Standards of Practice

Practice Guideline for Adult Antibiotic Prophylaxis during Vascular and Interventional Radiology Procedures

Aradhana M. Venkatesan, MD, Sanjoy Kundu, MD, David Sacks, MD, Michael J. Wallace, MD, Joan C. Wojak, MD, Steven C. Rose, MD, Timothy W.I. Clark, MD, MSc, B. Janne d’Othee, MD, MPH, Maxim Itkin, MD, Robert S. Jones, DO, MSc, FACP, Donald L. Miller, MD, Charles A. Owens, MD, Dheeraj K. Rajan, MD, LeAnn S. Stokes, MD, Timothy L. Swan, MD, Richard B. Towbin, MD, and John F. Cardella, MD

J Vasc Interv Radiol 2010; 21:1611–1630

Abbreviations: AHA ϭ American Heart Association, GI ϭ gastrointestinal, GU ϭ genitourinary, IE ϭ infective endocarditis, IR ϭ interventional radiology, IV ϭ intravenous, IVC ϭ inferior vena cava, RF ϭ radiofrequency, SIR ϭ Society of Interventional Radiology, TIPS ϭ transjugular intrahepatic portosystemic shunt, UAE ϭ uterine artery embolization

PREAMBLE complications from nosocomial infec- vide recommendations concerning only tion. This document summarizes the antibiotic prophylaxis, not treatment of There is a need for current, formal rec- findings from the available surgical and infectious complications. It is beyond ommendations in the interventional ra- IR literature on this topic. Anticipated the scope of this article to discuss the diology (IR) literature concerning the pathogens and corresponding antibiotic management of infectious complica- appropriate use of prophylactic antibi- coverage (dose and duration) are enu- tions sustained during vascular and in- otics for IR procedures. This is particu- merated for common vascular and non- terventional radiology procedures. larly important given the increasing in- vascular interventions in adults. Note The membership of the Society of In- cidence of antibiotic resistance and that this document is intended to pro- terventional Radiology (SIR) Standards of Practice Committee represents ex- perts in a broad spectrum of interven- tional procedures from the private and academic sectors of medicine. Gener- From the Center for Interventional Oncology, Radi- ventional Radiology (T.W.I.C.), New York Univer- ally, Standards of Practice Committee ology and Imaging Sciences (A.M.V.), National In- sity School of Medicine, NYU Medical Center, New members dedicate the vast majority of stitutes of Health Clinical Center; Department of York, New York; Departments of Radiology and Radiology and Radiologic Sciences (D.L.M.), Uni- Surgery (C.A.O.), University of Illinois Medical Cen- their professional time to performing in- formed Services University of the Health Sciences; ter, Chicago, Illinois; Department of Radiology and terventional procedures; as such, they Department of Radiology (D.L.M.), National Naval Radiological Sciences (L.S.S.), Vanderbilt University represent a valid broad expert constitu- Medical Center, Bethesda; Department of Radiol- Medical Center, Nashville, Tennessee; Department ency of the subject matter under consid- ogy, Division of Interventional Radiology (B.J.D.), of Radiology (T.L.S.), Marshfield Clinic, Marshfield, University of Maryland Medical Center, Baltimore, Wisconsin; and Department of Radiology (R.B.T.), eration for standards production. Maryland; Department of Medical Imaging (S.K.), Phoenix Children’s Hospital, Phoenix, Arizona. Re- Technical documents specifying Scarborough General Hospital; Division of Vascular ceived May 21, 2010; final revision received July 3, the exact consensus and literature re- and Interventional Radiology, Department of Med- 2010; accepted July 23, 2010. Address correspon- view methodologies as well as the ical Imaging (D.K.R.), University of Toronto, Uni- dence to A.M.V., c/o Debbie Katsarelis, 3975 Fair versity Health Network, Toronto, Ontario, Canada; Ridge Dr., Suite 400 N., Fairfax, VA 22033; E-mail: institutional affiliations and profes- Department of Interventional Radiology (D.S.) and [email protected] sional credentials of the authors of Section of Infectious Diseases (R.S.J.), Reading Hos- this document are available upon re- This document was written by the Standards of pital and Medical Center, West Reading; Depart- quest from SIR, 3975 Fair Ridge Dr., ment of Radiology, Division of Interventional Radi- Practice Committee for the Society of Interventional ology (M.I.), University of Pennsylvania Medical Radiology and endorsed by the Cardiovascular In- Suite 400 N., Fairfax, VA 22033. Center, Philadelphia; Department of Radiology terventional Radiological Society of Europe and Ca- (J.F.C.), Geisinger Health System, Danville, Pennsyl- nadian Interventional Radiology Association. vania; Department of Interventional Radiology None of the authors have identified a conflict of METHODOLOGY (M.J.W.), The University of Texas M.D. Anderson interest. Cancer Center, Houston, Texas; Department of Ra- SIR produces its Standards of diology (J.C.W.), Our Lady of Lourdes Medical Cen- ter, Lafayette, Louisiana; Department of Radiology © SIR, 2010 Practice documents using the follow- (S.C.R.), University of California San Diego Medical ing process. Standards documents of Center, San Diego, California; Department of Inter- DOI: 10.1016/j.jvir.2010.07.018 relevance and timeliness are concep-

1611 1612 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR tualized by the Standards of Practice biotics are given when they should for antibiotic prophylaxis and of Committee members. A recognized not be given; and (iii) how often is making recommendations regarding expert is identified to serve as the antibiotic timing within recom- the use of drugs based on studies of principal author for the standard. mended guidelines. Although prac- suboptimal design. However, these Additional authors may be assigned ticing physicians should strive to difficulties are far outweighed by the dependent upon the magnitude of achieve perfect outcomes, in practice potential improvements in safety the project. all physicians will fall short of ideal and treatment efficacy that may be An in-depth literature search is outcomes to a variable extent. There- gained by implementing the key les- performed using electronic medical fore, in addition to quality improve- sons learned from the existing peer- literature databases. Then a critical ment case reviews conducted after reviewed scientific literature that has review of peer-reviewed articles is individual complications, outcome evaluated antibiotic prophylaxis performed with regard to the study measure thresholds should be used during surgical and IR procedures. methodology, results, and conclu- to assess treatment safety and effi- Given the limited scientific founda- sions. The qualitative weight of these cacy in ongoing quality improve- tion, most of the recommendations articles is assembled into an evi- ment programs. For the purpose of presented in this document are in- dence table, which is used to write these guidelines, a threshold is a spe- tended to guide clinical practice the document such that it contains cific level of an indicator that, when rather than to mandate the use of evidence-based data with respect to reached or exceeded, should prompt specific algorithms. The authors content, rates, and thresholds. a review of departmental policies fully anticipate that these guidelines When the evidence of literature is and procedures to determine causes will be appropriately revised when weak, conflicting, or contradictory, and to implement changes, if neces- future studies of greater scientific consensus for the parameter is sary. Thresholds may vary from rigor are available. Levels of evi- reached by a minimum of 12 Stan- those listed here; for example, pa- dence have been assigned to the cur- dards of Practice Committee mem- tient referral patterns and selection rent recommendations within this bers using a Modified Delphi Con- factors may dictate a different clinical practice guideline that ad- sensus Method (Appendix A) (1–3). threshold value for a particular indi- here to definitions created by the For the purposes of these documents, cator at a particular institution. American College of Cardiology and consensus is defined as 80% Delphi Therefore, setting universal thresh- American Heart Association (AHA) participant agreement on a value or olds is difficult, and each department Guidelines Task Force (4,5). A sum- parameter. is urged to adjust its thresholds as mary of these definitions are pro- The draft document is critically needed to meet its specific quality vided