascular 97 . Huse DM , Cummins G , Taylor DC , Russell MW . Outpatient Project . Stroke . 2000 ; 31 ( 8 ): 1802 - 1811 . treatment of venous thromboembolism with low-molecular- 111 . Mehta RH , Stebbins A , Lopes RD , et al . Race, Bleeding, weight heparin: an economic evaluation . Am J Manag Care . and Outcomes in STEMI Patients Treated with Fibrinolytic 2002 ; 8 ( 1 suppl ): S10 - S16 . Therapy . Am J Med . 2011 ; 124 ( 1 ): 48 - 57 . 98 . Spyropoulos AC , Hurley JS , Ciesla GN , de Lissovoy G . 112 . Indications for fi brinolytic therapy in suspected acute myo- Management of acute proximal deep vein thrombosis: phar- cardial infarction: collaborative overview of early mortality macoeconomic evaluation of outpatient treatment with and major morbidity results from all randomised trials of www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e483S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians more than 1000 patients. Fibrinolytic Therapy Trialists’ 132 . Schweizer J , Elix H , Altmann E , Hellner G , Forkmann L . (FTT) Collaborative Group. Lancet . 1994 ; 343 ( 8893 ): Comparative results of thrombolysis treatment with rt-PA 311 - 322 . and urokinase: a pilot study . Vasa . 1998 ; 27 ( 3 ): 167 - 171 . 113 . Piazza G , Goldhaber SZ . Fibrinolysis for acute pulmonary 133 . Schweizer J , Kirch W , Koch R , et al . Short- and long-term embolism . Vasc Med . 2010 ; 15 ( 5 ): 419 - 428 . results after thrombolytic treatment of deep venous throm- 114 . Todd JL , Tapson VF . Thrombolytic therapy for acute pul- bosis . J Am Coll Cardiol . 2000 ; 36 ( 4 ): 1336 - 1343 . monary embolism: a critical appraisal . Chest . 2009 ; 135 ( 5 ): 134 . Tsapogas MJ , Peabody RA , Wu KT , Karmody AM , Devaraj 1321 - 1329 . KT , Eckert C . Controlled study of thrombolytic therapy in 115 . Kasirajan K , Gray B , Ouriel K . Percutaneous AngioJet deep vein thrombosis . Surgery . 1973 ; 74 ( 6 ): 973 - 984 . thrombectomy in the management of extensive deep venous 135 . Turpie AG , Levine MN , Hirsh J , et al . Tissue plasminogen thrombosis . J Vasc Interv Radiol . 2001 ; 12 ( 2 ): 179 - 185 . activator (rt-PA) vs heparin in deep vein thrombosis. Results 116 . V e d a n t h a m S , V e s e l y T M , P a r t i N , D a r c y M , H o v s e p i a n D M , of a randomized trial . Chest . 1990 ; 97 ( 4 suppl ): 172S - 175S . Picus D . Lower extremity venous thrombolysis with adjunctive 136 . Verhaeghe R , Besse P , Bounameaux H , Marbet GA . mechanical thrombectomy . J Vasc Interv Radiol . 2 0 0 2 ; 1 3 ( 1 0 ) : Multicenter pilot study of the effi cacy and safety of systemic 1001 - 1008 . rt-PA administration in the treatment of deep vein throm- 117 . Delomez M , Beregi JP , Willoteaux S , et al . Mechanical bosis of the lower extremities and/or pelvis . Thromb Res . thrombectomy in patients with deep venous thrombosis . 1989 ; 55 ( 1 ): 5 - 11 . Cardiovasc Intervent Radiol . 2001 ; 24 ( 1 ): 42 - 48 . 137 . Watz R , Savidge GF . Rapid thrombolysis and preservation 118 . Kinney TB , Valji K , Rose SC , et al . Pulmonary embolism of valvular venous function in high deep vein thrombosis. from pulse-spray pharmacomechanical thrombolysis of A comparative study between streptokinase and heparin clotted hemodialysis grafts: urokinase versus heparinized therapy . Acta Med Scand . 1979 ; 205 ( 4 ): 293 - 298 . saline . J Vasc Interv Radiol . 2000 ; 11 ( 9 ): 1143 - 1152 . 138 . Watson LI , Armon MP . Thrombolysis for acute deep 119 . Arnesen H , Heilo A , Jakobsen E , Ly B , Skaga E . A prospec- vein thrombosis. Cochrane Database Syst Rev . 2 0 0 4 ; ( 4 ) : tive study of streptokinase and heparin in the treatment of CD002783 . deep vein thrombosis . Acta Med Scand . 1978 ; 203 ( 6 ): 457 - 463 . 139 . Common HH , Seaman AJ , Rösch J , Porter JM , Dotter CT . 120 . Arnesen H , Høiseth A , Ly B . Streptokinase of heparin in the Deep vein thrombosis treated with streptokinase or hep- treatment of deep vein thrombosis. Follow-up results of a arin. Follow-up of a randomized study . Angiology . 1976 ; prospective study . Acta Med Scand . 1982 ; 211 ( 1-2 ): 65 - 68 . 27 ( 11 ): 645 - 654 . 121 . Browse NL , Thomas ML , Pim HP . Streptokinase and deep 140 . Comerota AJ , Gale SS . Technique of contemporary iliofem- vein thrombosis . BMJ . 1968 ; 3 ( 5620 ): 717 - 720 . oral and infrainguinal venous thrombectomy . J Vasc Surg . 122 . Duckert F , Müller G , Nyman D , et al . Treatment of deep 2006 ; 43 ( 1 ): 185 - 191 . vein thrombosis with streptokinase . BMJ . 1975 ; 1 ( 5956 ): 141 . Karp RB , Wylie EJ . Recurrent thrombosis after iliofemoral 479 - 481 . venous thrombectomy . Surg Forum . 1966 ; 17 : 147 . 123 . Elliot MS , Immelman EJ , Jeffery P , et al . A comparative 142 . Lansing AM , Davis WM . Five-year follow-up study of iliofem- randomized trial of heparin versus streptokinase in the oral venous thrombectomy . Ann Surg . 1968 ; 168 ( 4 ): 620 - 628 . treatment of acute proximal venous thrombosis: an interim 143 . Plate G , Akesson H , Einarsson E , Ohlin P , Eklöf B . report of a prospective trial . Br J Surg . 1979 ; 66 ( 12 ): 838 - 843 . Long-term results of venous thrombectomy combined 124 . Goldhaber SZ , Meyerovitz MF , Green D , et al . Randomized with a temporary arterio-venous fi stula . Eur J Vasc Surg . controlled trial of tissue plasminogen activator in prox- 1990 ; 4 ( 5 ): 483 - 489 . imal deep venous thrombosis. Am J Med . 1 9 9 0 ; 8 8 ( 3 ) : 144 . Plate G , Einarsson E , Ohlin P , Jensen R , Qvarfordt P , Eklöf B . 235 - 240 . Thrombectomy with temporary arteriovenous fi stula: the 125 . Kakkar VV , Flanc C , Howe CT , O’Shea M , Flute PT . treatment of choice in acute iliofemoral venous thrombosis . Treatment of deep vein thrombosis. A trial of heparin, strep- J Vasc Surg . 1984 ; 1 ( 6 ): 867 - 876 . tokinase, and arvin . BMJ . 1969 ; 1 ( 5647 ): 806 - 810 . 145 . Plate G , Eklöf B , Norgren L , Ohlin P , Dahlström JA . 126 . Kiil J , Carvalho A , Saksø P , Nielsen HO . Urokinase or hep- Venous thrombectomy for iliofemoral vein thrombosis— arin in the management of patients with deep vein thrombo- 10-year results of a prospective randomised study. Eur J sis? Acta Chir Scand . 1981 ; 147 ( 7 ): 529 - 532 . Vasc Endovasc Surg . 1997 ; 14 ( 5 ): 367 - 374 . 127 . Laiho MK , Oinonen A , Sugano N , et al . Preservation of 146 . Decousus H , Leizorovicz A , Parent F , et al . A clinical trial of venous valve function after catheter-directed and systemic vena caval fi lters in the prevention of pulmonary embolism thrombolysis for deep venous thrombosis . Eur J Vasc in patients with proximal deep-vein thrombosis. Prévention Endovasc Surg . 2004 ; 28 ( 4 ): 391 - 396 . du Risque d’Embolie Pulmonaire par Interruption Cave 128 . Marder VJ , Soulen RL , Atichartakarn V , et al . Quantitative Study Group . N Engl J Med . 1998 ; 338 ( 7 ): 409 - 415 . venographic assessment of deep vein thrombosis in the 147 . Mismetti P , Rivron-Guillot K , Quenet S , et al . A prospective evaluation of streptokinase and heparin therapy . J Lab Clin long-term study of 220 patients with a retrievable vena cava Med . 1977 ; 89 ( 5 ): 1018 - 1029 . fi lter for secondary prevention of venous thromboembolism . 129 . Porter JM , Seaman AJ , Common HH , Rösch J , Eidemiller Chest . 2007 ; 131 ( 1 ): 223 - 229 . LR , Calhoun AD . Comparison of heparin and strepto- 148 . Athanasoulis CA , Kaufman JA , Halpern EF , Waltman kinase in the treatment of venous thrombosis . Am Surg . AC , Geller SC , Fan CM . Inferior vena caval fi lters: review 1975 ; 41 ( 9 ): 511 - 519 . of a 26-year single-center clinical experience . Radiology . 130 . Robertson BR , Nilsson IM , Nylander G . Value of strep- 2000 ; 216 ( 1 ): 54 - 66 . tokinase and heparin in treatment of acute deep venous 149 . Cosmi B , Legnani C , Cini M , Guazzaloca G , Palareti G . thrombosis. A coded investigation . Acta Chir Scand . 1968 ; D-dimer levels in combination with residual venous obstruc- 134 ( 3 ): 203 - 208 . tion and the risk of recurrence after anticoagulation with- 131 . Schulman S , Granqvist S , Juhlin-Dannfelt A , Lockner D . drawal for a fi rst idiopathic deep vein thrombosis . Thromb Long-term sequelae of calf vein thrombosis treated with Haemost . 2005 ; 94 ( 5 ): 969 - 974 . heparin or low-dose streptokinase . Acta Med Scand . 1986 ; 150 . Streiff MB . Vena caval fi lters: a comprehensive review . 219 ( 4 ): 349 - 357 . Blood . 2000 ; 95 ( 12 ): 3669 - 3677 . e484S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 151 . Hajduk B , Tomkowski WZ , Malek G , Davidson BL . Vena vein thrombosis or pulmonary embolism and comparison cava fi lter occlusion and venous thromboembolism risk in of 6 and 12 weeks of therapy after isolated calf deep vein persistently anticoagulated patients: a prospective, observa- thrombosis . Circulation . 2001 ; 103 ( 20 ): 2453 - 2460 . tional cohort study . Chest . 2010 ; 137 ( 4 ): 877 - 882 . 168 . Pini M , Aiello S , Manotti C , et al . Low molecular weight hep- 152 . Fox MA , Kahn SR . Postthrombotic syndrome in relation to arin versus warfarin in the prevention of recurrences after vena cava fi lter placement: a systematic review . J Vasc Interv deep vein thrombosis . Thromb Haemost . 1994 ; 72 ( 2 ): 191 - 197 . Radiol . 2008 ; 19 ( 7 ): 981 - 985 . 169 . Schulman S , Rhedin AS , Lindmarker P , et al ; Duration of 153 . Usoh F , Hingorani A , Ascher E , et al . Prospective random- Anticoagulation Trial Study Group . A comparison of six ized study comparing the clinical outcomes between infe- weeks with six months of oral anticoagulant therapy after rior vena cava Greenfi eld and TrapEase fi lters . J Vasc Surg . a fi rst episode of venous thromboembolism . N Engl J Med . 2010 ; 52 ( 2 ): 394 - 399 . 1995 ; 332 ( 25 ): 1661 - 1665 . 154 . Nicholson W , Nicholson WJ , Tolerico P , et al . Prevalence of 170 . Boutitie F, Pinede L, Schulman S, et al. Infl uence of pre- fracture and fragment embolization of Bard retrievable vena ceding duration of anticoagulant treatment and initial pre- cava fi lters and clinical implications including cardiac perfo- sentation of venous thromboembolism on risk of recurrence ration and tamponade . Arch Intern Med . 2010 ; 170 ( 20 ): 1827 - after stopping therapy: analysis of individual participants’ 1831 . data from seven trials . BMJ . 2011 ;342:d3036. 155 . Dabbagh O , Nagam N , Chitima-Matsiga R , Bearelly S , 171 . Iorio A , Kearon C , Filippucci E , et al . Risk of recurrence Bearelly D . Retrievable inferior vena cava fi lters are not after a fi rst episode of symptomatic venous thromboembo- getting retrieved: where is the gap? Thromb Res . 2010 ; lism provoked by a transient risk factor: a systematic review . 126 ( 6 ): 493 - 497 . Arch Intern Med . 2010 ; 170 ( 19 ): 1710 - 1716 . 156 . Aissaoui N , Martins E , Mouly S , Weber S , Meune C . A 172 . Palareti G , Legnani C , Lee A , et al . A comparison of the meta-analysis of bed rest versus early ambulation in the safety and effi cacy of oral anticoagulation for the treatment management of pulmonary embolism, deep vein thrombo- of venous thromboembolic disease in patients with or with- sis, or both . Int J Cardiol . 2009 ; 137 ( 1 ): 37 - 41 . out malignancy . Thromb Haemost . 2000 ; 84 ( 5 ): 805 - 810 . 157 . Kahn SR , Shrier I , Kearon C . Physical activity in patients 173 . Lee AY , Levine MN , Baker RI , et al ; Randomized with deep venous thrombosis: a systematic review . Thromb Comparison of Low-Molecular-Weight Heparin versus Oral Res . 2008 ; 122 ( 6 ): 763 - 773 . Anticoagulant Therapy for the Prevention of Recurrent 158 . Hull R , Delmore T , Genton E , et al . Warfarin sodium ver- Venous Thromboembolism in Patients with Cancer (CLOT) sus low-dose heparin in the long-term treatment of venous Investigators . Low-molecular-weight heparin versus a coumarin thrombosis . N Engl J Med . 1979 ; 301 ( 16 ): 855 - 858 . for the prevention of recurrent venous thromboembolism in 159 . Levine MN , Hirsh J , Gent M , et al . Optimal duration of patients with cancer . N Engl J Med . 2003 ; 349 ( 2 ): 146 - 153 . oral anticoagulant therapy: a randomized trial comparing 174 . Ridker PM , Goldhaber SZ , Danielson E , et al ; PREVENT four weeks with three months of warfarin in patients with Investigators . Long-term, low-intensity warfarin therapy proximal deep vein thrombosis . Thromb Haemost . 1995 ; for the prevention of recurrent venous thromboembolism . 74 ( 2 ): 606 - 611 . N Engl J Med . 2003 ; 348 ( 15 ): 1425 - 1434 . 160 . Research Committee of the British Thoracic Society . Optimum 175 . Schulman S , Wåhlander K, Lundström T , Clason SB , duration of anticoagulation for deep-vein thrombosis and Eriksson H ; THRIVE III Investigators . Secondary preven- pulmonary embolism . Lancet . 1992 ; 340 ( 8824 ): 873 - 876 . tion of venous thromboembolism with the oral direct throm- 161 . Schulman S , Granqvist S , Holmström M , et al ; The Duration bin inhibitor ximelagatran. N Engl J Med . 2 0 0 3 ; 3 4 9 ( 1 8 ) : of Anticoagulation Trial Study Group . The duration of oral 1713 - 1721 . anticoagulant therapy after a second episode of venous 176 . Eichinger S , Minar E , Bialonczyk C , et al . D-dimer levels thromboembolism . N Engl J Med . 1997 ; 336 ( 6 ): 393 - 398 . and risk of recurrent venous thromboembolism . JAMA . 162 . Baglin T , Luddington R , Brown K , Baglin C . Incidence of 2003 ; 290 ( 8 ): 1071 - 1074 . recurrent venous thromboembolism in relation to clinical 177 . Palareti G , Legnani C , Cosmi B , et al . Predictive value of and thrombophilic risk factors: prospective cohort study . D-dimer test for recurrent venous thromboembolism after Lancet . 2003 ; 362 ( 9383 ): 523 - 526 . anticoagulation withdrawal in subjects with a previous idi- 163 . Christiansen SC , Cannegieter SC , Koster T , Vandenbroucke opathic event and in carriers of congenital thrombophilia . JP , Rosendaal FR . Thrombophilia, clinical factors, and Circulation . 2003 ; 108 ( 3 ): 313 - 318 . recurrent venous thrombotic events . JAMA . 2005 ; 293 ( 19 ): 178 . Shrivastava S , Ridker PM , Glynn RJ , et al . D-dimer, factor 2352 - 2361 . VIII coagulant activity, low-intensity warfarin and the risk 164 . Hansson PO , Sörbo J , Eriksson H . Recurrent venous throm- of recurrent venous thromboembolism . J Thromb Haemost . boembolism after deep vein thrombosis: incidence and risk 2006 ; 4 ( 6 ): 1208 - 1214 . factors . Arch Intern Med . 2000 ; 160 ( 6 ): 769 - 774 . 179 . Cosmi B , Legnani C , Tosetto A , et al ; PROLONG Investigators 165 . Heit JA , Mohr DN , Silverstein MD , Petterson TM , (on behalf of Italian Federation of Anticoagulation Clinics) . O’Fallon WM , Melton LJ III . Predictors of recurrence Usefulness of repeated D-dimer testing after stopping anti- after deep vein thrombosis and pulmonary embolism: a coagulation for a fi rst episode of unprovoked venous throm- population-based cohort study. Arch Intern Med . 2 0 0 0 ; boembolism: the PROLONG II prospective study . Blood . 160 ( 6 ): 761 - 768 . 2010 ; 115 ( 3 ): 481 - 488 . 166 . Palareti G , Legnani C , Cosmi B , Guazzaloca G , Pancani C , 180 . Douketis J , Tosetto A , Marcucci M , et al . Patient-level Coccheri S . Risk of venous thromboembolism recurrence: meta-analysis: effect of measurement timing, threshold, and high negative predictive value of D-dimer performed after patient age on ability of D-dimer testing to assess recur- oral anticoagulation is stopped. Thromb Haemost . 2 0 0 2 ; rence risk after unprovoked venous thromboembolism . Ann 87 ( 1 ): 7 - 12 . Intern Med . 2010 ; 153 ( 8 ): 523 - 531 . 167 . Pinede L , Ninet J , Duhaut P , et al ; Investigators of 181 . Verhovsek M , Douketis JD , Yi Q , et al. Systematic review: the “Durée Optimale du Traitement AntiVitamines K” D-dimer to predict recurrent disease after stopping antico- (DOTAVK) Study . Comparison of 3 and 6 months of oral agulant therapy for unprovoked venous thromboembolism. anticoagulant therapy after a fi rst episode of proximal deep Ann Intern Med . 2008 ;149(7):481-490. www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e485S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 182 . Kearon C , Gent M , Hirsh J , et al . A comparison of three 198 . Carrier M , Rodger MA , Wells PS , Righini M , LE Gal G . months of anticoagulation with extended anticoagulation for Residual vein obstruction to predict the risk of recurrent a fi rst episode of idiopathic venous thromboembolism . N venous thromboembolism in patients with deep vein throm- Engl J Med . 1999 ; 340 ( 12 ): 901 - 907 . bosis: a systematic review and meta-analysis . J Thromb 183 . Schulman S , Lindmarker P , Holmström M , et al . Post- Haemost . 2011 ; 9 ( 6 ): 1119 - 1125 . thrombotic syndrome, recurrence, and death 10 years 1 9 9 . E i c h i n g e r S , H e i n z e G , J a n d e c k L M , K y r l e P A . R i s k a s s e s s - after the fi rst episode of venous thromboembolism treated ment of recurrence in patients with unprovoked deep vein with warfarin for 6 weeks or 6 months . J Thromb Haemost . thrombosis or pulmonary embolism: the Vienna prediction 2006 ; 4 ( 4 ): 734 - 742 . model . Circulation . 2010 ; 121 ( 14 ): 1630 - 1636 . 184 . Schulman S , Svenungsson E , Granqvist S ; Duration of 200 . Stain M , Schönauer V , Minar E , et al . The post-thrombotic Anticoagulation Study Group . Anticardiolipin antibodies syndrome: risk factors and impact on the course of throm- predict early recurrence of thromboembolism and death botic disease . J Thromb Haemost . 2005 ; 3 ( 12 ): 2671 - 2676 . among patients with venous thromboembolism following 201 . Brandjes DP , Büller HR , Heijboer H , et al . Randomised anticoagulant therapy . Am J Med . 1998 ; 104 ( 4 ): 332 - 338 . trial of effect of compression stockings in patients with symp- 185 . Rodger MA , Kahn SR , Wells PS , et al . Identifying unpro- tomatic proximal-vein thrombosis. Lancet . 1997 ; 349 ( 9054 ): voked thromboembolism patients at low risk for recur- 759 - 762 . rence who can discontinue anticoagulant therapy . CMAJ . 202 . Prandoni P , Lensing AW , Prins MH , et al . Below-knee elas- 2008 ; 179 ( 5 ): 417 - 426 . tic compression stockings to prevent the post-thrombotic 186 . Eichinger S , Pabinger I , Stümpfl en A , et al . The risk of recur- syndrome: a randomized, controlled trial . Ann Intern Med . rent venous thromboembolism in patients with and without 2004 ; 141 ( 4 ): 249 - 256 . factor V Leiden . Thromb Haemost . 1997 ; 77 ( 4 ): 624 - 628 . 203 . Campbell IA , Bentley DP , Prescott RJ , Routledge PA , 187 . Ho WK , Hankey GJ , Quinlan DJ , Eikelboom JW . Risk of Shetty HG , Williamson IJ . Anticoagulation for three ver- recurrent venous thromboembolism in patients with common sus six months in patients with deep vein thrombosis or thrombophilia: a systematic review. Arch Intern Med . pulmonary embolism, or both: randomised trial . BMJ . 2006 ; 166 ( 7 ): 729 - 736 . 2007 ; 334 ( 7595 ): 674 - 680 . 188 . L i n d m a r k e r P , S c h u l m a n S , S t e n - L i n d e r M , W i m a n B , 204 . Agnelli G , Prandoni P , Santamaria MG , et al ; Warfarin E g b e r g N , J o h n s s o n H . T h e r i s k o f r e c u r r e n t v e n o u s t h r o m - Optimal Duration Italian Trial Investigators . Three months boembolism in carriers and non-carriers of the G1691A allele versus one year of oral anticoagulant therapy for idio- in the coagulation factor V gene and the G20210A allele in pathic deep venous thrombosis . N Engl J Med . 2001 ; 345 ( 3 ): the prothrombin gene. DURAC Trial Study Group. Dura tion 165 - 169 . of Anticoagulation . Thromb Haemost . 1999 ; 81 ( 5 ): 684 - 689 . 205 . Siragusa S , Malato A , Anastasio R , et al . Residual vein 189 . Segal JB , Brotman DJ , Necochea AJ , et al . Predictive thrombosis to establish duration of anticoagulation after value of factor V Leiden and prothrombin G20210A in a fi rst episode of deep vein thrombosis: the Duration of adults with venous thromboembolism and in family mem- Anticoagulation based on Compression UltraSonography bers of those with a mutation: a systematic review . JAMA . (DACUS) study . Blood . 2008 ; 112 ( 3 ): 511 - 515 . 2009 ; 301 ( 23 ): 2472 - 2485 . 206 . Prandoni P , Prins MH , Lensing AW , et al ; AESOPUS 190 . Kearon C , Julian JA , Kovacs MJ , et al ; ELATE Investigators . Investigators . Residual thrombosis on ultrasonography to Infl uence of thrombophilia on risk of recurrent venous guide the duration of anticoagulation in patients with deep thromboembolism while on warfarin: results from a ran- venous thrombosis: a randomized trial . Ann Intern Med . domized trial . Blood . 2008 ; 112 ( 12 ): 4432 - 4436 . 2009 ; 150 ( 9 ): 577 - 585 . 191 . McRae S , Tran H , Schulman S , Ginsberg J , Kearon C . Effect 207 . Farraj RS . Anticoagulation period in idiopathic venous of patient’s sex on risk of recurrent venous thromboembo- thromboembolism. How long is enough? Saudi Med J . lism: a meta-analysis . Lancet . 2006 ; 368 ( 9533 ): 371 - 378 . 2004 ; 25 ( 7 ): 848 - 851 . 192 . Douketis J , Tosetto A , Marcucci M , et al . Risk of recurrence 208 . Prandoni P , Noventa F , Ghirarduzzi A , et al . The risk of after venous thromboembolism in men and women: patient recurrent venous thromboembolism after discontinuing level meta-analysis . BMJ . 2011 ; 342 : d813 . anticoagulation in patients with acute proximal deep vein 193 . White RH , Zhou H , Romano PS . Incidence of idiopathic thrombosis or pulmonary embolism. A prospective cohort deep venous thrombosis and secondary thromboembolism study in 1,626 patients . Haematologica . 2007 ; 92 ( 2 ): 199 - 205 . among ethnic groups in California . Ann Intern Med . 1998 ; 209 . van Dongen CJ , Prandoni P , Frulla M , Marchiori A , Prins MH , 128 ( 9 ): 737 - 740 . Hutten BA . Relation between quality of anticoagulant treat- 194 . Agnelli G , Prandoni P , Becattini C , et al ; Warfarin Optimal ment and the development of the postthrombotic syndrome . Duration Italian Trial Investigators . Extended oral anticoag- J Thromb Haemost . 2005 ; 3 ( 5 ): 939 - 942 . ulant therapy after a fi rst episode of pulmonary embolism . 210 . Locadia M , Bossuyt PM , Stalmeier PF , et al . Treatment Ann Intern Med . 2003 ; 139 ( 1 ): 19 - 25 . of venous thromboembolism with vitamin K antagonists: 195 . Kearon C , Ginsberg JS , Anderson DR , et al ; SOFAST patients’ health state valuations and treatment preferences . Investigators . Comparison of 1 month with 3 months of anti- Thromb Haemost . 2004 ; 92 ( 6 ): 1336 - 1341 . coagulation for a fi rst episode of venous thromboembolism 211 . Hull RD , Pineo GF , Brant RF , et al ; LITE Trial Investigators . associated with a transient risk factor . J Thromb Haemost . Self-managed long-term low-molecular-weight heparin 2004 ; 2 ( 5 ): 743 - 749 . therapy: the balance of benefi ts and harms . Am J Med . 196 . Piovella F , Crippa L , Barone M , et al . Normalization rates 2007 ; 120 ( 1 ): 72 - 82 . of compression ultrasonography in patients with a fi rst 212 . Beyth RJ , Cohen AM , Landefeld CS . Long-term outcomes episode of deep vein thrombosis of the lower limbs: associ- of deep-vein thrombosis . Arch Intern Med . 1995 ; 155 ( 10 ): ation with recurrence and new thrombosis . Haematologica . 1031 - 1037 . 2002 ; 87 ( 5 ): 515 - 522 . 213 . Hull RD , Pineo GF , Brant R , et al. Home therapy of venous 197 . Prandoni P , Lensing AW , Prins MH , et al . Residual venous thrombosis with long-term LMWH versus usual care: thrombosis as a predictive factor of recurrent venous throm- patient satisfaction and post-thrombotic syndrome. Am J boembolism . Ann Intern Med . 2002 ; 137 ( 12 ): 955 - 960 . Med. 2009 ;122(8):762-769. e486S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 214 . Kahn SR , Ginsberg JS . Relationship between deep venous 229 . Varki A . Trousseau’s syndrome: multiple defi nitions and thrombosis and the postthrombotic syndrome . Arch Intern multiple mechanisms . Blood . 2007 ; 110 ( 6 ): 1723 - 1729 . Med . 2004 ; 164 ( 1 ): 17 - 26 . 230 . Falck-Ytter Y , Francis CW , Johanson NA , et al . Prevention of 215 . Hull R , Delmore T , Carter C , et al. Adjusted subcutaneous VTE in orthopedic surgery patients: antithrombotic therapy heparin versus warfarin sodium in the long-term treat- and prevention of thrombosis, 9th ed: American College of ment of venous thrombosis. N Engl J Med . 1 9 8 2 ; 3 0 6 ( 4 ) : Chest Physicians evidence-based clinical practice guidelines . 189 - 194 . Chest . 2012;141(2)(suppl):e278S-e325S. 216 . Monreal M , Lafoz E , Olive A , del Rio L , Vedia C . 231 . Robinson KS , Anderson DR , Gross M , et al . Ultrasonographic Comparison of subcutaneous unfractionated heparin with screening before hospital discharge for deep venous throm- a low molecular weight heparin (Fragmin) in patients with bosis after arthroplasty: the post-arthroplasty screening venous thromboembolism and contraindications to cou- study. A randomized, controlled trial . Ann Intern Med . marin . Thromb Haemost . 1994 ; 71 ( 1 ): 7 - 11 . 1997 ; 127 ( 6 ): 439 - 445 . 217 . Buller HR, Cohen AT, Davidson B, et al; van Gogh 232 . Kahn SR , Ginsberg JS . The post-thrombotic syndrome: Investigators. Idraparinux versus standard therapy for venous current knowledge, controversies, and directions for future thromboembolic disease . N Engl J Med . 2007 ; 357 (11): 1094 - research . Blood Rev . 2002 ; 16 ( 3 ): 155 - 165 . 1104 . 233 . Ginsberg JS , Hirsh J , Julian J , et al . Prevention and treat- 2 1 8 . I o r i o A , G u e r c i n i F , P i n i M . L o w - m o l e c u l a r - w e i g h t h e p a r i n ment of postphlebitic syndrome: results of a 3-part study . for the long-term treatment of symptomatic venous throm- Arch Intern Med . 2001 ; 161 ( 17 ): 2105 - 2109 . boembolism: meta-analysis of the randomized comparisons 234 . Partsch H , Kaulich M , Mayer W . Immediate mobilisation in with oral anticoagulants . J Thromb Haemost . 2003 ; 1 ( 9 ): acute vein thrombosis reduces post-thrombotic syndrome . 1906 - 1913 . Int Angiol . 2004 ; 23 ( 3 ): 206 - 212 . 219. van der Heijden JF , Hutten BA , Büller HR , Prins MH . 235 . Aschwanden M , Jeanneret C , Koller MT , Thalhammer C , Vitamin K antagonists or low-molecular-weight heparin for B u c h e r H C , J a e g e r K A . E f f e c t o f p r o l o n g e d t r e a t m e n t w i t h the long term treatment of symptomatic venous thromboem- compression stockings to prevent post-thrombotic sequelae: bolism. Cochrane Database Syst Rev . 2002 ;(4):CD002001. a randomized controlled trial . J Vasc Surg . 2 0 0 8 ; 4 7 ( 5 ) : 220 . Das SK , Cohen AT , Edmondson RA , Melissari E , Kakkar 1015 - 1021 . VV . Low-molecular-weight heparin versus warfarin for pre- 236 . Roumen-Klappe EM , den Heijer M , van Rossum J , et al . vention of recurrent venous thromboembolism: a random- Multilayer compression bandaging in the acute phase of ized trial . World J Surg . 1996 ; 20 ( 5 ): 521 - 526. deep-vein thrombosis has no effect on the development of 221 . Gonzalez-Fajardo JA , Arreba E , Castrodeza J , et al . the post-thrombotic syndrome . J Thromb Thrombolysis . Venographic comparison of subcutaneous low-molecular 2009 ; 27 ( 4 ): 400 - 405 . weight heparin with oral anticoagulant therapy in the long- 237 . Ten Cate-Hoek AJ , Ten Cate H , Tordoir J , Hamulyák K , term treatment of deep venous thrombosis . J Vasc Surg . Prins MH . Individually tailored duration of elastic compres- 1999 ; 30 ( 2 ): 283 - 292 . sion therapy in relation to incidence of the postthrombotic 222 . Hull R , Pineo G , Mah A , et al . Long-term low molecular syndrome . J Vasc Surg . 2010 ; 52 ( 1 ): 132 - 138 . weight heparin treatment versus oral anticoagulant therapy 238 . Frulla M , Marchiori A , Sartor D , et al . Elastic stockings, for proximal deep vein thrombosis [abstract] . Blood . hydroxyethylrutosides or both for the treatment of post- 2000 ; 96 : 449a . thrombotic syndrome. Thromb Haemost . 2005 ; 93 ( 1 ): 223 . Lopaciuk S , Bielska-Falda H , Noszczyk W , et al . Low molec- 183 - 185 . ular weight heparin versus acenocoumarol in the secondary 239 . Ginsberg JS , Magier D , Mackinnon B , Gent M , Hirsh J . prophylaxis of deep vein thrombosis . Thromb Haemost . Intermittent compression units for severe post-phlebitic 1999 ; 81 ( 1 ): 26 - 31 . syndrome: a randomized crossover study . CMAJ . 1999 ; 224 . López-Beret P , Orgaz A , Fontcuberta J , et al . Low molec- 160 ( 9 ): 1303 - 1306 . ular weight heparin versus oral anticoagulants in the long- 240 . O’Donnell MJ , McRae S , Kahn SR , et al . Evaluation of a term treatment of deep venous thrombosis . J Vasc Surg . venous-return assist device to treat severe post-thrombotic 2001 ; 33 ( 1 ): 77 - 90 . syndrome (VENOPTS). A randomized controlled trial . 225 . Veiga F , Escribá A , Maluenda MP , et al . Low molecular Thromb Haemost . 2008 ; 99 ( 3 ): 623 - 629 . weight heparin (enoxaparin) versus oral anticoagulant 241 . Lyseng-Williamson KA , Perry CM . Micronised purifi ed therapy (acenocoumarol) in the long-term treatment of fl avonoid fraction: a review of its use in chronic venous deep venous thrombosis in the elderly: a randomized trial . insuffi ciency, venous ulcers and haemorrhoids . Drugs . Thromb Haemost . 2000 ; 84 ( 4 ): 559 - 564 . 2003 ; 63 ( 1 ): 71 - 100 . 226 . Deitcher SR , Kessler CM , Merli G , Rigas JR , Lyons RM , 242 . S h o a b S S , P o r t e r J , S c u r r J H , C o l e r i d g e - S m i t h P D . Fareed J ; ONCENOX Investigators . Secondary prevention Endothelial activation response to oral micronised fl avo- of venous thromboembolic events in patients with active noid therapy in patients with chronic venous disease—a cancer: enoxaparin alone versus initial enoxaparin followed prospective study. Eur J Vasc Endovasc Surg . 1999 ; 17 ( 4 ): by warfarin for a 180-day period . Clin Appl Thromb Hemost . 313 - 318 . 2006 ; 12 ( 4 ): 389 - 396 . 243 . de Jongste AB , Jonker JJ , Huisman MV , ten Cate JW , 227 . Meyer G , Marjanovic Z , Valcke J , et al . Comparison of low- Azar AJ . A double blind three center clinical trial on the molecular-weight heparin and warfarin for the secondary short-term effi cacy of 0-(beta-hydroxyethyl)-rutosides in prevention of venous thromboembolism in patients with patients with post-thrombotic syndrome . Thromb Haemost . cancer: a randomized controlled study . Arch Intern Med . 1989 ; 62 ( 3 ): 826 - 829 . 2002 ; 162 ( 15 ): 1729 - 1735 . 244 . Monreal M , Callejas JM , Martorell A , et al . A prospective 228 . Romera A , Cairols MA , Vila-Coll R , et al . A randomised study of the long-term effi cacy of two different venoac- open-label trial comparing long-term sub-cutaneous low- tive drugs in patients with post-thrombotic syndrome . molecular-weight heparin compared with oral-anticoagulant Phlebology . 1994 ; 9 : 37 - 40 . therapy in the treatment of deep venous thrombosis . Eur J 245 . C a m p b e l l I A , Y e o h J , M e d l i c o t t S . D u r a t i o n o f h o s p i t a l Vasc Endovasc Surg . 2009 ; 37 ( 3 ): 349 - 356 . stay in patients with pulmonary venous thromboembolism: www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e487S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians a randomized comparison of unfractionated heparinversus 260 . Aujesky D , Obrosky DS , Stone RA , et al . Derivation and low molecular weight heparin [Abstract] . Thorax . 1 9 9 8 ; validation of a prognostic model for pulmonary embolism . 53 : 254 . Am J Respir Crit Care Med . 2005 ; 172 ( 8 ): 1041 - 1046 . 246 . Duroux P . A randomised trial of subcutaneous low molec- 261 . Jiménez D , Yusen RD , Otero R , et al . Prognostic models for ular weight heparin (CY 216) compared with intravenous selecting patients with acute pulmonary embolism for initial unfractionated heparin in the treatment of deep vein throm- outpatient therapy . Chest . 2007 ; 132 ( 1 ): 24 - 30 . bosis. A collaborative European multicentre study . Thromb 262 . Jiménez D , Aujesky D , Moores L , et al ; RIETE Investigators . Haemost . 1991 ; 65 ( 3 ): 251 - 256 . Simplifi cation of the pulmonary embolism severity index for 247 . Hull RD , Raskob GE , Pineo GF , et al . Subcutaneous prognostication in patients with acute symptomatic pulmo- low-molecular-weight heparin compared with continuous nary embolism . Arch Intern Med . 2010 ; 170 ( 15 ): 1383 - 1389 . intravenous heparin in the treatment of proximal-vein 263 . Jakobsson C , Jiménez D , Gómez V , Zamarro C , Méan M , thrombosis . N Engl J Med . 1992 ; 326 ( 15 ): 975 - 982 . Aujesky D . Validation of a clinical algorithm to identify low- 248 . Kirchmaier CM , Wolf H , Schäfer H , Ehlers B , Breddin risk patients with pulmonary embolism . J Thromb Haemost . HK; Certoparin-Study Group . Effi cacy of a low molecular 2010 ; 8 ( 6 ): 1242 - 1247 . weight heparin administered intravenously or subcutane- 264 . Chan CM , Woods C , Shorr AF . The validation and repro- ously in comparison with intravenous unfractionated hep- ducibility of the pulmonary embolism severity index . J arin in the treatment of deep venous thrombosis . Int Angiol . Thromb Haemost . 2010 ; 8 ( 7 ): 1509 - 1514 . 1998 ; 17 ( 3 ): 135 - 145 . 265 . Moores L , Aujesky D , Jiménez D , et al . Pulmonary 249 . Kuijer PM , Gallus AS , Cade J , et al . Randomized compar- Embolism Severity Index and troponin testing for the selec- ison of LMWH versus standard heparin in the initial treat- tion of low-risk patients with acute symptomatic pulmonary ment of pulmonary embolism [Abstract] . Thromb Haemost . embolism . J Thromb Haemost . 2010 ; 8 ( 3 ): 517 - 522 . 1995 ; 73 : 974 . 266 . Fraga M , Taffé P , Méan M , et al . The inter-rater reliability of 250 . Meyer G , Brenot F , Pacouret G , et al . Subcutaneous low- the Pulmonary Embolism Severity Index . Thromb Haemost . molecular-weight heparin fragmin versus intravenous 2010 ; 104 ( 6 ): 1258 - 1262 . unfractionated heparin in the treatment of acute non mas- 267 . Squizzato A , Galli M , Dentali F , Ageno W . Outpatient sive pulmonary embolism: an open randomized pilot study . treatment and early discharge of symptomatic pulmonary Thromb Haemost . 1995 ; 74 ( 6 ): 1432 - 1435 . embolism: a systematic review . Eur Respir J . 2009 ; 33 ( 5 ): 251 . Pérez de Llano LA , Baloira Villar A , Veres Racamonde A , 1148 - 1155 . Veiga F , Golpe Gómez R , Pajuelo Fernández F . Multicenter, 268 . Erkens PM , Gandara E , Wells P , et al . Safety of outpa- prospective study comparing enoxaparin with unfraction- tient treatment in acute pulmonary embolism . J Thromb ated heparin in the treatment of submassive pulmonary Haemost . 2010 ; 8 ( 11 ): 2412 - 2417 . thromboembolism [in Spanish] . Arch Bronconeumol . 2003 ; 269 . Kovacs MJ , Hawel JD , Rekman JF , Lazo-Langner A . 39 ( 8 ): 341 - 345 . Ambulatory management of pulmonary embolism: a 252 . Simonneau G , Sors H , Charbonnier B , et al . A comparison pragmatic evaluation. J Thromb Haemost . 2010 ; 8 ( 11 ): of low-molecular-weight heparin with unfractionated hep- 2406 - 2411 . arin for acute pulmonary embolism. The THESEE Study 270 . Zondag W , Mos IC , Creemers-Schild D , et al ; Hestia Study Group. Tinzaparine ou Heparine Standard: Evaluations Investigators . Outpatient treatment in patients with acute dans l’Embolie Pulmonaire . N Engl J Med . 1997 ; 337 ( 10 ): pulmonary embolism: the Hestia Study . J Thromb Haemost . 663 - 669 . 2011 ; 9 ( 8 ): 1500 - 1507 . 253 . The Columbus Investigators . Low-molecular-weight hep- 271 . Agterof MJ , Schutgens RE , Snijder RJ , et al . Out of hos- arin in the treatment of patients with . N Engl J Med . 1997 ; pital treatment of acute pulmonary embolism in patients 337 ( 10 ): 657 - 662 . with a low NT-proBNP level . J Thromb Haemost . 2010 ; 8 ( 6 ): 254 . Théry C , Simonneau G , Meyer G , et al . Randomized trial 1235 - 1241 . of subcutaneous low-molecular-weight heparin CY 216 272 . Baglin T . Fifty per cent of patients with pulmonary embo- (Fraxiparine) compared with intravenous unfractionated lism can be treated as outpatients . J Thromb Haemost . heparin in the curative treatment of submassive pulmo- 2010 ; 8 ( 11 ): 2404 - 2405 . nary embolism. A dose-ranging study. Circulation . 1 9 9 2 ; 273. Dong B , Jirong Y , Wang Q , Wu T . Thrombolytic treatment 85 ( 4 ): 1380 - 1389 . for pulmonary embolism. Cochrane Database Syst Rev. 255 . Quinlan DJ , McQuillan A , Eikelboom JW . Low-molecular- 2006 ;(2);CD004437. weight heparin compared with intravenous unfractionated 274 . Wood KE . Major pulmonary embolism: review of a patho- heparin for treatment of pulmonary embolism: a meta- physiologic approach to the golden hour of hemodynamically analysis of randomized, controlled trials . Ann Intern Med . signifi cant pulmonary embolism . Chest . 2002 ; 121 ( 3 ): 877 - 905 . 2004 ; 140 ( 3 ): 175 - 183 . 275 . McIntyre KM , Sasahara AA . Determinants of right ventric- 256 . Otero R , Uresandi F , Jiménez D , et al . Home treatment in ular function and hemodynamics after pulmonary embolism . pulmonary embolism . Thromb Res . 2010 ; 126 ( 1 ): e1 - e5 . Chest . 1974 ; 65 ( 5 ): 534 - 543 . 257 . Aujesky D , Roy PM , Verschuren F , et al . Outpatient ver- 276 . Goldhaber SZ , Visani L , De Rosa M . Acute pulmo- sus inpatient treatment for patients with acute pulmonary nary embolism: clinical outcomes in the International embolism: an international, open-label, randomised, non- Cooperative Pulmonary Embolism Registry (ICOPER) . inferiority trial . Lancet . 2011 ; 378 ( 9785 ): 41 - 48 . Lancet . 1999 ; 353 ( 9162 ): 1386 - 1389 . 258 . Uresandi F , Otero R , Cayuela A , et al . [A clinical predic- 277 . Laporte S , Mismetti P , Décousus H , et al ; RIETE tion rule for identifying short-term risk of adverse events Investigators . Clinical predictors for fatal pulmonary embo- in patients with pulmonary thromboembolism] . Arch lism in 15,520 patients with venous thromboembolism: Bronconeumol . 2007 ; 43 ( 11 ): 617 - 622 . fi ndings from the Registro Informatizado de la Enfermedad 259 . Wicki J , Perrier A , Perneger TV , Bounameaux H , Junod AF . TromboEmbolica venosa (RIETE) Registry . Circulation . Predicting adverse outcome in patients with acute pulmo- 2008 ; 117 ( 13 ): 1711 - 1716 . nary embolism: a risk score . Thromb Haemost . 2000 ; 84 ( 4 ): 2 7 8 . P o l l a c k C V , S c h r e i b e r D , G o l d h a b e r S Z , e t a l . C l i n i c a l c h a r - 548 - 552 . acteristics, management, and outcomes of patients diagnosed e488S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians with acute pulmonary embolism in the emergency depart- acute pulmonary embolism: results from a monocenter reg- ment: initial report of EMPEROR (Multicenter Emergency istry of 1,416 patients . Chest . 2008 ; 133 ( 2 ): 358 - 362 . Medicine Pulmonary Embolism in the Real World Registry) . 296 . Vanni S , Polidori G , Vergara R , et al . Prognostic value of J Am Coll Cardiol . 2011 ; 57 ( 6 ): 700 - 706 . ECG among patients with acute pulmonary embolism and 279 . Sanchez O , Trinquart L , Caille V , et al . Prognostic factors normal blood pressure . Am J Med . 2009 ; 122 ( 3 ): 257 - 264 . for pulmonary embolism: the prep study, a prospective 297 . Stein PD , Beemath A , Matta F , et al . Enlarged right ven- multicenter cohort study . Am J Respir Crit Care Med . tricle without shock in acute pulmonary embolism: progno- 2010 ; 181 ( 2 ): 168 - 173 . sis . Am J Med . 2008 ; 121 ( 1 ): 34 - 42 . 280 . Kucher N , Rossi E , De Rosa M , Goldhaber SZ . Massive pul- 298 . Fiumara K , Kucher N , Fanikos J , Goldhaber SZ . Predictors monary embolism . Circulation . 2006 ; 113 ( 4 ): 577 - 582 . of major hemorrhage following fi brinolysis for acute pulmo- 281 . Torbicki A , Perrier A , Konstantinides S , et al ; ESC nary embolism . Am J Cardiol . 2006 ; 97 ( 1 ): 127 - 129 . Committee for Practice Guidelines (CPG) . Guidelines on 299 . K a n t e r D S , M i k k o l a K M , P a t e l S R , P a r k e r J A , G o l d h a b e r S Z . the diagnosis and management of acute pulmonary embo- Thrombolytic therapy for pulmonary embolism. Frequency lism: the Task Force for the Diagnosis and Management of intracranial hemorrhage and associated risk factors . Chest . of Acute Pulmonary Embolism of the European Society of 1997 ; 111 ( 5 ): 1241 - 1245 . Cardiology (ESC) . Eur Heart J . 2008 ; 29 ( 18 ): 2276 - 2315 . 300 . The urokinase pulmonary embolism trial . A national cooper- 282 . Douketis JD , Leeuwenkamp O , Grobara P , et al . The inci- ative study . Circulation . 1973 ; 47 (2 suppl):II 1-II108 . dence and prognostic signifi cance of elevated cardiac tro- 301 . Dalla-Volta S , Palla A , Santolicandro A , et al . PAIMS 2: ponins in patients with submassive pulmonary embolism . alteplase combined with heparin versus heparin in the J Thromb Haemost . 2005 ; 3 ( 3 ): 508 - 513 . treatment of acute pulmonary embolism. Plasminogen 283 . Giannitsis E , Katus HA . Risk stratifi cation in pulmonary activator Italian multicenter study 2 . J Am Coll Cardiol . embolism based on biomarkers and echocardiography . 1992 ; 20 ( 3 ): 520 - 526 . Circulation . 2005 ; 112 ( 11 ): 1520 - 1521 . 302 . Dotter CT , Seamon AJ , Rosch J , et al . Streptokinase and 284 . Konstantinides S , Geibel A , Olschewski M , et al . Importance heparin in the treatment of pulmonary embolism: a random- of cardiac troponins I and T in risk stratifi cation of patients ized comparison . Vasc Surg . 1979 ; 13 : 42 - 52 . with acute pulmonary embolism . Circulation . 2002 ; 106 ( 10 ): 303 . PIOPED Investigators . Tissue plasminogen activator for the 1263 - 1268 . treatment of acute pulmonary embolism. A collaborative study 285 . Scridon T , Scridon C , Skali H , Alvarez A , Goldhaber SZ , by the PIOPED Investigators . Chest . 1990 ; 97 ( 3 ): 528 - 533 . Solomon SD . Prognostic signifi cance of troponin eleva- 304 . Jerjes-Sanchez C , Ramírez-Rivera A , de Lourdes García M, tion and right ventricular enlargement in acute pulmonary et al . Streprokinase and heparin versus heparin alone in embolism . Am J Cardiol . 2005 ; 96 ( 2 ): 303 - 305 . massive pulmonary embolism: a randomized controlled trial . 286 . Becattini C , Vedovati MC , Agnelli G . Prognostic value of J Thromb Thrombolysis . 1995 ; 2 ( 3 ): 227 - 229 . troponins in acute pulmonary embolism: a meta-analysis . 305 . Ly B , Arnesen H , Eie H , Hol R . A controlled clinical trial of Circulation . 2007 ; 116 ( 4 ): 427 - 433 . streptokinase and heparin in the treatment of major pulmo- 287 . Lankeit M , Friesen D , Aschoff J , et al . Highly sensitive tro- nary embolism . Acta Med Scand . 1978 ; 203 ( 6 ): 465 - 470 . ponin T assay in normotensive patients with acute pulmo- 306 . Marini C , Di Ricco G , Rossi G , Rindi M , Palla R , Giuntini C . nary embolism . Eur Heart J . 2010 ; 31 ( 15 ): 1836 - 1844 . Fibrinolytic effects of urokinase and heparin in acute pul- 288 . Lega JC , Lacasse Y , Lakhal L , Provencher S . Natriuretic monary embolism: a randomized clinical trial . Respiration . peptides and troponins in pulmonary embolism: a meta- 1988 ; 54 ( 3 ): 162 - 173 . analysis . Thorax . 2009 ; 64 ( 10 ): 869 - 875 . 307 . Tibbutt DA , Davies JA , Anderson JA , et al . Comparison 289 . Bova C , Pesavento R , Marchiori A , et al ; TELESIO Study by controlled clinical trial of streptokinase and heparin in Group . Risk stratifi cation and outcomes in hemodynam- treatment of life-threatening pulmonay embolism . BMJ . ically stable patients with acute pulmonary embolism: a 1974 ; 1 ( 5904 ): 343 - 347 . prospective, multicentre, cohort study with three months of 308 . Urokinase pulmonary embolism trial. Phase I results: a follow-up . J Thromb Haemost . 2009 ; 7 ( 6 ): 938 - 944 . cooperative study . JAMA . 1970 ; 214 (12): 2163 - 2172 . 290 . Stein PD , Matta F , Janjua M , Yaekoub AY , Jaweesh F , 309 . Becattini C , Agnelli G , Salvi A , et al ; TIPES Study Group . Alrifai A . Outcome in stable patients with acute pulmonary Bolus tenecteplase for right ventricle dysfunction in embolism who had right ventricular enlargement and/or hemodynamically stable patients with pulmonary embolism . elevated levels of troponin I. Am J Cardiol . 2 0 1 0 ; 1 0 6 ( 4 ) : Thromb Res . 2010 ; 125 ( 3 ): e82 - e86 . 558 - 563 . 310. Fasullo S , Scalzo S , Maringhini G , et al. Six-month echo- 291 . Sanchez O , Trinquart L , Colombet I , et al . Prognostic value cardiographic study in patients with submassive pul- of right ventricular dysfunction in patients with haemody- monary embolism and right ventricle dysfunction: namically stable pulmonary embolism: a systematic review . comparison of thrombolysis with heparin. Am J Med Sci . Eur Heart J . 2008 ; 29 ( 12 ): 1569 - 1577 . 2010;341(1):33-39. 292 . Goldhaber SZ . Echocardiography in the management of 311 . Goldhaber SZ , Haire WD , Feldstein ML , et al . Alteplase pulmonary embolism . Ann Intern Med . 2002 ; 136 ( 9 ): 691 - 700 . versus heparin in acute pulmonary embolism: randomised 293 . Schoepf UJ , Kucher N , Kipfmueller F , Quiroz R , Costello P , trial assessing right-ventricular function and pulmonary per- Goldhaber SZ . Right ventricular enlargement on chest com- fusion . Lancet . 1993 ; 341 ( 8844 ): 507 - 511 . puted tomography: a predictor of early death in acute pul- 312 . K o n s t a n t i n i d e s S , G e i b e l A , H e u s e l G , H e i n r i c h F , K a s p e r W ; monary embolism . Circulation . 2004 ; 110 ( 20 ): 3276 - 3280 . Management Strategies and Prognosis of Pulmonary 294 . ten Wolde M , Söhne M , Quak E , Mac Gillavry MR , Büller HR . Embolism-3 Trial Investigators . Heparin plus alteplase Prognostic value of echocardiographically assessed right compared with heparin alone in patients with submassive ventricular dysfunction in patients with pulmonary embo- pulmonary embolism . N Engl J Med . 2002 ; 347 ( 15 ): 1143 - 1150 . lism . Arch Intern Med . 2004 ; 164 ( 15 ): 1685 - 1689 . 313 . Levine M , Hirsh J , Weitz J , et al . A randomized trial of a 295 . F r é m o n t B , P a c o u r e t G , J a c o b i D , P u g l i s i R , C h a r b o n n i e r B , single bolus dosage regimen of recombinant tissue plasmin- de Labriolle A . Prognostic value of echocardiographic right/ ogen activator in patients with acute pulmonary embolism . left ventricular end-diastolic diameter ratio in patients with Chest . 1990 ; 98 ( 6 ): 1473 - 1479 . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e489S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 314 . Agnelli G , Becattini C , Kirschstein T . Thrombolysis vs hep- ment of massive pulmonary embolism hemodynamic reper- arin in the treatment of pulmonary embolism: a clinical cussions] . Arch Bronconeumol . 2001 ; 37 ( 2 ): 58 - 64 . outcome-based meta-analysis. Arch Intern Med . 2 0 0 2 ; 330 . Fava M , Loyola S , Flores P , Huete I . Mechanical fragmen- 162 ( 22 ): 2537 - 2541 . tation and pharmacologic thrombolysis in massive pulmo- 315 . Wan S , Quinlan DJ , Agnelli G , Eikelboom JW . Thrombolysis nary embolism . J Vasc Interv Radiol . 1997 ; 8 ( 2 ): 261 - 266 . compared with heparin for the initial treatment of pulmo- 331 . Schmitz-Rode T , Janssens U , Duda SH , Erley CM , Günther nary embolism: a meta-analysis of the randomized con- RW . Massive pulmonary embolism: percutaneous emer- trolled trials . Circulation . 2004 ; 110 ( 6 ): 744 - 749 . gency treatment by pigtail rotation catheter . J Am Coll 316 . Urokinase-streptokinase embolism trial. Phase 2 results. A Cardiol . 2000 ; 36 ( 2 ): 375 - 380 . cooperative study . JAMA . 1974 ; 229 ( 12 ): 1606 - 1613 . 332 . Schmitz-Rode T , Janssens U , Schild HH , Basche S , 317 . Goldhaber SZ , Agnelli G , Levine MN ; The Bolus Alteplase Hanrath P , Günther RW . Fragmentation of massive pul- Pulmonary Embolism Group . Reduced dose bolus alteplase monary embolism using a pigtail rotation catheter . Chest . vs conventional alteplase infusion for pulmonary embolism 1998 ; 114 ( 5 ): 1427 - 1436 . thrombolysis. An international multicenter randomized trial . 333 . Kucher N . Catheter embolectomy for acute pulmonary Chest . 1994 ; 106 ( 3 ): 718 - 724 . embolism . Chest . 2007 ; 132 ( 2 ): 657 - 663 . 318 . Goldhaber SZ , Kessler CM , Heit J , et al . Randomised con- 334 . Kuo WT , Gould MK , Louie JD , Rosenberg JK , Sze DY , trolled trial of recombinant tissue plasminogen activator Hofmann LV . Catheter-directed therapy for the treatment versus urokinase in the treatment of acute pulmonary embo- of massive pulmonary embolism: systematic review and lism . Lancet . 1988 ; 2 ( 8606 ): 293 - 298 . meta-analysis of modern techniques . J Vasc Interv Radiol . 319 . Goldhaber SZ , Kessler CM , Heit JA , et al . Recombinant 2009 ; 20 ( 11 ): 1431 - 1440 . tissue-type plasminogen activator versus a novel dosing 335 . Margheri M , Vittori G , Vecchio S , et al . Early and long- regimen of urokinase in acute pulmonary embolism: a ran- term clinical results of AngioJet rheolytic thrombectomy domized controlled multicenter trial . J Am Coll Cardiol . in patients with acute pulmonary embolism . Am J Cardiol . 1992 ; 20 ( 1 ): 24 - 30 . 2008 ; 101 ( 2 ): 252 - 258 . 320 . Meneveau N , Schiele F , Vuillemenot A , et al . Strepto kinase 336 . Zhou WZ , Shi HB , Yang ZQ , et al . Value of percutanous vs alteplase in massive pulmonary embolism. A randomized catheter fragmentation in the management of massive trial assessing right heart haemodynamics and pulmonary pulmonary embolism . Chin Med J (Engl) . 2009 ; 122 ( 15 ): vascular obstruction . Eur Heart J . 1997 ; 18 ( 7 ): 1141 - 1148 . 1723 - 1727 . 321 . Meyer G , Sors H , Charbonnier B , et al ; The European 337 . Leacche M , Unic D , Goldhaber SZ , et al . Modern surgical Cooperative Study Group for Pulmonary Embolism . Effects treatment of massive pulmonary embolism: results in 47 of intravenous urokinase versus alteplase on total pulmo- consecutive patients after rapid diagnosis and aggressive nary resistance in acute massive pulmonary embolism: a surgical approach . J Thorac Cardiovasc Surg . 2005 ; 129 ( 5 ): European multicenter double-blind trial . J Am Coll Cardiol . 1018 - 1023 . 1992 ; 19 ( 2 ): 239 - 245 . 338 . Meneveau N , Séronde MF , Blonde MC , et al . Management 322 . Sors H , Pacouret G , Azarian R , Meyer G , Charbonnier B , of unsuccessful thrombolysis in acute massive pulmonary Simonneau G . Hemodynamic effects of bolus vs 2-h embolism . Chest . 2006 ; 129 ( 4 ): 1043 - 1050 . infusion of alteplase in acute massive pulmonary embolism. 339 . Sukhija R , Aronow WS , Lee J , et al . Association of right A randomized controlled multicenter trial. Chest . 1 9 9 4 ; ventricular dysfunction with in-hospital mortality in patients 106 ( 3 ): 712 - 717 . with acute pulmonary embolism and reduction in mortality 323 . Tebbe U , Bramlage P , Graf A , et al . Desmoteplase in acute in patients with right ventricular dysfunction by pulmonary massive pulmonary thromboembolism . Thromb Haemost . embolectomy . Am J Cardiol . 2005 ; 95 ( 5 ): 695 - 696 . 2009 ; 101 ( 3 ): 557 - 562 . 340 . Fukuda I , Taniguchi S , Fukui K , Minakawa M , Daitoku K , 324 . Tebbe U, Graf A , Kamke W , et al . Hemodynamic effects Suzuki Y . Improved outcome of surgical pulmonary embo- of double bolus reteplase versus alteplase infusion in mas- lectomy by aggressive intervention for critically ill patients . sive pulmonary embolism. Am Heart J . 1999 ; 138 ( 1 Pt 1 ): Ann Thorac Surg . 2011 ; 91 ( 3 ): 728 - 732 . 39 - 44 . 341 . Beckman JA , Dunn K , Sasahara AA , Goldhaber SZ . 325 . Wang C , Zhai Z , Yang Y , et al ; China Venous Throm- Enoxaparin monotherapy without oral anticoagulation to boembolism (VTE) Study Group . Effi cacy and safety of low treat acute symptomatic pulmonary embolism. Thromb dose recombinant tissue-type plasminogen activator for the Haemost . 2003 ; 89 ( 6 ): 953 - 958 . treatment of acute pulmonary thromboembolism: a ran- 342 . K u c h e r N , Q u i r o z R , M c K e a n S , S a s a h a r a A A , G o l d h a b e r S Z . domized, multicenter, controlled trial . Chest . 2010 ; 137 ( 2 ): Extended enoxaparin monotherapy for acute symptomatic 254 - 262 . pulmonary embolism. Vasc Med . 2005 ; 10 ( 4 ): 251 - 256 . 326 . Wang C , Zhai Z , Yang Y , et al; China Venous Throm- 343 . Schulman S , Kearon C , Kakkar AK , et al ; RE-COVER boembolism Study Group . Effi cacy and safety of 2-hour Study Group . Dabigatran versus warfarin in the treat- urokinase regime in acute pulmonary embolism: a random- ment of acute venous thromboembolism . N Engl J Med . ized controlled trial . Respir Res . 2009 ; 10 : 128 . 2009 ; 361 ( 24 ): 2342 - 2352 . 327 . Meneveau N , Schiele F , Metz D , et al . Comparative effi cacy 344 . Gosselin MV , Rubin GD , Leung AN , Huang J , Rizk NW . of a two-hour regimen of streptokinase versus alteplase in Unsuspected pulmonary embolism: prospective detec- acute massive pulmonary embolism: immediate clinical and tion on routine helical CT scans. Radiology . 1998 ; 208 ( 1 ): hemodynamic outcome and one-year follow-up . J Am Coll 209 - 215 . Cardiol . 1998 ; 31 ( 5 ): 1057 - 1063 . 345 . O’Connell CL , Boswell WD , Duddalwar V , et al . 328 . Verstraete M , Miller GAH , Bounameaux H , et al . Intra- Unsuspected pulmonary emboli in cancer patients: clin- venous and intrapulmonary recombinant tissue-type plasmin- ical correlates and relevance. J Clin Oncol . 2 0 0 6 ; 2 4 ( 3 0 ) : ogen activator in the treatment of acute massive pulmonary 4928 - 4932 . embolism . Circulation . 1988 ; 77 ( 2 ): 353 - 360 . 346 . Sebastian AJ , Paddon AJ . Clinically unsuspected pulmonary 329 . de Gregorio M , Gimeno M , Alfonso R , et al . [Mechanical embolism—an important secondary fi nding in oncology CT . fragmentation and intrapulmonary fi brinolysis in the treat- Clin Radiol . 2006 ; 61 ( 1 ): 81 - 85 . e490S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 347 . Storto ML , Di Credico A , Guido F , Larici AR , Bonomo L . ACCP evidence-based clinical practice guidelines . Chest . Incidental detection of pulmonary emboli on routine MDCT 2004 ; 126 ( 1 suppl ): 63S - 71S . of the chest . AJR Am J Roentgenol . 2005 ; 184 ( 1 ): 264 - 267 . 365 . Mehta S , Helmersen D , Provencher S , et al ; for the Canadian 348 . Douma RA , Kok MG , Verberne LM , Kamphuisen PW , Thoracic Society Pulmonary Vascular Disease - CTEPH Büller HR . Incidental venous thromboembolism in cancer CPG Development Committee ; and the Canadian Thoracic patients: prevalence and consequence . Thromb Res . 2010 ; Society Canadian Respiratory Guidelines Committee . 125 ( 6 ): e306 - e309 . Diagnostic evaluation and management of chronic thrombo- 349 . Farrell C , Jones M , Girvin F , Ritchie G , Murchison JT . embolic pulmonary hypertension: A clinical practice guide- Unsuspected pulmonary embolism identifi ed using multi- line . Can Respir J . 2010 ; 17 ( 6 ): 301 - 334 . detector computed tomography in hospital outpatients . Clin 366 . Mayer E , Jenkins D , Lindner J , et al . Surgical management Radiol . 2010 ; 65 ( 1 ): 1 - 5 . and outcome of patients with chronic thromboembolic pulmo- 350 . Hui GC , Legasto A , Wittram C . The prevalence of symp- nary hypertension: results from an international prospective tomatic and coincidental pulmonary embolism on computed registry . J Thorac Cardiovasc Surg . 2011 ; 141 ( 3 ): 702 - 710 . tomography . J Comput Assist Tomogr . 2008 ; 32 ( 5 ): 783 - 787 . 367 . Pepke-Zaba J, Delcroix M, Lang I, et al. Chronic thrombo- 351 . den Exter PL , Hooijer J , Dekkers OM , Huisman MV . Risk embolic pulmonary hypertension (CTEPH). Results from of recurrent venous thromboembolism and mortality in an international prospective registry. Circulation . 1 2 4 ( 1 8 ) : patients with cancer incidentally diagnosed with pulmonary 1973-1981. embolism: a comparison with symptomatic patients . J Clin 368 . Feinstein JA , Goldhaber SZ , Lock JE , Ferndandes SM , Oncol . 2011 ; 29 ( 17 ): 2405 - 2409 . Landzberg MJ . Balloon pulmonary angioplasty for treat- 352 . Dentali F , Ageno W , Pierfranceschi MG , et al . Prognostic ment of chronic thromboembolic pulmonary hypertension . relevance of an asymptomatic venous thromboembolism Circulation . 2001 ; 103 ( 1 ): 10 - 13 . in patients with cancer. J Thromb Haemost . 2 0 1 1 ; 9 ( 5 ) : 369 . Bresser P , Pepke-Zaba J , Jaïs X , Humbert M , Hoeper MM . 1081 - 1083 . Medical therapies for chronic thromboembolic pulmonary 353 . O’Connell C , Razavi P , Ghalichi M , et al . Unsuspected hypertension: an evolving treatment paradigm . Proc Am pulmonary emboli adversely impact survival in patients Thorac Soc . 2006 ; 3 ( 7 ): 594 - 600 . with cancer undergoing routine staging multi-row detector 370 . Becattini C , Manina G , Busti C , Gennarini S , Agnelli G . computed tomography scanning . J Thromb Haemost . 2011 ; Bosentan for chronic thromboembolic pulmonary hyperten- 9 ( 2 ): 305 - 311 . sion: fi ndings from a systematic review and meta-analysis . 354 . Becattini C , Agnelli G , Pesavento R , et al . Incidence of Thromb Res . 2010 ; 126 ( 1 ): e51 - e56 . chronic thromboembolic pulmonary hypertension after a fi rst 371 . Decousus H , Epinat M , Guillot K , Quenet S , Boissier C , episode of pulmonary embolism . Chest . 2006 ; 130 ( 1 ): 172 - 175 . T a r d y B . S u p e r fi cial vein thrombosis: risk factors, diagnosis, 355 . Pengo V , Lensing AWA , Prins MH , et al ; Thromboembolic and treatment . Curr Opin Pulm Med . 2003 ; 9 ( 5 ): 393 - 397 . Pulmonary Hypertension Study Group . Incidence of chronic 372 . Wichers IM , Di Nisio M , Büller HR , Middeldorp S . thromboembolic pulmonary hypertension after pulmonary Treatment of superfi cial vein thrombosis to prevent deep embolism . N Engl J Med . 2004 ; 350 ( 22 ): 2257 - 2264 . vein thrombosis and pulmonary embolism: a systematic 356 . Ribeiro A , Lindmarker P , Johnsson H , Juhlin-Dannfelt A , review . Haematologica . 2005 ; 90 ( 5 ): 672 - 677 . Jorfeldt L . Pulmonary embolism: one-year follow-up with 373 . Quenet S , Laporte S , Décousus H , Leizorovicz A , Epinat M , echocardiography Doppler and fi ve-year survival analysis . Mismetti P ; STENOX Group . Factors predictive of venous Circulation . 1999 ; 99 ( 10 ): 1325 - 1330 . thrombotic complications in patients with isolated superfi - 357 . Dentali F , Donadini M , Gianni M , et al . Incidence of cial vein thrombosis . J Vasc Surg . 2003 ; 38 ( 5 ): 944 - 949 . chronic pulmonary hypertension in patients with previous 374 . Decousus H , Quéré I , Presles E , et al ; POST (Prospective pulmonary embolism . Thromb Res . 2009 ; 124 ( 3 ): 256 - 258 . Observational Superfi cial Thrombophlebitis) Study Group . 358 . K l i n e J A , S t e u e r w a l d M T , M a r c h i c k M R , H e r n a n d e z - N i n o J , Superfi cial venous thrombosis and venous thromboembo- Rose GA . Prospective evaluation of right ventricular func- lism: a large, prospective epidemiologic study . Ann Intern tion and functional status 6 months after acute submassive Med . 2010 ; 152 ( 4 ): 218 - 224 . pulmonary embolism: frequency of persistent or subsequent 375 . Di Nisio M , Wichers IM , Middeldorp S . Treatment for elevation in estimated pulmonary artery pressure . Chest . superfi cial thrombophlebitis of the leg . Cochrane Database 2009 ; 136 ( 5 ): 1202 - 1210 . Syst Rev . 2007 ;(2): CD004982 . 359 . Hoeper MM , Mayer E , Simonneau G , Rubin LJ . Chronic 376 . Superfi cial Thrombophlebitis Treated By Enoxaparin Study thromboembolic pulmonary hypertension. Circulation . Group . A pilot randomized double-blind comparison of a 2006 ; 113 ( 16 ): 2011 - 2020 . low-molecular-weight heparin, a nonsteroidal anti-infl am- 360 . Jamieson SW , Kapelanski DP , Sakakibara N , et al . matory agent, and placebo in the treatment of superfi cial Pulmonary endarterectomy: experience and lessons learned vein thrombosis . Arch Intern Med . 2003 ; 163 ( 14 ): 1657 - 1663 . in 1,500 cases . Ann Thorac Surg . 2003 ; 76 ( 5 ): 1457 - . 377 . Titon JP , Auger D , Grange P , et al . [Therapeutic man- 361 . Fedullo PF , Auger WR , Kerr KM , Rubin LJ . Chronic agement of superfi cial venous thrombosis with calcium thromboembolic pulmonary hypertension . N Engl J Med . nadroparin. Dosage testing and comparison with a non- 2001 ; 345 ( 20 ): 1465 - 1472 . steroidal anti-infl ammatory agent] . Ann Cardiol Angeiol 362 . Bernard J , Yi ES . Pulmonary thromboendarterectomy: (Paris) . 1994 ; 43 ( 3 ): 160 - 166 . a clinicopathologic study of 200 consecutive pulmonary 378 . Prandoni P , Tormene D , Pesavento R ; Vesalio Investigators thromboendarterectomy cases in one institution. Hum Group . High vs. low doses of low-molecular-weight hep- Pathol . 2007 ; 38 ( 6 ): 871 - 877 . arin for the treatment of superfi cial vein thrombosis of the 363 . Thistlethwaite PA , Kaneko K , Madani MM , Jamieson SW . legs: a double-blind, randomized trial . J Thromb Haemost . Technique and outcomes of pulmonary endarterectomy sur- 2005 ; 3 ( 6 ): 1152 - 1157 . gery . Ann Thorac Cardiovasc Surg . 2008 ; 14 ( 5 ): 274 - 282 . 379 . Marchiori A , Verlato F , Sabbion P , et al . High versus low 364 . Doyle RL , McCrory D , Channick RN , Simonneau G , Conte doses of unfractionated heparin for the treatment of super- J ; American College of Chest Physicians . Surgical treat- fi cial thrombophlebitis of the leg. A prospective, controlled, ments/interventions for pulmonary arterial hypertension: randomized study. Haematologica . 2002 ; 87 ( 5 ): 523 - 527 . www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e491S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians 380 . Belcaro G , Nicolaides AN , Errichi BM , et al. Superfi cial 400 . Karabay O , Yetkin U , Onol H . Upper extremity deep vein thrombophlebitis of the legs: a randomized, controlled, thrombosis: clinical and treatment characteristics . J Int Med follow-up study . Angiology . 1999 ; 50 ( 7 ): 523 - 529 . Res . 2004 ; 32 ( 4 ): 429 - 435 . 381 . Andreozzi GM , Signorelli S , Di Pino L , et al . Tolerability 401 . Kovacs MJ , Kahn SR , Rodger M , et al. A pilot study of cen- and clinical effi cacy of desmin in the treatment of superfi cial tral venous catheter survival in cancer patients using low- thrombovaricophlebitis . Angiology . 1996 ; 47 ( 9 ): 887 - 894 . molecular-weight heparin (dalteparin) and warfarin without 382 . Decousus H , Prandoni P , Mismetti P , et al ; CALISTO Study catheter removal for the treatment of upper extremity deep Group . Fondaparinux for the treatment of superfi cial-vein vein thrombosis (The Catheter Study). J Thromb Haemost . thrombosis in the legs. N Engl J Med . 2010 ; 363 ( 13 ): 1222 - 1232 . 2007 ;5(8):1650-1653. 383. Blondon M , Righini M , Bounameaux H , et al. Fondaparinux 402 . AbuRahma AF , Short YS , White JF III , Boland JP . for isolated superfi cial-vein thrombosis of the legs: a cost- Treatment alternatives for axillary-subclavian vein throm- effectiveness analysis. Chest . In press. 10.1378/chest.11-0625. bosis: long-term follow-up. Cardiovasc Surg . 1 9 9 6 ; 4 ( 6 ) : 384 . Lozano FS , Almazan A . Low-molecular-weight heparin 783 - 787 . versus saphenofemoral disconnection for the treatment of 403 . Horne MK III , Mayo DJ , Cannon RO III , Chen CC , above-knee greater saphenous thrombophlebitis: a prospec- Shawker TH , Chang R . Intraclot recombinant tissue tive study . Vasc Endovascular Surg . 2003 ; 37 ( 6 ): 415 - 420 . plasminogen activator in the treatment of deep venous 385 . Sullivan V , Denk PM , Sonnad SS , Eagleton MJ , Wakefi eld TW . thrombosis of the lower and upper extremities . Am J Med . Ligation versus anticoagulation: treatment of above-knee 2000 ; 108 ( 3 ): 251 - 255 . superfi cial thrombophlebitis not involving the deep venous 404 . Lee JT , Karwowski JK , Harris EJ , Haukoos JS , Olcott C system . J Am Coll Surg . 2001 ; 193 ( 5 ): 556 - 562 . IV . Long-term thrombotic recurrence after nonoperative 386 . Flinterman LE , Van Der Meer FJ , Rosendaal FR , Doggen CJ . management of Paget-Schroetter syndrome . J Vasc Surg . Current perspective of venous thrombosis in the upper 2006 ; 43 ( 6 ): 1236 - 1243 . extremity . J Thromb Haemost . 2008 ; 6 ( 8 ): 1262 - 1266 . 405 . Lokanathan R , Salvian AJ , Chen JC , Morris C , Taylor DC , 387 . L e c h n e r D , W i e n e r C , W e l t e r m a n n A , E i s c h e r L , Hsiang YN . Outcome after thrombolysis and selective tho- Eichinger S , Kyrle PA . Comparison between idiopathic racic outlet decompression for primary axillary vein throm- deep vein thrombosis of the upper and lower extremity bosis . J Vasc Surg . 2001 ; 33 ( 4 ): 783 - 788 . regarding risk factors and recurrence. J Thromb Haemost . 406 . Pegis JD , Papon X , Pasco A , Regnard O , Abraham P , Enon B . 2008 ; 6 ( 8 ): 1269 - 1274 . [In situ thrombolysis in the treatment of venous thrombo- 3 8 8 . S p e n c e r F A , E m e r y C , L e s s a r d D , G o l d b e r g R J ; W o r c e s t e r sis of effort in the arm [in French] . J Mal Vasc . 1997 ; 22 ( 3 ): Venous Thromboembolism Study . Upper extremity deep 187 - 192 . vein thrombosis: a community-based perspective . Am J 407 . Petrakis IE , Katsamouris A , Kafassis E , D’Anna M , Med . 2007 ; 120 ( 8 ): 678 - 684 . Sciacca V V . Two Different Therapeutic Modalities in the 389 . Owens CA , Bui JT , Knuttinen MG , Gaba RC , Carrillo TC . Treatment of the Upper Extremity Deep Vein Thrombosis: Pulmonary embolism from upper extremity deep vein throm- Preliminary Investigation With 20 Case Reports . Int J bosis and the role of superior vena cava fi lters: a review of Angiol . 2000 ; 9 ( 1 ): 46 - 50 . the literature . J Vasc Interv Radiol . 2010 ; 21 ( 6 ): 779 - 787 . 408 . Sabeti S , Schillinger M , Mlekusch W , Haumer M , Ahmadi 390 . Mai C , Hunt D . Upper-extremity deep venous thrombosis: R , M i n a r E . T r e a t m e n t o f s u b c l a v i a n - a x i l l a r y v e i n t h r o m b o - a review . Am J Med . 2011 ; 124 ( 5 ): 402 - 407 . sis: long-term outcome of anticoagulation versus systemic 391 . Becker DM , Philbrick JT , Walker FB IV . Axillary and sub- thrombolysis . Thromb Res . 2002 ; 108 ( 5-6 ): 279 - 285 . clavian venous thrombosis. Prognosis and treatment . Arch 409 . Schindler J , Bona RD , Chen HH , et al . Regional throm- Intern Med . 1991 ; 151 ( 10 ): 1934 - 1943 . bolysis with urokinase for central venous catheter-related 392 . Joffe HV , Kucher N , Tapson VF , Goldhaber SZ ; Deep Vein thrombosis in patients undergoing high-dose chemotherapy Thrombosis (DVT) FREE Steering Committee . Upper- with autologous blood stem cell rescue . Clin Appl Thromb extremity deep vein thrombosis: a prospective registry of Hemost . 1999 ; 5 ( 1 ): 25 - 29 . 592 patients . Circulation . 2004 ; 110 ( 12 ): 1605 - 1611 . 410 . Kahn SR , Elman EA , Bornais C , Blostein M , Wells PS . Post- 3 9 3 . P r a n d o n i P , P o l i s t e n a P , B e r n a r d i E , e t a l . U p p e r - e x t r e m i t y thrombotic syndrome, functional disability and quality of deep vein thrombosis. Risk factors, diagnosis, and complica- life after upper extremity deep venous thrombosis in adults . tions . Arch Intern Med . 1997 ; 157 ( 1 ): 57 - 62 . Thromb Haemost . 2005 ; 93 ( 3 ): 499 - 502 . 394 . Kucher N . Clinical practice. Deep-vein thrombosis of the 411 . Machleder HI . Evaluation of a new treatment strategy for upper extremities . N Engl J Med . 2011 ; 364 ( 9 ): 861 - 869 . Paget-Schroetter syndrome: spontaneous thrombosis of the 395 . Prandoni P , Bernardi E , Marchiori A , et al . The long term axillary-subclavian vein . J Vasc Surg . 1993 ; 17 ( 2 ): 305 - 315 . , clinical course of acute deep vein thrombosis of the arm: discussion 316-317. prospective cohort study . BMJ . 2004 ; 329 ( 7464 ): 484 - 485 . 412 . Schneider DB , Dimuzio PJ , Martin ND , et al . Combination 396 . Elman EE , Kahn SR . The post-thrombotic syndrome after treatment of venous thoracic outlet syndrome: open surgical upper extremity deep venous thrombosis in adults: a system- decompression and intraoperative angioplasty . J Vasc Surg . atic review . Thromb Res . 2006 ; 117 ( 6 ): 609 - 614 . 2004 ; 40 ( 4 ): 599 - 603 . 397 . Baskin JL , Pui CH , Reiss U , et al . Management of occlusion 413 . Spence LD , Gironta MG , Malde HM , Mickolick CT , and thrombosis associated with long-term indwelling central Geisinger MA , Dolmatch BL . Acute upper extremity deep venous catheters . Lancet . 2009 ; 374 ( 9684 ): 159 - 169 . venous thrombosis: safety and effectiveness of superior vena 398 . Saber W , Moua T , Williams EC , et al . Risk factors for caval fi lters . Radiology . 1999 ; 210 ( 1 ): 53 - 58 . catheter-related thrombosis (CRT) in cancer patients: a 414 . F e u g i e r P , A l e k s i c I , S a l a r i R , D u r a n d X , C h e v a l i e r J M . patient-level data (IPD) meta-analysis of clinical trials and Long-term results of venous revascularization for Paget- prospective studies . J Thromb Haemost . 2011 ; 9 ( 2 ): 312 - 319 . Schroetter syndrome in athletes. Ann Vasc Surg . 2 0 0 1 ; 399 . Savage KJ , Wells PS , Schulz V , et al . Outpatient use of low 15 ( 2 ): 212 - 218 . molecular weight heparin (Dalteparin) for the treatment 415 . Lee MC , Grassi CJ , Belkin M , Mannick JA , Whittemore AD , of deep vein thrombosis of the upper extremity . Thromb Donaldson MC . Early operative intervention after throm- Haemost . 1999 ; 82 ( 3 ): 1008 - 1010 . bolytic therapy for primary subclavian vein thrombosis: e492S Antithrombotic Therapy for VTE