in Appendix B. reviewed by the Standards of Prac- improvement program situation. tice Committee members, either by SIR is committed to the basic prin- telephone conference calling or face- ciples of outcome-focused, evidence- INTRODUCTION to-face meeting. The finalized draft based medicine. Ideally, every Stan- from the Committee is sent to the SIR dards of Practice Committee During the past two decades, sev- membership for further input/criti- recommendation would be based on eral authors have reported on the use cism during a 30-day comment pe- evidence derived from multiple pro- of prophylactic antibiotics for IR pro- riod. These comments are discussed spective randomized trials of ade- cedures (6–9). Spies et al (6) were the by the Standards of Practice Com- quate statistical power. Unfortu- first to provide a critical review on mittee, and appropriate revisions nately, currently there are no antibiotic prophylaxis in IR. Their made to create the finished stan- published multicenter randomized study underscored the lack of ran- dards document. Before its publica- trials that evaluate the clinical effi- domized controlled trials of antibi- tion the document is endorsed by the cacy and indications for antibiotic otic prophylaxis during IR proce- SIR Executive Council. prophylaxis during IR procedures. dures, indicating that the selection of The current guidelines were writ- Within the existing publications, sev- antibiotics for IR prophylaxis was ten to help determine appropriate eral major limitations are evident: (i) largely guided by the existing surgi- antibiotic prophylaxis for vascular lack of randomized controlled trials cal literature (6,9). There has been a and IR procedures. The elements of concerning the efficacy of IR antibi- lack of randomized controlled trials care necessitate knowledge of the otic prophylaxis and reliance on ex- in the radiology literature to provide following: (i) the anticipated patho- isting case series and retrospective scientific validation for the role and gens associated with the given inter- analyses; (ii) extreme variation in the effectiveness of antibiotic prophy- vention and (ii) pretreatment evalu- existing literature concerning patient laxis during IR procedures. Never- ation and patient-specific factors that selection parameters, definitions of theless, antibiotic prophylaxis has may be associated with higher likeli- short-term efficacy, and definitions become the perceived standard of hood of periprocedural infection. of infectious complications; and care for selected procedures, making The outcome measures or indicators (iii) absence of systematic assess- it possible that randomized con- for this process include (i) how often ment of long-term efficacy. SIR rec- trolled trials may never be per- antibiotics are given when they ognizes the potential pitfalls of de- formed (6,9). Extensive clinical data should be given; (ii) how often anti- veloping evidence-based standards on the use of antimicrobial agents in Volume 21 Number 11 Venkatesan et al • 1613 surgery exist. As noted by Spies et al responsibility ultimately lies with guide antibiotic prophylaxis for IR pro- (6), the analogous clinical consider- the individual IR physician, who cedures (6). ations in surgery serve as a basis for must determine for him- or herself Clean.—A procedure is regarded as developing an approach to the the appropriateness of a specific an- clean if the gastrointestinal (GI) tract, proper use of antibiotics in IR. Clin- tibiotic regimen for a specific patient genitourinary (GU) tract, or respiratory ical case series and retrospective and clinical scenario. As the antibi- tract is not entered, if inflammation is analyses in the IR literature concern- otic sensitivities of different patho- not evident, and if there is no break in ing the occurrence, prevention, and gens vary based on time and region, aseptic technique (18). An example of a management of infectious complica- the appropriate prophylactic antibi- clean procedure is routine diagnostic tions following specific IR proce- otic selection may differ depending angiography (7,18). dures are also available to guide clin- on the institution at which the phy- Clean-contaminated.—A procedure ical practice (10–14). This document sician practices. The IR physician is regarded as clean-contaminated if offers a logical basis for prophylactic should therefore periodically consult the GI, biliary, or GU tract is entered; use of antibiotics during IR proce- with his or her local hospital-based if inflammation is not evident; and dures based on the literature avail- infectious disease committee, con- there is no break in aseptic tech- able. Many current clinical and envi- sulting service, microbiology labora- nique. An example of a clean-con- ronmental factors mandate current tory, and/or pharmacy for recom- taminated procedure is nephrostomy recommendations for prophylaxis mended medications. tube placement in a patient with ster- during IR procedures. These include ile urine (7,18). the widening breadth of IR proce- Contaminated.—A procedure is re- dures and recent revisions to existing DEFINITIONS garded as contaminated if there is entry surgical practice guidelines from into an inflamed or colonized GI or GU which many IR prophylaxis guide- General Definitions tract without frank pus, or if a major break lines are derived (6,9,15–17). Addi- in aseptic technique occurs (18). tional factors include increasing an- This document uses definitions as Dirty.—A procedure is regarded tibiotic resistance, morbidity outlined by Spies et al (6) for coloni- as dirty if it involves entering an in- associated with nosocomial infec- zation, infection, clinical infection, fected purulent site such as an ab- tion, and increasing health care costs and prophylaxis, as follows: scess, a clinically infected biliary or (9,15–17). Colonization is the presence of a GU site, or perforated viscus (18). The specific antibiotic agent doses microorganism without host As noted later, antibiotic prophylaxis provided herein are recommenda- response. for IR procedures is generally recom- tions for normal adult patients. Al- Infection is the presence of an or- mended for procedures that are not though the doses and agents named ganism with host response. clean, or for procedures that are consid- here are suggestions meant to facili- Clinical infection represents infec- ered clean but, as a result of which, a tate the care of IR patients, it is the tions that produce local signs potentially significant volume of ne- ultimate responsibility of the indi- or symptoms (eg, fever, pain, crotic tissue is generated in potentially vidual interventional radiologist to leukocytosis). contaminated areas, eg, embolization pay close attention to individual pa- Prophylaxis is the administration of with intent to create infarction. Al- tient factors, including advanced pa- an antimicrobial in the periproce- though National Academy of Sciences/ tient age, existing medications, re- dural period in the absence of National Research Council surgical duced renal or hepatic function, clinical infection to prevent an in- wound classification may be used as a medication allergy history, the in- fectious complication. guide, it should be noted that the infec- tended procedure and its likely Bacteremia is the presence of bac- tious risks of open surgical procedures pathogens, and the dose and timing teria within the bloodstream differ from those of IR procedures (19). of the most recent antibiotic agent without clinical signs or symp- As noted by Zarrinpar et al (19), surgical dose. Antibiotic choices have been toms of infection (eg, fever, prophylaxis is typically directed at pre- listed as “first choice” if a specific pain, leukocytosis). venting infection of the wound by in- antibiotic agent has been consistently Septicemia is the presence of fected fluid or skin organisms. The described in the published literature pathogenic organisms or their small incision made during vascular in association with prophylaxis for a toxins in the bloodstream with and nonvascular IR procedures rarely given procedure. When no single concomitant systemic signs and serves as a site for clinically important agent has been uniformly reported as symptoms of sepsis (6). infection (19). Prophylaxis in IR is more the agent of choice for prophylaxis of commonly done to prevent infection re- a given procedure, no first-choice an- Procedure Classification