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians an effective treatment approach. J Vasc Surg . 1998 ; 27 ( 6 ): 434 . M o h i u d d i n S M , H i l l e m a n D E , D e s t a c h e C J , S t o y s i c h A M , 1101 - 1107 . , discussion 1107-1108. Gannon JM , Sketch MH Sr . Effi cacy and safety of early ver- 416 . Malcynski J , O’Donnell TF Jr , Mackey WC , Millan VA . sus late initiation of warfarin during heparin therapy in acute Long-term results of treatment for axillary subclavian vein thromboembolism . Am Heart J . 1992 ; 123 ( 3 ): 729 - 732 . thrombosis . Can J Surg . 1993 ; 36 ( 4 ): 365 - 371 . 435 . van Dongen CJJ , van den Belt AG , Prins MH , Lensing AW . 417 . Meier GH , Pollak JS , Rosenblatt M , Dickey KW , Gusberg RJ . Fixed dose subcutaneous low molecular weight heparins ver- Initial experience with venous stents in exertional axil- sus adjusted dose unfractionated heparin for venous thrombo- lary-subclavian vein thrombosis . J Vasc Surg . 1996 ; 24 ( 6 ): embolism . Cochrane Database Syst Rev . 2 0 0 4 ; ( 4 ) : C D 0 0 1 1 0 0 . 974 - 981 . , discussion 981-983. 436 . Fiessinger JN , Lopez-Fernandez M , Gatterer E , et al . Once- 418 . Sanders RJ , Cooper MA . Surgical management of subcla- daily subcutaneous dalteparin, a low molecular weight hep- vian vein obstruction, including six cases of subclavian vein arin, for the initial treatment of acute deep vein thrombosis . bypass . Surgery . 1995 ; 118 ( 5 ): 856 - 863 . Thromb Haemost . 1996 ; 76 ( 2 ): 195 - 199 . 419 . Sheeran SR , Hallisey MJ , Murphy TP , Faberman RS , 437 . Harenberg J , Schmidt JA , Koppenhagen K , Tolle A , Huisman MV , Sherman S . Local thrombolytic therapy as part of a multi- Büller HR ; EASTERN Investigators . Fixed-dose, body disciplinary approach to acute axillosubclavian vein throm- weight-independent subcutaneous LMW heparin versus ad- bosis (Paget-Schroetter syndrome) . J Vasc Interv Radiol . justed dose unfractionated intravenous heparin in the initial 1997 ; 8 ( 2 ): 253 - 260 . treatment of proximal venous thrombosis . Thromb Haemost . 420 . Urschel HC Jr , Patel AN . Paget-Schroetter syndrome 2000 ; 83 ( 5 ): 652 - 656 . therapy: failure of intravenous stents . Ann Thorac Surg . 438 . Lindmarker P , Holmström M , Granqvist S , Johnsson H , 2003 ; 75 ( 6 ): 1693 - 1696 . L o c k n e r D . C o m p a r i s o n o f o n c e - d a i l y s u b c u t a n e o u s Fragmin 421 . Yilmaz EN , Vahl AC , van Heek NT, Vermeulen, EGJ, with continuous intravenous unfractionated heparin in the Rauwerda JA. Long-term results of local thrombolysis fol- treatment of deep vein thrombosis . Thromb Haemost . 1994 ; lowed by fi rst rib resection: an encouraging clinical experi- 72 ( 2 ): 186 - 190 . ence in treatment of subclavian vein thrombosis . Vasc Surg . 439 . Prandoni P , Lensing AWA , Büller HR , et al . Comparison 2000 ; 34 (1): 17 - 23 . of subcutaneous low-molecular-weight heparin with intra- 422 . Usoh F , Hingorani A , Ascher E , et al . Long-term follow-up venous standard heparin in proximal deep-vein thrombosis . for superior vena cava fi lter placement . Ann Vasc Surg . Lancet . 1992 ; 339 ( 8791 ): 441 - 445 . 2009 ; 23 ( 3 ): 350 - 354 . 440 . Riess H , Koppenhagen K , Tolle A , et al ; TH-4 Study Group . 423 . Martinelli I , Battaglioli T , Bucciarelli P , Passamonti SM , Fixed-dose, body weight-independent subcutaneous low Mannucci PM . Risk factors and recurrence rate of primary molecular weight heparin Certoparin compared with adjusted- deep vein thrombosis of the upper extremities . Circulation . dose intravenous unfractionated heparin in patients with 2004 ; 110 ( 5 ): 566 - 570 . proximal deep venous thrombosis . Thromb Haemost . 2003 ; 424 . Berzaczy D , Popovic M , Reiter M , et al . Quality of life in 90 ( 2 ): 252 - 259 . patients with idiopathic subclavian vein thrombosis . Thromb 441 . Simonneau G , Charbonnier B , Decousus H , et al . Res . 2010 ; 125 ( 1 ): 25 - 28 . Subcutaneous low-molecular-weight heparin compared 425 . Dentali F , Ageno W , Witt D , et al ; WARPED consortium . with continuous intravenous unfractionated heparin in the Natural history of mesenteric venous thrombosis in patients treatment of proximal deep vein thrombosis . Arch Intern treated with vitamin K antagonists: a multi-centre, retrospec- Med . 1993 ; 153 ( 13 ): 1541 - 1546 . tive cohort study . Thromb Haemost . 2009 ; 102 ( 3 ): 501 - 504 . 442 . Othieno R , Abu Affan M , Okpo E . Home versus in-patient 426 . Kumar S , Sarr MG , Kamath PS . Mesenteric venous throm- treatment for deep vein thrombosis . Cochrane Database bosis . N Engl J Med . 2001 ; 345 ( 23 ): 1683 - 1688 . Syst Rev . 2007 ;( 3 ): CD003076 . 427 . Martinelli I , Franchini M , Mannucci PM . How I treat rare 443 . PREPIC Study Group . Eight-year follow-up of patients with venous thromboses . Blood . 2008 ; 112 ( 13 ): 4818 - 4823 . permanent vena cava fi lters in the prevention of pulmonary 428 . Plessier A , Darwish-Murad S , Hernandez-Guerra M , et al ; embolism: the PREPIC (Prevention du Risque d’Embolie European Network for Vascular Disorders of the Liver Pulmonaire par Interruption Cave) randomized study . (EN-Vie) . Acute portal vein thrombosis unrelated to cirrhosis: Circulation . 2005 ; 112 ( 3 ): 416 - 422 . a prospective multicenter follow-up study. Hepatology . 444 . Aschwanden M , Labs KH , Engel H , et al . Acute deep 2010 ; 51 ( 1 ): 210 - 218 . vein thrombosis: early mobilization does not increase the 429 . Amitrano L , Guardascione MA , Scaglione M , et al . frequency of pulmonary embolism . Thromb Haemost . 2001 ; Prognostic factors in noncirrhotic patients with splanchnic 85 ( 1 ): 42 - 46 . vein thromboses. Am J Gastroenterol . 2 0 0 7 ; 1 0 2 ( 1 1 ) : 445 . Blättler W , Partsch H . Leg compression and ambulation is 2464 - 2470 . better than bed rest for the treatment of acute deep venous 430 . Condat B , Pessione F , Hillaire S , et al . Current outcome of thrombosis . Int Angiol . 2003 ; 22 ( 4 ): 393 - 400 . portal vein thrombosis in adults: risk and benefi t of antico- 446 . Jünger M , Diehm C , Störiko H , et al . Mobilization versus agulant therapy . Gastroenterology . 2001 ; 120 ( 2 ): 490 - 497 . immobilization in the treatment of acute proximal deep 431 . Amitrano L , Guardascione MA , Menchise A , et al . Safety venous thrombosis: a prospective, randomized, open, mul- and effi cacy of anticoagulation therapy with low molecular ticentre trial . Curr Med Res Opin . 2006 ; 22 ( 3 ): 593 - 602 . weight heparin for portal vein thrombosis in patients with 447 . Partsch H , Blättler W . Compression and walking versus liver cirrhosis . J Clin Gastroenterol . 2010 ; 44 ( 6 ): 448 - 451 . bed rest in the treatment of proximal deep venous throm- 432 . Darwish Murad S , Plessier A , Hernandez-Guerra M , et al ; bosis with low molecular weight heparin. J Vasc Surg . EN-Vie (European Network for Vascular Disorders of the 2000 ; 32 ( 5 ): 861 - 869 . Liver) . Etiology, management, and outcome of the Budd- 448 . Schellong SM , Schwarz T , Kropp J , Prescher Y , Beuthien- Chiari syndrome . Ann Intern Med . 2009 ; 151 ( 3 ): 167 - 175 . Baumann B , Daniel WG . Bed rest in deep vein thrombosis 433 . Darwish Murad S , Valla DC , de Groen PC , et al . and the incidence of scintigraphic pulmonary embolism . Determinants of survival and the effect of portosystemic Thromb Haemost . 1999 ; 82 ( suppl 1 ): 127 - 129 . shunting in patients with Budd-Chiari syndrome . Hepatology . 449 . Hull RD , Pineo GF , Brant RF , et al ; LITE Trial Investigators . 2004 ; 39 ( 2 ): 500 - 508 . Long-term low-molecular-weight heparin versus usual care www.chestpubs.org CHEST / 141 / 2 / FEBRUARY, 2012 SUPPLEMENT e493S

Downloaded from chestjournal.chestpubs.org at ACCP HQ on February 9, 2012 © 2012 American College of Chest Physicians in proximal-vein thrombosis patients with cancer . Am J Med . 452 . Belcaro G , Laurora G , Cesarone MR , De Sanctis MT . 2006 ; 119 ( 12 ): 1062 - 1072 . Prophylaxis of recurrent deep venous thrombosis. A ran- 450 . Kakkar VV , Gebska M , Kadziola Z , Saba N , Carrasco P ; domized, prospective study using indobufen and gradu- Bemiparin Investigators . Low-molecular-weight heparin in ated elastic compression stockings . Angiology . 1993 ; 44 ( 9 ): the acute and long-term treatment of deep vein thrombosis . 695 - 699 . Thromb Haemost . 2003 ; 89 ( 4 ): 674 - 680 . 453 . Arpaia G , Cimminiello C , Mastrogiacomo O , de Gaudenzi E . 451 . Kolbach DN , Sandbrink MW , Neumann HA , Prins MH . Efﬁ cacy of elastic compression stockings used early or after Compression therapy for treating stage I and II (Widmer) resolution of the edema on recanalization after deep venous post-thrombotic syndrome . Cochrane Database Syst Rev . thrombosis: the COM.PRE Trial . Blood Coagul Fibrinolysis . 2003 ; ( 4 ): CD004177 . 2007 ; 18 ( 2 ): 131 - 137 .

e494S Antithrombotic Therapy for VTE

Antithrombotic Therapy for VTE Disease

Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Clive Kearon, MD, PhD; Elie A. Akl, MD, MPH, PhD; Anthony J. Comerota, MD; Paolo Prandoni, MD; Henri Bounameaux, MD; Samuel Z. Goldhaber, MD, FCCP; Michael Nelson, MD, FCCP; Philip Wells, MD; Michael K. Gould, MD, MS, FCCP; Francesco Dentali, MD; Mark Crowther, MD; and Susan R. Kahn, MD

1 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded ( http :// www from . chestpubs chestjournal.chestpubs.org . org / site / misc / reprints . xhtml at ACCP ). DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) per 1,000 (from 31 fewer to 144 more) per 1,000 (from 44 fewer to 142 more) per 1,000 (from 4 fewer to 178 fewer) Anticoagulation Risk Difference With Parenteral With t. t.

a Parenteral 33 per 1,000 16 fewer 50 per 1,000 16 fewer 200 per 1,000 134 fewer Risk With No Risk With Anticoagulation (95% CI) (0.05-5.37) (0.12-3.85) (0.11-0.98) Relative Effect Summary of Findings Anticoagulation With Parenteral With Study Event Rates (%) Anticipated Absolute Effects 2/60 (3.3) 1/60 (1.7) RR 0.5 3/60 (5) 2/60 (3.3) RR 0.67 With No With Parenteral 12/60 (20) 4/60 (6.7) RR 0.33 Anticoagulation

b,c b,d b,c due to imprecision due to imprecision due to imprecision of Evidence Overall Quality Overall Mortality (important outcome) Bias Major bleeding (critical outcome) Publication Undetected Moderate Undetected Moderate Undetected Moderate c d c Serious Serious Serious Recurrent VTE (critical outcome; assessed with symptomatic extension or recurrence) indirectness indirectness indirectness No serious No serious No serious Quality Assessment Quality Assessment [Section 2.1] Evidence Proﬁ le: Parenteral Anticoagulation vs No in Acute VTE [Section 2.1] Evidence Proﬁ risk ratio. risk ratio. inconsistency inconsistency inconsistency 5 No serious No serious No serious

RR RR b b b Table S1 — S1 Table 1 Bias Inconsistency Indirectness Imprecision Risk of risk of bias risk of bias risk of bias No serious No serious No serious 6 mo 6 mo 6 mo Study described as double blinded; outcome adjudicators blinded. None of the study participants were lost to follow-up. Intention-to-treat analysis. Study was stopped early for beneﬁ Study described as double blinded; outcome adjudicators blinded. None of the study participants were lost to follow-up. Intention-to-treat Low number of events caused by the early stoppage trial. Both groups treated with acenocoumarol. Both groups treated with acenocoumarol. t or appreciable harm. CI includes values suggesting no effect as well either appreciable beneﬁ 120 (1 study),

Participants (Studies), Follow-up Bibliography: Brandjes et al. 120 (1 study),

120 (1 study), a b c d

2 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

a-d (95% CI) (95% CI) With Early With (and Shorter per 1,000 (from 14 fewer to 23 more) per 1,000 (from 28 fewer to 35 more) per 1,000 (from 9 fewer to 66 more) Risk Difference Warfarin Initiation Warfarin Duration Heparin) Duration 2 fewer 8 fewer 14 more i i i Anticipated Absolute Effects Delayed Warfarin Warfarin Heparin) Risk With Risk With Initiation (and 24 per 1,000 47 per 1,000 Longer Duration Longer 16 per 1,000 (95% CI) (0.41-1.95) (0.4-1.74) (0.68-3.23) Relative Effect Relative Summary of Findings Warfarin Warfarin Heparin) Duration Initiation With Early With (and Shorter s 10 in Hull et al) and a lower number of days of hospital stay (9.1 vs 13.0 in Gallus; in 13.0 vs (9.1 stay hospital of days of number lower a and al) et Hull in 10 s Study Event Rates (%) 13/338 (3.8) 12/350 (3.4) RR 0.9 14/338 (4.1) 12/350 (3.4) RR 0.83 (and Longer ) e With Delayed With Warfarin Initiation Warfarin Duration Heparin) Duration vitamin K antagonist. See Table S1 legend for expansion of other abbreviation. legend for expansion of other abbreviation. S1 Table vitamin K antagonist. See

5 g,h g,h 10/338 (3.0) 15/350 (4.3) RR 1.48 g,k,l Overall due to imprecision due to imprecision Evidence Quality of Bias Mortality (important outcome because 34% of subjects had mural thrombus rather than VTE, in addition to major methodologic limitations). 5 Major bleeding (critical outcome) Recurrent VTE (critical outcome) Publication Undetected High

l unfractionated heparin; VKA unfractionated heparin; VKA 5 Undetected Moderate Undetected Moderate h h imprecision Serious Serious No serious Excluded Mohiuddin et al 4 indirectness indirectness indirectness No serious No serious No serious pulmonary embolism; UFH pulmonary embolism; UFH Quality Assessment 5 L e r o y e r e t a l . 3 inconsistency inconsistency inconsistency No serious No serious No serious H u l l e t a l ,

g g g,k of bias of bias of bias No serious risk No serious risk No serious risk [Section 2.4] Evidence Proﬁ le: Early Warfarin (and Shorter Duration Heparin) vs Delayed Warfarin (and Longer Duration Heparin) for Acute VTE (and Shorter Duration Heparin) vs Delayed Warfarin le: Early Warfarin [Section 2.4] Evidence Proﬁ low-molecular-weight heparin; PE heparin; PE low-molecular-weight 5

j f f Table S2 — S2 Table 3 mo 3 mo 3 mo VKA therapy delayed for 4 to 10 d. VKA therapy delayed for 4 to 10 d. v 5 al; et Gallus in 9.5 vs (4.1 therapy heparin of days of number lower a with associated was VKA of initiation early The t and clinically important harm. The 95% CI on relative effect includes both clinically important beneﬁ Patientsand investigators were not blinded in two studies (Gallus et al and Leroyer et al) and were blinded in one study (Hull et al). Concealment was not clearly described but was probable in the Differences in death, independently of differences in recurrent VTE and major bleeding, is unlikely. Differences in death, independently of differences recurrent VTE and major bleeding, is unlikely. It is unclear whether bleeding was assessed at 10 d in all subjects or just while heparin was being administered, which could yield a biased estimate in favor of short-duration therapy in one study (Hull et al). al). et (Hull study one in therapy short-duration of favor in estimate biased a yield could which administered, being was heparin while just or subjects all in d 10 at assessed was bleeding whether unclear is It VKA therapy started within 1 day of starting heparin therapy (UFH in two studies and LMWH in one study). VKA therapy started within 1 day of starting heparin (UFH in two studies and LMWH one study). et al), and few had PE (only included in Gallus al). most DVT were symptomatic (asymptomatic included in Hull some had isolated distal DVT; Most patients had proximal DVT, Outcome assessment was at hospital discharge in the study by Gallus et al (although there was also extended follow-up) and 3 mo in the studies by Hull et al and Leroyer et al. 3 mo in the studies by Hull et al and Leroyer al. Outcome assessment was at hospital discharge in the study by Gallus et al (although there also extended follow-up) and Event rate corresponds to the median event rate in the included studies. Event rate corresponds to the median event in included studies. rst 10 d) in two studies (Gallus et al, Hull al) and at 3 mo one study (Leroyer al). Bleeding was assessed early (in hospital or in the ﬁ effect of shorter therapy on risk bleeding is not a major concern. Because the shorter duration of heparin therapy is very unlikely to increase bleeding, wide 95% CIs around relative B i b l i o g r a p h y : G a l l u s e t a l , 688 (3 studies), 688 (3 studies),

LMWH LMWH

a b c d e f g h i j k l 11.7 vs 14.7 in Hull; 11.9 vs 16.0 in Leroyer et al). 11.7 vs 14.7 in Hull; 11.9 16.0 Leroyer et al).

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision three studies. Primary outcome appears to have been assessed after a shorter duration of follow-up in the shorter treatment arm of one study nal analysis postrandomization (Gallus). because of earlier discharge from the subjects in this study were excluded from the ﬁ hospital, and 20% of

688 (3 studies),

3 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) per 1,000 fewer 11 (from to 25 more) per 1,000 fewer 20 (from to 19 more) per 1,000 (from 7 fewer to 30 more) With LMWH With Risk Difference 3 more 5 fewer 4 more c c c UFH 33 per 1,000 42 per 1,000 16 per 1,000 Risk With SC Risk With Effect Relative (0.68-1.76) (0.52-1.45) (0.56-2.9) (95% CI) Summary of Findings Study Event Rates (%) Anticipated Absolute Effects 31/780 (4) 34/786 (4.3) RR 1.1 31/777 (4) 26/786 (3.3) RR 0.87 15/815 (1.8) 19/819 (2.3) RR 1.27

a,b a,b b,c due to imprecision due to imprecision due to imprecision of Evidence SC UFH With LMWH With Overall Quality Bias Publication Mortality (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) Undetected Moderate Undetected Moderate Undetected Moderate subcutaneous. See Table S1 and S2 legends for expansion of other abbreviations. S1 and S2 legends for expansion of other abbreviations. subcutaneous. See Table b b b 5 SC SC 24 Serious Serious Serious Kearon et al. 23 indirectness indirectness indirectness No serious No serious No serious [Section 2.5.1] Evidence Proﬁ le: LMWH vs SC UFH for Initial Anticoagulation of Acute VTE le: LMWH vs SC UFH for Initial Anticoagulation of Acute VTE [Section 2.5.1] Evidence Proﬁ Quality Assessment Prandoni et al, 2 inconsistency inconsistency inconsistency Table S3 — S3 Table No serious No serious No serious

a a a Faivre et al, 21 risk of bias risk of bias risk of bias No serious No serious No serious

3 mo 3 mo 3 mo Precision judged from the perspective of whether SC heparin is noninferior to LMWH. The total number of events and the total number of participants are relatively low. number of participants are relatively low. Precision judged from the perspective of whether SC heparin is noninferior to LMWH. The total number events and In the two largest trials (Prandoni et al and Kearon et al, 87% of patients), allocation was concealed, outcome adjudicators and data analysts were concealed, analysis was intention to treat, and there were there and treat, to intention was analysis concealed, were analysts data and adjudicators outcome concealed, was allocation patients), of 87% al, et Kearon and al et (Prandoni trials largest two the In Event rate corresponds to the median event in included studies. 1,566 (3 studies),

a b c

no losses to follow-up. no losses to follow-up. 1,563 (3 studies), 1,634 (4 studies), Bibliography: Lopaciuk et al, Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

4 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Comments Comments 81% and popliteal or more distal DVT in 19%. Primary outcome was repeat venography. of proximal and distal not reported). Primary outcome was repeat venography. 77%, asymptomatic or distal 19%. in PE and 4%, in DVT Seventy percent of patients were treated entirely as an outpatient (76% of DVT and 39% of PE) Postran- domization exclusions in 10 patients receiving UFH and one patient receiving LMWH. 65%, distal DVT in 18%, and PE in 17%. Population: Femoral DVT in Population: DVT (proportion

3/72 (4.2%) 1/35 12/360 (3.3%) RR 1.0 (0.5-2.2) 22/352 (6.3%) RR 0.8 (0.4-1.5) (0.2-2.2) (0.2-1.3) 1/72 (1.4%) 3/35 7/360 (1.9%) RR 0.7 12/352 (3.4%) RR 0.5 (0.5-2.2) (0.5-2.3) Recurrent DVT or PE Major Bleeding Mortality Total a 72/75 1/72 (1.4%) 74/7429/35 0/74 RR 3.1 (0.1-7.5) 0/74RR 3.1 (0.1-7.5) 1/35 30/33 0/74RR 7.2 (0.4-137) 1/33 RR 0.9 (0.1-14.5) 0/33 RR 6.6 (0.3-123) 0/33 RR 2.8 (0.1-67) 345/355 13/345 (3.8%) 6/348 (1.7%) 18/348 (5.2%) Population: Proximal DVT in 360/360 14/360 (3.9%) RR 1.1 352/353 12/352 (3.4%) RR 1.1 360/360 15/360 (4.2%) 5/360 (1.4%) 12/360 (3.3%) Population: Proximal DVT in Patients Analyzed in which it was 10 d. in which it was 10 d. 22 6 . 5 d tional ف 91) 5 units; units) 261) Fixed Dose Heparin. See Table S1 and S2 legends for expansion of other abbreviations. legends for expansion of other abbreviations. S2 and S1 Table Fixed Dose Heparin. See 5 50 kg: 12,500 units; 5 50 kg: 4,000 units; , , [Section 2.5.1] LMWH vs SC UFH for Initial Anticoagulation of Acute VTE: Clinical Description and Results [Section 2.5.1] LMWH vs SC UFH for Initial Anticoagulation of Acute VTE: Clinical Description and Results 6.5 d Interventions ف 70 kg: 6,000 70 kg: 17,500 units)

by 250 units/kg SC bid initially and adjusted to aPTT for 10 d Units/kg SC bid for 10 d by 250 units/kg SC bid and adjusted to aPTT for 10 d Units IV followed by 150 International Units/kg SC bid for 10 d followed by 250 units/kg SC bid (no adjustment) for 6.3 d or enoxaparin (n or enoxaparin (n 100 International Units/kg SC bid for 7.1 d 50-70 kg: 5,000 . followed by SC bid doses (initially 50-70 kg: 15,000 units; . adjusted to aPTT for Units/kg SC bid for UFH 5,000 units IV followed Fraxiparine 97 International CY222 2,000 International Dalteparin (n UFH IV ( Table S4 — S4 Table /1992 UFH 5,000 units IV followed /2004 21 23 /2006 (FIDO) UFH 333 units/kg SC 24 /1988 22 activated prothrombin time; FIDO activated prothrombin time; FIDO 5 (Galilei) Follow-up was for 3 mo except the study by Faivre et al Author/Year (Acronym) Author/Year a Lopaciuk et al Faivreet al Prandoni et al

Kearon et al aPTT Nadroparin 85 Interna 85 Nadroparin

5 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians cacy analysis Comments Comments from the efﬁ were 10 (2.8%) for the UFH and one (0.2%) for the LMWH group.