sulting from the communication that is tibiotic is listed. Instead, a list of created by a needle or catheter (19) be- common antibiotic choices are pro- The National Academy of Sciences/ tween an infected space and the blood- vided, which are described in the lit- National Research Council has divided stream. Data from studies of surgical erature. It is important to note that surgical wounds into four classes— wound prophylaxis may therefore not this document should not be consid- clean, clean-contaminated, contami- always be directly applicable to IR pro- ered a replacement for the close and nated, and dirty—each of which is as- cedures performed on the same organ, careful evaluation of individual pa- sociated with a different risk of infection and the interventional radiologist is en- tients by their IR practitioners. The (7,18). This classification may be used to couraged to consider the appropriate- 1614 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR ness of prophylaxis in accordance with which parenteral antibiotic treatment was presence of multiple-organ disease, the clinical circumstances of the individ- completely withheld (6,22). when included in a predictive model ual patient and site-specific flora (19). With respect to duration of ad- along with the level of wound con- ministration, several studies have tamination, allowed identification of COMPLICATIONS evaluated the efficacy of single ver- a subgroup of patients who would sus multiple dose protocols for pro- sustain 90% of wound infections (28). Complications can be grouped on phylaxis, demonstrating that a single Prophylactic antibiotic agents should the basis of outcome. Major compli- appropriately timed dose is as effec- be administered immediately before cations from antibiotic therapy result tive as a multiple-dose protocol transfer to the interventional suite, or by in admission to a hospital for ther- (9,23–26). Recommended IR practice the IR nurse when the patient arrives in apy (for outpatient procedures), an for most procedures warranting pro- the interventional suite. The latter is unplanned increase in the level of phylaxis involves administration of a optimal, particularly if there is a delay in care, prolonged hospitalization, per- single preprocedural dose of an an- the patient’s arrival, as the time of anti- manent adverse sequelae, or death. timicrobial agent just before com- biotic agent administration should not Minor complications result in no se- mencement of a procedure. Excep- be altered (9). In keeping with the cur- quelae; they may require nominal tions include procedures when a rent Joint Commission on the Accredi- therapy or a short hospital stay for patient undergoes instrumentation tation of Healthcare Organization 2009 observation (generally overnight; of an obstructed viscus, such as bili- Hospital Patient Safety Goals (29), Appendix C). The complication rates ary or kidney obstruction, in which which recapitulate the existing surgical described in the present document the risk of postprocedural bactere- literature, intravenous (IV) antimicro- refer to major complications. mia caused by intravasation of or- bial prophylaxis should be administered ganisms into the bloodstream re- within 1 hour before incision, with 2 TIMING AND mains present until the organ is hours allowed for the administration of ADMINISTRATION OF adequately drained (9). The antibi- vancomycin and fluoroquinolones, and PROPHYLAXIS otic agent used in this setting admit- with discontinuation of the prophylactic tedly spans the boundary between antimicrobial agent within 24 hours af- The basis for timing and duration of prophylaxis and treatment, and in ter intervention, unless clinical circum- antibiotic prophylaxis of IR procedures such patients, treatment should be stances (eg, obstructed viscus) warrant is derived from the surgical literature. continued until satisfactory drainage continuation of a prophylactic regimen The rationale for current surgical antibi- of the viscus is achieved (9). In per- as therapy. An accurate list of the pa- otic prophylaxis emerged from animal cutaneous drainage procedures, the tient’s current medications and known experiments performed in the 1960s by risk of bacteremia is attributed to allergies should also be reviewed to Burke (20), who demonstrated the great- communication of the fluid cavity safely prescribe any setting-specific an- est suppression of infection when anti- with the arterial, venous, and lym- tibiotics and to assess for the potential microbial agents were administered be- phatic systems (6,27). In the absence allergic or adverse drug reactions (29– fore inoculation. The likelihood of of data on this topic, Spies et al (6) 31). Appropriate antibiotic coverage postprocedure infection increased the have advocated continued antibiotic should be selected by an experienced IR longer the delay between initiation of therapy for 48 hours after such a pro- practitioner. When a patient is already antibiotic prophylaxis and the proce- cedure to reduce the risk of bactere- receiving antibiotic drugs, the timing of dure; an initial dose given 3 hours after mia and its consequences. Other the most recent dose of antibiotic agent a procedure resulted in no difference in cases that may warrant prolonged should be reviewed, as well as the ap- the likelihood of infection in treated an- prophylaxis include those that in- propriateness of the antibiotic relative to imals versus untreated controls (20). volve earlier surgical manipulation the procedure and the likely pathogens More recent clinical studies corroborate (eg, bilioenteric anastomosis before (9). If indicated, an additional dose or these findings, demonstrating a fivefold chemoembolization). different agent may be administered be- increased rate of infectious events when Additional factors to consider in- fore the procedure. If a procedure is antibiotic administration occurs more clude patient condition and proce- likely to be prolonged (ie, Ͼ2 h), a than 3 hours after a procedure (7,21). In dure duration (9). Investigators from supplemental dose of antibiotic a randomized double-blinded study of the Centers for Disease Control (7,28) agent should be considered, depend- patients undergoing colonic surgery, a have described three risk factors as- ing on the half-life of the agent being 16% frequency of wound infection oc- sociated with an increased risk of administered (9). curred when no prophylaxis was used; postoperative surgical wound infec- the incidences were 15% when prophy- tion, in addition to the level of Antibiotic Resistance laxis was given after the procedure and wound contamination. These factors only 6% when prophylaxis was given 1 were derived from information col- Patient morbidity and mortality re- hour before surgery (6,22). Overall, an lected on more than 58,000 operative sulting from infectious complications almost fourfold greater incidence of in- patients. A simple multivariate risk have been increasing during the past cisional infection was demonstrated index was created, which prospec- two decades (9). This, in large part, is when antibiotic agent was withheld ver- tively predicted surgical wound in- related to the increasing frequency and sus when it was given preoperatively (18). fection risk (28). Factors including a spectrum of antibiotic resistant infec- Initiation of antibiotic therapy postopera- procedure lasting more than 2 hours, tions (9,32). Factors facilitating selection tively yielded similar results to those in an abdominal operation, and the of antibiotic resistance include the over- Volume 21 Number 11 Venkatesan et al • 1615 use of antibiotic agents by health care ondary to several factors, including (9,50,51). Prophylaxis is similarly practitioners, the increased use of inva- the presence of a foreign body and recommended for peripheral en- sive devices, a greater number of sus- the presence of inflammation at the dografts, including those for superfi- ceptible hosts (ie, those in an immuno- site of stent implantation (7,40). Stent cial femoral artery recanalization compromised condition), and lapses in infection has been reported in the and dialysis access endografts, appropriate infection control (9). When aorta, iliac, renal, coronary, and sub- which are increasingly used in IR choosing an antibiotic agent for prophy- clavian arteries (9,41–48). When re- practice. laxis, one needs to consider the likely ported, complications are