ITT ITT Loss to Follow-up Analysis of UFH group who did not have repeat venography are assumed to have completed clinical follow-up Nil Three of CY222 group and six Not described ITT Nil ITT Postrandomization exclusions randomized controlled trial. See Table S1 , S2 , and S4 legends , S4 randomized for and S2 controlled expansion , trial. of Table See S1 other 5 Patients: CN Caregivers: PN Caregivers: CN Caregivers: CN Caregivers: CN Adjudicators: CY Adjudicators: CY Adjudicators: PN Adjudicators: CN Data Analysts: PY Data Analysts: PY Data Analysts: PN Data Analysts: PN probably yes; RCT 5 CY Concealed Blinding Randomization probably no; PY RCT RCT PN Patients: PN RCT PY Patients: CN RCT CY Patients: CN 5 Study Design 70 kg: 6.5 d . ف [Section 2.5.1] LMWH vs SC UFH for Initial Anticoagulation of Acute VTE: Methodologic Quality [Section 2.5.1] LMWH vs SC UFH for Initial Anticoagulation of Acute VTE: Methodologic Quality 50 kg: 12,500 intention to treat; PN , 5 6.5 d 261) or enoxaparin ف 5 50 kg: 4,000 units; Interventions , 91) 100 International Table S5 — S5 Table 5 70 kg: 17,500 units) adjusted 250 units/kg SC bid and adjusted to aPTT for 10 d units; 50-70 kg: 15,000 . to aPTT for Units/kg SC bid for 250 units/kg SC bid initially and adjusted to aPTT for 10 d IV followed by 150 International Units/kg SC bid for 10 d by 250 units/kg SC bid (no adjustment) for 6.3 d 50-70 kg: 5,000 units; 6,000 units) followed by SC bid doses (initially units/kg SC bid for 10 d (n Units/kg SC bid for 7.1 d Nadroparin 85 International CY222 2,000 International Units UFH 333 units/kg SC followed UFH IV ( Dalteparin (n Dalteparin (n certainly yes; ITT 5 /1992 UFH 5,000 units IV followed by /2004 21 23 /2006 /1988 UFH 5,000 units IV followed by 24 22 certainly no; CY 5 (FIDO) (Galilei) Kearon et al Author/Year (Acronym) Author/Year Lopaciuk et al

Prandoni et al

Faivre et al CN CN Fraxiparine 97 International abbreviations. abbreviations.

6 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

(from 2 fewer to 16 fewer) (from 6 fewer to 23 fewer) (from 8 fewer to 0 more) LMWH (95% CI) CI) (95% LMWH Risk Difference With With Difference Risk 10 fewer per 1,000 15 fewer per 1,000 5 fewer per 1,000 c c c UFH Anticipated Absolute Effects Risk With IV Risk With 15 per 1,000 46 per 1,000 55 per 1,000 Effect Relative (0.45-1) (0.66-0.95) (0.58-0.89) (95% CI) Summary of Findings Study Event Rates (%) 69/3,517 (2) 41/3,393 (1.2) RR 0.67 232/3,789 (6.1) 187/4,119 (4.5) RR 0.79 208/3,869 (5.4) 151/4,107 (3.7) RR 0.72 due to due to due to a,b a,b a,b,d Evidence IV UFH With LMWH With risk of bias, publication bias risk of bias, publication bias risk of bias, publication bias Overall Quality of Low Low Low

b b b Mortality (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) strongly suspected strongly suspected strongly suspected Reporting bias Reporting bias Reporting bias

d imprecision imprecision imprecision No serious No serious No serious t and no effect. t and no effect. See Table S1 and S2 legends for expansion of abbreviations. legends for expansion of abbreviations. S2 and S1 Table See [Section 2.5.1] Evidence Profi le: LMWH vs IV UFH for Initial Anticoagulation of Acute VTE le: LMWH vs IV UFH for Initial Anticoagulation of Acute VTE [Section 2.5.1] Evidence Profi 7,20 Quality Assessment indirectness indirectness indirectness No serious No serious No serious cant benefi Table S6 — S6 Table Included studies. 6 inconsistency inconsistency inconsistency No serious No serious No serious a a a Risk of Bias InconsistencySerious Indirectness Imprecision Publication Bias Serious Serious (17 studies), 3 mo (20 studies), 3 mo (17 studies), 3 mo Inverted funnel plot very suggestive of publication bias. Many of the included studies are of small size, and all are funded by industry. by industry. Inverted funnel plot very suggestive of publication bias. Many the included studies are small size, and all funded CI interval includes values suggesting signifi Of the 20 trials, allocation was concealed in nine and was unclear whether concealed in the remaining 11. Eighteen trials had blinded outcome assessors. Seven trials did not have any postrandomization Event rate corresponds to the median event in included studies.

a b c d Participants (Studies), Follow-up 7,908 exclusions or losses to follow-up. Ten trials reported the number of participants lost to follow-up, which ranged from 1.0% to 12.7%. One trial did not report the drop-outs. 12.7%. One trial did not report the drop-outs. trials reported the number of participants lost to follow-up, which ranged from 1.0% exclusions or losses to follow-up. Ten 6,910

7,976 Bibliography: van Dongen et al.

7 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) (from 15 fewer to 18 more) (from 6 fewer to 29 more) (from 7 fewer to 12 more) Risk Difference With Fondaparinux With 1 fewer per 1,000 2 fewer per 1,000 f g t. t. Anticipated Absolute Effects LMWH

Risk With Risk With 30 per 1,000 7 more per 1,000 12 per 1,000 41 per 1,000 2.0. 2.0. a-c . (95% CI) (0.8-1.97) (0.43-2.03) (0.64-1.45) Relative Effect Relative Summary of Findings RR 0.93 RR 0.96 f g With With Fondaparinux 12/1,098 (1.1) 43/1,098 (3.9) f g Study Event Rates (%) 33/1,107 (3.0) 41/1,098 (3.7) RR 1.25 45/1,107 (4.1) 13/1,107 (1.2) due due due d,e d,e d,e to imprecision to imprecision to imprecision of Evidence LMWH With Overall Quality 100 kg) SC once daily for at least 5 d and until VKAs induced an INR . 2.0. 2.0. . Mortality (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) Undetected Moderate Undetected Moderate Undetected Moderate e e e Serious Serious Serious indirectness indirectness indirectness 50 kg and 10.0 mg in patients weighing , No serious No serious No serious Quality Assessment [Section 2.5.1] Evidence Proﬁ le: Fondaparinux vs LMWH for Initial Anticoagulation of Acute DVT le: Fondaparinux vs LMWH for Initial Anticoagulation of Acute DVT [Section 2.5.1] Evidence Proﬁ inconsistency inconsistency inconsistency international normalized ratio. See Table S1 and S2 legends for expansion of other abbreviations. S1 and S2 legends for expansion of other abbreviations. international normalized ratio. See Table No serious No serious No serious Table S7 — S7 Table 5 INR INR

d d d 25 of bias of bias of bias No serious risk No serious risk No serious risk 3 mo 3 mo 3 mo Allocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Analysis excluded 0.6% of randomized patients. Not stopped early for benefi Allocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Analysis excluded 0.6% CI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. t or harm; relatively low number of events. CI includes values suggesting no effect and either benefi patients in the fondaparinux group and 13 Twelve in the LMWH group had a major bleeding during the initial period (7 d). Of these, two in the fondaparinux group and none in the LMWH group were Fondaparinux 7.5 mg (5.0 in patients weighing All patients had acute symptomatic DVT. Five fatal VTE in fondaparinux group and fi ve fatal VTE in LMWH group. ve fatal VTE in LMWH group. Five fatal VTE in fondaparinux group and fi

2,205 (1 study), 2,205 (1 study), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias a c d e f g fatal. 2,205 (1 study), Enoxaparin 1 mg/kg of body weight SC bid for at least 5 d and until VKAs induced an INR

Bibliography: Büller et al.

8 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians et al et 27 With LMWH With (from 13 fewer to 29 more) (from 24 fewer to 21 more) (from 6 fewer to 20 more) Once (95% CI) Once (95% CI) Risk Difference

Anticipated Absolute Effects a,b 31 per 1,000 2 more per 1,000 49 per 1,000 7 fewer per 1,000 12 per 1,000 2 more per 1,000 (0.57-1.94) (0.52-1.42) (95% CI) Bid Risk With (0.48-2.66) Relative Effect Relative Summary of Findings Once With LMWH With Study Event Rates (%) 9/772 (1.2) 10/750 (1.3) RR 1.13 20/647 (3.1) 20/614 (3.3) RR 1.05 32/647 (4.9) 26/614 (4.2) RR 0.86

28 28 due c,e due to due to c-e c,e,f inconsistency, inconsistency, imprecision inconsistency, imprecision to imprecision of Evidence Bid With Overall Quality Mortality (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) Undetected Low Undetected Low Undetected Moderate e e e Serious Serious Serious and in favor of once-daily dose Charbonnier et al. and in favor of once-daily dose Charbonnier et al. 16 16 indirectness indirectness indirectness No serious Quality Assessment [Section 2.5.2] Evidence Proﬁ le: LMWH Once vs Twice Daily for Initial Anticoagulation of Acute VTE le: LMWH Once vs Twice [Section 2.5.2] Evidence Proﬁ No serious No serious d f inconsistency Serious Serious No serious See Table S1, S2, and S5 legends for expansion of other abbreviations. S1, S2, and S5 legends for expansion of other abbreviations. See Table 26 Table S8 — S8 Table

c c c of bias of bias of bias No serious risk No serious risk No serious risk ve included studies, one included patients with PE and DVT, and four included only patients with DVT. All studies addressed the initial management of VTE. initial management of VTE. All studies addressed the and four included only patients with DVT. ve included studies, one patients with PE and DVT, ve included studies used four brands of LMWH (enoxaparin, tinzaparin, dalteparin, and nadroparin). In Merli et al, enoxaparin 1 mg/kg bid was compared with 1.5 mg/kg once daily. Holmström Holmström daily. once mg/kg 1.5 with compared was bid mg/kg 1 enoxaparin al, et Merli In nadroparin). and dalteparin, tinzaparin, (enoxaparin, LMWH of brands four used studies included ve 37%; point effect estimate in favor of bid dose Merli et al 65%; point effect estimate in favor of bid dose Merli et al 5

5 3 mo 3 mo 10 d

2 2 I Of the ﬁ I Imprecision judged relative to no difference. Imprecision judged relative to no difference. All included studies concealed allocation. Two studies had a double-blind design, and two others were single blind. One study did not mention blinding. ITT likely used in all studies. Participants were lost were Participants studies. all in used likely ITT blinding. mention not did study One blind. single were others two and design, double-blind a had studies Two allocation. concealed studies included All T h e ﬁ

1,261 (3 studies), 1,261 (3 studies), 1,522 (5 studies), to follow-up in only two studies (0.3% and 2.2%). to follow-up in only two studies (0.3% and 2.2%).

a b c d e f Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias Bibliography: van Dongen et al.

adjusted the dose to anti-Xa levels, which resulted in different daily doses after a number of days. In the remaining studies, the dose of the once-daily administration was double the dose of the twice-daily the of dose the double was administration once-daily the of dose the studies, remaining the In days. of number a after doses daily different in resulted which levels, anti-Xa to dose the adjusted administration (equal total daily dose).

9 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

h-k With Home With (from 14 fewer to 8 more) (from 25 fewer to 7 more) (from 7 fewer to 43 fewer) Risk Difference Treatment (95% CI) CI) (95% Treatment

See comment h-k Risk Anticipated Absolute Effects Treatment With Hospital With 21 per 1,000 7 fewer per 1,000 46 per 1,000 13 fewer per 1,000 74 per 1,000 29 fewer per 1,000 See comment See

a-d .001) at the end of the initial treatment. The 5

P Effect Relative (0.33-1.36) (0.45-1.15) (0.42-0.9) (95% CI) Summary of Findings … Not pooled t. Overall, the judgment was that these limitations would not eﬁ With Home With Treatment t. t. … .002) and social functioning ( 5 Study Event Rates (%)

Treatment P 18/851 (2.1) 12/857 (1.4) RR 0.67 39/851 (4.6) 28/857 (3.3) RR 0.72 63/851 (7.4) 39/857 (4.6) RR 0.61 With Hospital Hospital With

due a,b,e a,b,e,g due to due to showed potential benefi a,b,e,f i-k 35 assessed changes in quality of life using the Medical Outcome Study Short Form-36 in 300 patients of Evidence indirectness, imprecision due to indirectness indirectness, imprecision to indirectness 35 Overall Quality (mean hospital stay 2.7 in home arm vs 8.1 d in hospital arm) and Levine et al (2.1 vs 6.5 d). (mean hospital stay 2.7 in home arm vs 8.1 d arm) and Levine et al (2.1 6.5 d). 13 Bias Publication and O’Brien et al Mortality (important outcome) Undetected Moderate 13 Major bleeding (critical outcome) Recurrent VTE (critical outcome) Quality of life (important outcome) Undetected Low f See Table S1 and S2 for expansion of abbreviations. S1 and S2 for expansion of abbreviations. See Table imprecision 13,34,35 Serious No serious Seriousf Undetected Moderate Seriousk Undetected Low j a,b a,b a,b Serious Serious Serious Serious [Section 2.7] Evidence Profi le: Home Treatment vs Hospital Treatment of Acute DVT of Acute DVT vs Hospital Treatment le: Home Treatment [Section 2.7] Evidence Profi Quality of life.

i Quality Assessment 13,14,30-33 showed no effect, whereas Koopman et al Table S9 — S9 Table 34 inconsistency inconsistency inconsistency inconsistency No serious No serious No serious No serious They found that, the change in scores from baseline to day 7 was not signiﬁ cantly different between the treatment groups for seven of They the found eight that, domains. the The change one in exception scores from baseline to day 7 was not signiﬁ

; included studies. 14 e e e 29 of bias of bias of bias of bias No serious risk No serious risk No serious risk No serious risk reported evaluation of health-related quality of life using the EQ-5D. They found no differences in mean quality-of-life scores or in the proportion of patients showing improvement in 34 ), h 3 mo 3 mo 3 mo 3 mo Only one study (Boccalon et al) used LMWH in both treatment arms. Remaining studies used UFH in the inpatient arm and LMWH in the outpatient arm. in the outpatient arm. Only one study (Boccalon et al) used LMWH in both treatment arms. Remaining studies UFH the inpatient arm and rmed PE. Studies also excluded patients who were pregnant. DVT and excluded patients with suspected or confi All studies included patients with lower-extremity Bäckman et al Judged as precise based on the narrow CI around absolute effect. Judged as precise based on the narrow CI around absolute effect. of evidence by one level. we downgraded the quality together with the potential inconsistency, Not able to evaluate but imprecision is possible. Taken Out of six studies, allocation was clearly concealed in three (unclear in remaining three). Outcome adjudicators were blinded in the two largest studies (unclear in remaining). Four reported loss to follow-up follow-up to loss reported Four remaining). in (unclear studies largest two the in blinded were adjudicators Outcome three). remaining in (unclear three in concealed clearly was allocation studies, six of Out Four studies had partial hospital treatment of many in the home arm: Koopman et al (mean hospital stay 2.7 in home arm vs 8.1 d in hospital arm), Levine et al (2.1 vs 6.5 d), Boccalon et al (1 vs 9.6 d), and d), 9.6 vs (1 al et Boccalon d), 6.5 vs (2.1 al et Levine arm), hospital in d 8.1 vs arm home in 2.7 stay hospital (mean al et Koopman arm: home the in many of treatment hospital partial had studies Four Studies included in the systematic review should have recruited patients whose home circumstances were adequate. The CI includes both values suggesting benefi t and harm. t and harm. The CI includes both values suggesting benefi Potential inconsistency as Bäckman et al of the three studies had partial hospital treatment many in home arm: Koopman et al Two 1,708 (6 studies), 0 (3 studies

Bibliography: Othieno et al self-rated health state. Koopman et al evaluated health-related quality of life using the Medical Outcome Study Short Form-20 and an adapted version of the Rotterdam Symptom Checklist. The changes over time were similar in both groups except that the patients receiving LMWH had better scores for physical activity ( cance of results. O’Brien et al authors did not report enough data to assess precision and clinical signiﬁ

1,708 (6 studies), 1,708 (6 studies), participating in Levine et al.

a b c d e f g h i j k Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Ramacciotti et al (3 vs 7 d). In Daskalopoulos et al, there was no hospital stay at all in the home group. Chong et al did not report duration of hospital stay. report duration of hospital stay. Ramacciotti et al (3 vs 7 d). In Daskalopoulos al, there was no hospital stay at all in the home group. Chong did not was the domain of social functioning, where a greater improvement was observed for the outpatient group. was the domain of social functioning, where a greater improvement observed for outpatient group. cant in only a small (was study). signiﬁ ITT analysis was conducted in four (unclear in remaining two). No study was stopped early for ben warrant downgrading of quality because it has already been downgraded by at least one level based on other factors. warrant downgrading of quality because it has already been downgraded by at least one level based on other factors.

10 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

ab (Continued) Directed (95% CI) (95% CI) Thrombolysis With Catheter- With (from 23 fewer to 535 more) (from 123 fewer to 588 fewer) (from 39 fewer to 67 more) (from 48 fewer to 340 more) Risk Difference 318 fewer per 1,000 29 more per 1,000 j 34 fewer per 1,000 31 fewer per 1,000 e f f,h Anticipated Absolute Effects Directed Catheter- Thrombolysis Risk With No Risk With 39 per 1,000 48 per 1,000 29 per 1,000 588 per 1,000 (95% CI) (0-8.09) (0.01-2.71) (0.19-19.46) (0-0.79) Relative Effect Relative Summary of Findings RR 2.00 g Directed Thrombolysis With Catheter- With 0/77 (0) RR 0.14 e Study Event Rates (%) 3/76 (3.9) 1/76 (1.3) 0/77 (0) RR 0.35 0/76 (0) 1/77 (1.3) With No With Directed Catheter- 49/70 (70) 23/68 (33.8) RR 0.46 Thrombolysis due c,i due to due to due to c,d c,d c,d imprecision imprecision imprecision to indirectness of Evidence Overall Quality Mortality (important outcome) Undetected Moderate Undetected Low Undetected Low Undetected Low Nonfatal recurrent VTE (critical outcome) Nonfatal major bleeding (critical outcome) d d d imprecision Very serious Very serious Very serious Very No serious i indirectness indirectness indirectness No serious No serious No serious Serious Quality Assessment inconsistency inconsistency inconsistency inconsistency No serious No serious No serious No serious

c c c c [Section 2.9] Evidence Proﬁ le: Catheter-Directed Thrombolysis vs No Catheter-Directed Thrombolysis for Extensive Acute DVT of the Leg Thrombolysis vs No Catheter-Directed le: Catheter-Directed [Section 2.9] Evidence Proﬁ of bias of bias of bias of bias No serious risk No serious risk No serious risk No serious risk Postthrombotic syndrome (critical outcome; assessed with: complete lysis on venography (Elsharawy et al); patency on ultrasound and air plethysmography (Enden et al)) Postthrombotic syndrome (critical outcome; assessed with: complete lysis on venography (Elsharawy et al); patency ultrasound Table S10 — S10 Table 7 d 2 y 3 mo 3 mo 153 (2 studies), 138 (2 studies), 153 (1 study), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias 153 (2 studies),

11 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

Directed (95% CI) (95% CI) footnote Thrombolysis With Catheter- With Risk Difference See Anticipated Absolute Effects Directed Catheter- .078). They reported better overall overall better reported They .078). Thrombolysis Risk With No Risk With 5

P ... (95% CI) Relative Effect Relative Summary of Findings 1 y) who have a low risk of bleeding . . 1 y) who have a low risk of bleeding Ն 68 Directed Thrombolysis With Catheter- With Study Event Rates (%) .087) and improved Health Utilities Index ( tudy. tudy. 5

With No With Directed Catheter- P .006) compared with patients who were treated with anticoagulation alone. alone. anticoagulation with treated were who patients with compared .006) Thrombolysis 5

P postthrombotic syndrome. See Table S1 legend for expansion of other abbreviations. S1 legend for expansion of other abbreviations. postthrombotic syndrome. See Table 5 30 m,n of Evidence Continued Overall Quality

14 d, good functional status, life expectancy 39 , This probably underestimates PTS baseline risk given that overall, 52% of patients reported the current use of use current the reported patients of 52% overall, that given risk baseline PTS underestimates probably This 40 Table S10— Table Undetected Low .022), and fewer overall symptoms ( 5

P catheter-directed thrombolysis; PTS thrombolysis; PTS catheter-directed imprecision 5 No serious CDT CDT 38 indirectness No serious .033), less health distress ( 5

Quality Assessment P Comerota et al. 37 inconsistency t. t. No serious Enden et al, 36 .046), less stigma (

m,n P of bias

No serious risk 38 ), l Quality of life (important outcome; measured with the Medical Outcome Survey Short Form-12, Health Utilities Index MARK version 2/3 questionnaires; better indicated by lower values) Quality of life (important outcome; measured with the Medical Outcome Survey Short Form-12, Health Utilities Index MARK version 16 mo Participation rate was 65%. Participation rate was 65%. In selected patients with extensive acute proximal DVT (eg, iliofemoral DVT, symptoms for In selected patients with extensive acute proximal DVT (eg, iliofemoral DVT, t and harm. CI includes values suggesting both benefi rst 7 d. Most of bleeding events occur during the fi Recall was used to measure quality of life prior the thrombotic event; we did not consider these measurements. At the initial follow-up (mean, 16 mo), patients treated with CDT reported a trend toward a higher mental summary scale ( Three control patients died of cancer. Three control patients died of cancer. and nine patients had minor bleeding complications. CDT, one patient had “durable and partial impairment of sensibility the foot” immediately after receiving In the Enden et al study, (from 29 fewer to 138 fewer) over 2 y. Severe PTS: assuming the same RR of 0.46 and a baseline risk 13.8%, absolute reduction is 75 fewer severe PTS per 1,000 All patients were anticoagulated per protocol, but the intervention group received CDT in addition to anticoagulation. All patients were anticoagulated per protocol, but the intervention group received CDT in addition to anticoagulation. Allocation was concealed in Enden et al and unclear in Elsharawy et al. Outcome assessor blinded in both studies. Follow-up rates were 87% in Enden et al and 100% in Elsharawy et al. None of the Baseline risks for nonfatal recurrent VTE and major bleeding derived from Douketis et al. Surrogate outcome: absence of patency at 6 mo in Enden et al study; complete lysis Elsharawy s study. Outcomes) Thrombosis (Venous VETO the of ndings fi the on based is estimate This Camerota et al. 98 (1 study

role physical functioning (

a b c d e f g h i j compression stockings during study follow-up. k l m n o Bibliography: Elsharawy et al,

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias studies was stopped early for beneﬁ

12 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) (72%) a Results Results cant lysis: 69/87 (79%) 1 y: 78% at puncture site, no intervention/ no site, puncture at transfusion Femoral, 40% Partial lysis: 5/24 (21%) Bleeding: 6/24 (25%) Patency: 3 mo: 84% Partial lysis: 5/25 (20%) No lysis: 2/25 (8%) Complications: 1 small hematoma Signifi cant lysis: 18/25 Signifi Signifi Iliac (63%) Femoral (40%) No lysis: 18/87 (21%) PE: 1 (1%) 5/77 (6%); Bleeding: major, Patency at 1 y: 11/77 (14%) minor, Iliac: 63% 3 mo 1 y Early results: Outcomes Follow-up complications bleeding Clot lysis, Clot lysis, PE, Clot lysis, bleeding 13 mo (mean) cant lysis: 19/24 (79%) Signifi 50% received . Interventions (mean) infused over 30 h (mean), followed by heparin and then warfarin for 8-12 wk residual stenoses angioplasty (2) or stenting (14) urokinase infused for 75 h (mean) with 5,000 International Units bolus heparin plus infusion adjusted to aPTT (15), (52 limbs), stent (38), AVF surgical thrombectomy (13), mechanical thrombectomy (4), surgical bypass (3) with 1,000 U/h IV heparin, followed by heparin, adjusted to APTT thrombectomy (3) and stents (9) 4.9 million units urokinase Adjunctive therapy: limbs with 2,000-2,500 units/kg per h Adjunctive therapy: angioplasty 3 mg/h rt-PA (mean 86 mg) infused 3 mg/h rt-PA Adjunctive therapy: hydrodynamic 14 d 14 d Observational Studies of at Least 20 Patients) Observational Studies of at Least 20 Patients) Յ Յ 14 d (7) duration 14 d (18) duration Participants 14 d (16) . . Յ 14 d (8) duration . iliofemoral DVT (20) or iliofemoral DVT (69) or DVT or Type of Publication Type [Section 2.9] CDT vs No CDT for Extensive Acute DVT of the Leg: Clinical Description and Results (All Randomized Trials and Prospective and Prospective [Section 2.9] CDT vs No for Extensive Acute DVT of the Leg: Clinical Description and Results (All Randomized Trials /1997 Prospective study 24 patients with iliofemoral /1997 Prospective registry 77 patients (87 limbs) with 42 43 /1994 Prospective registry 21 patients (27 limbs) with 41 Table S11 — S11 Table Semba et al Author/Year Bjarnason et al et al Verhaeghe

13 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (83%) (Continued) (11%) 1%) b , ( b Results Results (1%) cant lysis: 1/33 (3%) cant lysis: 15/18 15/18 lysis: cant b 50% lysis: 54/312 (17%) Bleeding: 54/473 Femoral: 47% No lysis: 0 (0%) PE: 0 Bleeding: No lysis: 17/17 (100%) PE: 1/17 (6%) Bleeding: 0 Death: 2/473 6-mo patency: 8/33 (24%) Bleeding: 2/33 (6%) PE: 2/33 (6%) , PE: 6/473 6-mo patency: 15/18 (83%) Bleeding: 2/18 (11%) Patency at 1 y: Iliac: 64% Complete lysis: 11/18 (61%) Anticoagulation, 1 wk: Complete lysis: 0/17 (0%) 30-d signiﬁ signiﬁ 30-d 50%-100% lysis: 258/312 (83%) 1 wk and 6 mo 1 wk CDT, Anticoag: 6 mo Anticoagulation: 6 mo?CDT: CDT: 1 y Early results: Outcomes Follow-up bleeding bleeding bleeding, death Clot lysis, PE, Clot lysis, PE, Clot lysis, PE, 1 ف 50% received . Continued Interventions million units SK pulse-spray for 1 h, followed by 100,000 units/h SK infusion until complete lysis, no change in 12 h, or complication angioplasty/stent (1) 1,000-2000 units/h heparin infusion for 5-7 d. (mean) infused for 53.4 h (mean) in 297 limbs (no CDT) systemic infusion (54) dose 4,500 units urokinase followed by 4,500 units/kg per h for 24-48 h or 4-8-mg bolus of rt- PA followed by 2-4-mg/h infusion residual stenosis stents (10) Adjunctive therapy: Anticoagulation: 33 patients given 18 patients received CDT: 7.8 million units urokinase In 6 limbs, only systemic infusion Adjunctive therapy: stents (104), CDT: 18 patients given loading CDT: Adjunctive therapy: patients with Table S11— Table 10 d Յ 10 d , 14 d duration warfarin) orlysis Յ 10 d (99) duration 1 Participants . angio/stent (if needed). iliofemoral DVT given choice between conventional therapy (heparin 1 offered only to patients Lysis with DVT and no contraindications duration randomized to CDT or anticoagulation alone with lower limb DVT (188) or 35 patients with DVT 287 patients (312 infusions) multicenter registry Type of Publication Type RCT, single center RCT, Prospective /2001 Prospective study 51 patients (51 limbs) with /2002) /1999 45 36 44 (National Multicenter Registry) Mewissen et al Author/Year AbuRahma et al

Elsharawy et al

14 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Catheter- 5 Results Results stula; CaVenT stula; CaVenT 50%lysis: 44/50 (68%) PE: 0 Major bleeding: 0 (puncture site nerve damage: 0) 32/50 (64%) 19/53 (36%) Ն PE: 0 Major bleeding: 1/50 CDT, 6 mo CDT, Complete lysis: 13/18 (72%) No lysis: 0 (0%) Anticoagulation, 6 mo cant lysis:2/17 (12%) Signiﬁ No lysis: 7/17 (41%) Anticoagulation, 1 wk: not assessed Lysis 6 mo CDT, Iliofemoral patency: Anticoagulation, 6 mo Iliofemoral patency:

arteriovenous ﬁ 5 AVF AVF 46 1 wk and 6 mo 1 wk CDT, Outcomes Follow-up bleeding (PTS at 24 mo pending) streptokinase. See Table S2, S4, streptokinase. S5, See and Table S10 legends for expansion of other Clot lysis, PE, 5 Continued Interventions received 5,000 units heparin bolus, followed by heparin adjusted to aPTT maximum of 20 mg/24 h for 4 d. in four referral centers Treated practice, administered locally Anticoagulation: 17 patients CDT: tPA 0.01 mg/kg per h, tPA CDT: Anticoagulation alone: usual Table S11— Table recombinant tissue plasminogen activator; SK recombinant tissue plasminogen activator; SK 5 21 d , Participants duration Iliofemoral DVT 20 patients and retrospective studies are described in 3 Table of the eighth edition of these guidelines. , Type of Publication Type RCT, multicenter RCT, /2009 37 ve of 27 limbs treated with CDT; two could not be crossed with the guidewire. two could not be crossed with the guidewire. ve of 27 limbs treated with CDT; (CaVenT) Calculated from total number of patients in Venous Registry. Registry. Calculated from total number of patients in Venous Twenty-ﬁ Enden et al

Author/Year Thrombolysis in Directed Thrombosis; Acute Venous rt-PA Iliofemoral Vein a b

abbreviations. abbreviations. Early prospective observational studies with

15 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S12 —[Section 2.9] CDT vs No CDT for Extensive Acute DVT of the Leg: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to follow-up

Semba et al41 /1994 N/A N/A N/A N/A Verhaeghe et al42 /1997 N/A N/A N/A N/A Bjarnason et al43 /1997 N/A N/A N/A N/A Mewissen et al44 /1999 N/A N/A N/A N/A AbuRahma et al45 /2001 N/A N/A N/A N/A Elsharawy et al36 /2002 Computer-designated PN N for patients, caregivers, 0 cards and probably data analysts. Y for vascular imaging Enden et al37 /2009 Computer-designated Y No for patients and One loss to follow-up cards caregivers. Yes for (CDT), ﬁ ve vascular imaging. withdrawals, ﬁ ve postrandomization exclusions N 5 no; N/A 5 not applicable; Y 5 yes. See Table S10 legend for expansion of other abbreviation.

16 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Complete clot lysis: 3/5 (60%) Partial lysis: 1/5 (20%) No lysis: 1 (20%) PE: 0 Bleeding: 0 Complete clot lysis: 0/5 Partial lysis: 0/5 No lysis: 5/5 (100%)) PE: 0 Bleeding: 0 cant lysis: 5/8 (63%) Signifi Partial lysis: 2/8 (25%) No lysis: 1/8 (12%) Bleeding: Major: 2/8 (25%) Minor: 2/8 (25%) cant lysis: 1/8 (12%) Signifi Partial lysis: 2/8 (25%) No lysis: 5/8 (63%) Bleeding: Major: 1/8 (12%) Minor: 1/8 (12%) Complete clot lysis: 6/9 (67%) Partial lysis: 1/9 (11%) No lysis: 2/9 (22%) PE: 0 Bleeding: 4/10 (40%) Death: 2/9 (22%) Note: (1 patient excluded from treatment) Anticoagulation: Anticoagulation: Thrombolysis: 7-10 d Thrombolysis: 7 d Thrombolysis: 6-12 mo bleeding bleeding bleeding, death Clot lysis, PE, Clot lysis, Clot lysis, PE, 5 d 3 100 mg hydrocortisone then rst hour, for fi continued every 6 h for 3 d (5 patients) doses of heparin 5,000 units for 48 h followed by warfarin (5 patients) units over 90 min, then 100,000 units as maintenance dose for 22.5 h; heparin 500 mg given during 24 h, plus prednisone (8 patients) plus prednisone (8 patients) IV over 30 min; 900,000 units every 6 h (10 patients) Lysis: 600,000 units SK plus Lysis: Anticoagulation: 4-6 hourly Thrombolysis: SK 200,000 Anticoagulation: Heparin Thrombolysis: SK 500,000 units 4 d rmed , Compared Systemic Thrombolytic Therapy With No Thrombolytic Therapy) No Thrombolytic Therapy) Compared Systemic Thrombolytic Therapy With lower-extremity lower-extremity DVT confi by phlebography DVT DVT of 10 patients with 16 patients with 30 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results center center center RCT, single RCT, single RCT, [Section 2.10] Systemic Lysis vs No Systemic Lysis for Extensive Acute DVT of the Leg: Clinical Description and Results (Randomized Trials That Trials for Extensive Acute DVT of the Leg: Clinical Description and Results (Randomized vs No Systemic Lysis [Section 2.10] Systemic Lysis /1968 53 /1968 /1969 single RCT, 52 54 Table S13 — S13 Table Robertson et al Kakkar et al Browse et al Author/Year

17 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

s Complete/partial lysis: 10/19 (53%) No lysis: 9/19 (47%) PE: 0 Minor bleeding: 3 (16%) Complete/partial lysis: 1/15 (7%) No lysis: 14/15 (93%) PE: 1/15 (7%) Bleeding: 0 cant lysis: 39/92 (42%) Signifi Partial lysis: 23/92 (25%) No lysis: 30/92 (33%) PE: 7 (8%) Major bleeding: 58 (62%) Minor bleeding: 24 (26%) cant lysis: 0/42 (0%) Signifi Partial lysis: 4/42 (10%) No lysis: 38/42 (90%) PE: 5/42 (12%) Major bleeding: 2/42 (5%) Minor bleeding: 4/42 (10%) Complete lysis: 1/10 (10%) Partial lysis: 3/10 (30%) No lysis: 6/10 (60%) PE: 0 Bleeding: 0 Death: 0 Complete clot lysis: 2/9 (22%) Partial lysis: 2/9 (22%) No lysis: 5/9 (55%) PE: 1/10 Death: 2/9 (22%) Bleeding: 2/9 (22%) Note: (1 patient excluded from treatment) Anticoagulation: Anticoagulation: Arvin: Anticoagulation: Thrombolysis: 7 d ف :d Thrombolysis 7 bleeding bleeding Clot lysis, PE, Clot lysis, PE, Continued 5 d 3 rst dose Table S13— Table 5 d (10 patients) 3 80 units IV over 6 h; 80 units over 15 min; 40-80 units every 6 h 10,000 units IV over 5 min, then 10,000-15,000 units every 6 h (10 patients) initial dose of SK IV, initial dose of SK IV, then SK 100,000 units/h maintained and adjusted up to 72 h IV heparin for 1 wk 6-12 h post SK (19 patients) into affected limb, 7,000 units bolus then 1,500 units/h adjusted; continued for 7 d (15) SK calculated according to tolerance injected over 15-30 min; maintenance dose at 30 mL/h was two-thirds of fi (92 patients) heparin for initial dose followed by 25,000 units/24 h infusion (42 patients). Arvin: Arvin loading dose Anticoagulation: Heparin Thrombolysis: titrated Anticoagulation: Heparin IV Thrombolysis: initial dose Anticoagulation: 5,000 units 5 d , DVT of or subacute DVT 34 patients with 134 patients with acute Type of Type Publication Participants Interventions Outcomes Follow-up Results center study RCT, single RCT, Prospective /1973 /1975 55 56 Duckert et al

Author/Year et al Tsapogas

18 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Complete lysis: 6/23 (26%) Partial lysis: 15/23 (65%) No lysis: 2/23 (9%) PE: 0 Bleeding: 4/23 (17%) Death: 1 (4%) Complete lysis: 1/26 (4%) Partial lysis: 20/26 (77%) No lysis: 5/26 (19%) PE: 0 Bleeding: 1/26 (4%) Death: 0 cant lysis: 5/12 (42%) Signifi Partial lysis: 2/12 (16%) No lysis: 5/12 (42%) Death: 1/12 (8%) cant lysis: 0/12 (0%) Signifi Partial lysis: 3/12 (25%) No lysis: 9/12 (75%) Death: 0 cant lysis: 15/21 (71%) Signifi No lysis: 6/21 (29%) PE: 1/21 (5%) Bleeding: 2/21 (9%) cant lysis: 5/21 (24%) Signifi No lysis: 16/21 (76%) PE: 0 Bleeding: 2/21 (9%) Anticoagulation: Anticoagulation: Anticoagulation: Thrombolysis 10 d 5 d Thrombolysis: 21 d-6 y Thrombolysis: bleeding, death due to treatment death due to treatment bleeding Clot lysis, PE, Clot lysis, Clot lysis, PE, Continued Table S13— Table over 30 min, then 100,000 units/h titrated for 72 h followed by IV heparin titrated over 7 d (23 patients) 150 units/kg loading dose then titrated for 10 d (26 patients) 250,000 units SK for 20 min, followed by 100,000 units/h for 72 h (12 patients) heparin 150 units/kg IV, followed by titrated infusion for 72 h hydrocortisone prior to treatment of SK 250,000 units IV, then 100,000 International Units/h IV for 72-96 h (21 patients) 15,000 International Units IV bolus, then total of 30,000 International Units IV infusion for 72-90 h (21 patients) Thrombolysis: SK 250,000 units IV Anticoagulation: IV heparin Thrombolysis: initial dose of Anticoagulation: initial dose Cotreatment: 100 mg bolus Thrombolysis: loading dose Anticoagulation: heparin 14 d , 5 d. , DVT duration DVT proximal DVT of 50 patients with 24 patients with 42 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results center center center RCT, single RCT, single RCT, single RCT, /1978 /1977 59 /1975 58 57 Marder et al Arnesen et al Porter et al Author/Year

19 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) (60%) a (9%) b Signifi cant lysis:17/26 (65%) Signifi Partial lysis: 1/26 (4%) No lysis: 8/26 (31%) PE: 0 Bleeding: 2 (8%) cant lysis: 0/25 (0%) Signifi Partial lysis: 0/25 (0%) No lysis: 25/25 (100%) PE: 0 Bleeding: 2/21 (9%) Symptom-free: 12/20 cant lysis: 8/18 (44%) Signifi Partial lysis: 4/18 (22%) No lysis: 6/18 (34%) PE: 1/18 (5%) Minor bleeding: 3/18 (12%) cant lysis: 1/17 (6%) Signifi Partial lysis: 5/17 (29%) No lysis: 11/17 (65%) PE: 1/17 (6%) Minor bleeding: 2/17 (12%) Symptom-free: 2/21 Immediate: Thrombolysis: Anticoagulation: Long term: Thrombolysis: Anticoagulation: Treatment 2 Treatment 19 mo (mean) Immediate: 5 d Long term: 1-2 mo Thrombolysis: clot lysis, PE, bleeding symptom free bleeding Immediate: Long term: Clot lysis, PE, Continued Table S13— Table of SK 600,000 units infused over 30 min, followed by 100,000/h for 3 d; heparin for 4 d following SK (26 patients) 10,000 units IV initially, followed by 10,000 units IV daily for a 6-h infusion to maintain clotting time of 2.5-3 times normal for 7 d (25 patients) SK 250,000 units in 30 min, followed by maintenance (18 patients) 100,000 units/h 45,000 units daily with warfarin (17) Thrombolysis: loading dose Anticoagulation: heparin Thrombolysis: initial dose of Anticoagulation: heparin 8 d , clinical history of DVT DVT 51 patients with 35 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results center study RCT, single RCT, Prospective /1979 /1979 60 61 Watz et al Watz Elliot et al Author/Year

20 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Partial lysis: 1/11 (9%) No lysis: 10/11 (91%) PE: 0 Bleeding: 3/11 (27%) Partial lysis: 1/8 (12%) No lysis: 7/8 (88%) PE: 0 Bleeding: 3/9 (33%) Note: 1 patient excluded from group Normal legs: 13/17 (77%) PTS symptoms (moderate): 4/17 (24%) Normal legs: 6/18 (33%) PTS symptoms (moderate): 9/18 (50%) Complete lysis: 7/17 (41%) Bleeding:3/17 (18%) PE: 0 Complete lysis: 2/19 (10%) Bleeding: 1/19 (5%) PE: 0 refl ecting degree of thrombosis refl (maximum, 40 units: complete thrombosis). The unit scores ect the reduction refl in thrombosis postlysis. Anticoagulation: Anticoagulation: Anticoagulation: 2 wk Thrombolysis: 6.5 y Thrombolysis: 5 y Thrombolysis: 72 h Note: Authors assigned veins a relative value bleeding symptoms bleeding, PE bleeding Clot lysis, PE, Normal legs, PTS Clot lysis, Clot lysis, Continued Table S13— Table 200,000 units IV for 24 h; after 18 h, heparin loading dose of 15,000 units, then 40,000 units/d for 5 d (11 patients) 40,000 units/day 4 for 6 d (9 patients) examination by blinded evaluators International Units IV over 15 min, then 100,000 International Units over 12 h for up to 7 d, titrated; given with heparin 5,000 International Units IV over 12 h (17 patients) 5,000 International Units IV for 15 min then 30,000 International Units/d, titrated over 7 d (19 patients) with rt-PA 100 mg over with rt-PA IV 8 h (day 1), 50 mg rt-PA over 8 h (day 2); 10% dose as bolus (11 patients) Thrombolysis: urokinase Anticoagulation: heparin Phlebography and clinical Thrombolysis: SK 50,000 Anticoagulation: heparin Study A: open-label study 72 h 7 d 10 d , , , DVT of RCT DVT of DVT of 20 patients with 35/42 patients from 36 patients with calf 32 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results center to RCT of Arnesen (1978) center cohort study (A) and multicenter (B) RTC RCT, single RCT, Follow-up single RCT, /1989 Prospective /1986 51 49 /1982 63 /1981 62 Arnesen et al et al Verhaeghe Schulman et al Kiil et al Author/Year

21 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) 3.2 24.3 34.3 2.8 2 2 2 2 heparin: 1 Complete lysis: 2/32 (6%) Partial lysis: 18/32 (57%) No lysis: 12/32 (38%) Bleeding: 1/32 (3%) Complete lysis: 1/17 (6%) Partial lysis: 8/17 (48%) No lysis:8/17 (48%) Bleeding: 0 Partial lysis: 2/11 (18%) No lysis: 9/11 (89%) Bleeding: 0 Change in unit score: 100 mg: rt-PA Change in unit score: Bleeding: 6 Change in unit score: Bleeding: 3 Change in unit score: Bleeding: 0 rt-PA: rt-PA: rt-PA Anticoagulation: Note: 5/65 venograms not analyzed Study A Study A Study B 50 mg: rt-PA Placebo: 36 h bleeding Clot lysis, Continued Table S13— Table heparin: rt-PA as heparin: rt-PA 1 IV for 24 h, then heparin 100 units/kg bolus, then 1,000 units/h, adjusted (36 patients) in group 1 plus heparin concomitantly (17 patients) 100 units/kg bolus, then 1,000 units/h (12 patients) 100 mL containing rt-PA 100 mL containing rt-PA 100 mg infused over 8 h (day 1), IV of 100 mL containing 50 mg rt-PA infused over 8 h (day 2); 10% dose as bolus (8 patients) 50 mg containing rt-PA infused over 8 h on both days 1 and 2; 10% dose as bolus (6 patients) containing placebo infused over 8 h on both days 1 and 2 (7 patients) IV bolus then continuous infusion 1,000 units/h for up to 72 h rt-PA: rt-PA 0.05 mg/kg per h rt-PA rt-PA: rt-PA Anticoagulation: heparin rt-PA 50 mg: IV of 100 mL rt-PA Placebo: IV of 100 mL Co-treatment: heparin 5,000 units 14 d , (65 randomizations) with DVT of 64 patients Type of Type Publication Participants Interventions Outcomes Follow-up Results multicenter /1990 RCT, 64 Author/Year 100 mg: IV of Study B: rt-PA

Goldhaber et al

22 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) heparin: heparin heparin: heparin: 1 1 1 1 50% lysis: 7/12 (58%) 50% lysis: 2/12 (17%) 50% lysis: 2/12 (17%) 50% lysis: 6/29 (21%) 50% lysis: 7/29 (24%) 50% lysis: 2/30 (7%) 50 lysis:5/30 (17%) Lysis Lysis Ն , No lysis: 3/12 (25%) Bleeding: 4/12 (33%) , No lysis: 10/12 (83%) Bleeding: 1/12 (8%) Lysis Ն , No lysis: 15/29 (52%) Bleeding: 1/29 (3%) Ն , No lysis: 23/30 (77%) Bleeding: 1/30 (3%) Complete lysis: 6/22 (27%) Bleeding: 1/22 (5%) PTS symptoms: 14/22 (64%) Complete lysis: 11/22 (50%) Bleeding: 1/22 (5%) PTS symptoms: 9/22 (41%) Placebo Placebo Phase 2: Urokinase: 24-48 h Phase 1: 7 d and 1 y rt-PA: bleeding bleeding symptoms Clot lysis, 7 d: clot lysis, 1 y: PTS Continued heparin: heparin: 1 1 heparin Table S13— Table heparin heparin (30 pts) 1 1 (12 patients) 0.5 mg/kg one-chain rt-PA IV for 8 h and repeated in 24 h (29 patients) 5,000-unit IV bolus then 30,000 units/24 h, adjusted for 7-10 d two-chain rt-PA 0.5 mg/kg two-chain rt-PA IV for 4 h (12 patients) vein 4 h/d for 7 d; heparin IV given concomitantly; warfarin day 7-12 mo International Units/hr IV into pedal vein continuously 7 d; heparin IV 7 d; plasminogen monitored; warfarin day 7-12 mo Placebo Phase 2: Lysis Placebo Phase 1: Lysis Phase 1: Lysis rt-PA 20 mg IV into pedal rt-PA Urokinase 100,000 7 d 7 d , , DVT of DVT of 83 patients with 69 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results multicenter center /1998 single RCT, 65 /1990 RCT, 50 Schweizer et al Cotreatment:

Turpie et al Turpie Author/Year

23 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 50% lysis: 7/50 (14%) 50% lysis: 16/50 (32%) 50% lysis: 9/50 (18%) 50% lysis: 17/50 (34%) 50% lysis: 7/50 (14%) 50% lysis: 13/50 (26%) 50% lysis:10/50 (20%) 50% lysis: 13/50 (26%) Complete lysis: 10/50 (20%) Ն , No lysis: 13/50 (26%) Bleeding: 2/50 (4%) Complete lysis: 10/50 (20%) Ն , No lysis: 11/50 (22%) Bleeding: 1/50 (2%) Complete lysis: 20/50 (40%) Ն , No lysis: 8/50 (16%) Bleeding: 5/50 (10%) PE: 5/50 (10%) Complete lysis: 17/50 (34%) Ն , No lysis: 8/50 (16%) Bleeding: 4/50 (8%) PE: 4/50 (8%) Complete lysis: 0 Bleeding: 0 PTS Symptoms: 15/22 (68%) Urokinase: Systemic SK: Systemic urokinase: Anticoagulation: 1 y rt-PA: bleeding, mortality Clot lysis, Continued brinogen and Table S13— Table urokinase 100,000 International Units infused continuously; ﬁ plasminogen monitored; IV heparin given concomitantly for 6 h with heparin up to 7 d. Premedications: hydrocortisone 100 mg, ranitidine 50 mg, clemastine 2 mg International Units/d for 4 h up to 7 d; IV heparin given concomitantly adjusted. compression bandages, compression therapy, warfarin for 12 mo adjusted for 7 d; warfarin day 1-12 mo 20 mg/day for 4 h through pedal vein for 4-7 d; IV heparin given simultaneously at 1,000 International Units/h, adjusted rt-PA: locoregional rt-PA locoregional rt-PA rt-PA: 9 d , DVT of 250 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results multicenter /2000 RCT, 66 Four deaths, two PEs, two other causes. Four deaths, two PEs, other causes. Four deaths, other causes, two lost to follow-up. Four deaths, other causes, two lost to follow-up. Urokinase: locoregional S1, S2, and S11 legends for expansion of abbreviations. See Table Systemic SK: 3 million units/d Systemic urokinase: 5 million Anticoagulation: heparin IV, Cotreatment: bed rest, Anticoagulation: heparin IV Author/Year Schweizer et al a b

24 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S14 —[Section 2.10] Systemic Lysis vs No Systemic Lysis for Extensive Acute DVT of the Leg: Methods

Author/Year Randomization Allocation concealment Blinding Loss to follow-up

Browse et al52 /1968 N/A N/A N/A 0 Robertson et al53 /1968 Patients given consecutive Labels on SK and Y, data assessors; unclear 0 code numbers and divided heparin coded for patients, caregivers, into equal groups of 2—not analysts truly randomized Kakkar et al54 /1969 Sequential sealed envelope Adequate Y, patients No, caregivers, 0 assessors, analysts Tsapogas et al55 /1973 Sealed envelope Adequate Not blinded 0 Duckert et al56 /1975 N/A N/A N/A N/A Porter et al57 /1975 Assigned at random to one of N/A N/A N/A two groups Marder et al58 /1977 Assigned at random to one Unclear Not blinded 23 of two groups. Five-day follow-up venograms were not performed in 3 of the SK patients, so an additional 3 patients were added to SK group in nonrandomized fashion. Arnesen et al59 /1978 Assigned at random to either Adequate Y, radiologic assessors No, 0 group by sealed envelope patients, caregivers, analysts Elliot et al60 /1979 Assigned at random to one of Unclear Y, assessors N, patients, N/A two groups caregivers, analysts Watz et al 61 /1979 N/A N/A N/A N/A Kiil et al62 /1981 Assigned at random to either Unclear Y, assessors, analysts N, 0 group patients, caregivers Arnesen et al63 /1982 N/A N/A Evaluator blinded 7 Schulman et al49 /1986 Assigned at random to either Adequate Single blind 0 group by sealed envelope Verhaeghe et al51 /1989 Assigned at random to one of Unclear Y, patients, assessors N, 0 three groups caregivers, analysts Goldhaber et al64 /1990 Assigned at random to one of three 2:2:1 allocation Not blinded 0 groups through sealed envelope scheme Turpie et al 50 /1990 Assigned at random to one of two Unclear Y, patients, assessors N, 37 groups in each phase of study caregivers, analysts Schweizer et al65 /1998 Assigned at random to one of three Adequate Y, assessors N, patients, 1 groups caregivers, analysts Schweizer et al66 /2000 Assigned at random to one of ﬁ ve Unclear Single blind—not sure who 12 groups Y 5 yes. See Table S5 and S12 legends for expansion of other abbreviations.