major, Procedure classification: clean source and pathogen, and the prophy- usually involving arteritis with pseu- Organisms encountered: S aureus, S laxis should be specifically directed doaneurysm formation (9,49). Risk epidermidis against these organisms. The routine factors have included repeat punc- Routine prophylaxis recommended: yes use of broad spectrum antibiotics is not ture of the same vessel during a First-choice antibiotic agent:1gce- cost effective and promotes further an- short time interval or use of a vascu- fazolin IV tibiotic resistance (9,33). lar sheath that has been in place for Alternate choices: (ii) if penicillin-aller- more than 24 hours (9,49). Recom- gic, vancomycin or clindamycin mendations have been made in the PROPHYLAXIS FOR SPECIFIC Level of evidence:5,8 IR PROCEDURES literature for prophylactic antibiotic agent use if an arterial sheath is left Lower-extremity Superficial Venous Vascular Interventions: Angiogra- in overnight after stent placement, or Insufficiency Treatment phy, Angioplasty, Thrombolysis, for patients undergoing multiple en- Arterial Closure Device Placement, dovascular interventions (46). One Current therapies for treatment of Stent Placement advocated approach involves admin- lower-extremity superficial venous in- istration of prophylactic antibiotic sufficiency—including varicose veins— Diagnostic angiography, routine an- agents before a second procedure to include endovascular thermal ablation, gioplasty, and thrombolysis are consid- patients in whom repeat intervention sclerotherapy, and ambulatory phlebec- ered clean procedures, and with careful is performed within 7 days (38). Al- tomy (52). Although routine practice in- technique, antibiotic prophylaxis is un- though infection represents a statis- cludes the use of sterile technique before necessary (34). Bacteremia occurs after tically low-risk event, serious out- these procedures, there are insufficient 4%–8% of angiographic procedures and comes have been associated with this data in the literature to warrant routine is typically asymptomatic (6,34,35). complication. At the present time, antibiotic prophylaxis (52). Adverse When it does occur, the source is usu- prophylactic antibiotic agents are not events related to these therapies are typ- ally an angiographic catheter that was recommended for routine arterial ically minor, and include paresthesias contaminated before use (6,34). Re- stent placement. However, for those and superficial phlebitis, with more se- peated puncture of a site or repeated rious adverse events—primarily per- catheterization of an indwelling sheath patients deemed to be at high risk taining to endovascular thermal abla- is believed to increase the risk of (ie, cases of repeat intervention periprocedural infection (7,36). within 7 d, prolonged indwelling ar- tion—including deep vein thrombosis, Percutaneous vascular closure de- terial sheath, or prolonged duration neurologic injury, and skin burns (52). of procedure), prophylaxis may be vices, including collagen plug de- Procedure classification: clean given (9,46). vices and suture-mediated closure Organisms encountered: S aureus, S devices, are increasingly used during Procedure: clean epidermidis cardiac and peripheral vascular in- Organisms encountered: Staphylo- Routine prophylaxis recommended:no terventions (37,38). Infectious com- coccus aureus, Staphylococcus Level of evidence:8 plications have been reported after epidermidis the use of these devices; these in- Routine prophylaxis recommended:no Inferior Vena Cava Filter Placement clude groin cellulitis and femoral ar- Special considerations: 1 g cefazolin Infection after inferior vena cava tery endarteritis (37–39). Comorbid IV if there is a high risk of stent (IVC) filter placement has not been a factors described in patients with in- infection; or (ii) if the patient is significant problem in clinical prac- fectious complications include diabe- allergic to penicillin, vancomy- tice. One case of fatal septicemia af- tes mellitus, obesity, and placement cin or clindamycin ter placement of an IVC filter has of a percutaneous suture closure de- Level of evidence: 5, 8 (Appendix B) vice within the previous 6 months, been reported in the IR literature (7). warranting caution when consider- Endograft Placement In this instance, the IVC filter (LGM ing the use of a percutaneous closure VenaTech IVC Filter; L-G Medical, device in such patients (38). How- Prophylactic antibiotic agents are Evanston, Illinois) was placed ever, at the present time, there are given routinely for aortic endograft through the site of an indwelling insufficient data to suggest prophy- therapy at many institutions, al- central venous catheter (53). When laxis before placement of these de- though there is limited scientific ev- placing an IVC filter, a “fresh” ve- vices. idence to support this approach nous access site is recommended Stent infection is an uncommon (9,49). Prosthetic graft infection is an (7,53). However, routine filter place- but serious complication of endovas- uncommon event, but when it oc- ment in an uninfected patient does cular treatment (9). This is likely sec- curs, it carries a high mortality rate not warrant prophylaxis (7,54). 1616 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

Procedure classification: clean There was no difference in the cath- of penicillin allergy, vancomy- Routine prophylaxis recommended:no eter-related bacteremia rates be- cin or clindamycin Level of evidence:5,6 tween the two groups, and although Level of evidence:8 the impregnated catheter group had Central Venous Access a lower cumulative infection rate for Embolization and Chemoembolization The use of antibiotic prophylaxis dwell times less than 5 days, the dif- for central venous catheter place- ference between the cumulative in- Historic IR data have reported a ment is controversial. Central ve- fection rates was not statistically sig- higher incidence of transient bactere- nous catheter placement is classified nificant for dwell times of 6, 7, or 8 mia after embolization compared as a clean procedure. Therefore, pro- days (59). with diagnostic angiography (60). In phylaxis is not routinely recom- Some authors have suggested that a prospective study of 45 patients mended, although meticulous sterile routine antibiotic prophylaxis may undergoing embolization, 32% of pa- technique and appropriate preproce- be warranted before placement of tients who received no antibiotic dural preparation is mandatory (9). tunneled catheters with implantable treatment developed bacteremia, al- When catheter-related infections do ports, given that these patients are though none developed clinical sep- occur, these are most commonly often in an immunocompromised sis. The organisms isolated included caused by coagulase negative Staph- state, and given the severity of com- S epidermidis, Streptococcus species, ylococcus species (9). Migration of plications and difficulty in treating and Corynebacterium species, all of skin organisms into the catheter tract infected ports (9). The benefit of an- which are normal in mucosal or skin is the most common route of infec- tibiotic prophylaxis for central ve- flora (60). In contrast, no patients tion, with contamination of the cath- nous access remains unproven, and who received antibiotic prophylaxis eter hub contributing to colonization therefore the routine use of antibiotic had positive blood cultures (60). of long-term catheters (9). Most in- prophylaxis remains an area of con- Findings from this study suggest fections in patients with long-term troversy. Prevention is critical to that bacteremia is especially likely venous access are caused by Gram- minimize the likelihood of infections following embolization. Findings positive bacteria (9). associated with implantable ports, as from additional studies support the A recent version of the Centers for these are difficult to treat and poten- use of antibiotic prophylaxis before Disease Control Guidelines for the tially fatal (9). Preventive strategies embolization. Reed et al (61) re- Prevention of Intravascular Cathe- include meticulous attention to ster- ported the results of 494 hepatic che- ter-related Infections (9,55) under- moembolization procedures, among ile technique. The use of catheters scored the importance of sterile tech- which 14 were performed without with the fewest number of lumens nique, the use of 2% chlorhexidine prophylactic antibiotics. One