25 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) 1,000 (from 2 fewer to 75 more) 1,000 (from 16 fewer to 36 more) 1,000 (from 36 fewer to 273 more) With Systemic With Lysis (95% CI) (95% CI) Lysis Risk Difference Risk 24 more per 13 more per h h,j Lysis Risk With Risk With Anticipated Absolute Effects No Systemic (95% CI) Relative Effect Summary of Findings With With Lysis Systemic Study Event Rates (%) With No With 5/233 (2.1) 4/455 (0.9) RR 0.86 (0.27-2.68) 21 per 1,000 3 fewer per 2/233 (0.9) 10/454 (2.2) RR 1.28 (0.25-6.68) 48 per 1,000 10/234 (4.3) 38/454 (8.4) RR 1.84 (0.94-3.59) 29 per 1,000 Systemic Lysis due c,d,i due to due to b-e d,e,g to imprecision imprecision imprecision of Evidence Overall Quality Mortality (important outcome) Undetected Low Undetected Low e e Nonfatal recurrent VTE (critical outcome) Nonfatal major bleeding (critical outcome) Very serious Very Very serious Very Serious Undetected Moderate

c,d d c,d indirectness indirectness indirectness No serious No serious No serious Quality Assessment [Section 2.10] Evidence Proﬁ le: Systemic Lysis vs No Systemic Lysis for Extensive Acute DVT of the Leg for Extensive Acute DVT of the Leg vs No Systemic Lysis le: Systemic Lysis [Section 2.10] Evidence Proﬁ inconsistency inconsistency inconsistency No serious No serious No serious Table S15 — S15 Table

b g i of bias of bias of bias No serious risk No serious risk No serious risk

a f f 3 mo 3 mo 3 mo 688 (5 studies), 687 (3 studies), 688 (10 studies),

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias

26 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ). 48

o 1,000 (from 300 fewer to 24 more) With Systemic With Lysis (95% CI) (95% CI) Lysis Risk Difference Risk 171 fewer per n Lysis ). ). Risk With Risk With Anticipated Absolute Effects 51 No Systemic Verhaeghe et al Verhaeghe 50 (95% CI) Relative Effect Summary of Findings With With Lysis Systemic We identiﬁ ed no studies published since the identiﬁ We search date of the systematic 36 Study Event Rates (%) With No With 24/230 (10.4) 27/448 (6) RR 0.71 (0.49-1.04) 588 per 1,000 Systemic Lysis due to d,e,l,m risk of bias, imprecision of Evidence Continued Overall Quality ). Excluding this study from the analysis does not change the effect estimate. All studies had blinded 48 Allocation concealed in seven of 10 studies. Follow-up inadequate in one 10 studies (Common et al 39i This probably underestimates PTS baseline risk given that overall, 52% of patients reported the current use of use current the reported patients of 52% overall, that given risk baseline PTS underestimates probably This 40 Table S15— Table Quality of life not measured Postthrombotic syndrome (critical outcome) Undetected Low e4 the absolute reduction is 40 fewer severe PTS per 1,000 (from 70 fewer to 6 more) over 2 y. the absolute reduction is 40 fewer severe PTS per 1,000 (from 70 to 6 more) over 2 y. 40 Serious

d3 from the analysis because it used catheter directed thrombolysis. 36 indirectness No serious cant effect. cant effect. Quality Assessment ) consisted of patients with calf vein thrombosis. ) consisted of patients with calf vein thrombosis. 49 inconsistency We excluded We Elsharawy et al 47 No serious l,m Serious

k 2 y No use of a standardized validated tool reported. No use of a standardized validated tool reported. Allocation ve concealed studies. in Follow-up three ve inadequate of (Common in fi et one al of fi Interventions varied across studies with regard SK, to urokinase), agent dose, (eg, use tPA, of the pedal vein administration, duration of treatment, and we concomitant did drugs (eg, steroids). However, Baseline risks for nonfatal recurrent VTE and for major bleeding derived from Douketis et al. study. Outcomes) Thrombosis (Venous VETO the of ndings fi the on based is estimate This Severe PTS: assuming the same RR of 0.71 and a baseline risk 13.8%, CI included both no effect and a potentially signifi None of the studies used a placebo control. Allocation concealed in two of three studies. Follow-up adequate all All studies had blinded outcome assessors. Range of follow-up in included studies: 1 to 6 y. Range of follow-up in included studies, 1 to 72 mo. Range of follow-up in included studies, 1 to 72 mo. The population of one study (Schulman et al Range of follow-up in included studies, 1 to 30 d. Range of follow-up in included studies, 1 to 30 d. Only 4% of all major bleeding events were intracranial bleeds. Only 4% of all major bleeding events were intracranial bleeds. None of the studies used placebo control. Allocation concealed in two of studies. Follow-up adequate all All studies had blinded outcome assessors. 678 (2 studies), … Bibliography: et Watson al. … … … … … … … … - … … review. See Table S1, S10, and S11 legends for expansion of abbreviations. S1, S10, and S11 legends for expansion of abbreviations. See Table review. not downgrade for indirectness given that there is no standard regimen, and all analyses showed heterogeneity in results.

outcome assessors. None of the studies used a placebo control. outcome assessors. None of the studies used a placebo control. Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias a b c d e f g h et al; used placebo (Turpie studies Excluding this study from the analysis does not affect effect estimate. All studies had blinded outcome assessors. Two j k l m n compression stockings during study follow-up. o

27 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

k (95% CI) (95% CI) 1,000 (from 47 fewer to 372 more) 1,000 (from 59 fewer to 329 fewer) With Surgical With Thrombectomy Risk Difference 30 fewer per 218 fewer per j d,e

a Risk With Risk With Anticipated Absolute Effects No Surgical Thrombectomy 48 per 1,000 588 per 1,000 t. t.

i (95% CI) (0.02-8.75) (0.44-0.9) Relative Effect Summary of Findings Not estimable See comment … f With Surgical With Thrombectomy 0/24 (0) RR 0.37 d Study Event Rates (%) Surgical 1/27 (3.7) With No With 25/27 (92.6) 14/24 (58.3) RR 0.63 Thrombectomy due to due to b,c g,h risk of bias, imprecision risk of bias, imprecision of Evidence Overall Quality This probably underestimates PTS baseline risk given that overall, 52% of patients reported the current use of 40 Mortality not reported PTS (critical outcome) Quality of life not measured the absolute reduction is 51 fewer severe PTS per 1,000 (from 14 fewer to 77 fewer) over 2 y. the absolute reduction is 51 fewer severe PTS per 1,000 (from 14 to 77 fewer) over 2 y. 40 Nonfatal recurrent VTE (critical outcome) Nonfatal major bleeding (critical outcome) Undetected Low Undetected See comment 0/27 (0) 0/24 (0) c c Undetected Low h

39 Serious Very serious Very Very serious Very

c indirectness indirectness indirectness t. t. No serious No serious No serious Quality Assessment [2.11] Evidence Profi le: Surgical Thrombectomy Vs No for Extensive Acute DVT of the Leg [2.11] Evidence Profi inconsistency inconsistency inconsistency See Table S1, S5, and S10 legends for expansion of abbreviations. S1, S5, and S10 legends for expansion of abbreviations. See Table No serious No serious No serious b b g 67 Table S16 — S16 Table Serious Serious mo mo Not clear whether allocation was concealed. No blinding reported. Not clear whether analysis was ITT. Follow-up rate 88% at 6 mo. Study not stopped early for benefi Follow-up rate 88% at Not clear whether allocation was concealed. No blinding reported. analysis ITT. One event, which was a symptomatic PE. limitations. Few number of events. This warrants rating down the quality evidence by a second level when considered along with study Baseline risks for nonfatal recurrent VTE derived from Douketis et al. In addition to other study limitations, this outcome was assessed by those who did the surgery and anticoagulation. No standardized tool was used. One surgical patient had an amputation secondary to Severe PTS: assuming the same RR of 0.63 and a baseline risk 13.8% over 2 y, The study included patients with DVT with symptoms of leg swelling not exceeding 7 d and a proximal extension of the thrombus above the inguinal ligament, but not into the vena cava. above the inguinal ligament, but not into vena cava. The study included patients with DVT symptoms of leg swelling not exceeding 7 d and a proximal extension the thrombus CI includes values suggesting either harm or benefi No severe bleeding complications were recorded in either group. Three patients in thrombectomy group developed local wound hematoma. hematoma. No severe bleeding complications were recorded in either group. Three patients thrombectomy group developed local wound The RR is based on the 6-mo data. The RR is based on the 6-mo data. Thrombosis ndings Outcomes of study. the VETO (Venous This estimate is based on the fi 51 (1 study), 3 … 51 (1 study), 3 … … … … … ... … … … ...... … Bibliography: Plate et al. … … … … … … … … - … … 51 (1 study), 2 y Serious

a b c d e f g h i j k Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias

venous gangrene and was not counted in the PTS assessment. compression stockings during study follow-up. compression stockings during study follow-up.

28 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians .05) .025) , ,

P P .005) .005) (Continued) ,

P Results Results ciency: PTS sequelae: 25/27 (93%) Iliofemoral patency: 9/26 (35%) competence: 7/27 (26%) Valve (PE in 1 patient) Good: 75% Fair: 20% Poor: 5% PTS sequelae: 14/24 (58%, Iliofemoral patency: 16/21 (76%, competence: 13/23, (52% Valve (venous gangrene in 1 patient) Normal: 61% Postthrombotic: 23% Occluded: 39% Normal: 82% Abnormal: 18% Normal: 29% Abnormal: 71% Normal: 29% Abnormal: 71% PTS sequelae: 6/22 (27%) Iliac patency: 11/22 (50%) pressure: 60 mm Hg (mean) Venous Hematoma: 11% patency: 86% AVF PE: 4% infection: 26%: Wound Surgical: (iliofemoral): Venography IV pressure: Plethysmography: Foot volumetry: Follow-up 6 mo Medical: 56 d (mean) Patent iliac vein: 88% 9-10 mo insufﬁ Venous 5 y Medical: ciency: Outcomes patency, valve competence patency, PE, wound patency, AVF infection poor Good, fair, Normal, postthrombotic, occluded abnormal abnormal abnormal venous pressure PTS sequelae: iliofemoral hematoma, patency, Venous insufﬁ Venous Venography (vein segment): Venography pressure: Normal, Venous Plethysmography: Normal, Foot volumetry: Normal, PTS sequelae, iliac patency, PTS sequelae, iliac patency, Interventions followed by 500 units/kg per and oral 24 h adjusted to aPTT, anticoagulation (31 patients) closed with temporary AVF at 6-8 wk, heparin preoperatively and postoperatively plus warfarin postoperatively venous pressure, venous plethysmography, foot volumetry bectomy with temporary AVF bectomy with temporary AVF plus anticoagulation as above (27 patients) anticoagulation alone vs Surgical: operative venous thrombectomy plus anticoagulation Medical: 5,000-unit bolus heparin Iliofemoral venous thrombectomy Clinical PTS, venography, Prior treatment: Medical: iliofemoral venous thrombosis iliofemoral DVT (age of clot mean, 3 d) prior operative venous thrombectomy and AVF closed at 6-8 wk for iliofemoral DVT 19 surgical) available for evaluation at 5 y (All Randomized Trials and Prospective Observational Studies of at Least 20 Patients) and Prospective Observational Studies of at Least 20 Patients) (All Randomized Trials 57 patients (58 limbs) with 41/58 patients (22 medical, ) 67 up to RCT (Plate, 1984 Type of PublicationType Participants [Section 2.11] Surgical Thrombectomy vs No Surgical Thrombectomy for Extensive Acute DVT of the Leg: Clinical Description and Results [Section 2.11] Surgical Thrombectomy vs No for Extensive Acute DVT of the Leg: Clinical Description and Results /1986 Prospective registry 70 patients (71 legs) with /1986 68 69 /1984 multicenter RCT, 58 patients with acute /1990 Five-year follow- 67 70 Table S17 — S17 Table Einarsson et al Surgical: operative venous throm- Einarsson et al

Plate et al Author/Year Plate et al

29 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Results Results ux): 56% AVF closure ux): 56% AVF .05) ,

success rate: 87% P Iliofemoral: 88% Popliteal: 94% PE: 16% symptomatic, 31% asymptomatic Symptom-free: 81% Normal photoplethysmography (no reﬂ PTS sequelae: 2/19 (11%) Iliac patency: 15/19 (78%) pressure: 43 mm Hg (mean, Venous PTS sequelae: 15/17 (88%) Iliac patency: 7/17 (41%) pressure: 63 mm Hg (mean) Venous PTS sequelae: 7/13 (54%) Iliac patency: 10/12 (83%) pressure: 55 mm Hg (mean) Venous Patency:

Surgical: Surgical: See Table S1, S2, S4, S10, and S11 legends for 46 (mean) Follow-up 24 mo 10 y Medical: Outcomes symptoms, normal photoplethysmography, closure successful AVF venous pressure Patency, PE, clinical Patency, PTS sequelae, iliac patency, PTS sequelae, iliac patency, Continued Table S17— Table Interventions with temporary AVF with temporary AVF (closed 6-12 wk postoperative) percutaneous dilatation of severe iliac stenosis (3 patients) anticoagulation alone vs Surgical: operative venous thrombectomy plus anticoagulation Operative venous thrombectomy Prior treatment Medical: DVT of 1-14 d medical arm, 13 from surgical arm) available for evaluation 20 patients and retrospective studies are described in Table 5 of the eighth edition of these guidelines. 30/58 patients (17 from , up to RCT (Plate, 1984 [10] and 1990 [20]) Type of PublicationType Participants /1991 Prospective registry 48 patients with iliofemoral 71 /1997 follow- Ten-year 72 Adjunctive therapy: transvenous Plate et al Author/Year Neglén et al Early prospective observational studies with expansion of abbreviations. expansion of abbreviations.

30 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S18 —[Section 2.11 ] Surgical Thrombectomy vs No Surgical Thrombectomy for Extensive Acute DVT of the Leg: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up

Plate et al67 /1984 ND PN N, patients, caregivers, 7 assessors, and data analysts Einarsson et al68 /1986 N/A N/A N/A N/A Einarsson et al69 /1986 N/A N/A N/A N/A Plate et al70 /1990 N/A N/A N/A 17 Neglén et al71 /1991 N/A N/A N/A N/A Plate et al72 /1997 N/A N/A N/A 28 See Table S5 and S12 legends for expansion of abbreviations.

31 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

a,b (Continued) (from 19 fewer to 224 more) (from 89 fewer to 63 more) (from 113 fewer to 82 more) (from 21 fewer to 121 fewer) Risk Difference With Vena Cava Vena With Filters (95% CI) Filters (95% CI) Filters Vena Cava Vena Risk With No Risk With 273 per 1,000 82 more per 1,000 185 per 1,000 31 fewer per 1,000 515 per 1,000 26 fewer per 1,000 151 per 1,000 89 fewer per 1,000

e h i

g Effect Relative (95% CI) (0.78-1.16) (0.52-1.34) (0.2-0.86) (0.93-1.82) Summary of Findings With Vena Vena With Cava Filters Study Event Rates (%) Anticipated Absolute Effects Filters With No With Vena Cava Vena 103/200 (51.5) 98/200 (49.0) RR 0.95 31/168 (18.5) 26/169 (15.4) RR 0.83 24/159 (15.1) 9/145 (6.2) RR 0.41 41/150 (27.3) 57/160 (35.6) RR 1.3 due due due due c,d c,f c,f c,d to imprecision to imprecision to imprecision to imprecision of Evidence Overall Quality Bias Mortality (important outcome) Symptomatic PE (critical outcome) Publication Major bleeding (important outcome) Recurrent DVT (important outcome) Undetected Moderate Undetected Moderate Undetected Moderate Undetected Moderate d f f d Serious Serious Serious Serious indirectness indirectness indirectness indirectness No serious No serious No serious No serious Quality Assessment inconsistency inconsistency inconsistency inconsistency No serious No serious No serious No serious

c c c c [Section 2.13] Evidence Proﬁ le: Vena Cava Filter vs No Vena Cava Filter for Acute Proximal DVT of the Leg Treated With Anticoagulation With Cava Filter for Acute Proximal DVT of the Leg Treated Cava Filter vs No Vena le: Vena [Section 2.13] Evidence Proﬁ of bias of bias of bias of bias Table S19 — S19 Table Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision 400 (1 study), 8 y No serious risk 304 (1 study), 8 y No serious risk 337 (1 study), 8 y No serious risk 310 (1 study), 8 y No serious risk

32 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians study)

k (from 238 fewer to 91 more) Risk Difference With Vena Cava Vena With Filters (95% CI) Filters (95% CI) (prospective 74 Filters Vena Cava Vena Risk With No Risk With Mismetti et al 8 699 per 1,000 91 fewer per 1,000 Effect Relative (95% CI) (0.66-1.13) Summary of Findings With Vena Vena With Cava Filters Study Event Rates (%) Anticipated Absolute Effects Filters With No With 0/186 (0) 0/193 (0) … … … Vena Cava Vena 107/153 (69.9) 109/155 (70.3) RR 0.87 due to to due f due to Continued d,j risk of bias, imprecision imprecision of Evidence Overall Quality Moderate PTS (important outcome) Table 19— Table Quality of life not reported Bias Complications (important outcome) Publication Undetected Low Undetected d f 2 factorial design) followed by oral anticoagulation for at least 3 mo. 2 factorial design) followed by oral anticoagulation for at least 3 mo. 3 Serious Serious See Table S1, S2, S5, and S10 legends for expansion of abbreviations. S1, S2, S5, and S10 legends for expansion of abbreviations. See Table 73 indirectness indirectness No serious No serious Quality Assessment inconsistency inconsistency No serious The PREPIC Investigators. 8 No serious j of bias (retrospective study) reported an incidence of 0.3% for major complications among 1,731 patients receiving vena cava fi lters predominantly for secondary prevention of VTE. lters predominantly for secondary prevention of VTE. (retrospective study) reported an incidence of 0.3% for major complications among 1,731 patients receiving vena cava fi 75 Anticoagulation consisted of LMWH or UFH initially (according to a 2 Anticoagulation consisted of LMWH or UFH initially (according to a 2 t or appreciable harm. CI includes both negligible effect and appreciable benefi RR, 1.78 (95% CI, 1.09-2.94) at 2 y. RR, 1.0 (95% CI, 0.29-3.4) at 12 d; RR, 1.08 (95% CI, 0.73-1.58) at 2 y. RR, 1.0 (95% CI, 0.29-3.4) at 12 d; 1.08 0.73-1.58) 2 y. PE) RR, 0.23 (95% CI, 0.05-1.05) at 12 d (both symptomatic and asymptomatic PE). 0.54 0.21-1.41) 2 y (symptomatic No complications directly lter related or to its the insertion fi reported in the PREPIC (Prevention du Risque d’Embolie Pulmonaire par Interruption Cave) trial. Four types lters of were permanent used: vena c LGM Vena Tech cava Corporation), (B. fi Cardial eld Braun (C.R. (Boston Melsugen Bard, Scientifi AG), Inc), titanium and Greenfi Bird’s Nest (Cook Allocation concealed. Data collectors and outcome adjudicators blinded. ITT analysis. Data missing for 4% at 2 y and 1% at 8 y. Enrollment was stopped at 400 instead of targeted 800 due to slow Small number of events. Small number of events. RR, 1.5 (95% CI, 0.54-4.14) at 12 d. RR, 0.74 (95% CI, 0.41-1.36) at 2 y. RR, 1.5 (95% CI, 0.54-4.14) at 12 d. 0.74 0.41-1.36) 2 y. No standardized validated tool used to measure PTS. Group Incorporated). Group Incorporated).

379 (1 study), 2 y No serious risk reported an incidence of 3.2% (excluding lter tiltingAthanasoulis et al ﬁ and puncture site hematoma) among 220 patients receiving retrievable vena ltercava forﬁ secondary prevention of VTE, whereas while Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision a b c d e f g h i j k

recruitment. recruitment. … Bibliography: Decousus et al, … … … … … … … … - … … 308 (1 study), 8 y Serious

33 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians cantly

j With Early With Risk Difference 232 fewer to 12 more) 232 fewer to 12 more) 40 fewer to 124 more) 8 fewer to 70 more) Ambulation (95% CI) Ambulation (95% CI)

a,b Delayed Risk With Risk With Ambulation 400 per 1,000per 400 136 fewer per 1,000 (from Effect Relative Summary of Findings (95% CI) 5]) . With Early With Ambulation 17 36 - See footnote Study Event Rates (%) Anticipated Absolute Effects 9/11 (81.8) 14/26 (53.8) RR 0.66 (0.42-1.03) Moderate 2/186 (1.1) 3/199 (1.5) RR 1.3 (0.23-7.55) 11 per 1,000 3 more per 1,000 (from Ambulation With Delayed With 22/186 (11.8) 27/199 (13.6) RR 1.16 (0.66-2.05)1,000per 118 19 more per 1,000 (from due to due to due to due to d,e d-g h,i e,h risk of bias, imprecision risk of bias, imprecision risk of bias, indirectness risk of bias, imprecision of Evidence Overall Quality Low g .05 for stockings). .05 for stockings). Mortality (important outcome) ,

cant harms. cant harms. P Bias Publication Undetected Low Undetected Low

e Undetected Undetected Low e e PE (critical outcome; assessed with symptomatic or asymptomatic PE) imprecision serious PTS (important outcome; assessed with Villata-Prandoni score [value PTS (important outcome; assessed with Villata-Prandoni .001 for bandages, See Table S1, S2, S5, and S10 legends expansion of other abbreviations. S1, S2, S5, and S10 legends expansion of other abbreviations. See Table Serious Very Very Serious ,

78-82 77 P f t and values of clinically signifi No serious i indirectness indirectness indirectness [Section 2.14] Evidence Profi le: Early Ambulation vs Delayed for Acute DVT of the Leg [Section 2.14] Evidence Profi Quality Assessment No serious No serious Serious No serious cant benefi Included studies. 77 Table S20 — S20 Table inconsistency inconsistency inconsistency inconsistency Aissaoui et al. No serious No serious No serious No serious 76 d d h h Quality of life (important outcome; measured with quality of life questionnaire in chronic limb venous insuffi ciency [CIVIQ]; better indicated by lower values) Quality of life (important outcome; measured with quality questionnaire in chronic limb venous insuffi Bias Inconsistency Indirectness Imprecision Risk of Serious Serious

c 3 mo 4-12 d Concealmentallocationof reported oneinfourofstudies; blinding outcomeof assessors reported twoinfourofstudies; ITTanalysis reported twoinfourofstudies. Follow-up 97%-100%. twoInof Two of four Two eligible studies excluded patients with symptomatic PE; in the third 24% study, of participants had symptomatic PE at baseline. It was not clear whether the fourth study excluded patients Concealment of allocation not reported, outcome assessors not blinded for this outcome; 70% follow-up rate; compression stockings used on patients with early mobilization but in patients with delayed CI includes both values of clinically signifi Funnel plot reported as not asymmetrical by Aissaoui et al. In two of four eligible trials, all patients received early compression therapy (bandages or stockings). In the two other trials, only patients randomized to early ambulation received early compression Three studies reporting acute phase mortality reported no deaths. PE assessed as both symptomatic and asymptomatic PE. PE assessed as both symptomatic and asymptomatic PE. Psychologic and overall somatic quality of life did not differ signifi cantly between the treatment groups, whereas DVT-related items, especially those refl ecting the ease of locomotion, showed signifi showed locomotion, of ease the ecting refl those especially items, DVT-related whereas groups, treatment the between cantly signifi differ not did life of quality somatic overall and Psychologic 37 (1 study) 2 y Serious greater improvement with compression than bed rest (

Bibliography: Kahn et al, four trials, only patients randomized to early ambulation received early compression therapy (bandages or stockings). In the two other trials, all patients received early compression therapy. other trials, all patients received early compression therapy. four trials, only patients randomized to early ambulation received compression therapy (bandages or stockings). In the two 53 (1 study ), 2 y 53 (1 study Serious with symptomatic PE. with symptomatic PE.

385 (4 studies),

a b c d e f g h j

therapy. therapy. mobilization. mobilization. iNo explanation was provided. Participants (Studies), Follow-up 385 (4 studies),

34 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians .05), ,

(Continued) P cantly improved .01); near absence of pain ,

P .01) ,

P leg circumference and clinical scores signifi in groups A and B compared with group C bandages ( and stockings vs bed rest ( at 9 d achieved with bandages only follow-up scans were performed in 1,256/1,289 patients baseline PEs PE, group A: 2/15 (13%) PE, group B: 1/15 (7%) PE, group C: 1/15 (7%) Well-being/quality of life: Improved with stockings ( Well-being/quality rst 4 d w/ bandages Leg pain: Decreased faster during fi Note: initial lung scans were performed in 1,270/1,289 patients; Bed Rest: PE: 14/63 (22%) Note: new PEs were asymptomatic; 12/16 patients had 9 d distance, pain, Summary results between groups: Walking 9 d 10 d PE at admission: 629/1,270 (50%) 10 d: 77/1,256 (61%) 10 d Ambulation: PE: 10/59 (17%) 3 mo Ambulation: PE: 10/69 (14%) scan / scan at scan / / levels, leg circumference, clinical scores, PE, side effects and day 4 by well-being, and quality of DVT-related life, leg pain by visual analog scale, edema, clinical scores, thrombus progression admission and after 10 d of treatment, PE by distance, pain Walking New PE between baseline PE on Walking distance, Walking stockings: bandages: 4 h/d for d bandages: stockings: 1 1 Ն 1 1 elastic compression stockings plus walking exercises (15 patients) LMWH (15 patients) day 2, then ambulation and compression (64 patients) inelastic Unna boot bandages plus walking exercises (15 patients) under supervision, LMWH (69 patients) compression, and immediate ambulation elevation and compression (62 patients) fi rm inelastic bandages, fi ambulation (18 patients) elastic compression stockings, ambulation (18 patients) Ambulation Bed rest, no compression, Ambulation: leg elevation until Ambulation Ambulation All treated with LMWH, Ambulation Ambulation 14 d duration acute proximal DVT proximal DVT , acute DVT acute DVT proximal DVT 126 patients with 45 patients with 129 patients with 1,289 patients with 53 patients with [Section 2.14] Early Ambulation vs Delayed Ambulation for Acute DVT of the Leg: Clinical Description and Results [Section 2.14] Early Ambulation vs Delayed for Acute DVT of the Leg: Clinical Description and Results Type of Type Publication Participants Interventions Outcomes Follow-up Results center multicenter center study Table S21 — S21 Table /2001 single RCT, 78 /1999 single RCT, 82 /2003 RCT /2000 RCT, /2001 Prospective 79 81 83 Aschwanden et al Blättler et al Partsch et al Partsch et al Bed rest for 8 d with leg Bed rest for 4 d (60 patients) Bed rest: PE: 6/60 (10%) Author/Year Schellong et al

35 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians .01) ,

P .001) ,

P .01) ,

P .001) ,

P thrombus remnants between groups outcome vs bed rest group (mean score, 8.2; Ambulation: 7/52 (13%) Ambulation: 7/52 (13%) stockings vs bed rest ( rest ( vs bed rest ( Thrombus extension: No difference in thrombus regression of PE: No difference between groups Pain: Lower pain levels in mobile group vs bed rest (ns) DVT group, ambulate: PE: 4/988 (0.4%) PE group, bed rest: PE: 2/385 (0.5%) PE group, ambulate: PE: 2/227 (0.9%) Primary target variable: Bed rest: 14/50 (28%) Clinical scores: Improved with bandages and stockings vs bed Thrombus progression: Improved with bandages and stockings 2 y PTS scores: Ambulatory group (mean score, 5.1) had improved 5 d New PE bed rest: 8/50 (16%) ambulation: 2/52 (4%) 3 mo DVT group, bed rest: PE: 7/1050 (0.7%) ed rmed Continued (Villalta-Prandoni scale) (Villalta-Prandoni analog scale and modifi Lowenberg test rst 15 d of PE during fi therapy thrombosis, infection or serious adverse event PTS assessment Thrombus regression Pain assessment by visual Symptomatic, confi PE, progression of or new Table S21— Table rest vs anticoagulation and ambulation with compression bandages or stockings ambulated for 5 d, LMWH, compression bandages ambulation: 1,050 (52%) patients received bed rest, and 988 (48%) ambulated. All received LMWH. ambulation: 385 (63%) patients received bed rest, and 227 (37%) ambulated. All received LMWH. 5 d of strict bed rest, LMWH, compression bandages. Anticoagulation and bed DVT group, bed rest or PE group, bed rest or Bed rest: 50 patients received followed up 2 y post-RCT acute DVT (2,038 [77%]) or PE (612 [23%]) proximal DVT 37 patients 2,650 patients with 103 patients with Type of Type Publication Participants Interventions Outcomes Follow-up Results to RCT (77) open design ed by age stratifi study Prospective ventilation/perfusion. See Table S2, S5, and S10 legends for expansion of other abbreviations. S2, S5, and S10 legends for expansion of other abbreviations. ventilation/perfusion. See Table 5

/ /2005 85 cant; /2004 2-y follow-up /2006 multicenter RCT, 84 80 not signiﬁ 5 Partsch et al et al Trujillo-Santos Jünger et al

ns Ambulation: 52 patients Author/Year Bed rest only (17 patients) Edema: Marked reduction in leg size with bandages and

36 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S22 —[Section 2.14] Early Ambulation vs Delayed Ambulation for Acute DVT of the Leg: Methodologic Quality

Allocation Author/Year Randomization Concealment Blinding Loss to Follow-up

Schellong et al82 /1999 Patients randomized to 1 of 2 study groups Unclear Y, assessors N, patients, 4 caregivers, analysts Partsch et al81 /2000 Patients randomized to 1 of 3 study groups Unclear Y, assessors N, patients, 0 by sealed envelope caregivers, analysts Aschwanden et al78 /2001 Patients randomized to 1 of 2 study groups Sealed envelope Not blinded 5 Partsch et al83 /200 N/A N/A N/A N/A Blättler et al79 /2003 Patients randomized to 1 of 3 study groups Not speciﬁ ed Y, assessors N, patients, 0 by sealed envelope caregivers, analysts Partsch et al84 /2004 N/A N/A N/A 21 Trujillo-Santos et al85 /2005 N/A N/A N/A N/A Jünger et al80 /2006 Patients randomized to 1 of 2 study groups Unclear Y, analysts N, patients, by sealed envelope caregivers, assessors See Table S5 and S12 legends for expansion of abbreviations.

37 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians domized

a,b 9 fewer to 4 more) 18 fewer to 21 more) 25 more to 91 more) or 6 wk vs 3 mo of Risk Difference With 4 Risk Difference With Anticoagulation (95% CI) Anticoagulation (95% CI) Control Risk With Risk With Effect Relative (95% CI) Summary of Findings vs 3 or 6 mo of With 4 or 6 wk With Anticoagulation Study Event Rates (%) Anticipated Absolute Effects 55/998 (5.5) 57/1,100 (5.2) RR 0.97 (0.68-1.38) 55 per 1,000 2 fewer per 1,000 (from 70/1,090 (6.4) 127/1,095 (11.6) RR 1.83 (1.39-2.42) 64 per 1,000 53 more per 1,000 (from 13/1,090 (1.2) 7/1,095 (0.6) RR 0.54 (0.22-1.32) 12 per 1,000 5 fewer per 1,000 (from f e-g e,f,h See Table S1, S5, and S2 legends for expansion of abbreviations. S1, S5, and S2 legends for expansion of abbreviations. See Table 90 Overall quality of Evidence Control With Bias Mortality (important outcome) Publication Undetected High Undetected High Undetected High Major bleeding (critical outcome) Recurrent VTE (critical outcome)

g h British Thoracic Society. British Thoracic Society. 89 imprecision imprecision imprecision No serious No serious No serious Levine et al, 88 indirectness indirectness indirectness No serious No serious No serious

f f f Quality Assessment Schulman et al, 87 inconsistency inconsistency inconsistency rst provoked isolated distal DVT, proximal DVT or PE provoked in Kearon et al; fi rst isolated distal DVT in Pinede et al; fi rst isolated distal DVT, proximal DVT, or PE or DVT, proximal DVT, distal isolated rst fi al; et Pinede in DVT distal isolated rst fi al; et Kearon in provoked PE or DVT proximal DVT, distal isolated provoked rst No serious No serious No serious

Pinede et al, e 0%. 0%. 86 [Sections 3.1.1-3.1.4] Evidence Proﬁ le: Four or Six Weeks vs Three or Six Months as Minimum Duration of Anticoagulation for VTE le: Four or Six Weeks [Sections 3.1.1-3.1.4] Evidence Proﬁ 5

2 Bias Inconsistency Indirectness Imprecision I Risk of 1 y in all studied except for Schulman et al in which it was 2 y. 1 y in all studied except for Schulman et al which it was 2 y. risk of bias of bias of bias ف No serious No serious risk No serious risk ), c Table S23 — S23 Table

d d d 1-2 y 1-2 y 1-2 y Populations varied among studies: fi studies: among varied Populations Follow-up was for Differences in mortality are expected to be mediated by differences recurrent VTE and bleeding. Generally, study design was strong. No study stopped early for benefi t; two stopped early because of slow recruitment (Kearon et al, Pinede et al). In one study study (British design Thoracic was Society), strong. 44 No ran study stopped Generally, early for benefi who are known to have a very for low the risk subgroup of of recurrence, patients there with is isolated imprecision No distal and imprecision DVT, the for possibility overall that estimates. the However, Short vs longer duration of anticoagulation was 6 wk vs 6 mo for Schulman et al, 6 wk vs 3 mo for Pinede et al, and 4 wk vs 3 mo for the other three studies. 3 mo for the other three studies. Short vs longer duration of anticoagulation was 6 wk mo for Schulman et al, 3 Pinede and 4 randomized al et Levine and al et Kearon diagnosis; at randomized Society Thoracic British and al, et Schulman al, et Pinede studies: across varied treatment of start the to relative randomization of Timing No heterogeneity with 2,098 (5 studies), 2,185 (5 studies ), 2,185 (5 studies

2,185 (5 studies rmed) in British Thoracic Society. in Schulman et al; proximal DVT (21% had cancer) Levine and or PE (29% not objectively conﬁ Participants (Studies), Follow up Bibliography: Kearon et al, a b c e f g d h

to stop or continue treatment of 2 more months after the initial 4 wk treatment. shorter duration of anticoagulation is adequate and not associated with a clinically important higher risk recurrence. patients were excluded centrally as they did not satisfy eligibility criteria. Patients and caregivers were blinded in two studies (Kearon et al, Levine et al). Adjudicators of outcomes were blinded in all but appears to have a low unexplained drop-out frequency. one study (British Thoracic Society). All studies appeared to have used effective randomization concealment, ITT analysis, and

38 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) rst isolated calf DVT. rst isolated calf DVT. Treated for 1 mo. VTE was Treated asymptomatic in 9% and isolated calf DVT in 18%. One VTE occurred while on warfarin. 91%). Cancer in 21%. objectively diagnosed; proportion with a previous VTE not known. or PE. Only asked about bleeding while on VKAs. recurrent VTE among 116 patients with postoperative VTE. Proximal DVT (ﬁ rst episode in Proximal DVT (ﬁ

All bleeds were on VKA. Only 1

ed Population: fi (0.1-74.4) (0.4-2.5) (0.6-1.8) (0.7-1.4) 9/105 (9%) 1/81 (1%) RR 3.1 9/109 (8%) RR 1.0 28/354 (8) RR 1.1 17/454 (4%) RR (0.0-51.6) (0.1-70.2) (within 2 mo of randomization) RR 0.8 (0.2-3.0) (0.4-33.4) RR 4.9 (0.6-41.6) 0/81 RR 1.0 1/109 (1%) RR 2.9 4/354 (1%) 3/92 RR 3.4 5/454 (1%) (0.1-2.5) (0.2-1.4) 0.5 (0.3-0.9) (0.3-10.0) (0.4, 0.7) 12/105 (11%) 0/105 7/109 (6%) RR 0.6 Length Follow-up Recurrent DVT or PE Major Bleeding Mortality Total Comments Short (4 or 6 wk) vs intermediate (3 mo) durations of anticoagulation Analyzed No. Patients 358/358 1 y 28/358 (11%) 5/358 (1%) 26/358 (7%) Population: DVT or PE; only 71% 81/81105/105 11 mo 15 mo443/443 3/81 (4%) RR 0.6 2/105 (2%) 2 y 1/105 (1%) 80/443 (18%)109/113 1/443 Not specifi 22/443 (5%) First VTE: DVT (distal or proximal) 354/354 1 y 14/354 (4%) RR 92/92454/454 3/92 RR 1.7 43/454 (9%) RR 0.5 [Sections 3.1.1-3.1.4] Comparison of Durations of Anticoagulant Therapy for DVT and PE: Clinical Description and Results [Sections 3.1.1-3.1.4] Comparison of Durations Anticoagulant Therapy for DVT and PE: Clinical Description Results For 1 mo For 2 more mo. for 1.5 mo for 1.5 mo for 2 more mo. for 3 mo for 3 mo for 6 mo VKA (INR 2.0-3.0) VKA stopped (placebo) 84/84 11 moVKA (INR 2.0-3.0) 5/84 (6%)VKA (INR 2.0-2.85) 0/84VKA (INR 2.0-3.0) 0/84 rst DVT or PE. Population: fi Table S24 — S24 Table /1995 88 /2004 /2001 /1995 VKA stopped (placebo) 105/107 9 mo 86 87 89 /1992 90 (SOFAST) (DOTAVK) (DURAC 1) et al Kearon et al Author/Year (Acronym)Author/Year Intervention VKA (INR 2.0-3.0)

VKA (INR 2.0-3.0) VKA (INR 2.0-3.0) Pinede et al VKA (INR 2.0-2.85) British Thoracic Society Schulman et al Levine et al

39 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ed. (Continued) 3 mo. Among the 3 mo (see above) Ն Ն rst proximal DVT or Treated for Treated 4 groups, only 1 recurrent VTE while on VKA. imal DVT treated for 3 mo. One patient had recurrent VTE on VKA. Bleeding in the intervention group was while on VKA. (provoked, 24%; unprovoked, 76%) treated for 3 mo and residual DVT on baseline ultrasound with calf DVT not known. Only bleeding during treatment is reported; 20% of VTE outcomes were not objectively verifi for Treated PE. Recurrent VTE occurred after VKA in 26/28 of the short duration groups and 21/27 of the long duration groups. Population: fi rst unprovoked PE. Population: fi rst unprovoked prox- Population: fi Population: fi rst provoked PE. Population: fi Population: fi rst proximal DVT Population: fi

ed Population: fi ed (total of 3 non- VTE/bleed deaths) (0.6-2.5) (0.4-3.0) (0.4-2.8) (0.5-153) 7/91 (8%) 7/133 (5%) 0/70 (0%) 15/369 (4%) Population: DVT or PE; proportion 19/369 (5%) RR 1.3 8/90 (9%) RR 1.16 7/134 (5%) RR 1.0 4/75 (5%) RR 8.4 treatment) treatment) RR 16.5 (1.0-285) (0.5-21.9) (0.4-10.7) (0.4-4.4) (0.1-56) 1/91 (1%) 0/70 (0%) 8/380 (2%) (during 6 mo. 2/90 (2%) RR 2.0 4/134 (3%) RR 2.0 7/269 (3%)) RR 1.4 1/75 (1%) RR 1.9 Continued (0.6-1.5) (0.5-2.2) (0.6-1.7) (0.6-1.9) (0.2-1.7) Table S24— Table Length Follow-up Recurrent DVT or PE Major Bleeding Mortality Total Comments Different intermediate durations (6 or 12 mo vs 3 mo) of anticoagulation Analyzed No. Patients 91/91 2.6 y (mean) 11/91 (12%) 133/133 3.2 y (mean) 21/133 (16%) 2/133 (2%) 369/396 1 y380/414 31/369 (8%) 1 y 0/369 (during 3 mo. 29/380 (8%) RR0.9 90/9070/70 2.9 y (mean) 11/90 (12%) RR 1.0 2.8 y (mean) 7/70 (10%) 134/134 3.1 y (mean)270/270 21/134 (16%) RR 1.0 15 mo 21/270 (8%) 5/270 (2%) Not specifi 75/75 2.9 y (mean) 4/75 (5%) RR 0.5 269/269 23/269 (9%) RR 1.1 for 3 mo for 6 mo for 9 more mo for 9 mo for 3 mo for 3 more mo for 6 mo VKA stopped VKA (INR 2.0-3.0) VKA stopped VKA (INR 2.0-3.0) VKA (INR 2.0-3.0) VKA stopped 92/92 1.8 y 27/92 (29%) 1/92 (1%) Not specifi VKA stopped VKA (INR 2.0-3.0) /2007 VKA (INR 2.0.0-3.5) 91 /2008 /2003 /2001 /2001 94 92 93 87 (WODIT PE) (WODIT DVT) (DOTAVK) (DACUS) Author/Year (Acronym)Author/Year Intervention Campbell et al VKA (INR 2.0-3.5)

VKA (INR 1.0-3.0) Agnelli et al Pinede et al Siragusa et al Agnelli et al

40 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 3 Ն nite (Continued) rst unprovoked imal DVT or PE. Treated for imal DVT or PE. Treated mo. VKA stopped and D-dimer Eight control positive 1 mo later. patient. Restarted VKA, some cial phlebitis. One after superfi recurrent VTE in VKA group after VKA stopped. proximal DVT or PE (5% had previous provoked VTE). The recurrent VTE in the VKA patient was after stopping VKA. imal) or PE. All recurrent VTE nite VKA group were in the indefi after stopping VKAs. Bleeding rst 6 mo of VKA in 1 during the fi of 6 mo group and indefi group (only asked about bleeding while on VKAs). (24 mo group); 1 VTE on therapy (24 mo group)

(0.1-13.6) (0.0-3.3) (0.3-1.3) 1/120 (1%) RR 0.9 1/79 (1%) RR 0.3 10/116 (9%) RR 0.6 0/32 2.0-3.0) ف (0.1-62.6) (0.4-140) (0.9-11.3) (0.2-6.7) 1/120 (1%) RR 2.6 3/79 (4%) RR 7.4 10/116 (9%) RR 3.2 2/32 (6%) RR 1.0 Continued (0.0-0.4) (0.0-0.5) (0.0-0.4) (0.1-1.5) 2/103 (2%) RR 0.1 Table S24— Table 1.5 y) Length nite vs intermediate durations of anticoagulation (INR Follow-up Recurrent DVT or PE Major Bleeding Mortality Total Comments Indefi Analyzed No. Patients 120/122 (maximum, 79/79111/111 (maximum, 2 y) 1/79 (1%) RR 0.1 4 y 23/111 (2%)32/36 3/111 (3%) 3 y 16/111 (14%) 7/32 (22%) Second VTE: DVT (distal or prox- 2/32 (6%) 0/32 In total: 2 VTE after 24 mo 116/11632/36 3/116 (3%) RR 0.1 3/32 (3%) RR 0.4 nite VKA nitely (INR 2.0-3.0) (not blinded) more years 6 mo 6 mo indefi 24 mo Remain off (stop) VKA 103/105 1.4 y (mean) 18/120 (15%) 0/103VKA stopped (placebo) 83/83 10 mo (mean) 1/103 (1%) 17/83 (20%)VKA (INR 2.0-2.85) for 0/83 rst unprovoked prox- Population: fi 3/83 (4%) Population: fi /1997 97 /2006 /1999 95 /2004 VKA (INR 2.0-3.0) for 96 98 (LAFIT) (PROLONG) (DURAC 2) Palareti et al Author/Year (Acronym)Author/Year Intervention Restart indefi VKA (INR 2.0-3.0) for 2 Kearon et al Schulman et al VKA (INR 2.0-2.85) VKA (INR 2.0-2.85) for Farraj et al

41 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians rst 3 mo. VKA Ն Durée Optimale du Traitement Traitement du Optimale Durée 5 DVT or PE (fi rst episode in DVT or PE (fi for 31%). Treated (INR 2.0-3.0) (mean 12 mo). Five recurrent VTE in INR 1.5-1.9 and three in the INR 2.0-3.0 group after stopping VKAs. (distal or proximal) PE (fi episode in 38%). Eight recurrent VTE in the VKA group after stopping VKAs. or unprovoked (57%) proximal DVT treated for 3 mo. One VTE in each group while on VKAs. exible group was treated for The fl a mean of 4 mo (provoked) and 5 mo (unprovoked) longer.

(0.2-1.2) RR 0.5 (0.1-1.6) (0.7-3.2) f other abbreviations. f other abbreviations. 8/369 (2%) RR 0.5 4/255 (2%) 4/255 (6%) RR 1.5 Warfarin Optimal Duration Warfarin in Italian patients Trial with DVT; 5 Long-term Anticoagulation for a First episode of Idiopathic venous Thromboem- 5 (0.3-2.3) (0.5-12.7) RR 2.0 (0.4-10.8) 8/369 (2%) RR 0.9 5/255 (2%) RR 2.5 4/270/255 (1%) 1.5-2.0, after initial INR 2.0-3.0 in both groups) ف Continued (0.1-0.9) (0.2-0.7) (0.4-1.1) Duration of Anticoagulation based on Compression UltraSonography; DOTAVK DOTAVK UltraSonography; Compression on based Anticoagulation of Duration 6/369 (2%)RR 0.4 14/255 (5%) RR 0.4 32/270 (12%) RR 0.7 Table S24— Table First Acute Secondary Thrombosis; WODIT DVT 5 5 4.3 y) Length Follow-up Recurrent DVT or PE Major Bleeding Mortality Total Comments Anticoagulation for Thrombo-Embolism; LAFIT Anticoagulation for Thrombo-Embolism; LAFIT Low intensity (INR 1.5-1.9) vs conventional 2.0-3.0) 5 Analyzed No. Patients 253/253 2.1 y (mean) 37/253 (15%) 2/253 (1%) 8/253 (3%)369/369 Population: unprovoked DVT 268/268 33 mo 46/268 (17%) 2/268 (1%) 11/268 (4%) rst provoked (43%) Population: fi 270/270 nite vs intermediate durations of anticoagulation (INR Indefi Fixed vs fl exible duration, depending on presence of residual thrombosis follow-up ultrasound (INR 2.0-3.0) Fixed vs fl restarted (placebo) (blinded) and 3 more months if unprovoked thrombosis and until resolved or 9 more mo if provoked or 21 more mo if unprovoked Duration of Anticoagulation; ELATE Duration of Anticoagulation; ELATE VKA stopped or not VKA INR 1.5-1.9 369/369 2.4 y (mean) 16/369 (4%)VKA INR 2.0-3.0 9/369 (2%) 16/369 (4%) Population: unprovoked proximal VKA stopped if provoked Stopped if no residual vein 5 Prevention of Recurrent Venous Thromboembolism; SOFAST Prevention of Thromboembolism; Recurrent SOFAST Venous 5 /2009 Warfarin Optimal Duration Italian Trial in patients with Pulmonary Embolism. See Table S1, S2, and S7 legends for expansion o in patients with Pulmonary Embolism. See Table Optimal Duration Italian Trial Warfarin /2003 99 /2003 Ultrasound Findings to Adjust the Duration of Anticoagulation; DACUS DACUS Anticoagulation; of Duration the Adjust to Findings Ultrasound 5 101 100 5 (PREVENT) (ELATE) (AESOPUS) Author/Year (Acronym)Author/Year Intervention Prandoni et al WODIT PE bolism; PREVENT bolism; PREVENT AntiVitamines K; DURAC AntiVitamines K; DURAC VKA INR 1.5-2.0 AESOPUS 255/255 (maximum, Kearon et al

Ridker et al

42 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) recruitment. recruitment. Patient withdrawals: 4 in short- and 16 long-duration of 22 patients groups. Total dropped out. protein C found after randomization. withdrew consent shortly after randomization. excluded centrally as did not satisfy entry criteria. excluded centrally as did not satisfy entry criteria. Stopped early because of low recruitment. 2 patients in the intervention groups Five patients in the crossed over. intended group did not stop VKA. Four patients in the control groups restarted VKAs. t. Four patients in the beneﬁ intervention and 2 patients in the control group crossed over.