of these skin preparation, avoidance of rou- necessary for management of the pa- 14 cases resulted in fatal sepsis tine catheter change, and the use of tient reduces portals for colonization within 24 hours of intervention antiseptic or antibiotic agent–im- (9). For patients who require inter- (7,61). In a patient-by-patient analy- pregnated catheters only if previous mittent long-term venous access, im- sis of these data, of the 226 patients infection rates are high. van de We- plantable devices are recommended, as who received prophylaxis, six devel- tering et al (56) evaluated the effi- these are more resistant to contamina- oped hepatic abscesses but no cases cacy of administration of antibiotic tion (9). Such devices are now in use for of fatal sepsis occurred (61). In a sep- agents before insertion of a central dialysis access management in an at- arate clinical series of 410 emboliza- venous catheter with or without van- tempt to reduce infection rates. Al- tion cases (62), eight deaths were re- comycin/heparin flush technique in though the use of antibiotic prophylaxis ported, of which seven were caused the first 45 days after catheter place- may be considered in specific clinical by infectious complications in pa- ment to minimize Gram-positive scenarios (eg, immunocompromised pa- tients who had not received antibi- catheter-related infections in oncol- tients who require catheter placement otic prophylaxis. Prophylaxis tar- ogy patients (9), and antibiotic agent before chemotherapy and those with a geted against skin pathogens is administration before catheter inser- history of catheter infection), routine recommended before performing tu- tion followed by vancomycin/hepa- prophylaxis for these procedures is not mor and/or solid organ emboliza- rin flush was associated with a de- recommended. tion, including the liver, kidney, and creased incidence of Gram-positive spleen, when there is an intent to infections (56). Of note, this strategy Procedure classification: clean create infarction or a high likelihood was not advocated for general prac- Organisms encountered: S aureus, S of infarction, as this may result in a tice, but in the setting of recurrent epidermidis potentially significant volume of ne- catheter infection (9,56–58). Loo et al Routine prophylaxis recommended:no crotic tissue in potentially contami- (59) reported a survey of 196 consec- consensus nated areas (9,19,60). Routine pro- utive central venous catheters placed Special considerations (eg, immuno- phylaxis remains controversial in the in the intensive care unit in 151 pa- compromised patients who re- setting of embolization for the pur- tients, in which an antiseptic agent– quire catheter placement before poses of controlling bleeding from a impregnated catheter was alternated chemotherapy and those with a viscus or solid organ, such as in the on a bimonthly basis with a standard history of catheter infection): setting of trauma. Depending on the triple-lumen central venous catheter. 1 g cefazolin IV; or in the case clinical setting and treatment goals, Volume 21 Number 11 Venkatesan et al • 1617 the target organ, and likelihood of Routine prophylaxis recommended: United Kingdom has recommended additional pathogens, the antibiotic yes (if intent to create infarc- that prophylactic antibiotic treatment regimen should be adjusted accord- tion or high likelihood of solid should not be given at the time of the ingly (9). organ infarction) embolization as infectious complica- The IR literature regarding the use First-choice antibiotic agent: no con- tions are generally delayed (as long as of antibiotic prophylaxis for chemo- sensus 2–3 weeks after the procedure) (13, embolization is small. In the absence Common antibiotic choices: (i) 1.5–3 g 76,77). In the Ontario Uterine Fibroid of randomized clinical trials, the ef- ampicillin/sulbactam IV (hepatic Embolization Trial (77), which in- fectiveness of prophylaxis in this set- chemoembolization); (ii) 1 g cefa- cluded 555 patients, antibiotic prophy- ting is unproven, although several zolin and 500 mg metronidazole laxis (1 g cefazolin IV) was routinely clinical series (9,11,12,63–66) have IV (hepatic chemoembolization); administered at four hospitals and re- suggested that major infectious com- (iii) 2 g ampicillin IV and 1.5 served for patients at increased risk of plications may be sustained in this mg/kg gentamicin (hepatic che- infection at an additional four hospi- population. Many operators rou- moembolization); (iv)1gceftriax- tals. This trial reported only two infec- tinely administer antibiotic prophy- one IV (hepatic chemoembolization tion-related hysterectomies in 570 laxis for this procedure, including or renal or splenic emboliza- UAE procedures, with one occurring coverage for skin flora and for Gram- tion); (iv) if penicillin-allergic, in each cohort (13,77). A retrospective negative enteric organisms, even vancomycin or clindamycin review by Rajan et al (78) of 410 pa- though this practice has not been plus aminoglycoside tients undergoing UAE was performed to prospectively proven to be of benefit Special considerations: in patients identify risk factors for the development for all patients (9,11,12,61,62–66). for hepatic chemoembolization of intrauterine infection after UAE (13,78). Some practitioners recommend con- without an intact sphincter of Multiple variables were analyzed as pre- tinued administration of antibiotics Oddi (previous sphincterot- dictors for intrauterine infectious compli- for 3–7 days after chemoemboliza- omy, biliary drainage, history cations requiring medical and/or sur- tion to cover Gram-negative enteric of bilioenteric anastomosis), gical therapy, including the use of pathogens, although this too lacks consider tazobactam/piperacil- preprocedural antibiotics, embolic prospective validation (9,67). Pa- lin; also consider bowel prepa- agent used, quantity of embolic mate- tients without an intact sphincter of ration in this population rial, location of tumors (eg, submuco- Oddi as a result of earlier surgery or Level of evidence:4,7,8 sal, nonsubmucosal), and size and lo- sphincterotomy or biliary drainage cation of the dominant tumor (78). are at increased risk for subsequent Uterine Artery Embolization Intrauterine infectious complications abscess formation (67). The risk of requiring IV antibiotic therapy and/or postembolization infection appears Uterine artery embolization (UAE) surgery occurred in five patients to be reduced by the performance of has become a desirable alternative to (1.2%), with no specific risk factor for a bowel preparation the night before hysterectomy or myomectomy for the intrauterine infection after UAE iden- treatment and by ensuring coverage treatment of symptomatic leiomyomas tified (78). Rajan et al (78) concluded of Gram-positive and Gram-negative (13,70). The risk of infection has been that infection after embolization is rare aerobic and anaerobic organisms (eg, largely described in the literature as (13,78). In contrast to these studies, tazobactam/piperacillin). However, being low, and is reported to occur in there has been a single clinical series it remains to be seen how high the 0.2%–1% of patients (9,13,71,72). How- from the United Kingdom (9,79) that infectious risks are, despite adoption ever, several of the reported cases of described a 17% readmission rate for of an aggressive regimen (12). fatal sepsis occurring after UAE have infection after UAE (seven of 42 pa- Even fewer reports exist concern- occurred in patients who did not re- tients). There was one infection-re- ing infectious complications of ra- ceive prophylaxis (9,73–75). The most lated hysterectomy. Three of the pa- dioembolization. There have been in- common organisms causing infection tients had urinary tract infection, most dividual reports of hepatic abscess after UAE are common skin flora likely secondary to bladder catheter- after yttrium-90 radioembolization, (Staphylococcus or Streptococcus spe- ization; in the three remaining pa- but ongoing research is needed in cies), as is the case for other solid or- tients, no source of infection was iden- this population to better assess the gan embolization. At least one fatal tified (9,79). The antibiotic prophylaxis risks of infectious complications and case of sepsis after UAE has been re- regimen used in this clinical series in- to determine the efficacy of antibiotic ported in a patient who developed an cluded amoxicillin/clavulanate three prophylaxis (67–69). Escherichia coli urinary tract infection times daily and metronidazole twice Procedure classification: clean; clean- after the procedure and fatal subse- daily on the day of the procedure, fol- contaminated (bilioenteric surgery) quent E coli sepsis (9,74). lowed by a further 48-hour adminis- Organisms encountered: S aureus, The role of prophylactic antibiotics tration of the same antibiotic agent Streptococcus species, Corynebac- for patients undergoing UAE has been after the procedure. Patients were ad- terium species with or without en- debated in the literature (13). The joint mitted to the hospital for 3 days (79). teric flora (in cases of earlier working party for the Royal College of This multiple-day, multidrug ap- sphincter of Oddi manipulation/ Radiologists and the Royal College of proach to prophylaxis is largely the bilioenteric surgery) Obstetricians and Gynecologists in the exception for UAE prophylaxis in the 1618 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