ITT Stopped early because of slow ITT Stopped early because of slow ITT Five patient were excluded because ITT placebo and 4 warfarin patients Two ITT Forty-four randomized patients ITT Sixty-one randomized patients ITT patients in the control groups and Two ITT stopped early for lack of adequate Trial ed ed ed follow-up but partial follow-up achieved Placebo: 0/84 VKA: 0/81 Not specifi Probably low or nil of 44 patients Total dropped out during Placebo: 1/105 VKA: 6/109 (did not complete 11-mo follow-up) No VKA: 27/354 VKA: 30/358 3- mo VKA: 6/369 6- mo VKA: 4/380 Not specifi Probably low or nil Not specifi Probably low or nil Caregivers: CY Adjudications: CY Data Analysts: CY Caregivers: CN Adjudications: CY Data Analysts: PN Caregivers: CN Adjudications: VTE, CY Other: PN Data Analysts: PN Adjudications: VTE, CY Data Analysts: PN Caregivers: CN Adjudications: CN Data Analysts: PN Caregivers: CN Adjudications: CN Data Analysts: PN Caregivers: CN Adjudications: CY Data Analysts: PN Caregivers: CN Adjudications: CY Data Analysts: PN Concealed Blinding Loss to Follow-up Analysis Comments Randomization RCT CYRCT Patients: CY CYRCT Patients: CN CYRCT Patients: CN CYRCT Patients: CY Caregivers: CYRCT Patients: CN CYRCT Patients: CN CYRCT Patients: CN CY Patients: CN Study Design [Sections 3.1.1-3.1.4] Comparison of Durations of Anticoagulant Therapy for DVT and PE: Methodologic Quality [Sections 3.1.1-3.1.4] Comparison of Durations Anticoagulant Therapy for DVT and PE: Methodologic Quality VKA stopped (placebo) VKA (INR 2.0-3.0) for 2 more mo VKA (INR 2.0-3.0) for 1.5 and 3 mo VKA (INR 1.0-3.0) for 3 and 6 mo VKA (INR2.0-2.85) for 1.5 mo VKA (INR 2.0-2.85) for 6 mo VKA (INR 2.0-3.0) for 2 more mo. VKA (INR 2.0-3.0) For 1 mo (INR 2.0-3.0) for 3 mo VKA (INR 2.0-3.5) for 6 mo VKA stopped VKA (INR 2.0-3.0) for 9 more mo. VKA stopped VKA (INR 2.0-3.0) for 9 mo Table S25 — S25 Table /2007 VKA (INR 2.0.0-3.5) for 3 mo /1995 /1992 91 88 90 /2004 /2003 /2001 /2001 /1995 VKA stopped (placebo) 86 92 93 87 89 Society et al (SOFAST) (DOT AVK) (DURAC 1) (WODIT PE) (WODIT DVT) Kearon et al Author/Year (Acronym)Author/Year Intervention Pinede et al Schulman et al Levine et al British Thoracic Campbell et all Agnelli et al Agnelli et al

43 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians benefi t. After recurrent VTE, benefi patients were not followed. resulting for potential and follow-up shorter in underestimation of bleeding in the no-VKA group. t. Number of crossovers not benefi described. for each group because of poor compliance. postenrollment exclusions were stopped postrandomization. Trial early because recurrent VTE was higher than expected. lupus anticoagulant found after randomization. 7.7 mo in 6-mo group and 42.7 nite (48 mo) group. in indefi

ITT Four patients excluded because ITT Actual mean duration of VKA was ITT stopped early because of overall Trial ITT Crossover to INR 2.0-3.0 in 21 patients 1/369 INR Յ 3/105 1/369 , Յ follow-up, but partial follow-up achieved No VKA: VKA: 0/122 None of 14 patients Total dropped out during ITT stopped early because of overall Trial ed Not specifi Probably low or nil 0/320/32 (see comments) INR 1.5-1.9: 1.5-1.9: ITT Four postrandomization exclusions None ITT Not known whether the many 4 subjects in each group ITT Continued Caregivers: CN Adjudications: CY Data Analysts: PN Caregivers: CY Adjudications: CY Data Analysts: CY Caregivers: CN Adjudications: VTE, CY PN Other, Data Analysts: PN Caregivers: CY Adjudications: CY Data Analysts: CY Caregivers: CN Data Collectors: PN. Adjudicators: PN Data Analysts: PN Caregivers: CY Adjudications: CY Data Analysts: CY Caregivers: CN Adjudications: PY Data Analysts: PN Caregivers: CN Adjudications: VTE, CY Data Analysts: PY Table S25— Table Concealed Blinding Loss to Follow-up Analysis Comments Randomization RCT CYRCT Patients: CN CYRCT Patients: CY CY Patients: CN RCT CYRCT Patients: CY PYRCT Patients: CN CY Patients: CY RCT PY Patients: CN RCT CY Patients: CN Study Design nitely nitely VKA resolved or 9 more mo if provoked or 21 more mo if unprovoked thrombosis and until Remain off (stop) VKA Restart indefi (INR 2.0-3.0) (not blinded) VKA stopped (placebo) VKA (INR 2.0-3.0) for 2 more y VKA (INR 2.0-2.85) for 6 mo VKA (INR 2.0-2.85) indefi VKA stopped or not restarted (placebo) VKA INR 1.5-2.0 VKA (INR 2.0-3.0) for 6 mo VKA (INR 2.0-2.85) for 24 mo VKA INR 1.5-1.9 VKA INR 2.0-3.0 (blinded) VKA stopped VKA (INR 1.0-3.0) for 9 mo. VKA stopped if provoked and 3 more mo if unprovoked /1997 /2009 97 /2008 /2003 99 /2006 /2003 /1999 94 95 /2004 96 101 100 98 (LAFIT) (PROLONG) (DURAC 2) (PREVENT) (ELATE) (DACUS) (AESOPUS) Ridker et al Schulman et al Farraj et al Kearon et al Palareti et al Stopped if no residual vein Siragusa et al Author/Year (Acronym)Author/Year Intervention Kearon et al Prandoni et al

See Table S1, S2, S7, and S24 legends for expansion of abbreviations. S1, S2, S7, and S24 legends for expansion of abbreviations. See Table

44 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians t (Agnelli

(95% CI) (95% CI) a,b (from 36 fewer to 16 more) (from 2 more to 37 more) (from 8 fewer to 47 more) Risk Difference With 6 or 12 mo With 3 mo Anticipated Absolute Effects Risk With Risk With 9 per 1,000 13 more per 1,000 44 per 1,000 13 more per 1,000 115 per 1,000per 115 13 fewer per 1,000 Effect Relative (0.69-1.14) (1.2-5.16) (0.81-2.08) (95% CI) Summary of Findings 12 mo With 6 or With Study Event Rates (%) enrolled provoked and unprovoked proximal isolated DVT and distal DVT, 29/663 (4.4) 38/668 (5.7) RR 1.3 9/1,025 (0.9) 24/1,036 (2.3) RR 2.49 ) 118/1,025 (11.5) 105/1,036 (10.1) RR 0.89 g

d-e d

f Overall Quality due to imprecision due to imprecision of Evidence 3 mo With Bias Publication Undetected High Mortality (important outcome See Table S1, S2, and S5 legends for expansion of abbreviations. S1, S2, and S5 legends for expansion of abbreviations. See Table Major bleeding (critical outcome) Recurrent VTE (critical outcome) 94 Undetected Moderate Undetected Moderate d,e d Siragusa. 92 imprecision Serious No serious Serious 2,000. 2,000. Agnelli et al, , 93 indirectness indirectness indirectness No serious No serious No serious Quality Assessment Agnelli et al, 91 inconsistency inconsistency inconsistency No serious No serious No serious

c f [Sections 3.1.1-3.1.4] Evidence Profi le: Six or Twelve Months vs Three as Minimum Duration of Anticoagulation for VTE le: Six or Twelve [Sections 3.1.1-3.1.4] Evidence Profi Campbell et al, 87 ned the assessment of bleeding to when subjects were receiving anticoagulant therapy, which could have infl ated the increase in bleeding associated ned with the the assessment longer of duration bleeding of to when subjects which were could receiving have anticoagulant infl therapy, risk of bias risk of bias risk of bias No serious No serious No serious Table S26 — S26 Table 1-3 y 1-3 y 1-3 y Study populations varied across studies: Pinede et al enrolled provoked and unprovoked proximal DVT and PE; Campbell et al et Campbell PE; and DVT proximal unprovoked and provoked enrolled al et Pinede studies: across varied populations Study t or harm. CIs include both values suggesting no effect and either benefi Low number of events and a total participants Differences in mortality are expected to be mediated by differences recurrent VTE and bleeding. Timing of randomization Timing relative to the start of treatment and length of treatment in the non-3-mo group varied across studies: Pinede et al and Campbell et al randomized at diagnosis, and Agnelli et al study t; design two Generally, was stopped strong. early No because study of stopped slow early recruitment for (Campbell benefi et al, Pinede et al), and one stopped because of lack of benefi One study may have confi 2,061 (6 studies), 1,331 (5 studies), 2,061 (6 studies), PE; Agnelli et al (2003) had separate randomizations for provoked PE (3 vs 6 mo) and unprovoked (3 vs 12 mo); and Agnelli et al (2001) enrolled unprovoked proximal DVT. (2001) enrolled unprovoked proximal DVT. PE; Agnelli et al (2003) had separate randomizations for provoked PE (3 vs 6 mo) and unprovoked 12 mo); a b c d e f g et al [2001]). In one study (Campbell), 20% of VTE outcomes were not objectively confi rmed. Patients and caregivers were Adjudicators not of blinded et outcomes in al were any [2001]). blinded study. In in one all study but (Campbell), one 20% of VTE outcomes were not objectively confi frequency. study (Campbell). All studies used effective randomization concealment and ITT analysis appear to have a low unexplained drop-out

randomized after the initial 3 mo of treatment to stop, or continue, treatment. The longer duration of treatment was 6 mo in Pinede, Campbell, and Agnelli et al (2003) (provoked PE), and 12 mo in Agnelli in mo 12 and PE), (provoked (2003) al et Agnelli and Campbell, Pinede, in mo 6 was treatment of duration longer The treatment. continue, or stop, to treatment of mo 3 initial the after randomized (2001) and Agnelli (2003) (unprovoked PE).

Bibliography: Pinede et al, therapy (Campbell et al). therapy (Campbell et al). Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

45 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

a,b rst f-j

f-j k,l f-j

(Continued)

f-j k,l k,l With Extended With Risk Difference 62 fewer to 91 fewer) 31 fewer to 45 fewer) (from 93 fewer to 137 fewer) 0 more to 69 more) 44 fewer to 2 more) 0 more to 35 more) more to 17 more) Anticoagulation (95% CI) Anticoagulation (95% CI) Low risk of bleeding unprovoked distal DVT High risk of bleeding First unprovoked VTE Second unprovoked VTE Moderate risk of bleeding First VTE provoked by surgery First VTE provoked nonsurgical/ﬁ Risk With Risk With No Extended Anticoagulation 100 per 1,000 88 fewer per 1,000 (from 50 per 1,000 44 fewer per 1,000 (from 12 per 1,000 20 more per 1,000 (from 63 per 1,000 27 fewer per 1,000 (from 150 per 1,000 132 fewer per 1,000 6 per 1,000 10 more per 1,000 (from 10 per 1,000 10 per 1,000 3 per 1,000 5 more per 1,000 (from 0 (95% CI) (0.31-1.03) (0.09-0.38) (1.02-6.76) Summary of Findings Relative Effect With Extended With Anticoagulation Study Event Rates (%) Anticipated Absolute Effects With No With Extended Anticoagulation 38/599 (6.3) 16/585 (2.7) RR 0.57 7/599 (1.2%) 21/585 (3.6%) RR 2.63

c-e due to imprecision due to imprecision of Evidence Overall Quality Mortality (important outcome) Major bleeding at 1 y (critical outcome) Recurrent VTE at 1 y (critical outcome) Bias Publication Undetected High 102/599 (17) 21/585 (3.6) RR 0.12 Undetected Moderate d,e imprecision Serious Undetected Moderate Serious No serious indirectness indirectness indirectness No serious No serious No serious

c Quality Assessment inconsistency inconsistency inconsistency No serious No serious No serious [Sections 3.1.1-3.1.4] Extended Anticoagulation vs No Extended Anticoagulation for Different Groups of Patients With VTE and Without Cancer VTE and Without [Sections 3.1.1-3.1.4] Extended Anticoagulation vs No for Different Groups of Patients With risk of bias risk of bias risk of bias No serious No serious No serious Table S27 — S27 Table studies), 10-36 mo studies), 10-36 mo studies), 10-36 mo 1,184 (4

1,184 (4 Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision 1,184 (4

46 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

rst f-j

f-j k,l f-j

(Continued) f-j k,l k,l With Extended With Risk Difference (from 186 fewer to 273 fewer) (from 93 fewer to 137 fewer) (from 279 fewer to 409 fewer) 1 more to 346 more) 1 more to 173 more) 19 fewer to 27 fewer) 0 more to 87 more) Anticoagulation (95% CI) Anticoagulation (95% CI) Low risk of bleeding unprovoked distal DVT High risk of bleeding First unprovoked VTE Second unprovoked VTE Moderate risk of bleeding First VTE provoked by surgery First VTE provoked nonsurgical/ﬁ Risk With Risk With No Extended Anticoagulation 300 per 1,000 264 fewer per 1,000 150 per 1,000 132 fewer per 1,000 60 per 1,000 98 more per 1,000 (from 450 per 1,000 396 fewer per 1,000 30 per 1,000 49 more per 1,000 (from 30 per 1,000 26 fewer per 1,000 (from 15 per 1,000 24 more per 1,000 (from (95% CI) (0.09-0.38) (1.02-6.77) Summary of Findings Relative Effect With Extended With Anticoagulation Study Event Rates (%) Anticipated Absolute Effects With No With Extended Anticoagulation 7/599 (1.2%) 21/585 (3.6%) RR 2.63 Continued due to imprecision of Evidence Overall Quality Table S27— Table Burden of anticoagulation not reported Major bleeding at 5 y (critical outcome) Recurrent VTE at 5 y (critical outcome) Bias Publication Undetected High 102/599 (17%) 21/585 (3.6%) RR 0.12 imprecision No serious Serious Undetected Moderate indirectness indirectness No serious No serious Quality Assessment inconsistency inconsistency No serious No serious risk of bias risk of bias No serious No serious studies), 10-36 mo studies), 10-36 mo 1,184 (4

1,184 (4 Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

47 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

rst episode episode rst m n 2004) to 10-fold With Extended With Risk Difference Anticoagulation (95% CI) Anticoagulation (95% CI) See comment See comment m n rst year after discontinuation Ann Intern Med. Risk With Risk With No Extended Anticoagulation (95% CI) Summary of Findings Relative Effect . 2008) and threefold (Prandoni. With Extended With Anticoagulation Ann Intern Med t. t. t. t. Study Event Rates (%) Anticipated Absolute Effects With No With . 2010;170(19):1710-1716); we assumed a 0.5% yearly risk thereafter (3% over 5 y). y). 5 over (3% thereafter risk yearly 0.5% a assumed we 2010;170(19):1710-1716); . Extended Anticoagulation Continued See Table S1, S2, S7, and S10 legends for expansion of abbreviations. S1, S2, S7, and S10 legends for expansion of abbreviations. See Table 95 Arch Intern Med of Evidence PTS not reported Overall Quality Table S27— Table Bias Publication Palareti (PROLONG). 98 Farraj, 96 Kearon et al (LAFIT), Quality Assessment 97 because target INR was 1.75 (low intensity), which has been shown in an RCT to be less effective than a target of 2.5. because target INR was 1.75 (low intensity), which has been shown in an RCT to be less effective than a of 2.5. . 2008;179(5):417-426; 11.0% 19.6% and over over 29.1% 1 3 over y, y, 5 y in Prandoni et al assumed (2007). a We risk of 10% the ﬁ 100 CMAJ 2005) increase in PTS . . 2005) increase in PTS J Thromb Haemost. 0%. 0%. 5

2 Burden of anticoagulation: endured by all patients who continue extended-duration anticoagulation (100%) and applies to patients who stop anticoagulation (no extended-duration anticoagulation) who We excluded Ridker et al (PREVENT) We CI includes both values suggesting no effect and either appreciable harms or benefi can who recurrence for risk low at patients thromboembolism unprovoked Identifying al. et PS, Wells SR, Kahn MA, Rodger in y 1 over 9.3% VTE: unprovoked of episode rst fi with patients in risk Annual Baseline risk over 2 y of 58.8% for PTS and 13.8% for severe PTS (VETO [Venous Thrombosis Outcomes study]; Small number of events. Decision to rate down also takes into account that two studies were stopped early for benefi rst year (Iorio et al); we assumed 2.5% yearly thereafter (15% over 5 y). rst VTE provoked by non surgical factor: about 5% the fi Annual risk in patients with fi Annual risk of major bleeding is based on three risk levels: low, intermediate, and high. The corresponding 0.3%, 0.6%, and 1.2% risks are estimates based on control arms of included studies (see Table 3 ). ). 3 Table (see studies included of arms control on based estimates are risks 1.2% and 0.6%, 0.3%, corresponding The high. and intermediate, low, levels: risk three on based is bleeding major of risk Annual Studies vary in follow-up duration (10 mo to 3 y) and in duration of time-limited VKA (3 to 6 mo). Studies vary in follow-up duration (10 mo to 3 y) and of time-limited VKA (3 6 mo). I Annual risk of VTE recurrence after discontinuing oral anticoagulation therapy in patients with fi rst VTE provoked by surgery: 1% (Iorio A, Kearon C, Filippucci E, et al. Risk of recurrence after a fi Annual risk in patients with second episode of unprovoked VTE: we assumed an RR of 1.5 compared with a fi rst episode of unprovoked VTE: 15% the fi rst year after discontinuation, 7.5% yearly thereafter yearly 7.5% discontinuation, after year rst fi the 15% VTE: unprovoked of episode rst fi a with compared 1.5 of RR an assumed we VTE: unprovoked of episode second with patients in risk Annual Case fatality rate of recurrent VTE after discontinuing oral anticoagulation therapy: 3.6% (Carrier 2010). after discontinuing oral anticoagulation therapy). Case fatality rate of major bleeding during initial oral anticoagulation therapy: 11.3% (Carrier et al) (no data available for … … … … … … … … … … See comment Bibliography: Schulman et al (DURAC 2), discontinue anticoagulant therapy. discontinue anticoagulant therapy. subsequently experience a recurrent VTE (5%, 10%, 15% at 1 y; 15%, 30%, 45% 5 y). Dongen. (Van and 5% yearly thereafter (30% over 5 y). and 5% yearly thereafter (30% over 5 y). Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision … a b c …d e f g …h i (45% over 5 y). …j k l m …n … … … … … See comment

of symptomatic venous thromboembolism provoked by a transient risk factor: a systematic review. of symptomatic venous thromboembolism provoked by a transient risk factor: systematic review.

48 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

j g (Continued)

j (95% CI) (95% CI) With LMWH With (from 31 fewer to 21 more) (from 5 fewer to 16 fewer) (from 9 fewer to 4 more) (from 13 fewer to 43 fewer) (from 32 fewer to 108 fewer) (from 36 fewer to 16 more) Risk Difference cancer No cancer Metastatic cancer Metastatic cancer Nonmetastatic cancer VKA No cancer or nonmetastatic Risk With Risk With 164 per 1,000 7 fewer per 1,000 30 per 1,0001,000 per fewer 11 20 per 1,000 4 fewer per 1,000 80 per 1,0001,000 per fewer 30 200 per 1,0001,000 per fewer 76 80 per 1,0001,000 per fewer 15

Effect Relative (0.81-1.13) (0.55-1.2) (0.46-0.84) (95% CI) a-c Summary of Findings Study Event Rates (%) Anticipated Absolute Effects 53/1,351 (3.9%) 45/1,386 (3.2%) RR 0.81 202/1,231 (16.4) 204/1,265 (16.1) RR 0.96 105/1,349 (7.8) 67/1,378 (4.9) RR 0.62 due due due d,e f h,i to imprecision to imprecision to risk of bias of Evidence VKA With LMWH With Overall Quality Mortality (important outcome) Bias Major bleeding (critical outcome) Recurrent VTE (critical outcome) Publication Undetected Moderate Undetected Moderate Undetected Moderate e i imprecision Serious Serious No serious [Section 3.3] Evidence Proﬁ le: LMWH vs VKA for Long-term Treatment of VTE le: LMWH vs VKA for Long-term Treatment [Section 3.3] Evidence Proﬁ indirectness indirectness indirectness No serious No serious No serious Quality Assessment Table S28 — S28 Table inconsistency inconsistency inconsistency No serious No serious

d h No serious f risk of bias risk of bias No serious Serious No serious (7 studies), 6 mo (8 studies), 6 mo (8 studies), 6 mo 2,496 2,727

2,737 Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

49 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ).

103,110

n (95% CI) (95% CI) With LMWH With (from 12 fewer to 46 fewer) Risk Difference Moderate VKA Risk With Risk With 200 per 1,0001,000 per fewer 30 (Cesarone 2003 Circ abstract). PTS 800,000; requires antiplatelet therapy; 115 , … … Two of these studies Two enrolled only patients Effect Relative 109 (0.77-0.94) (95% CI) Kakkar et al, 114 Summary of Findings Romera et al adjusted to 2-y time frame. adjusted to 2-y time frame. 108 118 injection, no dietary interactions, no frequent blood testing/ monitoring Veiga et al, Veiga LMWH: daily 113 Meyer G et al,

k (separate data provided for cancer and non-cancer patients in one study 107 Study Event Rates (%) Anticipated Absolute Effects 103,106,109 medication, dietary dietary medication, interactions, frequent blood testing/monitoring, increased hospital/ clinic visits 31/44 (70.5%) 34/56 (60.7%) RR 0.85 Gonzalez-Fajardo et al, 112 Lopez-Beret et al, 106 due to Continued adjusted to a 6-mo time frame). adjusted to a 6-mo time frame). l,m 117 risk of bias, indirectness Das et al, of Evidence VKA With LMWH With Overall Quality 111 Table S28— Table Lopaciuk et al, Bias 105 75 y, cancer, metastatic disease; chronic renal or hepatic failure; platelet count cancer, 75 y, and Beyth et al, Publication Undetected Low . 116 Not applicableNot High daily Warfarin: Burden of anticoagulation (important outcome) Lee et al, 104 limitations imprecision No serious , and 3 enrolled both patients with and without cancer PTS (important outcome; assessed with self-reported leg symptoms and signs) Hull et al, 103 No serious m 108,102,105 limitations No serious Serious Hull et al, 102 Quality Assessment ) . 50% of the acute treatment dose during the extended phase of treatment. 50% of the acute treatment dose during extended phase treatment. 116 Ն limitations inconsistency No serious No serious l See Table S1, S2, S5, and S10 legends for expansion of abbreviations. S1, S2, S5, and S10 legends for expansion of abbreviations. See Table 104 3 enrolled only patients with cancer limitations Serious 104,107 (1 study), 2 y The association between leg symptoms and signs at 3 mo and long-term PTS is uncertain. The association between leg symptoms and signs at 3 mo long-term PTS is uncertain. The initial parenteral anticoagulation was similar in both arms for all except one study (Hull et al [2007]) in which patients randomized to LMWH received initially the whereas same patients LWMH, ect selective reporting of outcomes. One study did not report deaths, which is unusual and could reﬂ No study was blinded; diagnosis of major bleeding has a subjective component. by Kahn et al, Baseline risk assumes that patients all wear pressure stockings. Control event rate comes from observational studies in review CI includes both no effect and harm with LMWH. CI includes both no effect and harm with LMWH. Risk of recurrent VTE: low corresponds to patients without cancer (3% estimate taken from recent large of RCTs acute treatment), intermediate corresponds to patients with local or recently resected the burden of injections). cant difference in quality of life but suggested greater satisfaction with LMWH over VKA (questionnaire did not directly assess Hull et al reported no signiﬁ Limited to LMWH regimens that used of these studies enrolled three only enrolled and patients only three Two without patients enrolled cancer, with both cancer, patients with and without cancer (separate data provided for cancer and non- None of the studies were blinded, although the diagnosis of recurrent VTE has a subjective component and there could be a lower threshold for diagnosis of recurrent VTE in VKA-treated patients because patients VKA-treated in VTE recurrent of diagnosis for threshold lower a be could there and component subjective a has VTE recurrent of diagnosis the although blinded, were studies the of None Risk of bleeding: low corresponds to patients without risk factor for bleeding (ie, Patients and investigators not blinded. Self-reported leg symptoms signs after 3 mo of treatment. The 95% CIs for the RR for major bleeding includes a potentially clinically important increase or decrease with LMWH and may vary with the dose of LMWH used during the extended phase of therapy. therapy. of phase extended the during used LMWH of dose the with vary may and LMWH with decrease or increase important clinically potentially a includes bleeding major for RR the for CIs 95% The 100

data from: Hull et al. randomized to VKA received initially UFH. randomized to VKA received initially UFH. switching the treatment of such patients to LMWH is widely practiced. At the same time, there is reluctance to diagnose recurrent VTE in patients who are already on LMWH because there is no attractive no is there because LMWH on already are who patients in VTE recurrent diagnose to reluctance is there time, same the At practiced. widely is LMWH to patients such of treatment the switching alternative treatment option. without cancer, without cancer, Excluded studies (less than 50% of therapeutic dose LMWH during extended phase): Pini et al, … No serious Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Bibliography: Included studies: Deitcher et al, a b c d e f g h i j 2) (based on Prandoni et al history of bleeding without a reversible cause) (Table k l m n

cancer patients in one study). cancer patients in one study). or advanced locally with patients to high and 4%]), to increased is risk low if [particularly al et Prandoni with consistent be to appears and analysis this in studies six the across rate average on (based cancer distant metastatic cancer (Prandoni et al

50 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) t. t. (from 16 fewer to 1 more) (from 7 fewer to 4 more) (from 17 fewer to 1 more) Risk Difference With Rivaroxaban With 9 fewer per 1,000 4 fewer per 1,000

h i a-c Anticipated Absolute Effects Risk With Risk With LMWH and VKA Therapy 29 per 1,000 9 fewer per 1,000 30 per 1,000 11 per 1,000

j … … … (95% CI) (0.44-1.02) (0.44-1.04) (0.34-1.38) Relative Effect HR 0.68 HR 0.68 h i Summary of Findings With With medication, no dietary interactions, no frequent blood testing/ monitoring Rivaroxaban Rivaorxaban: daily 13/1,718 (0.8) i 36/1,731 (2.1) h Study Event Rates (%) Therapy and VKA With LMWH With medication, dietary interactions, frequent blood testing/monitoring, increased hospital/ clinic visits 49/1,718 (2.9) 38/1,731 (2.2) HR 0.67 51/1,718 (3) 19/1,711 (1.1)

e,f e,g e,g due to imprecision due to imprecision due to imprecision of Evidence Overall Quality Death (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) Bias Publication Undetected Moderate Undetected Moderate Undetected Moderate f g g Burden of anticoagulation (important outcome) not reported Serious Serious Serious indirectness indirectness indirectness No serious No serious No serious Quality Assessment hazard ratio. See Table S1, S2, S5, and S7 legends for expansion of other abbreviations. S1, S2, S5, and S7 legends for expansion of other abbreviations. hazard ratio. See Table [Section 3.3] Evidence Proﬁ le: Rivaroxaban vs LMWH and VKA Therapy for Short- and Long-term Treatment of VTE le: Rivaroxaban vs LMWH and VKA Therapy for Short- Long-term Treatment [Section 3.3] Evidence Proﬁ 5 inconsistency inconsistency inconsistency No serious No serious No serious 8 d and then VKA therapy targeted to INR 2.5 for 3, 6, or 12 mo. 8 d and then VKA therapy targeted to INR 2.5 for 3, 6, or 12 mo.

HR HR e e e ف 119 Table S29 — S29 Table risk of bias risk of bias risk of bias No serious No serious No serious

d d d nite or possible fatal VTE in rivaroxaban group and one in LMWH/VKA group. nite or possible fatal VTE in rivaroxaban group and one LMWH/VKA group.

6-12 mo 6-12 mo 6-12 mo Enoxaparin 1 mg/kg bid for ed to be 3 mo (12%), 6 (63%), or 12 (25%). Follow-up was prespecifi One defi Allocation was concealed. Patients, providers, and data collectors were not blinded, but outcome adjudicators were blinded. ITT analysis; 1.0% loss to follow-up. Not stopped early for benefi Allocation was concealed. Patients, providers, and data collectors were not blinded, but outcome adjudicators blinded. ITT t and harm. CI includes values suggesting benefi Rivaroxaban 15 mg bid for 3 wk and then 20 mg/d for a total of 3 (12%), 6 (63%), or 12 (25%) months. Rivaroxaban 15 mg bid for 3 wk and then 20 mg/d a total of (12%), 6 (63%), or 12 (25%) months. 6%; previous VTE, 19%). ed proximal DVT of the legs (unprovoked, 62%; cancer, Included patients had acute, symptomatic, objectively verifi CI includes values suggesting benefi t or no effect; relatively low number of events. t or no effect; relatively low number of events. CI includes values suggesting benefi Bleeds contributing to death: one in the rivaroxaban group and fi ve in the warfarin group. ve in the warfarin group. Bleeds contributing to death: one in the rivaroxaban group and fi Calculated from reported data. 3,449 (1 study),

Participants (Studies) Follow up Risk of Bias Inconsistency Indirectness Imprecision

a b c d e f g h i j 3,429 (1 study),

Bibliography: Einstein DVT. Bibliography: Einstein DVT. … … … … … … … daily Warfarin:

3,449 (1 study),

51 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

j per 1,000 (from 43 fewer to 64 fewer) (from 3 fewer to 15 more) (from 3 more to 16 more) Risk Difference Rivaroxaban With (95% CI) 58 fewer g 7 more per 1,000

i … … Placebo Anticipated Absolute Effects Risk With Risk With 3 per 1,000 2 fewer per 1,000 71 per 1,000 t. t.

a,b

f (95% CI) (0.58-42) (0.04-5.4) (0.09-0.39) medication, no dietary interactions, no frequent blood testing/ monitoring Relative Effect Summary of Findings ) increase in PTS with recurrent VTE in the ipsilateral leg. leg. ipsilateral the in VTE recurrent with PTS in increase ) HR 0.18 RR 4.9 g i 121 With With Rivaroxaban 8/602 (1.3) g 4/598 (0.7) i 0/590 (0) 2/594 (0.34) 1/602 (0.17) RR 0.49 ) to 10-fold (van Dongen et al 120

d,e d,h 42/594 (7.1) d due to imprecision due to imprecision of Evidence Placebo With Overall Quality Bias Publication Mortality (important outcome) Undetected High Major bleeding (critical outcome) Recurrent VTE (critical outcome) PTS (important outcome) not reported ). There is threefold (Prandoni et al 40 Undetected Moderate Undetected Moderate e h imprecision Burden of anticoagulation (important outcome) not reported Serious No serious Serious Study Event Rates (%) indirectness indirectness indirectness No serious No serious No serious [Section 3.3] Evidence Proﬁ le: Rivaroxaban vs Placebo for Extended Anticoagulation of VTE [Section 3.3] Evidence Proﬁ Quality Assessment inconsistency inconsistency inconsistency Table S30 — S30 Table No serious No serious No serious

d d d See Table S1, S2, S5, S10, and S29 legends for expansion of abbreviations. S1, S2, S5, S10, and S29 legends for expansion of abbreviations. See Table 119 risk of bias risk of bias risk of bias No serious No serious No serious

c c nite or possible fatal VTE in rivaroxaban group and one in LMWH/VKA group. nite or possible fatal VTE in rivaroxaban group and one LMWH/VKA group. 6 or 12 mo 6 or 12 mo 6 or 12 mo Included patients had acute, symptomatic, objectively verifi ed proximal DVT of the legs or PE (unprovoked, 73%; cancer, 5%; previous VTE, 19%). 5%; previous VTE, 19%). ed proximal DVT of the legs or PE (unprovoked, 73%; cancer, Included patients had acute, symptomatic, objectively verifi to follow-up. Not stopped early for benefi Allocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. ITT analysis; 0.2% loss t and harm. CI includes values suggesting benefi CI includes values suggesting benefi t or no effect; relatively low number of events. t or no effect; relatively low number of events. CI includes values suggesting benefi One defi Rivaroxaban 20 mg/d for 6 or 12 mo after initial long-term therapy. Rivaroxaban 20 mg/d for 6 or 12 mo after initial long-term therapy. ed to be 6 mo (60%) or 12 (40%). Follow-up was prespecifi Calculated from reported data with addition of one event to each event rate as event rate 0 in control group. Calculated from reported data with addition of one event to each rate as 0 in control group. Bleeds contributing to death: none in the rivaroxaban group and none in the warfarin group. Bleeds contributing to death: none in the rivaroxaban group and warfarin group. PTS: baseline risk over 2 y of 58.8% for PTS and 13.8% for severe PTS (Kahn et al 1,188 (1 study), 1,196 (1 study),

Bibliography: Einstein DVT. Bibliography: Einstein DVT.

a b c d e f g h i j

1,196 (1 study), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision … … … … … … … … … … … … … … … … … … … Rivaroxaban: daily

52 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) (from 8 fewer to 13 more) (from 10 fewer to 9 more) (from 7 fewer to 16 more) Risk Difference With Dabigatran With 3 fewer per 1,000 0 more per 1,000 f g t. t. Anticipated Absolute Effects Warfarin Risk With Risk With 19 per 1,000 17 per 1,000 0 fewer per 1,000 19 per 1,000 … … … (95% CI)

(0.45-1.48) (0.53-1.79) (0.65-1.84) a-c Relative Effect HR 0.82 HR 1.01 f g Summary of Findings With With medication, No dietary restrictions, no frequent blood testing/ monitoring Dabigatran Dabigatran: daily 20/1,274 (1.6) 30/1,274 (2.4) f g medication, restrictions, dietary frequent blood testing/monitoring, increased hospital/ clinic visits 21/1,265 (1.7) 21/1,274 (1.6)24/1,265 (1.9) HR 0.98 24/1,265 (1.9)

d,e d,e d,e due to imprecision due to imprecision due to imprecision of Evidence Warfarin With Overall Quality Death (important outcome) Recurrent VTE (critical outcome) Bias Publication Undetected Moderate Undetected Moderate Undetected Moderate e e e Burden of anticoagulation (important outcome) not reported Serious Serious Serious [Section 3.3] Dabigatran vs VKA Therapy for Long-term Treatment of VTE [Section 3.3] Dabigatran vs VKA Therapy for Long-term Treatment Study Event Rates (%) indirectness indirectness indirectness No serious No serious No serious Table S31 — S31 Table Quality Assessment inconsistency inconsistency inconsistency No serious No serious No serious See Table S2, S5, and S29 legends for expansion of abbreviations. S2, S5, and S29 legends for expansion of abbreviations. See Table

d d d 122 risk of bias risk of bias risk of bias No serious No serious No serious 6 mo 6 mo 6 mo Warfarin adjusted to achieve an INR of 2.0 to 3.0 for 6 mo after an initial treatment with LMWH or IV UFH. adjusted to achieve an INR of 2.0 3.0 for 6 mo after initial treatment with LMWH or IV UFH. Warfarin ed ITT analysis; 1.1% loss to follow-up. Not stopped early for benefi Allocation was concealed. Patients, providers, data collectors, and outcome adjudicators were blinded. Modifi CI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. t or harm; relatively low number of events. CI includes values suggesting no effect and either benefi One fatal major bleeding event in dabigatran group and one warfarin group. Dabigatran 150 mg bid taken orally for 6 mo after an initial treatment with LMWH or IV UFH. Dabigatran 150 mg bid taken orally for 6 mo after an initial treatment with LMWH or IV UFH. ed proximal DVT of the legs or PE. Included patients had acute, symptomatic, objectively verifi One fatal VTE in dabigatran group and three fatal VTEs in warfarin group. One fatal VTE in dabigatran group and three VTEs warfarin group. Bibliography: Schulman et al. 2,539 (1 study),

Major bleeding (critical outcome) 2,539 (1 study), a b c d e f g … … … … … … … daily Warfarin:

2,539 (1 study), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

53 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians cantly .01) ,

P .001) Severe PTS: 11% .001) , ,

No difference between groups better outcome with ambulation and bandaging or stockings compared with bed rest ( Mild-moderate PTS: 20% (RR, 0.42; 95% CI 0.27-0.66; P 0.25-0.95; CI, 95% 0.49; (RR, P PTS symptoms: 0% between groups PTS: 47% Severe 23% stockings: 13% PTS symptoms: 25% (CI, 15.6%-33.4%) 49% (CI, 38.7%-59.4%) 0.8 (95 % CI, 0.3-1.3) PTS score: Signifi Compression stockings: Control group: Mild-moderate Placebo: PTS symptoms: 4% Control group: PTS symptoms: Control group: 20% HR: then every 6 mo to a median of 76 mo 3 and 6 mo, 0-7 y (mean, 3 y) PTS compression 3-5 y Compression stockings: C4 on Ն of mild to moderate and severe PTS of mild to moderate and severe PTS CEAP) Leg circumferencePTS score (Villalta-Prandoni) Calf circumference: Cumulative incidence PTS symptoms 57 mo (mean) Compression stockings: Overall leg pain 2 y Leg pain: No difference Cumulative incidence PTS shin changes ( early 1 early ambulation ambulation (13 patients) (30 mm Hg) 1 (13 patients) no compression (11 patients) below-knee customized elastic compression stockings with ankle pressure 30-40 mm Hg (96 patients) below-knee elastic compression stockings 20-30 mm Hg (24 patients) LMWH followed by oral anticoagulation no intervention (98 patients) (23 patients) below-knee elastic compression stockings 30-40 mm Hg (90 patients) intervention (90 patients) below-knee, 26-36 mm Hg (84 patients) stockings (85 patients) Elastic stockings Bed rest for 9 d, Compression stockings: Compression stockings: All anticoagulated with Control group: Placebo: placebo stocking Inelastic bandages Compression stockings: Control group: no Compression stockings: Control group: stopped rst rst symptomatic [Section 4.1] Elastic Stocking for Prevention of PTS: Clinical Description and Results [Section 4.1] Elastic Stocking for Prevention of PTS: Clinical Description and Results with fi proximal DVT with valvular incompetence 1 y post-DVT with acute DVT followed long term episode of symptomatic, acute proximal DVT proximal DVT without PTS after 6 mo of compression stockings Table S32 — S32 Table RCT 194 patients RCT 47 asymptomatic patients RCT 180 patients with fi Type of Type Publication Participants Interventions Outcomes Follow-up Results /2008 single center RCT, rst or recurrent 169 fi 130 /2001 /2004 /1997 124 120 127 /2004 2-y follow-up to RCT 37 symptomatic patients 84

Ashwanden et al

Ginsberg et al Prandoni et al Author/Year Brandjes et al

S1, S2, S5, S10, and S29 legends for expansion of other abbreviations. See Table Partsch et al

54 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S33 —[Section 4.1] Elastic Stocking for Prevention of PTS: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up

Brandjes et al127 /1997 Y Y (sealed envelopes) Patients: N Intervention group: 4 lost Caregivers: N to follow-up, 19 died Assessors: Y Control group: 2 lost to Data analysis: PY follow-up, 18 died Ginsberg et al124 /2001 Y Probably, but not Patients: Y Intervention: lost to speciﬁ ed Caregivers: Y follow-up not reported, Assessors: Y 3 died data analysis: PY Control group: lost to follow-up not reported Prandoni et al120 /2004 Y Y Patients: N Intervention: 2 lost to caregivers: N follow-up, 6 died Assessors: Y Control group: 13 lost Data analysis: PY to follow-up Ashwanden et al130 /2008 Y Y Patients: N Intervention: Caregivers: N 19 (described) Assessors: N Control group: Data analysis: N 13 (described) See Table S5 legend for expansion of abbreviations.

55 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians rst

a,b (from 82 fewer to 141 more) (from 177 fewer to 316 fewer) Stockings (95% CI) Stockings (95% CI) Elastic Compression Risk Difference With Risk Difference With 259 fewer per 1,000 … 2 more per 1,000 k d,f j Anticipated Absolute Effects Stockings Risk With Risk With No Elastic Compression 479 per 1,000 210 per 1,000

e e (95% CI) (0.61-1.67) (0.34-0.63) ). ). Summary of Findings Relative Effect 129 Stockings With Elastic With Compression See Table S1 and S5 legends for expansion of abbreviations. S1 and S5 legends for expansion of abbreviations. See Table 41/210 (19.5) RR 0.46 d 126 … … … See comment Study Event Rates (%) Stockings Compression 26/188 (13.8) 26/186 (14) RR 1.01 91/211 (43.1) With No Elastic With

) and 29.1% conﬁ rmed VTE (Prandoni ) and 29.1% conﬁ c h,i 128

due to imprecision due to risk of bias of Evidence Overall Quality Overall This probably underestimates the PTS baseline risk given that overall, 52% of patients reported the current use current the reported patients of 52% overall, that given risk baseline PTS the underestimates probably This 40 Bias Recurrent VTE (critical outcome) Quality of life (important outcome) Publication Undetected Moderate because they respectively randomized patients 7 and 12 mo after their DVT rather than We at the time of the acute DVT. 125 PTS (critical outcome; assessed with Villalta Score) PTS (critical outcome; assessed with Villalta Undetected Moderate i imprecision No serious Serious and Belcaro et al 124 indirectness indirectness No serious No serious Quality Assessment used ﬂ at-knitted stockings (30-40 mm Hg of pressure at used the ﬂ ankle); stockings were started at hospital discharge an average of 1 wk after admission. In both studies, 120 inconsistency inconsistency No serious

[Section 4.1] Evidence Profi le: Elastic Compression Stockings vs No Elastic Compression Stockings To Prevent PTS of the Leg le: Elastic Compression Stockings vs No To [Section 4.1] Evidence Profi h We excluded Ginsberg et al We 123 because they randomized patients to receive stockings at the time of diagnosis of DVT vs 2 wk later. because they randomized patients to receive stockings at the time of diagnosis DVT vs 2 wk later. 126 No serious c risk of bias No serious Serious most events occurred during the fi rst 6 mo: the cumulative incidence of PTS in the control group was 40% after 6 mo, 47% after 1 y, and 49% after 2 y. and 49% after 2 y. rst 6 mo: the cumulative incidence of PTS in control group was 40% after mo, 47% 1 y, most events occurred during the fi Table S34 — S34 Table 120 excluded patients with recurrent ipsilateral DVT, preexisting leg ulcers, or signs of CVI, bilateral thrombosis, a short life expectancy, or a contraindication for use of stockings (eg, advanced-stage (eg, stockings of use for contraindication a or expectancy, life short a thrombosis, bilateral CVI, of signs or ulcers, leg preexisting DVT, ipsilateral recurrent with patients excluded used graded elastic compression stockings (40 mm Hg of pressure at the ankle, 36 mm Hg at the lower calf, and 21 mm Hg at the upper calf); stockings were applied 2 to 3 wk after the fi 120 127

) k 5 y 2 y Prandoni In Prandoni, than PTS. did not rate down the quality of evidence for recurrent VTE lack blinding because this a more objective outcome We Severe PTS: assuming the same RR of 0.46 and a baseline risk of 8.1% over 2 y, the absolute reduction is 44 fewer severe PTS per 1,000 (from 30 fewer to 53 fewer) over 2 y. 1,000 (from 30 fewer to 53 fewer) over 2 y. the absolute reduction is 44 fewer severe PTS per Severe PTS: assuming the same RR of 0.46 and a baseline risk 8.1% over 2 y, This is an important outcome that should be considered in future studies. The effect estimate shown here results from a meta-analysis (Mantel-Haenszel fi xed-effects model) of the two relevant trials. A fi xed-effects model was chosen because of the small number of studies available. available. studies of number small the of because chosen was model xed-effects fi A trials. relevant two the of model) xed-effects fi (Mantel-Haenszel meta-analysis a from results here shown estimate effect The Brandjes Patients were not blinded to the treatment assignment, and outcomes partly based on subjective report of symptoms. This estimate is based on the fi ndings of the VETO (Venous Thrombosis Outcomes) study. study. Outcomes) Thrombosis (Venous VETO the of ndings fi the on based is estimate This CI includes both negligible effect and appreciable benefi t or appreciable harm. t or appreciable harm. CI includes both negligible effect and appreciable benefi nite VTE (Heit This estimate is the mean of two estimates derived from studies: 12.4% probable/defi

374 (2 studies), 0 (0 Participants (Studies), Follow-up Risk of Bias421 (2 studies), Inconsistency Indirectness Imprecision Bibliography: Kolbach et al. stockings were used for 2 y. stockings were used for 2 y. a b varicosis. ciency). Brandjes et al excluded patients with short life, paralysis of the leg, bilateral thrombosis, leg ulcers, or extensive peripheral arterial insufﬁ c d e f g h i j k

of compression stockings during study follow-up. of compression stockings during study follow-up. also excluded Arpaia et al episode of proximal DVT. episode Prandoni of proximal DVT.

56 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

H R v s H R 1 (from 174 fewer to 179 more) Risk Difference With Compression With Stockings (95% CI) Stockings (95% CI)

a-c Anticipated Absolute Effects Stockings Compression Risk With No Risk With 579 per 1,000 23 fewer per 1,000 (0.70-1.31) (95% CI) Relative Effect Summary of Findings ) d Stockings With Compression With Study Event Rates (%) With No With Stockings 33/57 (57.9) 32/58 (55.2) RR 0.96

Compression h

due to e-g risk of bias, imprecision of Evidence Overall Quality Low g Ulceration not reported Quality of life not reported Recurrent VTE not reported Bias Publication ve clinical criteria, including success symptoms refers and to ulcer the development). absence Treatment of treatment failure. Undetected f Serious Symptomatic relief (critical outcome; assessed with treatment success indirectness No serious Quality Assessment See Table S1, S2, S5, and S10 legends for expansion of abbreviations. S1, S2, S5, and S10 legends for expansion of abbreviations. See Table 131 [Section 4.2.1] Evidence Proﬁ le: Compression Stockings vs No Compression Stockings for Patients With PTS le: Compression Stockings vs No for Patients With [Section 4.2.1] Evidence Proﬁ inconsistency ; Frulla. 2005 124 No serious e Table S35 — S35 Table Serious

12-26 mo Ginsberg et al: placebo stockings (calf or thigh length, depending on symptoms). Ginsberg et al: placebo stockings (calf or thigh length, depending on symptoms). ned a priori based on any of fi Ginsberg et al reported treatment failure (defi with an RR of 4 for skin complications. after Stroke) trial suggests that compression stockings is associated Indirect evidence from the CLOTS1 (Clots in Legs Or sTockings Ginsberg et al: Adequacy of sequence generation and allocation concealment were unclear; patients and outcome assessors were adequately blinded; unclear whether analysis followed the ITT principle; ed only one small study partially supported by industry (provision of graduated compression stockings). Publication bias not detected but ruled out given that we identifi Ginsberg et al: graduated compression stockings, 30-40 mm Hg (calf or thigh length, depending on symptoms). Patients were encouraged to wear stockings as much as possible during waking hours. Frulla hours. waking during possible as much as stockings wear to encouraged were Patients symptoms). on depending length, thigh or (calf Hg mm 30-40 stockings, compression graduated al: et Ginsberg symptoms and signs suggestive of PTS. Frulla (2005) included patients with clinical Ginsberg et al included patients with PTS 1 y after chronic, typical proximal DVT. Very small number of patients. small number of patients. Very Absence of ulcer included in the treatment success outcome in Ginsberg et al. Absence of ulcer included in the treatment success outcome Ginsberg et al. 115 (2 studies), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision … … … … … … ... … … … ...... … … … … … … … … … … … … ...... … … … … … … ......

a b c d e f g h i comparison). comparison). (2005): below-knee graded elastic compression stockings (ECS) (30-40 mm Hg at the ankle). Patients in both arms of the study received hydroxyethylrutosides (HR) (we considered the ECS ECS the considered (we (HR) hydroxyethylrutosides received study the of arms both in Patients ankle). the at Hg mm (30-40 (ECS) stockings compression elastic graded below-knee (2005): were concealment allocation and generation sequence but to, adhered was principle ITT complete. was follow-up blinded; were assessors outcome (2005): Frulla complete. was follow-up whether unclear and patients were not blinded. unclear,

Bibliography: Ginsberg 2001 Frulla used the Villalta scale. scale. Frulla used the Villalta

57 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

(Continued) a-c Compression Risk Difference Device (95% CI) Device (95% CI) quality of life in the intervention groups was 2.3 higher (1.04 lower to 5.64 higher) symptomatic relief in the intervention groups was 0.41 SDs higher (0.02 lower to 0.85 higher) With Intermittent With The mean The mean Anticipated Absolute Effects mean Device of life in the control groups was 50.2 symptomatic relief in the control groups was 0 Intermittent Compression Risk With No Risk With (95% CI) Summary of Findings Relative Effect … The d Device Compression With Intermittent With 41 d Study Event Rates (%) … 0 … The mean quality 41 Device With No With Intermittent Compression

e-i f-h,k,l

due to imprecision due to imprecision of Evidence Overall Quality Moderate Moderate i i,l Bias Publication Undetected Undetected h h Serious Serious

g g indirectness indirectness No serious No serious

f Quality Assessment Quality of life (critical outcome; measured with VEINES-QOL; range scores, 0-100; better indicated by higher values) inconsistency inconsistency No serious No serious

e k [Section 4.2.2] Evidence Proﬁ le: Intermittent Compression Device vs No Intermittent Compression Device for Patients With Severe PTS le: Intermittent Compression Device vs No for Patients With [Section 4.2.2] Evidence Proﬁ risk of bias risk of bias No serious No serious ), d ), d,j Table S36 — S36 Table Symptomatic relief (critical outcome; measured with symptom score (includes scoring of pain, swelling, and limitation of activity); range of scores, 10-70; better indicated by higher values) Symptomatic relief (critical outcome; measured with symptom score (includes scoring of pain, swelling, and limitation activity);

8 wk 8 wk 82 (2 studies Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision 0 (1 study

58 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Compression Risk Difference Device (95% CI) Device (95% CI) With Intermittent With Anticipated Absolute Effects Device Intermittent Compression Risk With No Risk With (95% CI) Summary of Findings Relative Effect Device Compression With Intermittent With

Study Event Rates (%) m

Device n With No With Intermittent Compression cial effect. cial effect. Continued

of Evidence Overall Quality Ulceration not reported Table S36— Table Recurrent VTE not reported Bias Publication See Table S1, S5, and S10 legends for expansion of other abbreviations. S1, S5, and S10 legends for expansion of other abbreviations. See Table 133 Quality Assessment O’Donnell, 2008. 132

0%. 0%. 5

2 O’Donnell et al indicated no cases of recurrent VTE by the end of this study but judged the follow-up period to be short. O’Donnell et al indicated no cases of recurrent VTE by the end this study but judged follow-up period to be short. Ginsberg et al: Extremity pump used bid for 20 min each session; 15 mm Hg (placebo pressure) for 1 venous mo. return O’Donnell lower-limb assist et device al: with Venowave no connection between Crossover RCTs. small number of patients. CI includes both values suggesting no effect and a benefi Very CLOTS1 the from evidence Indirect effects. side skin-related nonserious developed participants other Three ulceration. venous a developed group control the in patient one that indicated al et O’Donnell In both studies, sequence generation was adequate; patients were blinded. Analysis adhered to ITT principle, and there were no missing outcome data. In Ginsberg et al (but not in O’Donnell et al), out- Some concerns with indirecteness given relatively short follow-up (8 wk). outcome assessors were Sequence not generation blinded, was and adequate; it patients was were not blinded; clear analysis adhered to ITT principle; and there were no missing outcome data. However, Ginsberg et al: Extremity pump used bid for 20 min each session; 50 mm Hg (therapeutic pressure) for 1 mo. O’Donnell et al: Venowave lower-limb venous return assist device to wear for most of the day the of most for wear to device assist return venous lower-limb Venowave al: et O’Donnell mo. 1 for pressure) (therapeutic Hg mm 50 session; each min 20 for bid used pump Extremity al: et Ginsberg Patients with previous DVT symptoms of severe PTS. I Publication bias not detected but not ruled out given that we identifi ed only two small studies with one (Ginsberg et al) partially supported by industry (provision of devices). ed only two small studies with one (Ginsberg et al) partially supported by industry (provision of devices). Publication bias not detected but ruled out given that we identifi O’Donnell et al. ed only one small study. Publication bias not detected but ruled out given that we identifi come assessors were not blinded, and it was not clear whether allocation was concealed. come assessors were not blinded, and it was clear whether allocation concealed. Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision motor and planar sheet for 8 wk. (Clots in Legs after Or Stroke) sTockings trial suggests that compression stockings are associated with an RR of 4 for skin complications: Common side effects are attributed to were Venowave heat sensation, skin irritation, and increased sweating.

… a …b c d …e f g h …i j k …l m n … ... … … … ...... … Bibliography: Ginsberg 1999, … … … … … ... … … … ...... for 8 wk. for 8 wk. whether allocation was concealed.

59 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ) c e et al (95% CI) (from 71 fewer to 247 more) (from 15 fewer to 275 more) Risk Difference With Risk Difference With Venoactive Medication Venoactive

a,b Anticipated Absolute Effects Venoactive Venoactive Medication Risk With No Risk With 476 per 1,000 67 more per 1,000 61 per 1,000 63 more per 1,000 (95% CI) (0.85-1.52) (0.76-5.51) Summary of Findings Relative Effect Medication With Venoactive Venoactive With Study Event Rates (%) With No With Venoactive Venoactive Medication 39/82 (47.6) 44/81 (54.3) RR 1.14 5/82 (6.1%) 13/121 (10.7%) RR 2.04

d,f-h due to d-g of Evidence inconsistency, inconsistency, imprecision due to imprecision ts. ts. Overall Quality Overall

Low Moderate g g Ulceration not reported cant. cant. Quality of life not reported Recurrent VTE not reported Bias Side effects (critical outcome) Publication

Undetected 5 or decreased by 30% at 12 mo compared with baseline in Frulla et al ; improved tiredness of the leg at 8 wk in de Jongst f Undetected f , Serious Serious because it compared two venoactive medications. because it compared two venoactive medications. 135 indirectness indirectness No serious

h Quality Assessment See Table S1, S5, and S10 legends for expansion of abbreviations. S1, S5, and S10 legends for expansion of abbreviations. See Table No serious 134 e [Section 4.3] Evidence Proﬁ le: Venoactive Medication vs No Venoactive Medication for Patients With PTS Medication for Patients With Medication vs No Venoactive le: Venoactive [Section 4.3] Evidence Proﬁ Serious inconsistency No serious

d d Table S37 — S37 Table de Jongste 1989. 131 risk of bias risk of bias 7%. 7%. 77%. 77%. 5 5 Symptomatic relief (critical outcome; assessed with PTS score (Villalta scale)

2 2 Patients with PTS and history of DVT in PTS leg. Patients with PTS and history of DVT in leg. not did they patients, blinded al et Jongste de Although follow-up. complete had and assessors outcome blinded studies Both unclear. were concealment allocation and generation sequence studies, both In I I ed only two small studies, and it was unclear whether they were funded by industry. Publication bias not detected but ruled out given that we identifi Included studies assessed rutosides; we excluded Monreal et al effect the t; benefi most the showed report to chose we symptom The score. composite a report not did but cramps) and legs, restless feeling, swelling heaviness, (pain, symptoms other assessed Investigators Small number of patients. CI including both values suggesting harms and benefi … … … …203 (2 studies) No serious …Bibliography: Frulla 2005 … … … … … … … … … … … … ... … ... … ... … … … … … … … ... … ...... the ITT principle, it did not blind patients. adhere to the ITT principle and did not use a validated scale measure symptomatic relief. Although Frulla (2005) adhered 163 (2 studies ) 163 (2 studies No serious a b c d e f g h

estimates for the other symptoms ranged from 0.8 to 1.4, and none was statistically signiﬁ

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision

60 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians t. t. fatal. fatal. Risk (95% CI) Fondaparinux Difference With Difference With (from 25 fewer to 6 more) (from 12 fewer to 11 more) (from 8 fewer to 32 more) 2.0. 2.0. . 4 fewer per 1,000 13 fewer per 1,000 f g UFH Risk With Risk With

43 per 1,000 9 more per 1,000 50 per 1,000 23 per 1,000 a-c (0.82-1.74) (0.51-1.12) (0.49-1.49) (95% CI) Relative Effect RR 0.85 RR 0.75 Summary of Findings f g 2.0. 2.0. . With With Fondaparinux 22/1,103 (2.0) 42/1,103 (3.8) f g Study Event Rates (%) Anticipated Absolute Effects 48/1,110 (4.3) 57/1,103 (5.2) RR 1.20 56/1,110 (5.0) 26/1,110 (2.3) due due due d,e d,e d,e of Evidence UFH With to imprecision to imprecision to imprecision Overall Quality Mortality (important outcome) Major bleeding (critical outcome) Recurrent VTE (critical outcome) Undetected Moderate Undetected Moderate Undetected Moderate e e e Serious Serious Serious 50, 50 to 100, or > 100 kg, respectively) SC once daily given for at least 5 days and until the use of VKAs resulted in an INR 50, 50 to 100, or > 100 kg, respectively) SC once daily given for at least 5 days and until the use of VKAs resulted in an , indirectness indirectness indirectness No serious No serious No serious [Section 5.4] Evidence Proﬁ le: Fondaparinux vs IV UFH for Initial Anticoagulation of Acute PE [Section 5.4] Evidence Proﬁ Quality Assessment inconsistency inconsistency inconsistency Table S38 — S38 Table No serious No serious No serious See Table S1, S2, S4, and S7 legends for expansion of abbreviations. S1, S2, S4, and S7 legends for expansion of abbreviations. See Table

d d d 136 of bias of bias of bias No serious risk No serious risk No serious risk 3 mo 3 mo 3 mo UFH continuous IV infusion (ratio of the aPTT to a control value, 1.5-2.5) given for at least 5 days and until the use of VKAs resulted in an INR UFH continuous IV infusion (ratio of the aPTT to a control value, 1.5-2.5) given for at least 5 days and until use VKAs patients were lost to follow-up. Not stopped early for benefi Allocation was concealed. Patients, providers, and data collectors not blinded. Outcome adjudicators were blinded; 0.6% of randomized CI includes values suggesting no effect and values suggesting either benefi t or harm; relatively low number of events. t or harm; relatively low number of events. CI includes values suggesting no effect and either benefi Fourteen patients in the fondaparinux group and 12 in the LMWH group had a major bleeding during the initial period (6-7 d). Of these, one in the fondaparinux group and one in the UFH group were Fondaparinux (5.0, 7.5, or 10.0 mg in patients weighing All patients had acute symptomatic hemodynamically stable PE. Sixteen fatal VTE in fondaparinux group and 15 fatal VTE in UFH group. Sixteen fatal VTE in fondaparinux group and 15 UFH group. 2,213 (1 study),

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias a b c d e f g 2,213 (1 study), Bibliography: Büller et al.

2,213 (1 study),

61 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) in Otero et al. in Otero et al. 140 Risk Difference (from 22 fewer to 26 more) (from 7 fewer to 44 more) (from 2 fewer to 69 more) With Early Discharge With

a,b Standard Risk With Risk With Discharge 26 per 1,000 11 fewer per 1,000 9 per 1,000 2 more per 1,000 4 per 1,000 8 more per 1,000 (95% CI) (0.17-1.97) (0.25-6.03) (0.45-16.71) Relative Effect Summary of Findings Discharge With Early With ; low risk on clinical prediction rule by Uresandi et al 139 Study Event Rates (%) Anticipated Absolute Effects Discharge 6/228 (2.6) 4/243 (1.6) RR 0.58 2/228 (0.9) 3/243 (1.2) RR 1.23 1/228 (0.4) 4/243 (1.6) RR 2.74 With Standard With due due due c,d c,d c,d t. t. Overall Evidence Quality of to imprecision to imprecision to imprecision Quality of life not reported Mortality (critical outcome) Major bleeding (critical outcome) Publication Bias Nonfatal recurrent PE (critical outcome) Undetected Moderate Undetected Moderate Undetected Moderate d d d Serious Serious Serious indirectness indirectness indirectness No serious No serious No serious Quality Assessment [Section 5.5] Evidence Profi le: Early Discharge vs Standard Discharge in the Treatment of Acute PE le: Early Discharge vs Standard in the Treatment [Section 5.5] Evidence Profi See Table S1, S2, and S5 legends for expansion of abbreviations. S1, S2, and S5 legends for expansion of abbreviations. See Table 138 inconsistency inconsistency inconsistency Table S39 — S39 Table Aujesky et al. No serious No serious No serious c c c 137 Serious Serious Serious 3 mo 3 mo 3 mo The two RCTs included only patients with low risk: risk classes I or II on the Pulmonary Embolism Severity Index in Aujesky et al included only patients with low risk: risk classes I or II on the Pulmonary Embolism Severity Index in Aujesky et The two RCTs CI includes both values suggesting no effect and appreciable harm or benefi Mean length of hospital stay: 3.4 (SD 1.1) vs 9.3 (SD 5.7) in Otero et al and 0.5 (SD 1.0) vs 3.9 (SD 3.1) in Aujesky et al. Mean length of hospital stay: 3.4 (SD 1.1) vs 9.3 5.7) in Otero et al and 0.5 1.0) 3.9 3.1) Aujesky al. high unexpectedly was mortality short-term of rate the because early stopped study reported, assessors outcome of blinding no analysis, ITT follow-up, to lost patients no concealed, allocation al: et Otero 471 (2 studies), 471 (2 studies), … Bibliography: Otero et al, … … … … … … … … … … 471 (2 studies),

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision a b c d in the early discharge group (2 [2.8%] vs 0 [0%]). Aujesky et al: unclear whether allocation was concealed, three (1%) patients had missing outcome data, ITT analysis, outcome adjudicators blinded, no early stoppage.

62 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

i,j i,j m Administered Thrombolytic (from 56 fewer to 28 more) (from 34 fewer to 12 more) (from 0 more to 2 more) (from 7 fewer to 3 more) Risk Difference With Systemically With Therapy (95% CI) Therapy (95% CI) Low Low High

a-d Therapy Systemically Administered Risk With No Risk With Thrombolytic 89 per 1,000 27 fewer per 1,000 57 per 1,000 17 fewer per 1,000 1 per 1,000 1 more per 1,000 11 per 1,000 3 fewer per 1,000

l (95% CI) (0.37-1.31) (0.4-1.21) (1-2.68) Relative Effect Summary of Findings Therapy Administered Thrombolytic With Systemically With Study Event Rates (%) Anticipated Absolute Effects Therapy With No With 26/423 (6.1) 15/424 (3.5) RR 0.7 30/404 (7.4) 18/397 (4.5) RR 0.7 24/423 (5.7) 38/424 (9) RR 1.63 Systemically Administered Thrombolytic

e,f,h,k due due e-h e-h to risk of bias and imprecision to risk of bias and imprecision due to risk of bias and imprecision of Evidence Overall Quality Low Low Moderate h h h Mortality (critical outcome) Major bleeding (critical outcome) Recurrent PE (important outcome) Undetected Undetected Undetected g g k Serious Serious Serious indirectness indirectness indirectness No serious No serious No serious

f f f Quality Assessment [Section 5.6.1] Evidence Proﬁ le: Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With Acute PE le: Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With [Section 5.6.1] Evidence Proﬁ inconsistency inconsistency inconsistency No serious No serious No serious e e e Table S40 — S40 Table Serious Serious Serious 30 d 10 d 30 d

801 (9 studies), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias 847 (12 studies),

847 (12 studies),

63 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ). See Table Table See ). 154

m Administered Thrombolytic (from 0 more to 104 more) Risk Difference

With Systemically With Therapy (95% CI) Therapy (95% CI) 155 High Fassulo et al et Fassulo 153 Therapy Systemically Administered Risk With No Risk With Thrombolytic 62 per 1,000 39 more per 1,000 (95% CI) Relative Effect Summary of Findings Therapy Administered Thrombolytic With Systemically With In that case, the absolute number of death associated with thrombolytics 156 Study Event Rates (%) Anticipated Absolute Effects ) and two recently published studies (Becattini C et al, et C (Becattini studies published recently two and ) Therapy With No With 152 Systemically Administered Thrombolytic Agnelli et al et Agnelli 151 Continued of Evidence Overall Quality Wan et al, et Wan 150 Table S40— Table 11), only three were clearly restricted to patients with right ventricular dysfunction; the rest either did not specify related eligibility related specify not did either rest the dysfunction; ventricular right with patients to restricted clearly were three only 11), 5 The median risk of bleeding over the ﬁ rst 10 d reported in the eligible trials was 3.1%. In that case, the absolute number of major bleeds major of number absolute the case, that In 3.1%. was trials eligible the in reported d 10 rst ﬁ the over bleeding of risk median The 157 Quality Assessment extracted from three systematic reviews (Dong et al, et (Dong reviews systematic three from extracted 141-149 0%. 0%. 5

2 Major bleeding risk stratifi cation derived from the RIETE cohort. RIETE the from derived cation stratifi risk bleeding Major Thrombolysis was in addition to anticoagulation (most of the studies used heparin followed by warfarin; three studies used warfarin only). only). Thrombolysis was in addition to anticoagulation (most of the studies used heparin followed by warfarin; three warfarin Studies included patients at low risk of bleeding. Inverted funnel plots suggested possible publication bias in favor of thrombolytics. Report of methodologic quality was poor in most studies. Of the 12 eligible studies, allocation was concealed in fi ve, three were single blinded (outcome assessor), six were double blinded, and three were three and blinded, double were six assessor), (outcome blinded single were three ve, fi in concealed was allocation studies, eligible 12 the Of studies. most in poor was quality methodologic of Report t and no effect or harm; small number of events. CI includes values suggesting both benefi CI includes values suggesting both harm and no effect; small number of events. Included studies used different thrombolytic agents with varying doses and durations of administration; no statistical heterogeneity was noted. was noted. Included studies used different thrombolytic agents with varying doses and durations of administration; no statistical heterogeneity criteria. eligibility. related specify not did or patients such of number a included either rest the whereas them, excluded six (shock); compromise hemodynamic with patients exclusively included study One I Recurrent PE stratifi cation based on the simplifi ed Pulmonary Embolism Severity Index validated in the RIETE (Registro Informatizado de la Enfermedad Tromboembólica) cohort. ed Pulmonary Embolism Severity Index validated in the RIETE (Registro Informatizado de la Enfermedad Tromboembólica) cation based on the simplifi Recurrent PE stratifi Some studies suggest that the baseline risk of mortality in patients with hemodynamic instability is high as 30% et (Wood al). major bleeding with thrombolytics use. Indirect evidence from studies of thrombolysis for myocardial infarction and acute stroke provide more-precise estimates increase criteria or included both patients with and without right ventricular dysfunction. As a result, it was not possible to perform reliable categorization of studies to conduct subgroup analyses based on the pres- the on based analyses subgroup conduct to studies of categorization reliable perform to possible not was it result, a As dysfunction. ventricular right without and with patients both included or criteria ence or absence of right ventricular dysfunction hemodynamic compromise.

Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision Publication Bias a b c d e f g h i j would be 90 fewer per 1,000 (from 189 to 93 more). k l m

(n compromise hemodynamic with patients to restricted not studies Of increased is evidence of quality therapy, thrombolytic with bleeding in increase the For t. beneﬁ for early stopped was studies the of None data. outcome missing on report not did studies Most blinded. not indirectness of this evidence to patients with PE is minor. from low to moderate because there is high quality evidence of this association in patients with myocardial infarction and the studies earlier Nine Bibliography: S1 and S2 legends for expansion of abbreviations. S1 and S2 legends for expansion of abbreviations. with thrombolysis would be 20 per 1,000 (from 0 more to 52 more). with thrombolysis would be 20 per 1,000 (from 0 more to 52 more).

64 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Comments Comments available bleeding required a blood transfusion. complete the treatment regimen were excluded from the analysis. received nonrandomized Uncertain if therapy. deaths were in patients who were randomized or not randomized. Trial stopped early for Trial t. All patients had beneﬁ cardiogenic shock at randomization. Heparin-treated patients appear to have failed heparin therapy before randomization, whereas the SK patients had not. at 60 h of treatment warfarin warfarin treatment of h 60 at mo. 6 for mg 25 dose initial angiographic reperfusion (not clearly stated). graphic reperfusion. Some 6-mo follow-up data Patients reporting major

Five of the 25 patients All hydrocortisone 100 mg and and mg 100 hydrocortisone All Primary outcome was RR (95% CI) RR 9.00 (0.64-126.85) RR 1.88 (0.19-18.60) RR 2.55 (0.26-24.56) Total Mortality (%) Total Heparin: 0/12 Seven patients who failed to Heparin: 2/16 (12.5%) Heparin: 2/11 (18.2%) RR (95% CI) RR 0.92 (0.06-12.95) RR 1.25 (0.33-4.68) RR 0.64 (0.14-2.86) Major Bleeding (%) Heparin: 0/4 (0%) Heparin: 4/4 (100%) Heparin: 4/16 (25.0%) Heparin: 2/11 (18.2%) RR 9.00 (0.64-126.85) RR 2.82 (0.12-64.39) RR 2.55 (0.26-24.56) Recurrent DVT and PE (%) RR (95% CI) Heparin: 0/12 Heparin: 1/12 (8.3%) heparin vs 1 SK Length of Follow-up Analyzed No. Patients (70.6%) Heparin: 4/4 Heparin: 4/4 (100%) Heparin: 12/17 Heparin: 16/16 Heparin: 1/16 (6.3%) Heparin: 11/11 Heparin: 2/11 (18.2%) SK: 4/4 In hospital SK: 0/4 (0%) SK: 0/4 (0%) SK: 0/4 (0%) Primary outcome not stated. SK: 11/13 (84.6%) 72 h SK: 0/11 SL: 1/11 (9.1%) SK: 0/11 SK: 15/15 In hospital SK: 0/15 SK: 3/15 (20.0%) SK: 1/15 (6.7%) All: warfarin/VKA. SK: 14/14 10 d SK: 1/14 (7.1%) SK: 4/14 (28.6%) SK: 1/14 (7.1%) Primary outcome was angio-

1 Interventions followed by 1,000 units/h intrapulmonary followed by 2,500 units for 72 h for 2-7 d followed by 1,250 units/h for 7 d 1 h followed by a bolus of heparin 10,000 units constant infusion of 1,000 units/h intrapulmonary followed by 100,000 units for 72 h 100,000 units/h for 18-72 h by 100,000 units/h for 72 h [Section 5.6.1] Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With Acute PE: Clinical Description and Results Acute PE: Clinical Description and Results [Section 5.6.1] Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With SK 1,500,000 units over /1974 SK 600,000 units /1979 SK 250,000 units followed by 148 Table S41 — S41 Table 143 /1978 SK 250,000 units followed /1995 146 145 et al Ly et al Ly

Dotter et al Jerjes-Sanchez Heparin 10,000 units Heparin 5,000 units Heparin 1,500 units per kg Heparin 15,000 units Author/Year Tibbutt et al Tibbutt

65 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

1 (Continued) Comments Comments receive heparin. continued for 1 y. of 5 d. includes moderate severe bleeding. available up to 12 mo. or need for escalation of therapy (later decision could be made after unblinding). scan perfusion. angiographic reperfusion. echocardiographic right ventricular function.

Thrombolysis arms did not All: heparin for a minimum The major bleeding reported Angiographic follow-up data Primary outcome was lung Primary outcome was

placebo: 1 RR (95% CI) 3/138 (2.2%) RR 1.56 (0.36-6.83) 0/10 (0.21-78.76) RR 1.23 (0.43-3.49) 0/10 (10.0%) RR 0.24 (0.01-4.84) Total Mortality (%) Total Heparin Low-dose urokinase: Heparin: 0/16 RR 4.05 Urokinase: 6/82 (7.3%) Heparin: 7/78 (8.9%) Alteplase: 4/118 (3.4%) Primary outcome was death High-dose urokinase: alteplase: 2/20 rt-PA Heparin: 2/55 (3.6%) placebo: 1 RR (95% CI) 5/138 (3.6%) RR 0.23 (0.03-1.97) 0/10 RR 1.20 (0.23-6.34) (45.1%) RR 0.60 (0.39-0.92) (0.8%) 0/10 (15.0%) (1.8%) RR 3.59 (0.39-33.33) Major Bleeding (%) Heparin Low-dose urokinase: Heparin: 2/16 (12.5%) Urokinase: 37/82 Heparin: 21/78 (26.9%) High-dose urokinase: alteplase: 3/20 rt-PA Heparin: 1/55 placebo: 1 Continued heparin vs 1 4/138 (2.9%) RR 1.17 (0.30-4.57) 0/10 (0.11-55.89) (14.6%) RR 1.31 (0.66-2.63) 0/10 (5.0%) RR 0.11 (0.01-1.91) Recurrent DVT and PE (%) RR (95% CI) Heparin Low-dose urokinase: Heparin: 15/78 (19.2%) rt-PA: 0/46rt-PA: 3/46 (6.5%) rt-PA: 0/46 rt-PA: Primary outcome was Urokinase vs heparin Table S41— Table rt-PA (alteplase) rt-PA 14-21 d Length of Follow-up 2 wk Urokinase: 12/82 7 d High-dose urokinase: 30 d alteplase: 1/20 rt-PA placebo: 1 Analyzed No. Patients 138/138 urokinase: 10/10 urokinase: 10/10 20/20 Heparin Heparin: 10/10 Heparin: 0/10 Heparin: 0/10 Heparin: 0/10 All patients: OACs Low-dose Heparin: 16/16 Heparin: 0/16 RR 2.43 Urokinase 82/82 Heparin: 78/78 Alteplase: 118/118 30 d Alteplase: 4/118 (3.4%) Alteplase: 1/118 High-dose alteplase: rt-PA 46/46rt-PA: In hospital Heparin: 55/55 Heparin: 5/55 (9.1%) placebo 1 Interventions heparin 1,000 1 7 d followed by OAC by 1,000 units/h units over 12 h daily for 3 d by 1,750 units/h for 7-10 d CTA units/lb followed by CTA units/lb per h 2,000 CTA 75 units/lb followed by 10 units/lb per h by alteplase 90 mg over 2 h units/h 3,300,000 units over 12 h 90 mg over 2 h 2 h followed by heparin 1,000 units/h by 1,000 units/h Urokinase: infusion of 2,000 Heparin: infusion of Alteplase 100 mg followed /1992 10 mg followed by rt-PA /1993 alteplase 100 mg over rt-PA 142 158 /1988 High dose: urokinase 147 /1970 /2002 141 159 et al et al et al Dalla-Volta Goldhaber et al Heparin 30,000 units/d for Marini et al Heparin 5,000 units followed Low dose: urokinase 800,000 Heparin 10,000 units followed Konstantinides Author/Year

Heparin 5,000 units followed UPET Study Group

66 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Comments Comments by attending physician in both groups after treatment dysfunction. Primary outcome was echocardiographic changes. fatal. One additional fatal and nonfatal PE in heparin arm by 180 d. bleeds. scan reperfusion (serial angiographic and lung scans were assessed). dysfunction. Primary outcome was echocardiographic changes. bleeding; one intracranial bleed (tenectaplase). One death occurred 19 d No fatal or intracranial

Urokinase Pulmonary Embolism Trial. See Table Table See Trial. Embolism Pulmonary Urokinase RR (95% CI) 5 (0.10-54.06) Total Mortality (%) Total rt-PA: 1/33 (3.0%)rt-PA: Primary outcome was lung Placebo: 0/25 RR 2.29 Placebo: 0/4 Heparin doses determined RR (95% CI) (0.07-30.59) Major Bleeding (%) rt-PA: 0/33 rt-PA: Placebo: 0/25 Continued heparin vs Recurrent DVT and PE (%) RR (95% CI) 1 Alteplase: 0/37 Alteplase: 2/37 (5.4%) Alteplase: 0/37 All had right ventricular rt-PA: 0/33 rt-PA: Placebo: 0/25 Table S41— Table Tenectaplase Tenectaplase Prospective Investigation of Pulmonary Embolism Diagnosis; UPET UPET Diagnosis; Embolism Pulmonary of Investigation Prospective 5 Length of Follow-up Placebo: 35/35 Placebo: 3/35 (8.5%) Placebo: 1/35 (2.9%) Placebo: 5/35 (14.2%) Three recurrent PE were heparin 1 Analyzed No. Patients 5,000 units followed by 1,000 units/h Placebo Placebo: 25/25 Placebo: 4/4 Placebo: 0/4 Placebo: 0/4 RR 1.50 28/28Tenectaplase: 30 d 2/28 Tenectaplase: Placebo: 32/32 2/28 Tenectaplase: 0/28 Tenectaplase: All had right ventricular Placebo: 1/32 Placebo: 1/32 Placebo: 1/32 No fatal PE or major oral anticoagulant; PIOPED PIOPED anticoagulant; oral 5 2 mg/kg ف heparin heparin Interventions 1 1 5,000 units followed by 30,000/d (doses determined by physician) bolus (80 International Units/kg and 18 International Units/kg per h) Placebo plus heparin rt-PA 40-80 mg at 1 mg/minrt-PA 9/9 rt-PA: Placebo 7 d 0/9 rt-PA: 1/9 (11.1%) rt-PA: 0/9 rt-PA: Primary outcome not stated Placebo /1990 /2010 Tenectaplase: 144 /2011 Alteplase 100 mg over 2 h Alteplase: 37/37 10 d /1990 0.6mg/kg over 2 min rt-PA 33/33 rt-PA: 10 d 153 154 160 Committee on Thrombolytic Agents; OAC OAC Agents; Thrombolytic on Committee 5 Investigators Fasullo et al Becattini et al PIOPED Author/Year Levine et al S1, S2, and S11 legends for expansion of other abbreviations. S1, S2, and S11 legends for expansion of other abbreviations. CTA

67 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Comments and at 60 h of treatment, warfarin initial dose 25 mg for 6 mo. complete the treatment regimen were excluded from the analysis. patients, and uncertain if deaths occurred in those who were randomized or not randomized.

ITT As treatedIncluded 5 nonrandomized As treated heparin: 0/4 (0%) Lost to Follow-up Analysis 1 Heparin: 0/78 (0%) Heparin: 0/4 (0%) SK: 0/11 (0%) Per protocolHeparin: 0/12 (0%) All hydrocortisone 100 mg SK: 0/14 (0%) Seven patients who failed to Heparin: 0/11 (0%) SK plus heparin: 0/15 (0%) ITT SK Urokinase: 0/82 (0%) Caregivers: PN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: PN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: PN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: PN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: CN Data Collectors: CY Adjudicators: CY Data Analysts: CY heparin vs 1 Urokinase vs heparin SK Concealed Blinding Randomize RCT PY Patients: PY RCT CY Patients: PY RCT PY Patients: PY RCT CY Patients: PY RCT CY Patients: CY Study Design Intervention constant infusion of [Section 5.6.1] Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With Acute PE: Methodologic Quality Acute PE: Methodologic Quality [Section 5.6.1] Systemic Thrombolytic Therapy vs Anticoagulation Alone in Patients With followed by 100,000 units for 72 h 100,000 units/h for 72 h 100,000 units/h for 18-72 h by a bolus of heparin 10,000 units 1 1,000 units/h by 1,000 units/h units/lb followed by CTA units/lb per h Heparin: 2,000 CTA infusion of 75 units/lb followed by 10 units/lb per h followed by 2,500 units for 72 h 1250 U/h for 7days SK 250,000 units followed by Urokinase: infusion of 2,000 Heparin 15,000 U followed by /1995 SK 1,500,000 units over 1 h followed 145 Table S42 — S42 Table /1974 SK 600,000 units intrapulmonary /1979 SK 250,000 units followed by /1970 148 143 141,149 /1978 146 Group Tibbutt et al Tibbutt UPET Study Author/Year Dotter et al Jerjes-Sanchez et al Heparin 1,500 units per kg for 2-7 d Heparin: 0/16 (0%) Heparin 10,000 units followed al Lyet Heparin 5,000 units intrapulmonary

68 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Comments 5,000 units followed by continuous infusion at starting dose of 30,000 rst 24 h. for the ﬁ

ITT All: UFH initial bolus of ITT Lost to Follow-up Analysis 0/10 (0%) 0/10 (0%) Low-dose urokinase: Heparin: 0/10 (0%) Heparin: 0/16 (0%)Heparin: 0/55 (0%) 0/33(0%) rt-PA: Heparin: 0/25(0%) UFH: 0/138 (0%) High-dose urokinase: 0/20 (0%)rt-PA: ITT 0/46 (0%)rt-PA: ITT rt-PA: 0/118 (0%)rt-PA: ITT All: UFH 5,000 units. Caregivers: CY Data Collectors: CY Adjudicators: CY Data Analysts: CY Caregivers: CN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: CN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: CN Data Collectors: PN Adjudicators: PN Data Analysts: PN Caregivers: CN Data Collectors: PY Adjudicators: PN Data Analysts: PY heparin vs —Continued 1 Table S42 Table Randomize Concealed Blinding rt-PA (alteplase) rt-PA RCT PY Patients: CY RCT PY Patients: PN RCT PY Patients: PN RCT CY Patients: PN RCT CY Patients: PY Study Design Intervention heparin 5,000 units 1 heparin 1,000 units/h 1 over 12 h over 12 h daily for 3 d followed by OAC by 90 mg over 2 h by 1,750 units/h for 7-10 d by heparin 1,000 units/h units/h followed by 30,000/d by alteplase 90 mg over 2 h by 1,000 units/h Placebo /2002 (alteplase) 100 mg followed rt-PA 159 /1992 (alteplase) 10 mg followed rt-PA /1993 100 mg over 2 h followed rt-PA 142 158 /1990 0.6 mg/kg over 2 min rt-PA /1988 High dose: urokinase 3,300,000 units 160 147 Goldhaber et al Konstantinides et al Dalla-Volta et al Dalla-Volta Levine et al Marini et al Low dose: urokinase 800,000 units Heparin 30,000 units/d for 7 d Heparin 10,000 units followed Heparin 5,000 units followed by 1,000 Heparin 5,000 units followed Author/Year

69 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Comments by attending physician. echocardiographic changes. echocardiographic changes.

ITT Primary outcome was Lost to Follow-up Analysis rt-PA: 0/9(0%)rt-PA: Heparin: 0/4(0%) ITT 0/37 rt-PA: All: heparin doses determined Heparin: 0/35 Tenectaplase: 0/28(0%)Tenectaplase: ITTHeparin: 0/30(0%) Primary outcome was Caregivers: CY Data Collectors: PY Adjudicators: PY Data Analysts: PY Caregivers: CY Data Collectors: PY Adjudicators: CY Data Analysts: PY Caregivers: CY Data Collectors: CY Adjudicators: CY Data Analysts: PY heparin vs —Continued 1 Table S42 Table Randomize Concealed Blinding rt-PA (alteplase) rt-PA RCT CY Patients: CY RCT CY Patients: CY Study Design 2 mg/kg bolus ف Intervention heparin 5,000 units heparin (80 International 1 1 determined by physician) followed by 1,000 units/h Units/kg and 18 International Units/kg per h) rt-PA 40-80mg at 1 mg/minrt-PA RCT PY Patients: CY Placebo Placebo /1990 /2010 Tenectaplase: 144 /2011 Alteplase 100 mg over 2 h 153 154 Investigators Author/Year Becattini et al PIOPED

Fasullo et al Placebo plus heparin (doses S1, S2, S5, S11, and S41 legends for expansion of other abbreviations. See Table

70 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (95% CI) (95% CI) (from 3 fewer to 83 more) 1,000 (from 16 fewer to 31 fewer) 1,000 (from 6 fewer to 16 fewer) (from 1 fewer to 10 more) Risk Difference With Fondaparinux Fondaparinux With 27 fewer per 13 fewer per h h 4 more per 1,000 h Risk With No With Risk Fondaparinux 4 per 1,000 33 per 1,000 19 per 1,000 1 per 1,000 0 fewer per 1,000 cial vein thrombosis. S1 See Table

e superﬁ 5 (95% CI)

(0.18-21.87) (0.06-0.53) (0.14-0.68) (0.06-15.86) a-c Relative Effect Relative Summary of Findings With With Fondaparinux te was 98%. ITT cacy analysis outcomes. for Not effi stopped early for Study Event Rates (%) Anticipated Absolute Effects With No With 1/1,500 (0.1) 2/1,502 (0.1) RR 1.99 1/1,488 (0.1) 1/1,499 (0.1) RR 0.99 Fondaparinux 22/1,500 (1.5) 4/1,502 (0.3) RR 0.18 26/1,500 (1.7) 8/1,502 (0.5) RR 0.31 due due d-g d,e,i due to d d to imprecision imprecision to imprecision of Evidence Overall Quality VTE (critical outcome) Bias Quality of life not measured Mortality (important outcome) Publication Undetected High Major bleeding (critical outcome) SVT recurrence (important outcome) Undetected Moderate Undetected High Undetected Moderate e,f f,g e,i imprecision Serious No serious Serious Serious Comparison of ARIXTRA™ in lower LImb Superfi cial Thrombophlebitis with Comparison placebo; of SVT ARIXTRA™ in lower LImb Superfi [Section 8.1] Evidence Profi le: Fondaparinux vs Placebo for Acute SVT le: Fondaparinux vs Placebo for Acute SVT [Section 8.1] Evidence Profi 5

162 indirectness indirectness indirectness indirectness CALISTO CALISTO No serious No serious No serious No serious 161 Table S43 — S43 Table Quality Assessment inconsistency inconsistency inconsistency inconsistency No serious No serious No serious No serious

d d d d risk of bias risk of bias risk of bias risk of bias No serious No serious No serious No serious t. t. 3 mo 3 mo 47 d 3 mo Patients in the two treatment groups benefi ted from close clinical monitoring with adequate diagnostic procedures in the event of new and persistent symptoms. ted from close clinical monitoring with adequate diagnostic procedures in the event of new and persistent symptoms. Patients in the two treatment groups benefi Allocation concealed. Outcome adjudicators, steering committee, patients, providers, and data collectors blinded. Follow-up ra Baseline risk derived from a large prospective cohort study. CI includes values suggesting large benefi t and values suggesting large harm. t and values suggesting large harm. CI includes values suggesting large benefi Small number of events. Fondaparinux 2.5 mg for 45 d. Fondaparinux 2.5 mg for 45 d. Patients with infusion-related SVT were excluded if from CALISTO. We rated down by only one level because the low event rate and large sample size. rated down by only one level because the low event rate and large sample size. We The upper limit of the CI for absolute effect (10 more bleeds) is not low enough to suggest a clear balance of benefi ts vs harms. ts vs harms. The upper limit of the CI for absolute effect (10 more bleeds) is not low enough to suggest a clear balance benefi … Bibliography: CALISTO study by Decousus et al. … … … … … … … … … … 3,002 (1 study), 3,002 (1 study), 2,987 (1 study), benefi

a b c d e f g h i

3,002 (1 study), Participants (Studies), Follow-up Risk of Bias Inconsistency Indirectness Imprecision and S5 legends for expansion of other abbreviations. and S5 legends for expansion of other abbreviations.

71 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

Results ns for all comparisons of active ns for all comparisons of active RR 0.25 (95% CI, 0.03-2.24) (PE 0); RR 0.26 (95% CI, 0.03-2.33) RR 0.57 (95% CI, 0.11-3.02) treatment vs placebo Placebo: 33/112 (29.5%) RR 0.28 (95% CI, 0.14-0.55) RR 0.19 (95% CI, 0.08-0.44) (95% CI, 0.25-0.80) Placebo: 5/112 (4.5%) (PE 0) RR 1.22 (95% CI, 0.38-3.89) (PE 0); RR 0.85 (95% CI, 0.23-3.06) RR 0.91 (95% CI, 0.25-3.28) treatment vs placebo 37/112 (33.0%) RR 0.44 (95% CI, 0.26-0.74) 5 5

Enoxaparin 40 mg: 1/110 (0.9%) (PE 0); Enoxaparin 1.5 mg/kg: 1/106 (0.9%) 2/99 (2.0%) (PE 1); Tenoxicam: P SVT recurrence/extension day 12: Enoxaparin 40 mg: 9/110 (8.3%); Enoxaparin 1.5 mg/kg: 6/106 (5.7%); 13/99 (13.1%); RR 0.45 Tenoxicam: VTE 3 mo: Enoxaparin 40 mg: 6/110 (5.7%); (PE 2); Enoxaparin 1.5mg/kg: 4/106 (3.9%) 4/99 (4.3%) (PE 1); Tenoxicam: P SVT recurrence/extension 3 mo: Placebo: Enoxaparin 40 mg: 16/110 (14.5%); 3 mo VTE day 12: Placebo: 4/112 (3.6%) (PE 0) Follow-up c Screening ultrasound or symptomatic recurrence: VTE SVT recurrence/extension to saphenofemoral junction extension Day 12 (end of treatment): 3-mo VTE SVT recurrence/ Major bleeding Death Outcomes b Intervention bandages or compression stockings for at least 15 d Enoxaparin 40 mg SC daily Enoxaparin 1.5 mg/kg SC daily 20 mg po daily Tenoxicam, Placebo once daily All given for 8-12 d All patients prescribed elastic rmed 5 cm length) Ն ultrasound-confi acute symptomatic SVT ( of the lower extremity 436 patients with Participants a Type of Type Publication [Section 8.1] Superfi cial Vein Thrombosis Treatment: Clinical Description and Results (Randomized Trials Comparing Treatments) Comparing Treatments) Clinical Description and Results (Randomized Trials Thrombosis Treatment: cial Vein [Section 8.1] Superfi multicenter Parallel RCT, Parallel RCT,

163 Table S44 — S44 Table Group/2003 STENOX Study Author/Year

72 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

Results ns .01 in favor of nadroparin vs naproxen; RR 0.46 (95% CI, 0.27-0.77) (95% CI, 0.27-0.78) nadroparin: 1/38 (2.6%) RR 1.90 (95% CI, 0.18-20.1) (95% CI, 0.06-15.02) Fixed-dose nadroparin: 2/36 (5.6%) RR 0.18 (95% CI, 0.01-3.64) 0.19 (0.01-3.73) improvement in score from day 0 to 7: Fixed-dose nadroparin: 79.1% improved this difference was maintained at 8 wk 5 Յ

8-wk extension of thrombus or new SVT: 8-wk extension of thrombus or new SVT: Enoxaparin 1.5 mg/kg: 16/106 (15.1%); 15/99 (15.2%); RR 0.46 Tenoxicam: Major bleeding: 0 Death: Weight-based nadroparin: 2/40 (5%); Weight-based P nadroparin: 0/40 (0%); Weight-based Naproxen: 0/39 (0%); RR (95% CI) PE, or major bleeding in any group No DVT, Intensity of symptoms/signs: Overall nadroparin: 63.0% improved Weight-based Naproxen: 46.4% improved P 3 mo 3 mo follow-up: 8 wk Day 7 extension of thrombus: Fixed-dose Follow-up c ushing, edema, asymptomatic or symptomatic SVT extension, asymptomatic or symptomatic DVT, symptomatic PE at 3 mo signs (warmth, fl pain on palpation) thrombus at day 7 and 8 wk Composite outcome of Changes in symptoms and clinical Major bleeding Echocardiographic extension of Continued Outcomes b Table S44— Table Intervention nadroparin (190 anti-Xa International Units/kg for 10 d followed by 95 anti-Xa International Units/kg for 20 d) International Units/kg SC daily stockings for 7 d anti-Xa International Units daily High-dose weight-adjusted All patients wore compression Nadroparin fi xed dose, 6,150 Nadroparin fi rmed rmed ultrasound-confi acute SVT of the greater saphenous vein ultrasound-confi SVT of the lower extremities 164 patients with 117 patients with Participants a Type of Type Publication multicenter multicenter Parallel RCT, Parallel RCT, /1994 Parallel RCT, /2005 164 165 Investigators Group et al Titon Author/Year Prandoni for Vesalio Nadroparin 31.5 anti-Xa Naproxen 500 mg po daily given for 6 d Treatments DVT PE Naproxen: 1/39 (2.6%); RR 0.97

73 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians ns) (Continued) 5

P ns) ns) ns)

5 5

P P P Results Low dose: 6/30 (20%) High dose: 1/30 (3.3%); RR 0.17 (95% CI, 0.02-1.30; follow-up period: Low dose: 11/30 (36.7%) High dose: 8/30 (26.7%); RR 0.73 (95% CI, 0.34, 1.55; (95% CI, 0.01-1.98; treatment period: Low dose: 7/30 (23.3%) High dose: 3/30 (10%); RR 0.40 (95% CI, 0.11-1.40; while on treatment) Low dose: 5/81 (6.2%) 5/81 dose: Low treatment) on while (all occurred while on treatment) RR 2.56 (95% CI, 0.51-12.83) (3 symptomatic events; 1 [PE] occurred while on treatment) Low dose: 2/81 (2.5%) (both symptomatic DVT) RR 0.51 (95% CI, 0.10-2.72) symptoms and signs similar both groups Major bleeding: 0 Death: 0 4/30 (13.3%) (3 asymptomatic DVT, 1 PE) 4/30 (13.3%) (3 asymptomatic DVT, SVT High dose: 2/83 (2.4%) (1 occurred Extension/recurrence SVT during

Overall VTE during follow-up period: Overall extension/recurrence SVT during VTE High dose: 4/83 (4.8%) Rate of improvement in clinical Follow-up c symptoms and signs at 1 mo of thrombosis Major bleedingHIT Death High dose: 0/30 (0%); RR 0.11 Major bleeding Death Improvement in clinical VTEExtension/Recurrence 6 mo VTE during treatment period: Low dose: Continued Outcomes b ammatory Table S44— Table nadroparin Intervention International Units for 1 wk then 10,000 International Units for 3 wk) or local antiinfl drugs permitted but use not described (2,850 anti-Xa International Units for 30 d) and aspirin use discouraged International Units bid SC for 4 wk) Use of concomitant systemic No placebo group NSAIDS Low-dose UFH (5,000 rmed ultrasound-confi rst acute SVT of fi greater saphenous vein 60 patients with Participants a Type of Type Publication single center /2002 Parallel RCT, 166

High-dose UFH (12,500 Marchiori et al Low-dose Author/Year

74 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

Results .05 for groups C, D, E, F vs A or B ns B: 6/78 (7.7%); RR 0.46 (95% CI, 0.18-1.15) C: 1/70 (1.4%); RR 0.09 (95% CI, 0.01-0.64) D: 2/71 (2.8%); RR 0.17 (95% CI, 0.04-0.72) E: 1/76 (1.3%); RR 0.08 (95% CI, 0.01-0.59) F: 5/71 (7%); RR 0.42 (95% CI, 0.16-1.13) any group B: 2/78 (2.5%); RR 0.33 (95% CI, 0.07-1.60) C: 2/70 (2.8%); RR 0.37 (95% CI, 0.08-1.78) D: 0/71 (0%); RR 0.08 (95% CI, 0.0-1.47) E: 0/76 (0%); RR 0.08 (95% CI, 0.0-1.38) F: 0/71 (0%); RR 0.08 (95% CI, 0.0-1.47) A: 32/78 (41%) B: 11/78 (14.1%); RR 0.34 (95% CI, 0.19-0.63) C: 0/70; RR 0.02 (95% CI, 0.00-0.27) D: 4/71 (5.6%); RR 0.14 (95% CI, 0.05-0.37) E: 4/76 (5.2%); RR 0.13 (95% CI, 0.05-0.35) F: 5/71 (7.0%); RR 0.17 (95% CI, 0.07-0.42) Enoxaparin group: 3/30 (10%); RR 3.0 (95% CI, 0.33-27.24) , 5

not stated P Extension at 6 mo: A: 13/78 (16.7%) No major bleeding, HIT, or death in No major bleeding, HIT, P New DVT at 3 mo: A: 6/78 (7.7%) P Follow-up c New DVT at 3 mo Extension of SVT at 3 moExtension of SVT at 6 mo 6 mo Extension of thrombus at 3 mo: Recurrence/extension of SVT 6 moVTE Complications of surgery Recurrent SVT Surgical group: 1/30 (3.3%) Continued ed Outcomes b Table S44— Table ush ligation Intervention and simple fl and complete stripping and perforator ligation and low-dose SC heparin and LMWH and VKA anticoagulants not specifi alone under local anesthesia with short-term use of a compression bandage B. Elastic compression stockings D. Elastic compression stockings E. Elastic compression stockings Elastic compression stockings F. Doses and duration of A. Elastic compression stockings Saphenofemoral disconnection rmed rmed ultrasound-confi SVT and large varicose veins or venous incompetence ultrasound-confi above-knee internal saphenous SVT 562 patients with 60 patients with Participants a Type of Type Publication multicenter single center /1999 Parallel RCT, /2003 Parallel RCT, 167 168

C. Elastic compression stockings Belcaro et al Author/Year Lozano et al

75 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

Results symptomatic PE) Enoxaparin group: 0/30 (0%); RR 0.20 (95% CI, 0.01-4.0) 2/30 (6.7%) patient; mean, 1.6 d in hospital 300/patient; Enoxaparin group: $ 0 d in hospital 18/148 (12%) Medical group: 10/71 (14%); RR 1.16 (95% CI, 0.56-2.38) group: 2/88 (2.2%) RR 0.66 (95% CI, 0.14-3.13) group: 0/17 (0%); RR 1.10 (95% CI, 0.06-21.98) (hematoma, seroma, infection) Wound infections: Surgical group: Wound VTE Surgical group: 2/30 (6.7%) (both Major bleeding: 0 Death: 0 1,400/ Cost of treatment: Surgical group: $ SVT progression: Surgical group: Surgical group: 7/204 (3.4%) Medical DVT: PE: Surgical group: 2/98 (2.0%) Medical Surgical complications: 6/78 (7.7%) Bleeding complications: 0/17 (0%) group: 6 d to 14 mo group: 4-6 mo Surgical Medical Follow-up c Surgical complications Bleeding complications Major bleeding Death Costs SVT progression DVT PE Continued Outcomes b 246) 88) 5 5 Table S44— Table Intervention followed by VKA for 6 wks-6 mo) (n bid for 1 wk then once daily for 3 wk wear elastic compression stockings and used acetaminophen for pain vein at saphenofemoral junction with or without vein stripping (n Anticoagulation (IV heparin Outpatient enoxaparin 1 mg/kg No placebo/control group All patients were instructed to Ligation of greater saphenous rmed above-knee conﬁ SVT Patients with objectively Participants a Type of Type Publication of 6 studies (includes Belcaro (14 patients) and 5 small case series) /2001 Systematic review 169

Sullivan et al

Author/Year

76 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians gures in gures cant (Continued) cant difference

Results 75%) reported good or very . article by reviewer 47% Group B: 83% C: 79% in article by reviewer heparin: 0 LMWH: patients ( no signifi cacy, good treatment effi difference between groups 10/21 (47.6%) LMWH: 9/23 (39.1%); RR 0.82 (95% CI, 0.42-1.62) 1/23 (4.3%); RR 0.30 (95% CI, 0.03-2.70) in enoxaparin group no pain at 21 d, signifi between groups each time point; no erythema at 21 d, cant difference between groups no signifi not stated

P Pain by visual analog scale, day 21: Topical Pain by visual analog scale, day 21: Topical cacy assessment: Majority of Subjective effi heparin: Thrombus regression: Topical heparin: 3/21 (14.3%) LMWH: Topical DVT: Adverse events: Allergic reaction in 1 patient Death: 0 Improvement noted at each time point; Follow-up c cacy assessment by investigator and patient regression day 21 PainIncrease in functional abilityLocal edema 30 d Data extrapolated from graphs and fi Resolution of pain, day 30: Group A: Pain by visual analog scale (0-10) 21 dArea of erythema Data extrapolated from graphs and fi DVTAdverse events Death Area of erythema: Improvement noted at Continued Outcomes b Table S44— Table Intervention SC once daily SC bid intramuscular once daily spray gel (4 sprays of 458 International Units tid) A: Dermatan sulfate 100 mg Topical liposomal heparin Topical rmed of the lower limbs ultrasound-confi SVT 56 patients with SVT 46 patients with Participants a Type of Type Publication multicenter multicenter Parallel RCT, Parallel RCT, /2005 Parallel RCT, 170 /1996 171 et al Górski et al Author/Year B: Dermatan sulfate 100 mg C: Dermatan sulfate 200 mg Andreozzi Enoxaparin 40 mg SC once daily Subjective effi given for 7-14 d Treatment Duplex assessment for thrombus

77 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians cant ammatory drug;

Results nonsteroidal antiinﬂ 5 .05 groups B and C vs group A activities, day 30: Group A: 44% B: 67% Group C: 84% improvement in all 3 groups; no signiﬁ differences between groups ,

Increase in ability to perform normal daily P Local edema, day 30: Progressive Follow-up c heparin-induced thrombocytopenia; NSAID heparin-induced thrombocytopenia; NSAID 5 Continued Outcomes b Table S44— Table Intervention Treatment given for 30 d Treatment Participants a Type of Type Publication cial Thrombophlebitis Treated by Enoxaparin. See Table S1, S2, S5, S21, and S43 legends for expansion of other abbreviations. S1, S2, S5, S21, and S43 legends for expansion of other abbreviations. by Enoxaparin. See Table cial Thrombophlebitis Treated Superﬁ 5 Drugs: VKA, UFH, LMWH, NSAIDs, aspirin, topical treatments, surgery vs placebo, no treatment, each other or different durations or regimens of the same agent. or regimens of the same agent. Drugs: VKA, UFH, LMWH, NSAIDs, aspirin, topical treatments, surgery vs placebo, no treatment, each other or different durations Study design: RCT, cohort. cohort. Study design: RCT, Outcomes: extension of thrombus, symptomatic relief, DVT and PE, major bleeding, surgical complications, death. The CALISTO study that compared S42. fondaparinux HIT with no 28 fondaparinux and is Table described in Table

STENOX STENOX Author/Year a b c

78 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S45 —[Section 8.1] SVT Treatment: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up/ ITT STENOX Study Central randomization Visually identical drugs Investigators, patients, and 9 lost to follow-up/ ITT Group163 /2003 and packaging; triple assessors blinded dummy design Titon et al164 /1994 Randomized to one of Open label Not blinded 8 lost to follow-up/ not three treatment groups; specifi ed method of randomization not specifi ed Prandoni for Vesalio group Computer-generated random Double dummy Patients and adjudicators of 0 lost to follow-up/ ITT et al165 /2005 number sequence assigned outcome events blind to each patient to determine treatment group Marchiori et al166 /2002 Randomized to treatment Not specifi ed Assessors blinded 0 lost to follow-up/ not group by computer- specifi ed generated list Belcaro et al167 /1999 Not specifi ed Not blinded Not blinded 118 lost to follow-up/not specifi ed Lozano et al168 /2003 Method not specifi ed Not blinded Not blinded 3 lost to follow-up/not specifi ed Sullivan et al169 /2001 Review of six studies; N/A N/A N/A includes one RCT (Belcaro [14 patients]) and fi ve small case series Andreozzi et al170 /1996 Patients randomly assigned Open label Not blinded Not specifi ed/not to one of three therapeutic specifi ed groups (method not specifi ed) Górski et al171 /2005 Performed according to Open trial Not blinded 6 lost to follow-up/ITT a prespecifi ed randomization list; treatment allocated according to next number on list; no stratifi cation was performed See Table S5, S12, S43, and S44 legends for expansion of abbreviations.