IR literature; as in other clinical set- used as a bridge to transplantation in laxis regimens by Dravid et al (8) tings, multidrug regimens and pro- patients with severe complications of reported that a majority of respon- longed antibiotic administration have portal hypertension (13). With the in- dents (69%) use antibiotic prophy- been critiqued for eliminating normal troduction of covered stents and their laxis in patients undergoing TIPS Gram-positive organisms, thereby al- improved long-term patency, the em- creation; an infection rate of 13% was lowing Gram-negative organisms to ployment of TIPS has expanded reported by respondents in this sur- proliferate (13,80). Notwithstanding (13,81). Two separate infectious com- vey. The authors concluded that pro- little evidence from randomized con- plications from TIPS have been de- phylaxis is indicated for TIPS proce- trolled clinical trials, most clinical se- scribed: (i) periprocedural sepsis with- dures, and suggested a regimen of ries reporting on UAE have routinely out stent infection and (ii) TIPS stent cefoxitin 1 g IV every 6 hours for 48 administered a single dose of prepro- infection (9,13). hours (8). As noted by Beddy et al cedural antibiotic drugs, with the cur- Periprocedural sepsis has been (13), some interventionalists use rent antibiotic of choice being1gof described since the early days of ceftriaxone1gIVonce daily for 48 cefazolin (9). Cefazolin is inexpensive TIPS procedures and has been re- hours rather than cefoxitin, given its and has activity against the likely ported to occur in as many as 17% of enhanced activity against E coli, En- source of pathogens during solid or- patients (13,82). The causative organ- terobacter species, Gram-negative gan embolization, ie, skin pathogens isms are usually skin flora (eg, Staph- bacteria, anaerobes, and enterococci, (Staphylococcus or Streptococcus spe- ylococcus or Streptococcus species) and its once-daily dosing regimen. cies), as well as activity against E coli (13). Deibert et al (83) studied the Ampicillin/sulbactam also has im- (9). Other options that have been effectiveness of a single dose of a proved coverage against Enterococcus described include gentamicin plus species and can be used for TIPS pro- clindamycin, ampicillin, ampicillin/ second-generation cephalosporin to prevent post-TIPS infection. Patients phylaxis (13). sulbactam, or vancomycin in the pen- Infection of the TIPS stent itself icillin-allergic patient (9,72,80). who underwent 105 transjugular in- terventions were randomized to re- been more recently described, re- Patients with a history of hydrosal- portedly occurring in 1.7%–5.1% of pinx may warrant special consideration ceive no antibiotic treatment (46 in- terventions) or 2 g cefotiam (56 cases (9,82–86). Sanyal et al (82) were before UAE, although there is no pub- the first to describe the clinical pic- lished consensus regarding which interventions), which was adminis- ture of infection of the TIPS stent. agents should be used. Practitioners ex- tered at the beginning of the proce- Their diagnosis was based on the perienced in performing UAE in pa- dure (9,82). Post-TIPS infection was presence of fever with positive blood tients with a history of hydrosalpinx rec- defined by an increase in white cultures and (i) the presence of stent ommend prophylaxis with doxycycline blood cell count (Ͼ15,000/␮L), fever thrombus or vegetations or (ii) per- 100 mg twice daily for 7 days before the (Ͼ38.5°C), or a positive blood cul- sistent bacteremia in a patient with a procedure (J. Spies, MD, personal com- ture result (83). Patients who did not TIPS and no other detectable source munication, October 2009). receive cefotiam had an infection of infection (9,82). Clinical signs and rate of 20%, versus an infection rate Procedure classification: clean; clean- symptoms reported in this setting in- of 14% in patients treated with anti- contaminated cluded fever, tender hepatomegaly, biotics (83). The difference in infec- Organisms encountered: S aureus, S hypoxemia, septic pulmonary em- tion rate between the two groups did epidermidis, Streptococcus species boli, septic shock, neutrophilia, and not reach statistical significance, sug- with or without E coli subsequent development of necrotiz- gesting that routine use of prophy- Routine prophylaxis recommended: yes ing fasciitis (9,82). The causative or- lactic antibiotics for TIPS creation First-choice antibiotic agent:no ganisms included oral and enteric may not be of value (9,83). As noted consensus aerobic Gram-negative bacteria and Common antibiotic choices: (i) 1 g ce- by Ryan et al (9), the effectiveness of Candida species (9). Most patients re- fazolin IV; (ii) 900 mg clindamy- cefotiam for TIPS prophylaxis may sponded to antibiotic therapy (9,82). cin IV plus 1.5 mg/kg gentami- have been limited by its limited ac- DeSimone et al (84) studied 99 TIPSs cin; (iii) 2 g ampicillin IV; tivity against enterococcal species. A created during an 8-year period and (iv) 1.5–3 g ampicillin/sulbac- more appropriate agent may have identified five patients with no other tam IV; (v) if penicillin-allergic, been one with enhanced coverage alternative source of bacteremia who can use vancomycin against Enterococcus species, which were presumed to have TIPS stent infec- Special considerations: if history of may have resulted in a different out- tions. Patients developed bacteremia a hydrosalpinx, 100 mg doxycy- come in this study (9). As patients median of 100 days after TIPS place- cline twice daily for 7 days undergoing TIPS creation typically ment (range, 6–732 d), well beyond the Level of evidence:4,5,8 have multisystem disease and do not effective period of antibiotic prophy- tolerate the sequelae of infection, laxis. Bacteremia resolved in all cases Transjugular Intrahepatic many interventionalists administer Portosystemic Shunt (TIPS) Creation after treatment with IV antibiotic agents antibiotic prophylaxis in this setting, (84). Although acute infection related to As noted by Beddy et al (13), tran- despite a relative lack of evidence TIPS creation appears to be uncommon, sjugular intrahepatic portosystemic from randomized controlled clinical and the value of prophylactic antibiotic shunt (TIPS) creation was originally trials (8). The survey of IR prophy- agents has not been demonstrated via Volume 21 Number 11 Venkatesan et al • 1619 randomized clinical trials, most investi- otic-resistant peristomal infections has Routine prophylaxis recommended:for gators continue to administer antibiotic been described, most commonly methi- pull technique, yes; for push tech- prophylaxis to patients before TIPS pro- cillin-resistant S aureus, attributed to na- nique, no consensus cedures (8,9,13). sopharyngeal colonization, which trav- First-choice antibiotic agent: pull tech- els along the pull-type percutaneous nique, 1 g cefazolin IV Procedure classification: clean; clean- endoscopic gastrostomy to the peris- Special considerations: (i) second-genera- contaminated tomal wound during gastrostomy tion cephalosporin (head/neck can- Organisms encountered: skin flora placement (91–94). In principle, the per- cer); (ii) second-generation cephalo- (S epidermidis, S aureus), Coryne- cutaneous push technique avoids tube sporin followed by an oral course bacterium species, biliary patho- passage through the oropharynx and of a first-generation cephalosporin