79 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued) Results Results 35%: 16/36 (45%) Ն 35%: 17/36 (47%) None: 3/36 (8%) bleeding) 32/36 (89%) , or bleeding) treatment with UFH vs LMWH (2 arm, 1 leg) 2.0%, 4.2%, 7.7% PE, congestive heart failure [numbers not provided]) PE: 0/46 Major bleeding: 1/46 (2%) (on VKA) Death: 7/46 (15%) (none from PE or day 10: Lysis, 0/36 Recurrent DVT: PE: 0/36 Death: 9/36 (25%) (none due to PE Recurrent VTE: 3/53 (5.7%) Cumulative incidence 1, 2, and 5 y: Death: 11/53 (20.8%) (due to cancer, 1 y cant symptom relief, day 7: Signiﬁ 3 mo 1/46 (2%) Recurrence/extension DVT: Median of 48 mo Results not presented according to initial Follow-up c recurrence/extension recurrence/extension of DVT ultrasound Symptomatic PE Major bleeding Death Symptom relief of thrombus on Lysis Recurrent DVT PE Death Recurrent VTE Death Outcomes b Intervention (200 International Units/kg) and VKA with target INR of 2.0-3.0 International Units/kg for up to 7 d, then VKA (started on day 3; target INR 2-2.5) for mean of 4.7 mo (81% received UFH, 19% received LMWH) then VKA (median, 3 mo) Dalteparin daily for 5-7 d Duration of VKA not provided Nadroparin SC bid, 86 anti-Xa Therapeutic-dose heparin rst UEDVT and Prospective Observational Studies of at Least 20 Patients) and Prospective Observational Studies of at Least 20 Patients) (includes 16 with CVC) (includes 13 with CVC) (included 6 with CVC) 46 outpatients with UEDVT 36 inpatients with UEDVT 53 patients with ﬁ Participants a two center single center number of centers not stated Type of Publication Type Prospective cohort, Prospective cohort, Prospective cohort, [Section 9.1] Initial Treatment of Acute UEDVT With Anticoagulant Therapy: Clinical Description and Results (Randomized Trials [None Performed] Anticoagulant Therapy: Clinical Description and Results (Randomized Trials of Acute UEDVT With [Section 9.1] Initial Treatment /2004 /2003 120 /1999 173 172 Table S46 — S46 Table Karabay et al Prandoni et al Author/Year Savage et al

80 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Results Results to infuse: 0/74 to infuse: 0/74 PE: 0/74 Major bleeding: 3/74 (4%) 1 major bleed) Death: 7/74 (6 cancer, Catheter failure due to DVT or inability 3 mo Recurrent VTE: 0/74 Follow-up c to DVT or inability to infuse Recurrent VTE PE Major bleeding Death Catheter failure due Outcomes b Continued Intervention Table S46— Table (200 International Units/kg) and VKA to achieve target INR of 2.0-3.0 Dalteparin daily for 5-7 d conﬁ rmed UEDVT (all had conﬁ CVC) 74 cancer patients with Participants a upper-extremity DVT. See Table S2, S7, and S41 legends for expansion of other abbreviations. S2, S7, and S41 legends for expansion of other abbreviations. See Table DVT. upper-extremity 5 multicenter Type of Publication Type Prospective cohort, /2007 174 central venous catheter; UEDVT central venous catheter; UEDVT 5 Drugs: IV UFH or LMWH followed by OACs. Drugs: IV UFH or LMWH followed by OACs. Study design: prospective cohort studies. Study design: prospective cohort studies. and early symptom relief. Outcomes: recurrent DVT and PE, major bleeding, total mortality, CVC CVC

Author/Year a b c Kovacs et al

81 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S47 —[Section 9.1] Initial Treatment of Acute UEDVT With Anticoagulant Therapy: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up/ ITT

Savage et al172 /1999 N/A N/A N/A 1 lost to follow-up/N/A Karabay et al173 /2003 N/A N/A N/A 0 lost to follow-up/N/A Prandoni et al120 /2004 N/A N/A N/A 2 lost to follow-up/N/A Kovacs et al174 /2006 N/A N/A N/A 0 lost to follow-up/N/A See Table S12 and S46 legends for expansion of abbreviations.

Table S48 —[Section 9.2] Initial Treatment of Acute UEDVT With Thrombolytic Therapy: Clinical Description and Results (Randomized Trials [None Performed] and Prospective Observational Studies of at Least 10 Patients)

Author/Year Type of Publicationa Participants Interventionb Outcomesc Follow-up Results

Horne et al175 /2000 Prospective cohort, 18 patients with Catheter-directed Immediate patency 6 mo Immediate patency: single center axillary or rt-PA (2 mg/cm 10/18 (56%) subclavian DVT of thrombus to maximum of 20 mg) then VKA for 3 mo Establishment of Antegrade ﬂ ow: 11/18 antegrade ﬂ ow (61%) Bleeding events Bleeds (all minor): 5/18 (28%) Lee et al176 /2006 Prospective case 35 patients Oral VKA for mean Recurrent DVT 54 mo Ipsilateral recurrent series, single with primary of 5.2 mo DVT: 8/35 (23%) center UEDVT who had complete resolution of acute symptoms with CDT (n 5 29) or IV heparin (n 5 6) Early prospective observational studies with , 10 patients and retrospective studies are described in Table 3 of the eighth edition of these guidelines.46 See Table S2, S10, S11, and S46 legends for expansion of abbreviations. a Study design: retrospective and prospective cohort studies. b Drugs: thrombolytic therapy compared with different types of lytic therapy or with anticoagulants. c Outcomes: recurrent DVT and PE, vein patency, major bleeding, total mortality, and PTS of the arm.

Table S49 —[Section 9.2] Initial Treatment of Acute UEDVT With Thrombolytic Therapy: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up

Horne et al175 /2000 No N/A N/A Not speciﬁ ed Lee et al176 /2006 No N/A N/A 0 lost to follow-up See Table S12 legend for expansion of abbreviation.

82 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians (Continued)

35%: 16/36 (45%) Results Ն 35%: 17/36 (47%) (on VKA) or bleeding) 32/36 (89%) , None: 3/36 (8%) PE or bleeding) overall (all UEDVT) 2.4% (95% CI, 1.2%-4.0%) (results not provided by treatment group) to initial treatment with UFH vs LMWH PE: 0 Major bleeding: 1/46 (2%) Death: 7/46 (15%) (none from PE day 10: Lysis, 0 Recurrent DVT: PE: 0 Death: 9/36 (25%) (none due to PTS: 0 Annual incidence recurrent VTE: 3 mo Recurrence/extension: 1/46 (2%) 1 y cant symptom relief, day 7: Signifi Median of 5.1 y Recurrent VTE: 12/98 (12%) Follow-up b extension of DVT ultrasound anticoagulants stopped Symptomatic recurrence/ PE Major bleeding Death Symptom relief of thrombus on Lysis Recurrent DVT PE Death PTS Recurrent VTE after Recurrent VTE Median of 48 mo Results not presented according Outcomes a 3 mo Յ Intervention Units/kg daily for 5-7 d and VKA to achieve target INR of 2.0-3.0 for 3 mo not provided anti-Xa International Units/kg for up to 7 d, then VKA (started on day 3; target INR 2-2.5) for mean of 4.7 mo (77 patients), heparin SC (14 patients), or antiplatelet agents (7 patients) for (81% received UFH, 19% received LMWH) then VKA (median, 3 mo) Dalteparin 200 International Duration of VKA Nadroparin SC bid, 86 VKA for mean 6 mo Therapeutic-dose heparin rmed Participants rmed UEDVT rmed UEDVT [Section 9.3] Long-term Treatment of Acute UEDVT: Clinical Description and Results Clinical Description and Results of Acute UEDVT: [Section 9.3] Long-term Treatment confi (includes 16 with CVC) confi (includes 13 with CVC) primary UEDVT (none with CVC) rst UEDVT (included fi 6 with CVC) 46 outpatients with 36 inpatients with 98 patients with 53 patients with confi Table S50 — S50 Table Type of Publication Type two center single center prospective follow-up of cases, single center number of centers not stated Prospective cohort, Prospective cohort, Case-control study with /2004 /2004 Prospective cohort, /2003 177 178 /1999 173 172 Savage et al Author/Year

Karabay et al Martinelli et al Prandoni et al

83 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians

Results (2 arm, 1 leg) 2.0%, 4.2%, 7.7% PE, congestive heart failure [breakdown not provided]) inability to infuse: 0 inability to infuse: 0 Recurrent VTE: 3/53 (5.7%) Cumulative incidence 1, 2, and 5 y: cancer, to (due (20.8%) 11/53 Death: PTS: 13/53 (24.5%); 2 y Cumulative incidence: 27.3% PE: 0 Major bleeding: 3 (4%) 1 major bleed) Death: 7 (6 cancer, Catheter failure due to DVT or See Table S2, S7, S10, and S46 legends for expansion for legends S46 and S10, S7, S2, Table See 3 mo Recurrent VTE: 0 46 Follow-up b DVT or inability to infuse Death PTS Recurrent VTE PE Major bleeding Death Catheter failure due to Outcomes a Continued Intervention Units/kg daily for 5-7 d and VKA to achieve target INR of 2.0-3.0 Table S50— Table Dalteparin 200 International Participants rmed UEDVT conﬁ (all had CVC) 74 cancer patients with 20 patients, and retrospective studies, are described in Table 3 of the Eight edition of these guidelines these of edition Eight the of 3 Table in described are studies, retrospective and patients, 20 , Type of Publication Type multicenter Prospective cohort, /2007 174 Outcomes: recurrent DVT and PE, major bleeding, total mortality, and PTS of the arm. and PTS of the arm. Outcomes: recurrent DVT and PE, major bleeding, total mortality, Drugs: VKA, UFH, LMWH vs placebo, control or each other. Drugs: VKA, UFH, LMWH vs placebo, control or each other. Kovacs et al Author/Year a b

Early prospective observational studies with with studies observational prospective Early of abbreviations. of abbreviations.

84 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Table S51 —[Section 9.3] Long-term Treatment of Acute UEDVT: Methodologic Quality

Author/Year Randomization Allocation Concealment Blinding Loss to Follow-up

Savage et al172 /1999 N/A N/A N/A 1 lost to follow-up Karabay et al173 /2003 N/A N/A N/A 0 lost to follow-up Martinelli et al177 /2004 N/A N/A N/A Not speciﬁ ed Prandoni et al178 /2004 N/A N/A N/A 2 patients lost to follow-up Kovacs et al174 /2007 N/A N/A N/A 0 lost to follow-up See Table S12 and S46 legends for expansion of abbreviations.

85 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians References for proximal deep-vein thrombosis. N Engl J Med . 1 9 9 6 ; 1 . Brandjes DPM , Heijboer H , Büller HR , de Rijk M , Jagt H , 334 ( 11 ): 677 - 681 . ten Cate JW . Acenocoumarol and heparin compared with 15 . Lindmarker P , Holmström M , Granqvist S , Johnsson H , acenocoumarol alone in the initial treatment of proximal-vein Lockner D . Comparison of once-daily subcutaneous Frag- thrombosis . N Engl J Med . 1992 ; 327 ( 21 ): 1485 - 1489 . min with continuous intravenous unfractionated heparin in 2 . Gallus AS , Jackaman J , Tillett J , Mills W , Wycherley A . the treatment of deep vein thrombosis. Thromb Haemost . S a f e t y a n d e f fi cacy of warfarin started early after submas- 1994 ; 72 ( 2 ): 186 - 190 . sive venous thrombosis or pulmonary embolism. Lancet . 16 . Merli G , Spiro TE , Olsson CG , et al ; Enoxaparin Clinical 1986 ; 2 ( 8519 ): 1293 - 1296 . Trial Group . Subcutaneous enoxaparin once or twice daily 3 . Hull RD , Raskob GE , Rosenbloom D , et al . Heparin for compared with intravenous unfractionated heparin for treat- 5 days as compared with 10 days in the initial treatment of ment of venous thromboembolic disease . Ann Intern Med . proximal venous thrombosis. N Engl J Med . 1 9 9 0 ; 3 2 2 ( 1 8 ) : 2001 ; 134 ( 3 ): 191 - 202 . 1260 - 1264 . 17 . Prandoni P , Lensing AWA , Büller HR , et al . Comparison 4 . Leroyer C , Bressollette L , Oger E , et al ; The ANTENOX of subcutaneous low-molecular-weight heparin with intra- Study Group . Early versus delayed introduction of oral venous standard heparin in proximal deep-vein thrombosis . vitamin K antagonists in combination with low-molecular- Lancet . 1992 ; 339 ( 8791 ): 441 - 445 . weight heparin in the treatment of deep vein thrombosis. a 18 . Riess H , Koppenhagen K , Tolle A , et al ; TH-4 Study Group . randomized clinical trial . Haemostasis . 1998 ; 28 ( 2 ): 70 - 77 . Fixed-dose, body weight-independent subcutaneous low 5 . Mohiuddin SM , Hilleman DE , Destache CJ , Stoysich AM , molecular weight heparin Certoparin compared with adjusted- Gannon JM , Sketch MH Sr . Effi cacy and safety of early ver- dose intravenous unfractionated heparin in patients with sus late initiation of warfarin during heparin therapy in acute proximal deep venous thrombosis . Thromb Haemost . 2003 ; thromboembolism . Am Heart J . 1992 ; 123 ( 3 ): 729 - 732 . 90 ( 2 ): 252 - 259 . 6 . van Dongen CJJ , van den Belt AG , Prins MH , Lensing AW . 19 . Simonneau G , Charbonnier B , Decousus H , et al . Sub- Fixed dose subcutaneous low molecular weight heparins cutaneous low-molecular-weight heparin compared with versus adjusted dose unfractionated heparin for venous continuous intravenous unfractionated heparin in the treat- thromboembolism . Cochrane Database Syst Rev . 2004 ;( 4 ): ment of proximal deep vein thrombosis . Arch Intern Med . CD001100 . 1993 ; 153 ( 13 ): 1541 - 1546 . 7 . Breddin HK , Hach-Wunderle V , Nakov R , Kakkar VV; 20 . Simonneau G , Sors H , Charbonnier B , et al . A comparison CORTES Investigators. Clivarin: Assessment of Regression of of low-molecular-weight heparin with unfractionated hep- Thrombosis, Effi cacy, and Safety . Effects of a low-molecular- arin for acute pulmonary embolism. The THESEE Study weight heparin on thrombus regression and recurrent Group. Tinzaparine ou Heparine Standard: Evaluations dans thromboembolism in patients with deep-vein thrombosis . l’Embolie Pulmonaire. N Engl J Med . 1997 ; 337 ( 10 ): 663 - 669 . N Engl J Med . 2001 ; 344 ( 9 ): 626 - 631 . 21 . Lopaciuk S , Meissner AJ , Filipecki S , et al . Subcutaneous 8 . Decousus H , Leizorovicz A , Parent F , et al . A clinical trial of low molecular weight heparin versus subcutaneous unfrac- vena caval fi lters in the prevention of pulmonary embolism tionated heparin in the treatment of deep vein thrombosis: a in patients with proximal deep-vein thrombosis. Prévention Polish multicenter trial . Thromb Haemost . 1992 ; 68 ( 1 ): 14 - 18 . du Risque d’Embolie Pulmonaire par Interruption Cave Study 22 . Faivre R , Neuhart Y , Kieffer Y , et al . A new treatment of deep Group . N Engl J Med . 1998 ; 338 ( 7 ): 409 - 415 . venous thrombosis: low molecular weight heparin fractions. 9 . Fiessinger JN , Lopez-Fernandez M , Gatterer E , et al . Once- Randomized study [in French]. Presse Med . 1988 ; 17 ( 5 ): daily subcutaneous dalteparin, a low molecular weight hep- 197 - 200 . arin, for the initial treatment of acute deep vein thrombosis . 23 . Prandoni P , Carnovali M , Marchiori A ; Galilei Investiga- Thromb Haemost . 1996 ; 76 ( 2 ): 195 - 199 . tors . Subcutaneous adjusted-dose unfractionated heparin 10 . Harenberg J , Schmidt JA , Koppenhagen K , Tolle A , vs fi xed-dose low-molecular-weight heparin in the initial Huisman MV , Büller HR ; EASTERN Investigators . Fixed- treatment of venous thromboembolism. Arch Intern Med . dose, body weight-independent subcutaneous LMW heparin 2004 ; 164 ( 10 ): 1077 - 1083 . versus adjusted dose unfractionated intravenous heparin in 24 . Kearon C , Ginsberg JS , Julian JA , et al ; Fixed-Dose Heparin the initial treatment of proximal venous thrombosis . Thromb (FIDO) Investigators . Comparison of fi xed-dose weight- Haemost . 2000 ; 83 ( 5 ): 652 - 656 . adjusted unfractionated heparin and low-molecular-weight 11 . Hull RD , Raskob GE , Pineo GF , et al . Subcutaneous low- heparin for acute treatment of venous thromboembolism . molecular-weight heparin compared with continuous intra- JAMA . 2006 ; 296 ( 8 ): 935 - 942 . venous heparin in the treatment of proximal-vein thrombosis . 25 . Büller HR , Davidson BL , Decousus H , et al ; Matisse Investi- N Engl J Med . 1992 ; 326 ( 15 ): 975 - 982 . gators . Fondaparinux or enoxaparin for the initial treatment 12 . Kirchmaier CM , Wolf H , Schäfer H , Ehlers B , Breddin HK ; of symptomatic deep venous thrombosis: a randomized trial . C e r t o p a r i n - S t u d y G r o u p . E f fi cacy of a low molecular weight Ann Intern Med . 2004 ; 140 ( 11 ): 867 - 873 . heparin administered intravenously or subcutaneously in 26 . van Dongen CJ , MacGillavry MR , Prins MH . Once versus comparison with intravenous unfractionated heparin in the twice daily LMWH for the initial treatment of venous treatment of deep venous thrombosis. Int Angiol . 1 9 9 8 ; thromboembolism. Cochrane Database Syst Rev . 2 0 0 5 ; 17 ( 3 ): 135 - 145 . ( 3 ): CD003074 . 13 . Koopman MMW , Prandoni P , Piovella F , et al ; The Tasman 27 . Holmoström M , Berglund MC , Granquist S , Bratt G , Study Group. Treatment of venous thrombosis with intra- Törnebohm E , Lockner D . Fragmin once or twice daily venous unfractionated heparin administered in the hospi- subcutaneously in the treatment of deep venous thrombosis tal as compared with subcutaneous low-molecular-weight of the leg . Thromb Res . 1992 ; 67 ( 1 ): 49 - 55 . heparin administered at home. N Engl J Med . 1 9 9 6 ; 3 3 4 ( 1 1 ) : 28 . Charbonnier BA , Fiessinger JN , Banga JD , Wenzel E , 682 - 687 . d’Azemar P , Sagnard L . Comparison of a once daily with a 14 . Levine M , Gent M , Hirsh J , et al . A comparison of low- twice daily subcutaneous low molecular weight heparin reg- molecular-weight heparin administered primarily at home imen in the treatment of deep vein thrombosis. FRAXODI with unfractionated heparin administered in the hospital group . Thromb Haemost . 1998 ; 79 ( 5 ): 897 - 901 .

86 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 29 . Othieno R , Abu Affan M , Okpo E . Home versus in-patient 45 . AbuRahma AF , Perkins SE , Wulu JT , Ng HK . Iliofemoral treatment for deep vein thrombosis. Cochrane Database deep vein thrombosis: conventional therapy versus lysis and Syst Rev . 2007 ;( 3 ): CD003076 . percutaneous transluminal angioplasty and stenting. Ann 30 . Boccalon H , Elias A , Chalé JJ , Cadène A , Gabriel S . Clinical Surg . 2001 ; 233 ( 6 ): 752 - 760 . outcome and cost of hospital vs home treatment of proximal 46 . Kearon C , Kahn SR , Agnelli G , Goldhaber S , Raskob GE , deep vein thrombosis with a low-molecular-weight heparin: Comerota AJ ; American College of Chest Physicians . Anti- the Vascular Midi-Pyrenees study . Arch Intern Med . 2000 ; thrombotic therapy for venous thromboembolic disease: 160 ( 12 ): 1769 - 1773 . American College of Chest Physicians Evidence-Based 31 . Chong BH , Brighton TA , Baker RI , Thurlow P , Lee CH ; Clinical Practice Guidelines (8th Edition). Chest . 2 0 0 8 ; ASTH DVT Study Group . Once-daily enoxaparin in the 133 ( 6 suppl ): 454S - 545S . outpatient setting versus unfractionated heparin in hospital 47 . Watson LI , Armon MP . Thrombolysis for acute deep for the treatment of symptomatic deep-vein thrombosis . vein thrombosis. Cochrane Database Syst Rev . 2 0 0 4 ; ( 4 ) : J Thromb Thrombolysis . 2005 ; 19 ( 3 ): 173 - 181 . CD002783 . 32 . Daskalopoulos ME , Daskalopoulou SS , Tzortzis E , et al . 48 . Common HH , Seaman AJ , Rösch J , Porter JM , Dotter CT . Long-term treatment of deep venous thrombosis with a Deep vein thrombosis treated with streptokinase or hep- low molecular weight heparin (tinzaparin): a prospective arin. Follow-up of a randomized study . Angiology . 1 9 7 6 ; randomized trial. Eur J Vasc Endovasc Surg . 2005 ; 29 ( 6 ): 27 ( 11 ): 645 - 654 . 638 - 650 . 49 . Schulman S , Granqvist S , Juhlin-Dannfelt A , Lockner D . 33 . Ramacciotti E , Araújo GR , Lastoria S , et al ; CLETRAT Long-term sequelae of calf vein thrombosis treated with Investigators . An open-label, comparative study of the effi cacy heparin or low-dose streptokinase . Acta Med Scand . 1986 ; and safety of once-daily dose of enoxaparin versus unfrac- 219 ( 4 ): 349 - 357 . tionated heparin in the treatment of proximal lower limb 50 . Turpie AG , Levine MN , Hirsh J , et al . Tissue plasminogen deep-vein thrombosis . Thromb Res . 2004 ; 114 ( 3 ): 149 - 153 . activator (rt-PA) vs heparin in deep vein thrombosis. Results 34 . Bäckman K , Carlsson P , Kentson M , Hansen S , Engquist L , of a randomized trial . Chest . 1990 ; 97 ( 4 suppl ): 172S - 175S . Hallert C . Deep venous thrombosis: a new task for primary 51 . Verhaeghe R , Besse P , Bounameaux H , Marbet GA . Mul- health care. A randomised economic study of outpatient ticenter pilot study of the effi cacy and safety of systemic and inpatient treatment . Scand J Prim Health Care . 2 0 0 4 ; rt-PA administration in the treatment of deep vein throm- 22 ( 1 ): 44 - 49 . bosis of the lower extremities and/or pelvis . Thromb Res . 35 . O’Brien B , Levine M , Willan A , et al . Economic evalua- 1989 ; 55 ( 1 ): 5 - 11 . tion of outpatient treatment with low-molecular-weight 52 . Browse NL , Thomas ML , Pim HP . Streptokinase and deep heparin for proximal vein thrombosis. Arch Intern Med . vein thrombosis . BMJ . 1968 ; 3 ( 5620 ): 717 - 720 . 1999 ; 159 ( 19 ): 2298 - 2304 . 53 . Robertson BR , Nilsson IM , Nylander G . Value of strep- 36 . Elsharawy M , Elzayat E . Early results of thrombolysis tokinase and heparin in treatment of acute deep venous vs anticoagulation in iliofemoral venous thrombosis. A ran- thrombosis. A coded investigation . Acta Chir Scand . 1968 ; domised clinical trial . Eur J Vasc Endovasc Surg . 2002 ; 24 ( 3 ): 134 ( 3 ): 203 - 208 . 209 - 214 . 54 . Kakkar VV , Flanc C , Howe CT , O’Shea M , Flute PT . Treat- 37 . Enden T , Kløw NE , Sandvik L , et al ; CaVenT study group . ment of deep vein thrombosis. A trial of heparin, streptoki- Catheter-directed thrombolysis vs. anticoagulant therapy nase, and arvin . BMJ . 1969 ; 1 ( 5647 ): 806 - 810 . alone in deep vein thrombosis: results of an open ran- 55 . Tsapogas MJ , Peabody RA , Wu KT , Karmody AM , domized, controlled trial reporting on short-term patency . Devaraj KT , Eckert C . Controlled study of thrombolytic J Thromb Haemost . 2009 ; 7 ( 8 ): 1268 - 1275 . therapy in deep vein thrombosis. Surgery . 1973 ; 74 ( 6 ): 973 - 984 . 38 . Comerota AJ , Throm RC , Mathias SD , Haughton S , 56 . Duckert F , Müller G , Nyman D , et al . Treatment of deep Mewissen M . Catheter-directed thrombolysis for iliofemoral vein thrombosis with streptokinase. BMJ . 1 9 7 5 ; 1 ( 5 9 5 6 ) : deep venous thrombosis improves health-related quality of 479 - 481 . life. J Vasc Surg . 2000 ; 32 ( 1 ): 130 - 137 . 57 . Porter JM , Seaman AJ , Common HH , Rösch J , Eidemiller LR , 39 . Douketis JD , Foster GA , Crowther MA , Prins MH , Calhoun AD . Comparison of heparin and streptokinase in Ginsberg JS . Clinical risk factors and timing of recurrent the treatment of venous thrombosis . Am Surg . 1975 ; 41 ( 9 ): venous thromboembolism during the initial 3 months of 511 - 519 . anticoagulant therapy. Arch Intern Med . 2 0 0 0 ; 1 6 0 ( 2 2 ) : 58 . Marder VJ , Soulen RL , Atichartakarn V , et al . Quantitative 3431 - 3436 . venographic assessment of deep vein thrombosis in the eval- 40 . Kahn SR , Shrier I , Julian JA , et al . Determinants and time uation of streptokinase and heparin therapy . J Lab Clin Med . course of the postthrombotic syndrome after acute deep 1977 ; 89 ( 5 ): 1018 - 1029 . venous thrombosis . Ann Intern Med . 2008 ; 149 ( 10 ): 698 - 707 . 59 . Arnesen H , Heilo A , Jakobsen E , Ly B , Skaga E . A prospec- 41 . Semba CP , Dake MD . Iliofemoral deep venous thrombosis: tive study of streptokinase and heparin in the treatment aggressive therapy with catheter-directed thrombolysis . of deep vein thrombosis. Acta Med Scand . 1 9 7 8 ; 2 0 3 ( 6 ) : Radiology . 1994 ; 191 ( 2 ): 487 - 494 . 457 - 463 . 42 . Verhaeghe R , Stockx L , Lacroix H , Vermylen J , Baert AL . 60 . Elliot MS , Immelman EJ , Jeffery P , et al . A comparative Catheter-directed lysis of iliofemoral vein thrombosis with randomized trial of heparin versus streptokinase in the use of rt-PA . Eur Radiol . 1997 ; 7 ( 7 ): 996 - 1001 . treatment of acute proximal venous thrombosis: an interim 43 . Bjarnason H , Kruse JR , Asinger DA , et al . Iliofemoral deep report of a prospective trial. Br J Surg . 1979 ; 66 ( 12 ): 838 - 843 . venous thrombosis: safety and effi cacy outcome during 61 . Watz R , Savidge GF . Rapid thrombolysis and preservation 5 years of catheter-directed thrombolytic therapy . J Vasc of valvular venous function in high deep vein thrombosis. Interv Radiol . 1997 ; 8 ( 3 ): 405 - 418 . A comparative study between streptokinase and heparin 44 . Mewissen MW , Seabrook GR , Meissner MH , Cynamon J , therapy . Acta Med Scand . 1979 ; 205 ( 4 ): 293 - 298 . Labropoulos N , Haughton SH . Catheter-directed thrombol- 62 . Kiil J , Carvalho A , Saksø P , Nielsen HO . Urokinase or heparin ysis for lower extremity deep venous thrombosis: report of a in the management of patients with deep vein thrombosis? national multicenter registry . Radiology . 1999 ; 211 ( 1 ): 39 - 49 . Acta Chir Scand . 1981 ; 147 ( 7 ): 529 - 532 .

87 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 63 . Arnesen H , Høiseth A , Ly B . Streptokinase of heparin in venous thrombosis: a prospective, randomized, open, multi- the treatment of deep vein thrombosis. Follow-up results centre trial . Curr Med Res Opin . 2006 ; 22 ( 3 ): 593 - 602 . of a prospective study. Acta Med Scand . 1982 ; 211 ( 1-2 ): 81 . Partsch H , Blättler W . Compression and walking versus 65 - 68 . bed rest in the treatment of proximal deep venous throm- 64 . Goldhaber SZ , Meyerovitz MF , Green D , et al . Randomized bosis with low molecular weight heparin . J Vasc Surg . controlled trial of tissue plasminogen activator in proximal 2000 ; 32 ( 5 ): 861 - 869 . deep venous thrombosis . Am J Med . 1990 ; 88 ( 3 ): 235 - 240 . 82 . Schellong SM , Schwarz T , Kropp J , Prescher Y , Beuthien- 65 . Schweizer J , Elix H , Altmann E , Hellner G , Forkmann L . Baumann B , Daniel WG . Bed rest in deep vein thrombosis Comparative results of thrombolysis treatment with rt-PA and the incidence of scintigraphic pulmonary embolism . and urokinase: a pilot study . Vasa . 1998 ; 27 ( 3 ): 167 - 171 . Thromb Haemost . 1999 ; 82 ( Suppl 1 ): 127 - 129 . 66 . Schweizer J , Kirch W , Koch R , et al . Short- and long-term 83 . Partsch H . Therapy of deep vein thrombosis with low results after thrombolytic treatment of deep venous throm- molecular weight heparin, leg compression and immediate bosis . J Am Coll Cardiol . 2000 ; 36 ( 4 ): 1336 - 1343 . ambulation . Vasa . 2001 ; 30 ( 3 ): 195 - 204 . 67 . Plate G , Einarsson E , Ohlin P , Jensen R , Qvarfordt P , 84 . Partsch H , Kaulich M , Mayer W . Immediate mobilisation in E k l ö f B . T h r o m b e c t o m y w i t h t e m p o r a r y a r t e r i o v e n o u s fi stula: acute vein thrombosis reduces post-thrombotic syndrome . the treatment of choice in acute iliofemoral venous throm- Int Angiol . 2004 ; 23 ( 3 ): 206 - 212 . bosis . J Vasc Surg . 1984 ; 1 ( 6 ): 867 - 876 . 85 . Trujillo-Santos J , Perea-Milla E , Jiménez-Puente A , et al ; 68 . Einarsson E , Albrechtsson U , Eklöf B . Thrombectomy and RIETE Investigators . Bed rest or ambulation in the initial temporary AV-fi stula in iliofemoral vein thrombosis. Tech- treatment of patients with acute deep vein thrombosis or nical considerations and early results. Int Angiol . 1 9 8 6 ; pulmonary embolism: fi ndings from the RIETE registry . 5 ( 2 ): 65 - 72 . Chest . 2005 ; 127 ( 5 ): 1631 - 1636 . 69 . Einarsson E , Albrechtsson U , Eklöf B , Norgren L . 86 . Kearon C , Ginsberg JS , Anderson DR , et al ; SOFAST Follow-up evaluation of venous morphologic factors and Investigators . Comparison of 1 month with 3 months of anti- function after thrombectomy and temporary arteriovenous coagulation for a fi rst episode of venous thromboembolism fi stula in thrombosis of iliofemoral vein. Surg Gynecol Obstet . associated with a transient risk factor . J Thromb Haemost . 1986 ; 163 ( 2 ): 111 - 116 . 2004 ; 2 ( 5 ): 743 - 749 . 70 . Plate G , Akesson H , Einarsson E , Ohlin P , Eklöf B . Long- 87 . Pinede L , Ninet J , Duhaut P , et al ; Investigators of the term results of venous thrombectomy combined with a tem- “Durée Optimale du Traitement AntiVitamines K” (DOTAVK) porary arterio-venous fi stula. Eur J Vasc Surg . 1 9 9 0 ; 4 ( 5 ) : Study . Comparison of 3 and 6 months of oral anticoagulant 483 - 489 . therapy after a fi rst episode of proximal deep vein throm- 71 . Neglén P , al-Hassan HK , Endrys J , Nazzal MM , Christenson JT , bosis or pulmonary embolism and comparison of 6 and Eklof B . Iliofemoral venous thrombectomy followed by per- 12 weeks of therapy after isolated calf deep vein thrombosis . cutaneous closure of the temporary arteriovenous fi stula . Circulation . 2001 ; 103 ( 20 ): 2453 - 2460 . Surgery . 1991 ; 110 ( 3 ): 493 - 499 . 88 . Schulman S , Rhedin AS , Lindmarker P , et al ; Duration 72 . Plate G , Eklöf B , Norgren L , Ohlin P , Dahlström JA . of Anticoagulation Trial Study Group . A comparison of Venous thrombectomy for iliofemoral vein thrombosis— six weeks with six months of oral anticoagulant therapy after 10-year results of a prospective randomised study . Eur J a fi rst episode of venous thromboembolism . N Engl J Med . Vasc Endovasc Surg . 1997 ; 14 ( 5 ): 367 - 374 . 1995 ; 332 ( 25 ): 1661 - 1665 . 73 . PREPIC Study Group . Eight-year follow-up of patients with 89 . Levine MN , Hirsh J , Gent M , et al . Optimal duration of permanent vena cava fi lters in the prevention of pulmonary oral anticoagulant therapy: a randomized trial comparing embolism: the PREPIC (Prevention du Risque d’Embolie four weeks with three months of warfarin in patients with Pulmonaire par Interruption Cave) randomized study . proximal deep vein thrombosis . Thromb Haemost . 1 9 9 5 ; Circulation . 2005 ; 112 ( 3 ): 416 - 422 . 74 ( 2 ): 606 - 611 . 74 . Mismetti P , Rivron-Guillot K , Quenet S , et al . A prospective 90 . Research Committee of the British Thoracic Society . Opti- long-term study of 220 patients with a retrievable vena cava mum duration of anticoagulation for deep-vein thrombosis fi lter for secondary prevention of venous thromboembolism . and pulmonary embolism . Lancet . 1992 ; 340 ( 8824 ): 873 - 876 . Chest . 2007 ; 131 ( 1 ): 223 - 229 . 91 . Campbell IA , Bentley DP , Prescott RJ , Routledge PA , 75 . Athanasoulis CA , Kaufman JA , Halpern EF , Waltman AC , Shetty HG , Williamson IJ . Anticoagulation for three versus Geller SC , Fan CM . Inferior vena caval fi lters: review of a six months in patients with deep vein thrombosis or pul- 26-year single-center clinical experience . Radiology . 2000 ; monary embolism, or both: randomised trial. BMJ . 2007 ; 216 ( 1 ): 54 - 66 . 334 ( 7595 ): 674 - 680 . 76 . Kahn SR , Shrier I , Kearon C . Physical activity in patients 92 . Agnelli G , Prandoni P , Becattini C , et al ; Warfarin Optimal with deep venous thrombosis: a systematic review . Thromb Duration Italian Trial Investigators . Extended oral anticoag- Res . 2008 ; 122 ( 6 ): 763 - 773 . ulant therapy after a fi rst episode of pulmonary embolism . 77 . Aissaoui N , Martins E , Mouly S , Weber S , Meune C . A Ann Intern Med . 2003 ; 139 ( 1 ): 19 - 25 . meta-analysis of bed rest versus early ambulation in the 93 . Agnelli G , Prandoni P , Santamaria MG , et al ; Warfarin management of pulmonary embolism, deep vein thrombo- Optimal Duration Italian Trial Investigators . Three months sis, or both . Int J Cardiol . 2009 ; 137 ( 1 ): 37 - 41 . versus one year of oral anticoagulant therapy for idiopathic 78 . Aschwanden M , Labs KH , Engel H , et al . Acute deep deep venous thrombosis . N Engl J Med . 2001 ; 345 ( 3 ): 165 - 169 . vein thrombosis: early mobilization does not increase the 94 . Siragusa S , Malato A , Anastasio R , et al . Residual vein frequency of pulmonary embolism . Thromb Haemost . 2001 ; thrombosis to establish duration of anticoagulation after 85 ( 1 ): 42 - 46 . a fi rst episode of deep vein thrombosis: the Duration of 79 . Blättler W , Partsch H . Leg compression and ambulation is Anticoagulation based on Compression UltraSonography better than bed rest for the treatment of acute deep venous (DACUS) study . Blood . 2008 ; 112 ( 3 ): 511 - 515 . thrombosis . Int Angiol . 2003 ; 22 ( 4 ): 393 - 400 . 95 . Palareti G , Cosmi B , Legnani C , et al ; PROLONG Investi- 80 . Jünger M , Diehm C , Störiko H , et al . Mobilization versus gators . D-dimer testing to determine the duration of antico- immobilization in the treatment of acute proximal deep agulation therapy . N Engl J Med . 2006 ; 355 ( 17 ): 1780 - 1789 .

88 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 96 . Kearon C , Gent M , Hirsh J , et al . A comparison of after deep vein thrombosis . Thromb Haemost . 1 9 9 4 ; 7 2 ( 2 ) : three months of anticoagulation with extended antico- 191 - 197 . agulation for a fi rst episode of idiopathic venous thrombo- 112 . Das SK , Cohen AT , Edmondson RA , Melissari E , Kakkar VV . embolism . N Engl J Med . 1999 ; 340 ( 12 ): 901 - 907 . Low-molecular-weight heparin versus warfarin for preven- 97 . Schulman S , Granqvist S , Holmström M , et al ; The Dura- tion of recurrent venous thromboembolism: a randomized tion of Anticoagulation Trial Study Group . The duration of trial . World J Surg . 1996 ; 20 ( 5 ): 521 - 526 . oral anticoagulant therapy after a second episode of venous 113 . Gonzalez-Fajardo JA , Arreba E , Castrodeza J , et al . thromboembolism . N Engl J Med . 1997 ; 336 ( 6 ): 393 - 398 . Venographic comparison of subcutaneous low-molecular 98 . Farraj RS . Anticoagulation period in idiopathic venous weight heparin with oral anticoagulant therapy in the long- thromboembolism. How long is enough? Saudi Med J . 2004 ; term treatment of deep venous thrombosis. J Vasc Surg . 25 ( 7 ): 848 - 851 . 1999 ; 30 ( 2 ): 283 - 292 . 99 . Prandoni P , Prins MH , Lensing AW , et al ; AESOPUS Inves- 114 . Veiga F , Escribá A , Maluenda MP , et al . Low molecular tigators . Residual thrombosis on ultrasonography to guide weight heparin (enoxaparin) versus oral anticoagulant the duration of anticoagulation in patients with deep venous therapy (acenocoumarol) in the long-term treatment of thrombosis: a randomized trial. Ann Intern Med . 2 0 0 9 ; deep venous thrombosis in the elderly: a randomized trial . 150 ( 9 ): 577 - 585 . Thromb Haemost . 2000 ; 84 ( 4 ): 559 - 564 . 100 . Ridker PM , Goldhaber SZ , Danielson E , et al ; PREVENT 115 . Kakkar VV , Gebska M , Kadziola Z , Saba N , Carrasco P ; Investigators . Long-term, low-intensity warfarin therapy Bemiparin Investigators . Low-molecular-weight heparin in for the prevention of recurrent venous thromboembolism . the acute and long-term treatment of deep vein thrombosis . N Engl J Med . 2003 ; 348 ( 15 ): 1425 - 1434 . Thromb Haemost . 2003 ; 89 ( 4 ): 674 - 680 . 101 . Kearon C . Natural history of venous thromboembolism . 116 . Prandoni P , Lensing AW , Piccioli A , et al . Recurrent venous Circulation . 2003 ; 107 ( suppl ): I-22-I30 . thromboembolism and bleeding complications during 102 . Deitcher SR , Kessler CM , Merli G , Rigas JR , Lyons RM , anticoagulant treatment in patients with cancer and venous Fareed J ; ONCENOX Investigators . Secondary prevention thrombosis . Blood . 2002 ; 100 ( 10 ): 3484 - 3488 . of venous thromboembolic events in patients with active 117 . Beyth RJ , Cohen AM , Landefeld CS . Long-term outcomes cancer: enoxaparin alone versus initial enoxaparin followed of deep-vein thrombosis. Arch Intern Med . 1 9 9 5 ; 1 5 5 ( 1 0 ) : by warfarin for a 180-day period . Clin Appl Thromb Hemost . 1031 - 1037 . 2006 ; 12 ( 4 ): 389 - 396 . 118 . Kahn SR , Ginsberg JS . Relationship between deep venous 103 . Hull RD , Pineo GF , Brant RF , et al ; LITE Trial Investiga- thrombosis and the postthrombotic syndrome . Arch Intern tors . Self-managed long-term low-molecular-weight heparin Med . 2004 ; 164 ( 1 ): 17 - 26 . therapy: the balance of benefi ts and harms . Am J Med . 2007 ; 119 . Bauersachs R , Berkowitz SD , Brenner B , et al ; EINSTEIN 120 ( 1 ): 72 - 82 . Investigators . Oral rivaroxaban for symptomatic venous 104 . Hull RD , Pineo GF , Brant R , et al . Home therapy of venous thromboembolism . N Engl J Med . 2010 ; 363 ( 26 ): 2499 - 2510 . thrombosis with long-term LMWH versus usual care: 120 . Prandoni P , Lensing AW , Prins MH , et al . Below-knee elas- patient satisfaction and post-thrombotic syndrome . Am J Med . tic compression stockings to prevent the post-thrombotic 2009 ; 122 ( 8 ): 762 - 769 . syndrome: a randomized, controlled trial . Ann Intern Med . 105 . Lee AY , Levine MN , Baker RI , et al ; Randomized Compar- 2004 ; 141 ( 4 ): 249 - 256 . ison of Low-Molecular-Weight Heparin versus Oral Anti- 121 . van Dongen CJ , Prandoni P , Frulla M , Marchiori A , coagulant Therapy for the Prevention of Recurrent Venous Prins MH , Hutten BA . Relation between quality of antico- Thromboembolism in Patients with Cancer (CLOT) Inves- agulant treatment and the development of the postthrom- tigators . Low-molecular-weight heparin versus a coumarin botic syndrome. J Thromb Haemost . 2005 ; 3 ( 5 ): 939 - 942 . for the prevention of recurrent venous thromboembolism 122 . Schulman S , Kearon C , Kakkar AK , et al ; RE-COVER in patients with cancer. N Engl J Med . 2 0 0 3 ; 3 4 9 ( 2 ) : 1 4 6 - 1 5 3 . Study Group . Dabigatran versus warfarin in the treatment 106 . Lopaciuk S , Bielska-Falda H , Noszczyk W , et al . Low molec- of acute venous thromboembolism. N Engl J Med . 2 0 0 9 ; ular weight heparin versus acenocoumarol in the secondary 361 ( 24 ): 2342 - 2352 . prophylaxis of deep vein thrombosis. Thromb Haemost . 123 . Kolbach DN , Sandbrink MW , Neumann HA , Prins MH . 1999 ; 81 ( 1 ): 26 - 31 . Compression therapy for treating stage I and II (Widmer) 107 . López-Beret P , Orgaz A , Fontcuberta J , et al . Low molec- post-thrombotic syndrome. Cochrane Database Syst Rev . ular weight heparin versus oral anticoagulants in the long- 2003 ;( 4 ): CD004177 . term treatment of deep venous thrombosis . J Vasc Surg . 124 . Ginsberg JS , Hirsh J , Julian J , et al . Prevention and treat- 2001 ; 33 ( 1 ): 77 - 90 . ment of postphlebitic syndrome: results of a 3-part study . 108 . Meyer G , Marjanovic Z , Valcke J , et al . Comparison of low- Arch Intern Med . 2001 ; 161 ( 17 ): 2105 - 2109 . molecular-weight heparin and warfarin for the secondary 125 . Belcaro G , Laurora G , Cesarone MR , De Sanctis MT . Pro- prevention of venous thromboembolism in patients with phylaxis of recurrent deep venous thrombosis. A randomized, cancer: a randomized controlled study . Arch Intern Med . prospective study using indobufen and graduated elastic 2002 ; 162 ( 15 ): 1729 - 1735 . compression stockings . Angiology . 1993 ; 44 ( 9 ): 695 - 699 . 109 . Romera A , Cairols MA , Vila-Coll R , et al . A randomised 126 . A r p a i a G , C i m m i n i e l l o C , M a s t r o g i a c o m o O , d e G a u d e n z i E . open-label trial comparing long-term sub-cutaneous low- Effi cacy of elastic compression stockings used early or after molecular-weight heparin compared with oral-anticoagulant resolution of the edema on recanalization after deep venous therapy in the treatment of deep venous thrombosis . Eur J thrombosis: the COM.PRE Trial . Blood Coagul Fibrinolysis . Vasc Endovasc Surg . 2009 ; 37 ( 3 ): 349 - 356 . 2007 ; 18 ( 2 ): 131 - 137 . 110 . Hull RD , Pineo GF , Brant RF , et al ; LITE Trial Investigators . 127 . Brandjes DP , Büller HR , Heijboer H , et al . Randomised trial Long-term low-molecular-weight heparin versus usual care of effect of compression stockings in patients with symp- in proximal-vein thrombosis patients with cancer . Am J Med . tomatic proximal-vein thrombosis . Lancet . 1997 ; 349 ( 9054 ): 2006 ; 119 ( 12 ): 1062 - 1072 . 759 - 762 . 111 . Pini M , Aiello S , Manotti C , et al . Low molecular weight 128 . Heit JA , Mohr DN , Silverstein MD , Petterson TM , heparin versus warfarin in the prevention of recurrences O’Fallon WM , Melton LJ III Predictors of recurrence after

89 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians deep vein thrombosis and pulmonary embolism: a population- ized controlled trial. J Thromb Thrombolysis . 1 9 9 5 ; 2 ( 3 ) : based cohort study . Arch Intern Med . 2000 ; 160 ( 6 ): 761 - 768 . 227 - 229 . 129 . Prandoni P , Noventa F , Ghirarduzzi A , et al . The risk of 146 . Ly B , Arnesen H , Eie H , Hol R . A controlled clinical trial of recurrent venous thromboembolism after discontinuing streptokinase and heparin in the treatment of major pulmo- anticoagulation in patients with acute proximal deep vein nary embolism . Acta Med Scand . 1978 ; 203 ( 6 ): 465 - 470 . thrombosis or pulmonary embolism. A prospective cohort 147 . Marini C , Di Ricco G , Rossi G , Rindi M , Palla R , Giuntini C . study in 1,626 patients. Haematologica . 2 0 0 7 ; 9 2 ( 2 ) : 1 9 9 - 2 0 5 . Fibrinolytic effects of urokinase and heparin in acute pul- 130 . Aschwanden M , Jeanneret C , Koller MT , Thalhammer C , monary embolism: a randomized clinical trial . Respiration . Bucher HC , Jaeger KA . Effect of prolonged treatment 1988 ; 54 ( 3 ): 162 - 173 . with compression stockings to prevent post-thrombotic 148 . Tibbutt DA , Davies JA , Anderson JA , et al . Comparison by sequelae: a randomized controlled trial . J Vasc Surg . 2008 ; controlled clinical trial of streptokinase and heparin in 47 ( 5 ): 1015 - 1021 . treatment of life-threatening pulmonay embolism. BMJ . 131 . Frulla M , Marchiori A , Sartor D , et al . Elastic stockings, 1974 ; 1 ( 5904 ): 343 - 347 . hydroxyethylrutosides or both for the treatment of post- 149 . Urokinase Pulmonary Embolism Trial Study Group. Uro- thrombotic syndrome. Thromb Haemost . 2 0 0 5 ; 9 3 ( 1 ) : kinase pulmonary embolism trial: phase I results. JAMA . 183 - 185 . 1970 ; 214 ( 12 ): 2163 - 2172 . 132 . Ginsberg JS , Magier D , Mackinnon B , Gent M , Hirsh J . 150 . Dong B , Jirong Y , Wang Q , et al . Thrombolytic treatment Intermittent compression units for severe post-phlebitic for pulmonary embolism. Cochrane Database Syst Rev . syndrome: a randomized crossover study. CMAJ . 1 9 9 9 ; 2006 ; 2 : CD004437 . 160 ( 9 ): 1303 - 1306 . 151 . Wan S , Quinlan DJ , Agnelli G , Eikelboom JW . Thrombolysis 133 . O’Donnell MJ , McRae S , Kahn SR , et al . Evaluation of a compared with heparin for the initial treatment of pulmo- venous-return assist device to treat severe post-thrombotic nary embolism: a meta-analysis of the randomized con- syndrome (VENOPTS). A randomized controlled trial . trolled trials . Circulation . 2004 ; 110 ( 6 ): 744 - 749 . Thromb Haemost . 2008 ; 99 ( 3 ): 623 - 629 . 152 . Agnelli G , Becattini C , Kirschstein T . Thrombolysis vs 134 . de Jongste AB , Jonker JJ , Huisman MV , ten Cate JW , heparin in the treatment of pulmonary embolism: a clin- Azar AJ . A double blind three center clinical trial on the ical outcome-based meta-analysis . Arch Intern Med . 2002 ; short-term effi cacy of 0-(beta-hydroxyethyl)-rutosides in 162 ( 22 ): 2537 - 2541 . patients with post-thrombotic syndrome . Thromb Haemost . 153 . Becattini C , Agnelli G , Salvi A , et al ; TIPES Study Group . 1989 ; 62 ( 3 ): 826 - 829 . Bolus tenecteplase for right ventricle dysfunction in hemo- 135 . Monreal M , Callejas JM , Martorell A , et al . A prospective dynamically stable patients with pulmonary embolism . study of the long-term effi cacy of two different venoactive Thromb Res . 2010 ; 125 ( 3 ): e82 - e86 . drugs in patients with post-thrombotic syndrome . Phlebology . 154 . Fasullo S , Scalzo S , Maringhini G , et al . Six-month echo- 1994 ; 9 ( 1 ): 37 - 40 . cardiographic study in patients with submassive pulmonary 136 . Büller HR , Davidson BL , Decousus H , et al ; Matisse Inves- embolism and right ventricle dysfunction: comparison of tigators . Subcutaneous fondaparinux versus intravenous thrombolysis with heparin . Am J Med Sci . 2011 ; 341 ( 1 ): 33 - 39 . unfractionated heparin in the initial treatment of pulmonary 155 . J i m é n e z D , A u j e s k y D , M o o r e s L , e t a l ; R I E T E I n v e s t i - embolism . N Engl J Med . 2003 ; 349 ( 18 ): 1695 - 1702 . gators. Simplifi cation of the pulmonary embolism severity 137 . Otero R , Uresandi F , Jiménez D , et al . Home treatment in index for prognostication in patients with acute symptom- pulmonary embolism . Thromb Res . 2010 ; 126 ( 1 ): e1 - e5 . atic pulmonary embolism . Arch Intern Med . 2010 ; 170 ( 15 ): 138 . Aujesky D , Roy PM , Verschuren F , et al . Outpatient versus 1383 - 1389 . inpatient treatment for patients with acute pulmonary 156 . Wood KE . Major pulmonary embolism: review of a path- embolism: an international, open-label, randomised, non- ophysiologic approach to the golden hour of hemodynam- inferiority trial . Lancet . 2011 ; 378 ( 9785 ): 41 - 48 . ically signifi cant pulmonary embolism . Chest . 2002 ; 121 ( 3 ): 139 . Aujesky D , Obrosky DS , Stone RA , et al . Derivation and 877 - 905 . validation of a prognostic model for pulmonary embolism . 157 . Ruíz-Giménez N , Suárez C , González R , et al ; RIETE Inves- Am J Respir Crit Care Med . 2005 ; 172 ( 8 ): 1041 - 1046 . tigators . Predictive variables for major bleeding events in 140 . U r e s a n d i F , O t e r o R , C a y u e l a A , e t a l . [ A c l i n i c a l p r e d i c t i o n patients presenting with documented acute venous throm- rule for identifying short-term risk of adverse events in patients boembolism. Findings from the RIETE Registry . Thromb with pulmonary thromboembolism] . Arch Bronconeumol . Haemost . 2008 ; 100 ( 1 ): 26 - 31 . 2007 ; 43 ( 11 ): 617 - 622 . 158 . Goldhaber SZ , Haire WD , Feldstein ML , et al . Alteplase 141 . Urokinase Pulmonary Embolism Trial . A national co-operative versus heparin in acute pulmonary embolism: randomised study . Circulation . 1973 ; 47 ( Suppl II ): 1 . trial assessing right-ventricular function and pulmonary 142 . Dalla-Volta S , Palla A , Santolicandro A , et al . PAIMS 2: perfusion . Lancet . 1993 ; 341 ( 8844 ): 507 - 511 . alteplase combined with heparin versus heparin in the treat- 159 . Konstantinides S , Geibel A , Heusel G , Heinrich F , ment of acute pulmonary embolism. Plasminogen acti- K a s p e r W ; M a n a g e m e n t S t r a t e g i e s a n d P r o g n o s i s o f P u l m o - vator Italian multicenter study 2 . J Am Coll Cardiol . 1992 ; nary Embolism-3 Trial Investigators . Heparin plus alteplase 20 ( 3 ): 520 - 526 . compared with heparin alone in patients with submas- 143 . Dotter CT , Seamon AJ , Rosch J , et al . Streptokinase and sive pulmonary embolism. N Engl J Med . 2 0 0 2 ; 3 4 7 ( 1 5 ) : heparin in the treatment of pulmonary embolism: a ran- 1143 - 1150 . domized comparison . Vasc Surg . 1979 ; 13 ( 1 ): 42 - 52 . 160 . Levine M , Hirsh J , Weitz J , et al . A randomized trial of a 144 . PIOPED Investigators . Tissue plasminogen activator for single bolus dosage regimen of recombinant tissue plasmin- the treatment of acute pulmonary embolism. A collabora- ogen activator in patients with acute pulmonary embolism . tive study by the PIOPED Investigators . Chest . 1990 ; 97 ( 3 ): Chest . 1990 ; 98 ( 6 ): 1473 - 1479 . 528 - 533 . 161 . Decousus H , Prandoni P , Mismetti P , et al ; CALISTO Study 145 . Jerjes-Sanchez C , Ramírez-Rivera A , Arriaga-Nava R , et al ; Group. Fondaparinux for the treatment of superfi cial- de Lourdes García M . Streprokinase and heparin versus vein thrombosis in the legs. N Engl J Med . 2 0 1 0 ; 3 6 3 ( 1 3 ) : heparin alone in massive pulmonary embolism: a random- 1222 - 1232 .