gens, enteric Gram-negative rods, thereby prevents the deposition of mi- (head/neck cancer); (iii) if penicil- anaerobes, Enterococcus species croorganisms at the peristomal site. lin-allergic, can use vancomycin or Routine prophylaxis recommended:yes However, there are some data to sug- clindamycin First-choice antibiotic agent:no Level of evidence:2,3,5 consensus gest that, in head and neck cancer, there Common antibiotic choices: (i) 1 g is an increased risk of peristomal infec- ceftriaxone IV or (ii) 1.5–3 g am- tion and susceptibility to gastrostomy site infections irrespective of whether LIVER AND BILIARY picillin/sulbactam IV; (iii) if INTERVENTIONS penicillin-allergic, can use van- bacteria are introduced through the comycin or clindamycin and transoral or transesophageal route Biliary Drainage aminoglycoside (14,95). Cantwell et al (14) described a Level of evidence:2,4,5 15% rate of peristomal infection in a se- The normal unobstructed biliary ries of 57 patients with head and neck tree typically contains no bacteria (13). cancer undergoing percutaneous gas- However, in the setting of biliary dis- ϭ NONVASCULAR trostomy (n 53) or gastrojejunostomy ease, the biliary tract should be viewed ϭ INTERVENTIONS (n 4) tube placement (via the push or as contaminated (13,97). The incidence introducer technique). All instances of of infectious complication after biliary Fluoroscopically Guided peristomal infection occurred in patients drainage procedures has been re- Gastrostomy and Gastrojejunostomy who had not received antibiotic prophy- ported to vary between 24% and 46%, Tube Placement laxis (n ϭ 20). Patients receiving pro- with most interventional radiologists phylaxis (n ϭ 37) were administered 1 g routinely using antibiotic prophylaxis Fluoroscopically guided percutane- cefazolin IV and twice-daily cephalexin for biliary drainage (13). In the survey ous radiologic gastrostomy was first de- 500 mg for 5 days orally or via gastros- of IR prophylaxis regimens by Dravid scribed in 1981 (14,85). Since that time, tomy (n ϭ 35), or clindamycin 600 mg et al (8), 89% of respondents indicated technical modifications have been IV and 600 mg twice daily orally or via that they always used prophylaxis for made, including use of larger-bore tubes gastrostomy for 5 days (n ϭ 2) (14). De- biliary drainage, with 53% of nonusers (20–24 F vs initial use of 9–16 F tubes) spite these findings, there remains con- reporting infective complications after and the introduction of gastropexy de- troversy with regard to prophylaxis for biliary drainage. Preprocedural biliary vices in 1986 (13,14,86). At present, the push (ie, introducer) gastrostomy place- cultures have been reported to be use- two most common methods of gastros- ment, as additional published data sug- ful in planning the antibiotic strategy, tomy tube placement are the “push” gest no added benefit to antibiotic pro- although few interventionalists rou- method, which involves percutaneous phylaxis in this setting (95). Shastri et al tinely perform them (13). When ob- introducer technique through the ante- (96) performed a prospective random- tained, the most common isolates in- rior abdominal wall, and the more com- ized double-blind placebo controlled clude Enterococcus species, as well as monly performed “pull” method, which trial in which 97 patients undergoing yeast, Gram-negative aerobic bacilli, traverses the oropharynx. Infectious introducer percutaneous endoscopic and Streptococcus viridans (9,98). Se- complications of gastrostomy tube gastrostomy placement for malignant lected strains associated with mortal- placement are most commonly peris- oropharyngeal cancers were random- ity include E coli and Clostridium spe- tomal infections. The pull placement ized to receive placebo or prophylaxis (2 cies, which have been isolated in as technique has been associated with a g ceftriaxone IV). Infection rates were many as 75% of cases of fatal biliary high incidence of peristomal infectious low in both groups, with clinically sig- sepsis (9,99). Risk factors for bacterial complications (4%–30%), resulting in nificant wound infection observed in colonization in patients with biliary recommendations for routine prophy- one patient in each group during the obstruction undergoing percutaneous lactic antibiotics for this procedure (87– immediate 7-day postprocedure fol- biliary drainage include periproce- 89). A prospective, randomized trial low-up (96). dural fever, previous biliary instru- of prophylactic antibiotic agents (90) mentation, and bilioenteric anastomo- showed that cefazolin (1 g IV) reduced Procedure classification: clean-con- sis (9,98). the rate of infection from 28.6% to 7.4% taminated. Although a majority of intervention- in the setting of endoscopically placed Organisms encountered: skin flora (S epi- alists report antibiotic prophylaxis be- (ie, pull-type) gastrostomy tubes. Of in- dermidis, S aureus), Corynebacterium fore biliary drainage, there is variation terest, an increasing incidence of antibi- species regarding the regimen advocated. Clark 1620 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR et al (100) performed a prospective cies; Klebsiella, Pseudomonas, and ported in as many as 7% of patients study of 480 biliary procedures, ranging Bacteroides species, particularly undergoing nephrostomy drainage from simple tube injection to primary in cases of advanced biliary dis- for pyonephrosis (7,103–106). Coch- biliary drainage tube placement, in ease, including hepatolithiasis ran et al (104) stratified patients into which all patients received 1 g cefotetan Routine prophylaxis recommended: yes a high-risk group (aforementioned IV for prophylaxis. Forty-two patients First-choice antibiotic agent:no risk factors) and a low-risk group developed an increased white blood cell consensus (none of the aforementioned risk fac- count or fever, with seven (2%) devel- Common antibiotic choices: (i) 1 g ceftri- tors) (104). In the low-risk group, oping overt sepsis; greater infection risk axone IV; (ii)1.5–3 g ampicillin/sul- 14% developed evidence of sepsis was observed in patients who had un- bactam IV; (iii)1gcefotetan IV plus when prophylactic antibiotic drugs dergone biliary intervention previously 4 g mezlocillin IV; (iv) 2 g ampicillin were not given; 10% developed evi- (9,100). As a result, these authors have IV plus 1.5 mg/kg gentamicin IV; dence of sepsis when prophylactic advocated the addition of4gIVme- (v) if penicillin-allergic, can use antibiotic drugs were given (no sta- zlocillin to cefotetan in patients with a vancomycin or clindamycin and tistical difference). In the high-risk history of biliary intervention (100). aminoglycoside group, these figures were 50% and Other investigators advocate prophy- Level of evidence:5 9%, respectively (104). Despite these laxis with third-generation cephalospo- results, many authors recommend rins, given these antibiotic agents’ en- GU PROCEDURES the use of a prophylactic antibiotic hanced biliary excretion compared with agent for all patients, irrespective of second-generation agents (9). Ryan et al Percutaneous Nephrostomy Tube risk, with a majority of intervention- (9) advocate the use of 1 g ceftriaxone Placement, Tube Exchange, Ureteral alists reporting routine antibiotic IV, citing its long duration of activity Stents prophylaxis for GU procedures, and and its favorable dosing schedule. Am- nonusers reporting a 40% incidence picillin/sulbactam is considered by Percutaneous nephrostomy and ureteric stent implantation are com- of infective complications (8,13). An many the optimal antibiotic agent for exception to this practice includes prophylaxis in the setting of biliary in- mon IR procedures (13,102). In gen- eral, procedures performed in the the routine tube change in unin- tervention, given its activity against En- fected and unobstructed immuno- terococcus species (9). GU tract are regarded as clean- competent cases, in which routine As noted by Ryan et al (9), patients contaminated when they are per- prophylaxis has not been described with advanced biliary disease, including formed on an unobstructed system, in the literature, and is not war- those with hepatolithiasis, are at signif- with no previous infection and no ranted (7). Patients who have signs icant risk of liver abscess formation, sec- history of intervention (7). Factors of infection, or who are classified ondary biliary cirrhosis, portal hyper- predisposing to