90 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians 162 . Decousus H , Quéré I , Presles E , et al ; POST (Prospective 170 . Górski G , Szopiński P , Michalak J , et al . Liposomal heparin Observational Superfi cial Thrombophlebitis) Study Group . spray: a new formula in adjunctive treatment of superfi cial S u p e r fi cial venous thrombosis and venous thromboembolism: venous thrombosis . Angiology . 2005 ; 56 ( 1 ): 9 - 17 . a large, prospective epidemiologic study . Ann Intern Med . 171 . Andreozzi GM , Signorelli S , Di Pino L , et al . Tolerability 2010 ; 152 ( 4 ): 218 - 224 . and clinical effi cacy of desmin in the treatment of superfi cial 163 . Superfi cial Thrombophlebitis Treated By Enoxaparin thrombovaricophlebitis . Angiology . 1996 ; 47 ( 9 ): 887 - 894 . Study Group . A pilot randomized double-blind comparison 172 . Savage KJ , Wells PS , Schulz V , et al . Outpatient use of low of a low-molecular-weight heparin, a nonsteroidal anti- molecular weight heparin (Dalteparin) for the treatment infl ammatory agent, and placebo in the treatment of super- of deep vein thrombosis of the upper extremity. Thromb fi cial vein thrombosis. Arch Intern Med . 2 0 0 3 ; 1 6 3 ( 1 4 ) : Haemost . 1999 ; 82 ( 3 ): 1008 - 1010 . 1657 - 1663 . 173 . Karabay O , Yetkin U , Onol H . Upper extremity deep vein 164 . Titon JP , Auger D , Grange P , et al. [Therapeutic man- thrombosis: clinical and treatment characteristics . J Int Med agement of superfi cial venous thrombosis with calcium Res . 2004 ; 32 ( 4 ): 429 - 435 . nadroparin. Dosage testing and comparison with a non- 174 . Kovacs MJ , Kahn SR , Rodger M , et al . A pilot study of cen- steroidal anti-infl ammatory agent]. Ann Cardiol Angeiol tral venous catheter survival in cancer patients using low (Paris) . 1994 ; 43 ( 3 ): 160 - 166 . molecular weight heparin (dalteparin) and warfarin without 165 . Prandoni P , Tormene D , Pesavento R ; Vesalio Investigators catheter removal for the treatment of upper extremity deep Group . High vs. low doses of low-molecular-weight hep- vein thrombosis (the catheter study) . J Thromb Haemost . arin for the treatment of superfi cial vein thrombosis of the 2007 ; 5 ( 8 ): 1650 - 1653 . legs: a double-blind, randomized trial . J Thromb Haemost . 175 . Horne MK III , Mayo DJ , Cannon RO III , Chen CC , 2005 ; 3 ( 6 ): 1152 - 1157 . Shawker TH , Chang R . Intraclot recombinant tissue plas- 166 . Marchiori A , Verlato F , Sabbion P , et al . High versus low minogen activator in the treatment of deep venous throm- doses of unfractionated heparin for the treatment of super- bosis of the lower and upper extremities . Am J Med . 2000 ; fi cial thrombophlebitis of the leg. A prospective, controlled, 108 ( 3 ): 251 - 255 . randomized study . Haematologica . 2002 ; 87 ( 5 ): 523 - 527 . 176 . Lee JT , Karwowski JK , Harris EJ , Haukoos JS , Olcott C IV 167 . Belcaro G , Nicolaides AN , Errichi BM , et al . Superfi cial Long-term thrombotic recurrence after nonoperative man- thrombophlebitis of the legs: a randomized, controlled, agement of Paget-Schroetter syndrome . J Vasc Surg . 2006 ; follow-up study . Angiology . 1999 ; 50 ( 7 ): 523 - 529 . 43 ( 6 ): 1236 - 1243 . 168 . Lozano FS , Almazan A . Low-molecular-weight heparin 177 . Martinelli I , Battaglioli T , Bucciarelli P , Passamonti SM , versus saphenofemoral disconnection for the treatment of Mannucci PM . Risk factors and recurrence rate of primary above-knee greater saphenous thrombophlebitis: a prospec- deep vein thrombosis of the upper extremities . Circulation . tive study . Vasc Endovascular Surg . 2003 ; 37 ( 6 ): 415 - 420 . 2004 ; 110 ( 5 ): 566 - 570 . 169 . Sullivan V , Denk PM , Sonnad SS , Eagleton MJ , Wakefi eld TW . 178 . Prandoni P , Bernardi E , Marchiori A , et al . The long Ligation versus anticoagulation: treatment of above-knee term clinical course of acute deep vein thrombosis of superfi cial thrombophlebitis not involving the deep venous the arm: prospective cohort study. BMJ . 2 0 0 4 ; 3 2 9 ( 7 4 6 4 ) : system . J Am Coll Surg . 2001 ; 193 ( 5 ): 556 - 562 . 484 - 485 .

91 © 2012 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest PhysiciansDownloaded (http://www.chestpubs.org/site/misc/reprints.xhtml). from chestjournal.chestpubs.org at ACCP DOI: HQ 10.1378/chest.11-2301 on February 9, 2012 © 2012 American College of Chest Physicians Antithrombotic Therapy for VTE Disease : Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Clive Kearon, Elie A. Akl, Anthony J. Comerota, Paolo Prandoni, Henri Bounameaux, Samuel Z. Goldhaber, Michael E. Nelson, Philip S. Wells, Michael K. Gould, Francesco Dentali, Mark Crowther and Susan R. Kahn Chest 2012;141; e419S-e494S DOI 10.1378/chest.11-2301

This information is current as of February 9, 2012

Supplementary Material View e-supplements related to this article at: http://chestjournal.chestpubs.org/content/suppl/2012/02/03/141.2_suppl.e419S.DC1.html Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/141/2_suppl/e419S.full.html References This article cites 441 articles, 172 of which can be accessed free at: http://chestjournal.chestpubs.org/content/141/2_suppl/e419S.full.html#ref-list-1 Cited Bys This article has been cited by 4 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/141/2_suppl/e419S.full.html#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Selected Abstracts

Abstract 1 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Samantha MacLean, Sohail Mulla, Elie A. Akl, Milosz Jankowski, Per Olav Vandvik, Shanil Ebrahim, Shelley McLeod, Neera Bhatnagar, and Gordon H. Guyatt Patient Values and Preferences in Decision Making for Antithrombotic Therapy: A Systematic Review:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e1S-e23S; doi:10.1378/chest.11-2290 Abstract Full Text Full Text (PDF) Abstract 2 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: David A. Garcia, Trevor P. Baglin, Jeffrey I. Weitz, and Meyer Michel Samama Parenteral Anticoagulants:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e24S-e43S; doi:10.1378/chest.11-2291 Abstract Full Text Full Text (PDF) Abstract 3 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Walter Ageno, Alexander S. Gallus, Ann Wittkowsky, Mark Crowther, Elaine M. Hylek, and Gualtiero Palareti Oral Anticoagulant Therapy:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e44S-e88S; doi:10.1378/chest.11-2292 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 4 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: John W. Eikelboom, Jack Hirsh, Frederick A. Spencer, Trevor P. Baglin, and Jeffrey I. Weitz Antiplatelet Drugs:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e89S-e119S; doi:10.1378/chest.11-2293 Abstract Full Text Full Text (PDF) Abstract 5 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Jeffrey I. Weitz, John W. Eikelboom, and Meyer Michel Samama New Antithrombotic Drugs:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e120S-e151S; doi:10.1378/chest.11-2294 Abstract Full Text Full Text (PDF) Abstract 6 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Anne Holbrook, Sam Schulman, Daniel M. Witt, Per Olav Vandvik, Jason Fish, Michael J. Kovacs, Peter J. Svensson, David L. Veenstra, Mark Crowther, and Gordon H. Guyatt Evidence-Based Management of Anticoagulant Therapy:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e152S-e184S; doi:10.1378/chest.11-2295 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 7 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Gordon H. Guyatt, John W. Eikelboom, Michael K. Gould, David A. Garcia, Mark Crowther, M. Hassan Murad, Susan R. Kahn, Yngve Falck-Ytter, Charles W. Francis, Maarten G. Lansberg, Elie A. Akl, and Jack Hirsh Approach to Outcome Measurement in the Prevention of Thrombosis in Surgical and Medical Patients:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e185S-e194S; doi:10.1378/chest.11-2289 Abstract Full Text Full Text (PDF) Abstract 8 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES::

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 1 di 19 Selected Abstracts 07/02/12 22:14

Susan R. Kahn, Wendy Lim, Andrew S. Dunn, Mary Cushman, Francesco Dentali, Elie A. Akl, Deborah J. Cook, Alex A. Balekian, Russell C. Klein, Hoang Le, Sam Schulman, and M. Hassan Murad Prevention of VTE in Nonsurgical Patients:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e195S-e226S; doi:10.1378/chest.11-2296 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 9 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Michael K. Gould, David A. Garcia, Sherry M. Wren, Paul J. Karanicolas, Juan I. Arcelus, John A. Heit, and Charles M. Samama Prevention of VTE in Nonorthopedic Surgical Patients:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e227S-e277S; doi:10.1378/chest.11-2297 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 10 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Yngve Falck-Ytter, Charles W. Francis, Norman A. Johanson, Catherine Curley, Ola E. Dahl, Sam Schulman, Thomas L. Ortel, Stephen G. Pauker, and Clifford W. Colwell, Jr Prevention of VTE in Orthopedic Surgery Patients:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e278S-e325S; doi:10.1378/chest.11-2404 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 11 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: James D. Douketis, Alex C. Spyropoulos, Frederick A. Spencer, Michael Mayr, Amir K. Jaffer, Mark H. Eckman, Andrew S. Dunn, and Regina Kunz Perioperative Management of Antithrombotic Therapy:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e326S-e350S; doi:10.1378/chest.11-2298 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 12 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Shannon M. Bates, Roman Jaeschke, Scott M. Stevens, Steven Goodacre, Philip S. Wells, Matthew D. Stevenson, Clive Kearon, Holger J. Schunemann, Mark Crowther, Stephen G. Pauker, Regina Makdissi, and Gordon H. Guyatt Diagnosis of DVT:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e351S-e418S; doi:10.1378/chest.11-2299 Abstract Full Text Full Text (PDF) Online Data Supplements Abstract 13 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Clive Kearon, Elie A. Akl, Anthony J. Comerota, Paolo Prandoni, Henri Bounameaux, Samuel Z. Goldhaber, Michael E. Nelson, Philip S. Wells, Michael K. Gould, Francesco Dentali, Mark Crowther, and Susan R. Kahn Antithrombotic Therapy for VTE Disease:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e419S-e494S; doi:10.1378/chest.11-2301 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 14 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Lori-Ann Linkins, Antonio L. Dans, COL Lisa K. Moores, Robert Bona, Bruce L. Davidson, Sam Schulman, and Mark Crowther Treatment and Prevention of Heparin-Induced Thrombocytopenia:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e495S-e530S; doi:10.1378/chest.11-2303 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 15 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: John J. You, Daniel E. Singer, Patricia A. Howard, Deirdre A. Lane, Mark H. Eckman, Margaret C. Fang, Elaine M. Hylek, Sam Schulman, Alan S. Go, Michael Hughes, Frederick A. Spencer, Warren J. Manning, Jonathan L. Halperin, and Gregory Y. H. Lip Antithrombotic Therapy for Atrial Fibrillation:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e531S-e575S; doi:10.1378/chest.11-2304 Abstract Full Text Full Text (PDF) Online Data Supplement http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 2 di 19 Selected Abstracts 07/02/12 22:14

Abstract 16 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Richard P. Whitlock, Jack C. Sun, Stephen E. Fremes, Fraser D. Rubens, and Kevin H. Teoh Antithrombotic and Thrombolytic Therapy for Valvular Disease:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e576S-e600S; doi:10.1378/chest.11-2305 Abstract Full Text Full Text (PDF) Online Data Supplements Abstract 17 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Maarten G. Lansberg, Martin J. O’Donnell, Pooja Khatri, Eddy S. Lang, Mai N. Nguyen-Huynh, Neil E. Schwartz, Frank A. Sonnenberg, Sam Schulman, Per Olav Vandvik, Frederick A. Spencer, Pablo Alonso-Coello, Gordon H. Guyatt, and Elie A. Akl Antithrombotic and Thrombolytic Therapy for Ischemic Stroke:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e601S-e636S; doi:10.1378/chest.11-2302 Abstract Full Text Full Text (PDF) Online Data Supplements Abstract 18 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Per Olav Vandvik, A. Michael Lincoff, Joel M. Gore, David D. Gutterman, Frank A. Sonnenberg, Pablo Alonso-Coello, Elie A. Akl, Maarten G. Lansberg, Gordon H. Guyatt, and Frederick A. Spencer Primary and Secondary Prevention of Cardiovascular Disease:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e637S-e668S; doi:10.1378/chest.11-2306 Abstract Full Text Full Text (PDF) Online Data Supplement Abstract 19 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Pablo Alonso-Coello, Sergi Bellmunt, Catherine McGorrian, Sonia S. Anand, Randolph Guzman, Michael H. Criqui, Elie A. Akl, Per Olav Vandvik, Maarten G. Lansberg, Gordon H. Guyatt, and Frederick A. Spencer Antithrombotic Therapy in Peripheral Artery Disease:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e669S-e690S; doi:10.1378/chest.11-2307 Abstract Full Text Full Text (PDF) Online Data Supplements Abstract 20 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Shannon M. Bates, Ian A. Greer, Saskia Middeldorp, David L. Veenstra, Anne-Marie Prabulos, and Per Olav Vandvik VTE, Thrombophilia, Antithrombotic Therapy, and Pregnancy:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e691S-e736S; doi:10.1378/chest.11-2300 Abstract Full Text Full Text (PDF) Online Data Supplements Abstract 21 ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE- BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES:: Paul Monagle, Anthony K. C. Chan, Neil A. Goldenberg, Rebecca N. Ichord, Janna M. Journeycake, Ulrike Nowak-Göttl, and Sara K. Vesely Antithrombotic Therapy in Neonates and Children:: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines Chest February 2012 141:2 suppl e737S-e801S; doi:10.1378/chest.11-2308 Abstract Full Text Full Text (PDF) Online Data Supplements

Abstract 1 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Patient Values and Preferences in Decision Making for Antithrombotic Therapy: A Systematic Review Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: Development of clinical practice guidelines involves making trade-offs between desirable and undesirable consequences of alternative management strategies. Although the relative value of health states to patients should provide the basis for these trade-offs, few guidelines have systematically summarized the relevant evidence. We conducted a systematic review relating to values and preferences of patients considering antithrombotic therapy. http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 3 di 19 Selected Abstracts 07/02/12 22:14

Methods: We included studies examining patient preferences for alternative approaches to antithrombotic prophylaxis and studies that examined, in the context of antithrombotic prophylaxis or treatment, how patients value alternative health states and experiences with treatment. We conducted a systematic search and compiled structured summaries of the results. Steps in the process that involved judgment were conducted in duplicate.

Results: We identified 48 eligible studies. Sixteen dealt with atrial fibrillation, five with VTE, four with stroke or myocardial infarction prophylaxis, six with thrombolysis in acute stroke or myocardial infarction, and 17 with burden of antithrombotic treatment.

Conclusion: Patient values and preferences regarding thromboprophylaxis treatment appear to be highly variable. Participant responses may depend on their prior experience with the treatments or health outcomes considered as well as on the methods used for preference elicitation. It should be standard for clinical practice guidelines to conduct systematic reviews of patient values and preferences in the specific content area.

Accepted August 31, 2011.

Full Text

Abstract 2 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Parenteral Anticoagulants Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

This article describes the pharmacology of approved parenteral anticoagulants. These include the indirect anticoagulants, unfractionated heparin (UFH), low-molecular-weight heparins (LMWHs), fondaparinux, and danaparoid, as well as the direct thrombin inhibitors hirudin, bivalirudin, and argatroban. UFH is a heterogeneous mixture of glycosaminoglycans that bind to antithrombin via a unique pentasaccharide sequence and catalyze the inactivation of thrombin, factor Xa, and other clotting enzymes. Heparin also binds to cells and plasma proteins other than antithrombin causing unpredictable pharmacokinetic and pharmacodynamic properties and triggering nonhemorrhagic side effects, such as heparin-induced thrombocytopenia (HIT) and osteoporosis. LMWHs have greater inhibitory activity against factor Xa than thrombin and exhibit less binding to cells and plasma proteins than heparin. Consequently, LMWH preparations have more predictable pharmacokinetic and pharmacodynamic properties, have a longer half-life than heparin, and are associated with a lower risk of nonhemorrhagic side effects. LMWHs can be administered once daily or bid by subcutaneous injection, without coagulation monitoring. Based on their greater convenience, LMWHs have replaced UFH for many clinical indications. Fondaparinux, a synthetic pentasaccharide, catalyzes the inhibition of factor Xa, but not thrombin, in an antithrombin-dependent fashion. Fondaparinux binds only to antithrombin. Therefore, fondaparinux-associated HIT or osteoporosis is unlikely to occur. Fondaparinux exhibits complete bioavailability when administered subcutaneously, has a longer half-life than LMWHs, and is given once daily by subcutaneous injection in fixed doses, without coagulation monitoring. Three additional parenteral direct thrombin inhibitors and danaparoid are approved as alternatives to heparin in patients with HIT.

Accepted August 31, 2011.

Full Text

Abstract 3 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Oral Anticoagulant Therapy Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: The objective of this article is to summarize the published literature concerning the pharmacokinetics and pharmacodynamics of oral anticoagulant drugs that are currently available for clinical use and other aspects related to their management.

Methods: We carried out a standard review of published articles focusing on the laboratory and clinical characteristics of the vitamin K antagonists; the direct thrombin inhibitor, dabigatran etexilate; and the direct factor Xa inhibitor, rivaroxaban

Results: The antithrombotic effect of each oral anticoagulant drug, the interactions, and the monitoring of anticoagulation intensity are described in detail and discussed without providing specific recommendations. Moreover, we describe and discuss the clinical applications and optimal dosages of oral anticoagulant therapies, practical issues related to their initiation and monitoring, adverse events such as bleeding and other potential side effects, and available strategies for reversal.

Conclusions: There is a large amount of evidence on laboratory and clinical characteristics of vitamin K antagonists. A growing body of evidence is becoming available on the first new oral anticoagulant drugs available for clinical use, dabigatran and rivaroxaban.

Accepted August 31, 2011.

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 4 di 19 Selected Abstracts 07/02/12 22:14

Full Text

Abstract 4 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antiplatelet Drugs Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

The article describes the mechanisms of action, pharmacokinetics, and pharmacodynamics of aspirin, dipyridamole, cilostazol, the thienopyridines, and the glycoprotein IIb/IIIa antagonists. The relationships among dose, efficacy, and safety are discussed along with a mechanistic overview of results of randomized clinical trials. The article does not provide specific management recommendations but highlights important practical aspects of antiplatelet therapy, including optimal dosing, the variable balance between benefits and risks when antiplatelet therapies are used alone or in combination with other antiplatelet drugs in different clinical settings, and the implications of persistently high platelet reactivity despite such treatment.

Accepted August 31, 2011.

Full Text

Abstract 5 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES New Antithrombotic Drugs Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

This article focuses on new antithrombotic drugs that are in or are entering phase 3 clinical testing. Development of these new agents was prompted by the limitations of existing antiplatelet, anticoagulant, or fibrinolytic drugs. Addressing these unmet needs, this article (1) outlines the rationale for development of new antithrombotic agents; (2) describes the new antiplatelet, anticoagulant, and fibrinolytic drugs; and (3) provides clinical perspectives on the opportunities and challenges faced by these novel agents.

Accepted August 31, 2011.

Full Text

Abstract 6 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Evidence-Based Management of Anticoagulant Therapy Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: High-quality anticoagulation management is required to keep these narrow therapeutic index medications as effective and safe as possible. This article focuses on the common important management questions for which, at a minimum, low-quality published evidence is available to guide best practices.

Methods: The methods of this guideline follow those described in Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

Results: Most practical clinical questions regarding the management of anticoagulation, both oral and parenteral, have not been adequately addressed by randomized trials. We found sufficient evidence for summaries of recommendations for 23 questions, of which only two are strong rather than weak recommendations. Strong recommendations include targeting an international normalized ratio of 2.0 to 3.0 for patients on vitamin K antagonist therapy (Grade 1B) and not routinely using pharmacogenetic testing for guiding doses of vitamin K antagonist (Grade 1B). Weak recommendations deal with such issues as loading doses, initiation overlap, monitoring frequency, vitamin K supplementation, patient self-management, weight and renal function adjustment of doses, dosing decision support, drug interactions to avoid, and prevention and management of bleeding complications. We also address anticoagulation management services and intensive patient education.

Conclusions: We offer guidance for many common anticoagulation-related management problems. Most anticoagulation management questions have

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 5 di 19 Selected Abstracts 07/02/12 22:14 not been adequately studied.

Accepted August 31, 2011.

Full Text

Abstract 7 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Approach to Outcome Measurement in the Prevention of Thrombosis in Surgical and Medical Patients Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

This article provides the rationale for the approach to making recommendations primarily used in four articles of the Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines: orthopedic surgery, nonorthopedic surgery, nonsurgical patients, and stroke. Some of the early clinical trials of antithrombotic prophylaxis with a placebo or no treatment group used symptomatic VTE and fatal PE to measure efficacy of the treatment. These trials suggest a benefit of thromboprophylaxis in reducing fatal PE. In contrast, most of the recent clinical trials comparing the efficacy of alternative anticoagulants used a surrogate outcome, asymptomatic DVT detected at mandatory venography. This outcome is fundamentally unsatisfactory because it does not allow a trade-off with serious bleeding; that trade-off requires knowledge of the number of symptomatic events that thromboprophylaxis prevents. In this article, we review the merits and limitations of four approaches to estimating reduction in symptomatic thrombosis: (1) direct measurement of symptomatic thrombosis, (2) use of asymptomatic events for relative risks and symptomatic events from randomized controlled trials for baseline risk, (3) use of baseline risk estimates from studies that did not perform surveillance and relative effect from asymptomatic events in randomized controlled trials, and (4) use of available data to estimate the proportion of asymptomatic events that will become symptomatic. All approaches have their limitations. The optimal choice of approach depends on the nature of the evidence available.

Accepted August 31, 2011.

Full Text

Abstract 8 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Prevention of VTE in Nonsurgical Patients Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: This guideline addressed VTE prevention in hospitalized medical patients, outpatients with cancer, the chronically immobilized, long- distance travelers, and those with asymptomatic thrombophilia.

Methods: This guideline follows methods described in Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

Results: For acutely ill hospitalized medical patients at increased risk of thrombosis, we recommend anticoagulant thromboprophylaxis with low- molecular-weight heparin (LMWH), low-dose unfractionated heparin (LDUH) bid, LDUH tid, or fondaparinux (Grade 1B) and suggest against extending the duration of thromboprophylaxis beyond the period of patient immobilization or acute hospital stay (Grade 2B). For acutely ill hospitalized medical patients at low risk of thrombosis, we recommend against the use of pharmacologic prophylaxis or mechanical prophylaxis (Grade 1B). For acutely ill hospitalized medical patients at increased risk of thrombosis who are bleeding or are at high risk for major bleeding, we suggest mechanical thromboprophylaxis with graduated compression stockings (GCS) (Grade 2C) or intermittent pneumatic compression (IPC) (Grade 2C). For critically ill patients, we suggest using LMWH or LDUH thromboprophylaxis (Grade 2C). For critically ill patients who are bleeding or are at high risk for major bleeding, we suggest mechanical thromboprophylaxis with GCS and/or IPC at least until the bleeding risk decreases (Grade 2C). In outpatients with cancer who have no additional risk factors for VTE we suggest against routine prophylaxis with LMWH or LDUH (Grade 2B) and recommend against the prophylactic use of vitamin K antagonists (Grade 1B).

Conclusions: Decisions regarding prophylaxis in nonsurgical patients should be made after consideration of risk factors for both thrombosis and bleeding, clinical context, and patients’ values and preferences.

Accepted August 31, 2011.

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 6 di 19 Selected Abstracts 07/02/12 22:14

Full Text

Abstract 9 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Prevention of VTE in Nonorthopedic Surgical Patients Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: VTE is a common cause of preventable death in surgical patients.

Methods: We developed recommendations for thromboprophylaxis in nonorthopedic surgical patients by using systematic methods as described in Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines. Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

Results: We describe several alternatives for stratifying the risk of VTE in general and abdominal-pelvic surgical patients. When the risk for VTE is very low (< 0.5%), we recommend that no specific pharmacologic (Grade 1B) or mechanical (Grade 2C) prophylaxis be used other than early ambulation. For patients at low risk for VTE (∼1.5%), we suggest mechanical prophylaxis, preferably with intermittent pneumatic compression (IPC), over no prophylaxis (Grade 2C). For patients at moderate risk for VTE (∼3%) who are not at high risk for major bleeding complications, we suggest low-molecular-weight heparin (LMWH) (Grade 2B), low-dose unfractionated heparin (Grade 2B), or mechanical prophylaxis with IPC (Grade 2C) over no prophylaxis. For patients at high risk for VTE (∼6%) who are not at high risk for major bleeding complications, we recommend pharmacologic prophylaxis with LMWH (Grade 1B) or low-dose unfractionated heparin (Grade 1B) over no prophylaxis. In these patients, we suggest adding mechanical prophylaxis with elastic stockings or IPC to pharmacologic prophylaxis (Grade 2C). For patients at high risk for VTE undergoing abdominal or pelvic surgery for cancer, we recommend extended-duration, postoperative, pharmacologic prophylaxis (4 weeks) with LMWH over limited-duration prophylaxis (Grade 1B). For patients at moderate to high risk for VTE who are at high risk for major bleeding complications or those in whom the consequences of bleeding are believed to be particularly severe, we suggest use of mechanical prophylaxis, preferably with IPC, over no prophylaxis until the risk of bleeding diminishes and pharmacologic prophylaxis may be initiated (Grade 2C). For patients in all risk groups, we suggest that an inferior vena cava filter not be used for primary VTE prevention (Grade 2C) and that surveillance with venous compression ultrasonography should not be performed (Grade 2C). We developed similar recommendations for other nonorthopedic surgical populations.

Conclusions: Optimal thromboprophylaxis in nonorthopedic surgical patients will consider the risks of VTE and bleeding complications as well as the values and preferences of individual patients.

Accepted August 31, 2011.

Full Text

Abstract 10 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Prevention of VTE in Orthopedic Surgery Patients Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: VTE is a serious, but decreasing complication following major orthopedic surgery. This guideline focuses on optimal prophylaxis to reduce postoperative pulmonary embolism and DVT.

Results: In patients undergoing major orthopedic surgery, we recommend the use of one of the following rather than no antithrombotic prophylaxis: low-molecular-weight heparin; fondaparinux; dabigatran, apixaban, rivaroxaban (total hip arthroplasty or total knee arthroplasty but not hip fracture surgery); low-dose unfractionated heparin; adjusted-dose vitamin K antagonist; aspirin (all Grade 1B); or an intermittent pneumatic compression device (IPCD) (Grade 1C) for a minimum of 10 to 14 days. We suggest the use of low-molecular-weight heparin in preference to the other agents we have recommended as alternatives (Grade 2C/2B), and in patients receiving pharmacologic prophylaxis, we suggest adding an IPCD during the hospital stay (Grade 2C). We suggest extending thromboprophylaxis for up to 35 days (Grade 2B). In patients at increased bleeding risk, we suggest an IPCD or no prophylaxis (Grade 2C). In patients who decline injections, we recommend using apixaban or dabigatran (all Grade 1B). We suggest against using inferior vena cava filter placement for primary prevention in patients with contraindications to both pharmacologic and mechanical thromboprophylaxis (Grade 2C). We recommend against Doppler (or duplex) ultrasonography screening before hospital discharge (Grade 1B). For patients with isolated lower-extremity injuries requiring leg immobilization, we suggest no thromboprophylaxis (Grade 2B). For patients undergoing knee arthroscopy without a history of VTE, we suggest no thromboprophylaxis (Grade 2B).

Conclusions: Optimal strategies for thromboprophylaxis after major orthopedic surgery include pharmacologic and mechanical approaches. http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 7 di 19 Selected Abstracts 07/02/12 22:14

Accepted August 31, 2011.

Full Text

Abstract 11 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Perioperative Management of Antithrombotic Therapy Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: This guideline addresses the management of patients who are receiving anticoagulant or antiplatelet therapy and require an elective surgery or procedure.

Methods: The methods herein follow those discussed in the Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines. Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines article of this supplement.

Results: In patients requiring vitamin K antagonist (VKA) interruption before surgery, we recommend stopping VKAs 5 days before surgery instead of a shorter time before surgery (Grade 1B). In patients with a mechanical heart valve, atrial fibrillation, or VTE at high risk for thromboembolism, we suggest bridging anticoagulation instead of no bridging during VKA interruption (Grade 2C); in patients at low risk, we suggest no bridging instead of bridging (Grade 2C). In patients who require a dental procedure, we suggest continuing VKAs with an oral prohemostatic agent or stopping VKAs 2 to 3 days before the procedure instead of alternative strategies (Grade 2C). In moderate- to high-risk patients who are receiving acetylsalicylic acid (ASA) and require noncardiac surgery, we suggest continuing ASA around the time of surgery instead of stopping ASA 7 to 10 days before surgery (Grade 2C). In patients with a coronary stent who require surgery, we recommend deferring surgery > 6 weeks after bare-metal stent placement and > 6 months after drug-eluting stent placement instead of undertaking surgery within these time periods (Grade 1C); in patients requiring surgery within 6 weeks of bare-metal stent placement or within 6 months of drug-eluting stent placement, we suggest continuing antiplatelet therapy perioperatively instead of stopping therapy 7 to 10 days before surgery (Grade 2C).

Conclusions: Perioperative antithrombotic management is based on risk assessment for thromboembolism and bleeding, and recommended approaches aim to simplify patient management and minimize adverse clinical outcomes.

Accepted August 31, 2011.

Full Text

Abstract 12 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Diagnosis of DVT Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: Objective testing for DVT is crucial because clinical assessment alone is unreliable and the consequences of misdiagnosis are serious. This guideline focuses on the identification of optimal strategies for the diagnosis of DVT in ambulatory adults.

Results: We suggest that clinical assessment of pretest probability of DVT, rather than performing the same tests in all patients, should guide the diagnostic process for a first lower extremity DVT (Grade 2B). In patients with a low pretest probability of first lower extremity DVT, we recommend initial testing with D-dimer or ultrasound (US) of the proximal veins over no diagnostic testing (Grade 1B), venography (Grade 1B), or whole-leg US (Grade 2B). In patients with moderate pretest probability, we recommend initial testing with a highly sensitive D-dimer, proximal compression US, or whole-leg US rather than no testing (Grade 1B) or venography (Grade 1B). In patients with a high pretest probability, we recommend proximal compression or whole-leg US over no testing (Grade 1B) or venography (Grade 1B).

Conclusions: Favored strategies for diagnosis of first DVT combine use of pretest probability assessment, D-dimer, and US. There is lower-quality evidence available to guide diagnosis of recurrent DVT, upper extremity DVT, and DVT during pregnancy.

Accepted August 31, 2011.

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 8 di 19 Selected Abstracts 07/02/12 22:14

Full Text

Abstract 13 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic Therapy for VTE Disease Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: This article addresses the treatment of VTE disease.

Methods: We generated strong (Grade 1) and weak (Grade 2) recommendations based on high-quality (Grade A), moderate-quality (Grade B), and low- quality (Grade C) evidence.

Results: For acute DVT or pulmonary embolism (PE), we recommend initial parenteral anticoagulant therapy (Grade 1B) or anticoagulation with rivaroxaban. We suggest low-molecular-weight heparin (LMWH) or fondaparinux over IV unfractionated heparin (Grade 2C) or subcutaneous unfractionated heparin (Grade 2B). We suggest thrombolytic therapy for PE with hypotension (Grade 2C). For proximal DVT or PE, we recommend treatment of 3 months over shorter periods (Grade 1B). For a first proximal DVT or PE that is provoked by surgery or by a nonsurgical transient risk factor, we recommend 3 months of therapy (Grade 1B; Grade 2B if provoked by a nonsurgical risk factor and low or moderate bleeding risk); that is unprovoked, we suggest extended therapy if bleeding risk is low or moderate (Grade 2B) and recommend 3 months of therapy if bleeding risk is high (Grade 1B); and that is associated with active cancer, we recommend extended therapy (Grade 1B; Grade 2B if high bleeding risk) and suggest LMWH over vitamin K antagonists (Grade 2B). We suggest vitamin K antagonists or LMWH over dabigatran or rivaroxaban (Grade 2B). We suggest compression stockings to prevent the postthrombotic syndrome (Grade 2B). For extensive superficial vein thrombosis, we suggest prophylactic-dose fondaparinux or LMWH over no anticoagulation (Grade 2B), and suggest fondaparinux over LMWH (Grade 2C).

Conclusion: Strong recommendations apply to most patients, whereas weak recommendations are sensitive to differences among patients, including their preferences.

Accepted August 31, 2011.

Full Text

Abstract 14 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Treatment and Prevention of Heparin-Induced Thrombocytopenia Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: Heparin-induced thrombocytopenia (HIT) is an antibody-mediated adverse drug reaction that can lead to devastating thromboembolic complications, including pulmonary embolism, ischemic limb necrosis necessitating limb amputation, acute myocardial infarction, and stroke.

Methods: The methods of this guideline follow the Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

Results: Among the key recommendations for this article are the following: For patients receiving heparin in whom clinicians consider the risk of HIT to be > 1%, we suggest that platelet count monitoring be performed every 2 or 3 days from day 4 to day 14 (or until heparin is stopped, whichever occurs first) (Grade 2C). For patients receiving heparin in whom clinicians consider the risk of HIT to be < 1%, we suggest that platelet counts not be monitored (Grade 2C). In patients with HIT with thrombosis (HITT) or isolated HIT who have normal renal function, we suggest the use of argatroban or lepirudin or danaparoid over other nonheparin anticoagulants (Grade 2C). In patients with HITT and renal insufficiency, we suggest the use of argatroban over other nonheparin anticoagulants (Grade 2C). In patients with acute HIT or subacute HIT who require urgent cardiac surgery, we suggest the use of bivalirudin over other nonheparin anticoagulants or heparin plus antiplatelet agents (Grade 2C).

Conclusions: Further studies evaluating the role of fondaparinux and the new oral anticoagulants in the treatment of HIT are needed.

Accepted August 31, 2011.

Full Text

Abstract 15 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_supp…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 9 di 19 Selected Abstracts 07/02/12 22:14

CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic Therapy for Atrial Fibrillation Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: The risk of stroke varies considerably across different groups of patients with atrial fibrillation (AF). Antithrombotic prophylaxis for stroke is associated with an increased risk of bleeding. We provide recommendations for antithrombotic treatment based on net clinical benefit for patients with AF at varying levels of stroke risk and in a number of common clinical scenarios.

Methods: We used the methods described in the Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines article of this supplement.

Results: For patients with nonrheumatic AF, including those with paroxysmal AF, who are (1) at low risk of stroke (eg, CHADS2 [congestive heart failure, hypertension, age ≥ 75 years, diabetes mellitus, prior stroke or transient ischemic attack] score of 0), we suggest no therapy rather than antithrombotic therapy, and for patients choosing antithrombotic therapy, we suggest aspirin rather than oral anticoagulation or combination therapy with aspirin and clopidogrel; (2) at intermediate risk of stroke (eg, CHADS2 score of 1), we recommend oral anticoagulation rather than no therapy, and we suggest oral anticoagulation rather than aspirin or combination therapy with aspirin and clopidogrel; and (3) at high risk of stroke (eg, CHADS2 score of ≥ 2), we recommend oral anticoagulation rather than no therapy, aspirin, or combination therapy with aspirin and clopidogrel. Where we recommend or suggest in favor of oral anticoagulation, we suggest dabigatran 150 mg bid rather than adjusted-dose vitamin K antagonist therapy.

Conclusions: Oral anticoagulation is the optimal choice of antithrombotic therapy for patients with AF at high risk of stroke (CHADS2 score of ≥ 2). At lower levels of stroke risk, antithrombotic treatment decisions will require a more individualized approach.

Accepted August 31, 2011.

Full Text

Abstract 16 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic and Thrombolytic Therapy for Valvular Disease Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: Antithrombotic therapy in valvular disease is important to mitigate thromboembolism, but the hemorrhagic risk imposed must be considered.

Methods: The methods of this guideline follow those described in Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines. Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

Results: In rheumatic mitral disease, we recommend vitamin K antagonist (VKA) therapy when the left atrial diameter is > 55 mm (Grade 2C) or when complicated by left atrial thrombus (Grade 1A). In candidates for percutaneous mitral valvotomy with left atrial thrombus, we recommend VKA therapy until thrombus resolution, and we recommend abandoning valvotomy if the thrombus fails to resolve (Grade 1A). In patients with patent foramen ovale (PFO) and stroke or transient ischemic attack, we recommend initial aspirin therapy (Grade 1B) and suggest substitution of VKA if recurrence (Grade 2C). In patients with cryptogenic stroke and DVT and a PFO, we recommend VKA therapy for 3 months (Grade 1B) and consideration of PFO closure (Grade 2C). We recommend against the use of anticoagulant (Grade 1C) and antiplatelet therapy (Grade 1B) for native valve endocarditis. We suggest holding VKA therapy until the patient is stabilized without neurologic complications for infective endocarditis of a prosthetic valve (Grade 2C). In the first 3 months after bioprosthetic valve implantation, we recommend aspirin for aortic valves (Grade 2C), the addition of clopidogrel to aspirin if the aortic valve is transcatheter (Grade 2C), and VKA therapy with a target international normalized ratio (INR) of 2.5 for mitral valves (Grade 2C). After 3 months, we suggest aspirin therapy (Grade 2C). We recommend early bridging of mechanical valve patients to VKA therapy with unfractionated heparin (DVT dosing) or low-molecular-weight heparin (Grade 2C). We recommend long-term VKA therapy for all mechanical valves (Grade 1B): target INR 2.5 for aortic (Grade 1B) and 3.0 for mitral or double valve (Grade 2C). In patients with mechanical valves at low bleeding risk, we suggest the addition of low- dose aspirin (50-100 mg/d) (Grade 1B). In valve repair patients, we suggest aspirin therapy (Grade 2C). In patients with thrombosed prosthetic valve, we recommend fibrinolysis for right-sided valves and left-sided valves with thrombus area < 0.8 cm2 (Grade 2C). For patients with left-sided prosthetic valve thrombosis and thrombus area ≥ 0.8 cm2, we recommend early surgery (Grade 2C).

Conclusions: These antithrombotic guidelines provide recommendations based on the optimal balance of thrombotic and hemorrhagic risk.

Accepted August 31, 2011.

Full Text

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 10 di 19 Selected Abstracts 07/02/12 22:14

Abstract 17 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic and Thrombolytic Therapy for Ischemic Stroke Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Objectives: This article provides recommendations on the use of antithrombotic therapy in patients with stroke or transient ischemic attack (TIA).

Methods: We generated treatment recommendations (Grade 1) and suggestions (Grade 2) based on high (A), moderate (B), and low (C) quality evidence.

Results: In patients with acute ischemic stroke, we recommend IV recombinant tissue plasminogen activator (r-tPA) if treatment can be initiated within 3 h (Grade 1A) or 4.5 h (Grade 2C) of symptom onset; we suggest intraarterial r-tPA in patients ineligible for IV tPA if treatment can be initiated within 6 h (Grade 2C); we suggest against the use of mechanical thrombectomy (Grade 2C) although carefully selected patients may choose this intervention; and we recommend early aspirin therapy at a dose of 160 to 325 mg (Grade 1A). In patients with acute stroke and restricted mobility, we suggest the use of prophylactic-dose heparin or intermittent pneumatic compression devices (Grade 2B) and suggest against the use of elastic compression stockings (Grade 2B). In patients with a history of noncardioembolic ischemic stroke or TIA, we recommend long-term treatment with aspirin (75-100 mg once daily), clopidogrel (75 mg once daily), aspirin/extended release dipyridamole (25 mg/200 mg bid), or cilostazol (100 mg bid) over no antiplatelet therapy (Grade 1A), oral anticoagulants (Grade 1B), the combination of clopidogrel plus aspirin (Grade 1B), or triflusal (Grade 2B). Of the recommended antiplatelet regimens, we suggest clopidogrel or aspirin/extended-release dipyridamole over aspirin (Grade 2B) or cilostazol (Grade 2C). In patients with a history of stroke or TIA and atrial fibrillation we recommend oral anticoagulation over no antithrombotic therapy, aspirin, and combination therapy with aspirin and clopidogrel (Grade 1B).

Conclusions: These recommendations can help clinicians make evidence-based treatment decisions with their patients who have had strokes.

Accepted August 31, 2011.

Full Text

Abstract 18 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Primary and Secondary Prevention of Cardiovascular Disease Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: This guideline focuses on long-term administration of antithrombotic drugs designed for primary and secondary prevention of cardiovascular disease, including two new antiplatelet therapies.

Results: We present 23 recommendations for pertinent clinical questions. For primary prevention of cardiovascular disease, we suggest low-dose aspirin (75-100 mg/d) in patients aged > 50 years over no aspirin therapy (Grade 2B). For patients with established coronary artery disease, defined as patients 1-year post-acute coronary syndrome, with prior revascularization, coronary stenoses > 50% by coronary angiogram, and/or evidence for cardiac ischemia on diagnostic testing, we recommend long-term low-dose aspirin or clopidogrel (75 mg/d) (Grade 1A). For patients with acute coronary syndromes who undergo percutaneous coronary intervention (PCI) with stent placement, we recommend for the first year dual antiplatelet therapy with low-dose aspirin in combination with ticagrelor 90 mg bid, clopidogrel 75 mg/d, or prasugrel 10 mg/d over single antiplatelet therapy (Grade 1B). For patients undergoing elective PCI with stent placement, we recommend aspirin (75-325 mg/d) and clopidogrel for a minimum duration of 1 month (bare-metal stents) or 3 to 6 months (drug-eluting stents) (Grade 1A). We suggest continuing low-dose aspirin plus clopidogrel for 12 months for all stents (Grade 2C). Thereafter, we recommend single antiplatelet therapy over continuation of dual antiplatelet therapy (Grade 1B).

Conclusions: Recommendations continue to favor single antiplatelet therapy for patients with established coronary artery disease. For patients with acute coronary syndromes or undergoing elective PCI with stent placement, dual antiplatelet therapy for up to 1 year is warranted.

Accepted August 31, 2011.

Full Text

Abstract 19 of 21

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 11 di 19 Selected Abstracts 07/02/12 22:14

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic Therapy in Peripheral Artery Disease Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: This guideline focuses on antithrombotic drug therapies for primary and secondary prevention of cardiovascular disease as well as for the relief of lower-extremity symptoms and critical ischemia in persons with peripheral arterial disease (PAD).

Results: The most important of our 20 recommendations are as follows. In patients aged ≥ 50 years with asymptomatic PAD or asymptomatic carotid stenosis, we suggest aspirin (75-100 mg/d) over no therapy (Grade 2B) for the primary prevention of cardiovascular events. For secondary prevention of cardiovascular disease in patients with symptomatic PAD (including patients before and after peripheral arterial bypass surgery or percutaneous transluminal angioplasty), we recommend long-term aspirin (75-100 mg/d) or clopidogrel (75 mg/d) (Grade 1A). We recommend against the use of warfarin plus aspirin in patients with symptomatic PAD (Grade 1B). For patients undergoing peripheral artery percutaneous transluminal angioplasty with stenting, we suggest single rather than dual antiplatelet therapy (Grade 2C). For patients with refractory claudication despite exercise therapy and smoking cessation, we suggest addition of cilostazol (100 mg bid) to aspirin (75-100 mg/d) or clopidogrel (75 mg/d) (Grade 2C). In patients with critical limb ischemia and rest pain unable to undergo revascularization, we suggest the use of prostanoids (Grade 2C). In patients with acute limb ischemia due to acute thrombosis or embolism, we recommend surgery over peripheral arterial thrombolysis (Grade 1B).

Conclusions: Recommendations continue to favor single antiplatelet therapy for primary and secondary prevention of cardiovascular events in most patients with asymptomatic PAD, symptomatic PAD, and asymptomatic carotid stenosis. Additional therapies for relief of limb symptoms should be considered only after exercise therapy, smoking cessation, and evaluation for peripheral artery revascularization.

Accepted August 31, 2011.

Full Text

Abstract 20 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES VTE, Thrombophilia, Antithrombotic Therapy, and Pregnancy Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: The use of anticoagulant therapy during pregnancy is challenging because of the potential for both fetal and maternal complications. This guideline focuses on the management of VTE and thrombophilia as well as the use of antithrombotic agents during pregnancy.

Results: We recommend low-molecular-weight heparin for the prevention and treatment of VTE in pregnant women instead of unfractionated heparin (Grade 1B). For pregnant women with acute VTE, we suggest that anticoagulants be continued for at least 6 weeks postpartum (for a minimum duration of therapy of 3 months) compared with shorter durations of treatment (Grade 2C). For women who fulfill the laboratory criteria for antiphospholipid antibody (APLA) syndrome and meet the clinical APLA criteria based on a history of three or more pregnancy losses, we recommend antepartum administration of prophylactic or intermediate-dose unfractionated heparin or prophylactic low-molecular-weight heparin combined with low-dose aspirin (75-100 mg/d) over no treatment (Grade 1B). For women with inherited thrombophilia and a history of pregnancy complications, we suggest not to use antithrombotic prophylaxis (Grade 2C). For women with two or more miscarriages but without APLA or thrombophilia, we recommend against antithrombotic prophylaxis (Grade 1B).

Conclusions: Most recommendations in this guideline are based on observational studies and extrapolation from other populations. There is an urgent need for appropriately designed studies in this population.

Accepted August 31, 2011.

Full Text

Abstract 21 of 21

ANTITHROMBOTIC THERAPY AND PREVENTION OF THROMBOSIS, 9TH ED: AMERICAN COLLEGE OF CHEST PHYSICIAN EVIDENCE-BASED http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 12 di 19 Selected Abstracts 07/02/12 22:14

CLINICAL PRACTICE GUIDELINES ONLINE ONLY ARTICLES Antithrombotic Therapy in Neonates and Children Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

Background: Neonates and children differ from adults in physiology, pharmacologic responses to drugs, epidemiology, and long-term consequences of thrombosis. This guideline addresses optimal strategies for the management of thrombosis in neonates and children.

Methods: The methods of this guideline follow those described in the Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines.

Results: We suggest that where possible, pediatric hematologists with experience in thromboembolism manage pediatric patients with thromboembolism (Grade 2C). When this is not possible, we suggest a combination of a neonatologist/pediatrician and adult hematologist supported by consultation with an experienced pediatric hematologist (Grade 2C). We suggest that therapeutic unfractionated heparin in children is titrated to achieve a target anti-Xa range of 0.35 to 0.7 units/mL or an activated partial thromboplastin time range that correlates to this anti-Xa range or to a protamine titration range of 0.2 to 0.4 units/mL (Grade 2C). For neonates and children receiving either daily or bid therapeutic low-molecular-weight heparin, we suggest that the drug be monitored to a target range of 0.5 to 1.0 units/mL in a sample taken 4 to 6 h after subcutaneous injection or, alternatively, 0.5 to 0.8 units/mL in a sample taken 2 to 6 h after subcutaneous injection (Grade 2C).

Conclusions: The evidence supporting most recommendations for antithrombotic therapy in neonates and children remains weak. Studies addressing appropriate drug target ranges and monitoring requirements are urgently required in addition to site- and clinical situation-specific thrombosis management strategies.

Accepted August 31, 2011.

Full Text

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 13 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 14 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 15 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 16 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 17 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 18 di 19 Selected Abstracts 07/02/12 22:14

http://chestjournal.chestpubs.org/gca?gca=chest%3B141%2F2_sup…hest%3B141%2F2_suppl%2Fe737S&submit=Get+All+Checked+Abstracts Pagina 19 di 19