infection include ad- into a high risk group as mentioned tension, and death from sepsis or vanced age, diabetes, bladder dys- earlier, should be treated with an ap- hepatic failure. In these patients, the function, indwelling catheter, earlier boundary between prophylaxis and manipulation, ureterointestinal anas- propriate antibiotic agent before in- therapy is becomes blurred. Effective tomosis, bacteriuria, and stones tervention (9). As in biliary interven- antimicrobial therapy is as crucial a (7,9,13,103,104). In these patients, the tions, antibiotic treatment should be component of therapy as relief of bile GU tract should be managed as con- continued to treat obstructed sys- stasis by IR manipulation or clearance of taminated (7). The GU tract is re- tems until relief of obstruction is biliary stone burden (9). Sheen-Chen et garded as dirty if clinical infection is achieved. The results of urine cul- al (101) demonstrated the presence of present; in these instances, antibiotic tures should be used to tailor ongo- bacteria in the bile of all patients with therapy rather than prophylaxis is ing therapy (9). hepatolithiasis, most commonly Gram- the appropriate aim (7). The most common organisms to infect the GU Procedure classification: clean-con- negative bacteria such as Klebsiella spe- taminated; contaminated cies, E coli, and Pseudomonas, Enterococ- tract are Gram-negative rods (E coli, Proteus species, and Klebsiella spe- Organisms encountered: E coli and cus, and Bacteroides species, the latter Proteus, Klebsiella, and Enterococ- being the most frequently found anaer- cies) and Enterococcus species (7). If culture results are not available, am- cus species obes (9,101). As is the case for other IR Routine prophylaxis recommended: interventions, antibiotic treatments picillin combined with gentamicin or sulbactam is appropriate, as is ceftri- yes (except for routine tube should be adjusted according to the re- change in uninfected patients) sults of bacteriologic cultures, with anti- axone (7,13). Cronan et al (102) de- scribed an incidence of bacteremia of First-choice antibiotic: no consensus biotic therapy continued in the setting of 17% in patients undergoing nephros- Common antibiotic choices: (i) 1 g ce- obstruction, until relief of the obstruc- tomy tube exchange, with no differ- fazolin IV; (ii) 1 g ceftriaxone IV; tion has been achieved (9,13,101). ence in incidence between patients (ii) 1.5–3 g ampicillin/sulbac- Procedure classification: clean-con- receiving preprocedural antibiotic tam IV; (iv) 2 g ampicillin IV taminated; contaminated agents compared with those who did and 1.5 mg/kg gentamicin IV; Organisms encountered: Enterococcus not (102). Serious infectious compli- (v) If penicillin-allergic, can use species, Candida species, Gram- cations are encountered in patients vancomycin or clindamycin and negative aerobic bacilli, S viri- undergoing treatment for urinary aminoglycoside dans, E coli, and Clostridium spe- obstruction, with septic shock re- Level of evidence:4,5 Volume 21 Number 11 Venkatesan et al • 1621

TUMOR ABLATION risk. In the absence of definitive sci- absence of existing culture data. In the entific evidence, many practitioners otherwise asymptomatic patient, there Percutaneous tumor ablation, in- continue to empirically use prophy- is potential benefit to avoiding unnec- cluding radiofrequency (RF) abla- Ͻ laxis (13,64,112). essary wide-spectrum antibiotic cover- tion, has effectively treated small ( age by awaiting the results of culture Procedure classification (site-dependent): 3 cm) liver lesions, and has shown and sensitivity of drainage specimens clean; clean-contaminated (eg, bil- promise in the treatment of lung, re- when available. The most common ioenteric anastomosis/bypass) nal, and adrenal tumors (13,107– bacteria found in intraabdominal ab- Organisms encountered (organ-depen- 110). The issue of prophylactic anti- scesses are Gram-negative rods and dent): generally S aureus, S epider- biotic agents for tumor ablation is anaerobes, particularly E coli, Bacte- midis, Streptococcus species with controversial, with some operators roides fragilis, and Enterococcus species or without E coli; in cases of pre- administering them universally and (7,8,113). Pyogenic liver abscesses are vious bilioenteric anastomosis, others only in selected cases (13). most often caused by Enterobacter spe- consider organisms similar to There have been no randomized con- cies and anaerobes (7,114). Older anti- those for liver/biliary interven- trolled trials on antibiotic agent use biotic regimens used in this setting tion, eg, E coli, Proteus species, in patients undergoing RF ablation; at have included ampicillin, gentamicin, Klebsiella species, and Enterococ- present, most of the published data per- and metronidazole, but more common cus species taining to this topic relate to the per- current regimes include second- or Routine prophylaxis recommended:no sonal experience of various groups third-generation cephalosporins, such consensus (13). Infections after tumor ablation as cefoxitin1gIVevery 6 hours, ceftri- First-choice antibiotic agent:no have been mostly described in the axone1gIVevery 24 hours, or ampi- consensus setting of hepatic RF ablation, with cillin/sulbactam 3 g IV every 6 hours. Common antibiotic choices: (i) 1.5 g complications including cholangitis A combination of clindamycin and ampicillin/sulbactam IV (liver); and liver abscess being rare (re- gentamicin may be used if the patient Ͻ (ii) 1 g ceftriaxone IV (renal); ported in 1.5% of cases) but poten- has a severe penicillin allergy (8). tially severe (13,111,112). Instances (iii) 1 g cefazolin IV (bone); alter- in which patients are at an increased nate choices (site-dependent), Procedure classification: dirty risk of hepatic abscess warrant care- (iv) if penicillin-allergic, can sub- Organisms encountered: skin flora, S epi- ful evaluation, and include histories stitute vancomycin or clindamy- dermidis, S aureus, Corynebacterium of bilioenteric anastomosis, biliary cin for Gram-positive coverage; species; intracavitary pathogens, stent placement, and sphincterot- aminoglycoside for Gram-nega- Gram-negative bacteria, Enterococ- omy, all of which lead to retrograde tive coverage cus species, E coli, B fragilis, other enteric bacterial communication Level of evidence:8 anaerobes with the biliary tract (13). As noted Routine prophylaxis recommended: by Choi et al (111), the mechanism of yes abscess formation is thought to re- PERCUTANEOUS ABSCESS First-choice antibiotic agent:no sult from bacterial colonization and DRAINAGE consensus Common antibiotic choices: (i) 1–2 g growth in the zone of ablation (111). Patients referred for percutaneous cefoxitin IV every 6 hours; Before RF ablation of hepatic lesions abscess drainage are typically already (ii) 1–2 g cefotetan IV every 12 h; in routine cases, investigators have being treated with antibiotic drugs (7). (iii) 1 g ceftriaxone IV every 24 h; recommended 1.5 g IV ampicillin/ When patients are referred for percu- (iv) 3 g ampicillin/sulbactam IV sulbactam (13). There remains no taneous abscess drainage and are not every 6 h; (v) if penicillin-aller- consensus on the effectiveness of already being treated with antibiotic gic, can use vancomycin or clin- prophylactic antibiotic agents for pa- drugs, the IR physician should evalu- damycin for Gram-positive tients undergoing RF ablation of ate the appropriateness of empiric an- coverage; aminoglycoside for liver, lung, adrenal, renal, or other tibiotic therapy before obtaining cul- Gram-negative coverage solid lesions (13). Indeed, the occur- ture data via catheter drainage. In the Level of evidence:8 rence of delayed abscess formation absence of published data on this sub- weeks after tumor ablation despite ject, it is the opinion of the authors that prophylactic antibiotic coverage empiric antibiotic coverage before ab- PERCUTANEOUS BIOPSY makes superinfection of the thermal scess drainage should be reserved for lesion a possibility, given that ther- patients who present with clinical As noted by McDermott et al (7), mally damaged tissue may provide signs and symptoms of infection (eg, image-guided biopsies do not require an especially favorable environment fever, leukocytosis) at the time of the antibiotic prophylaxis unless per- for bacterial growth (109). This phe- drainage procedure. Admittedly, anti- formed via the transrectal route. The nomenon highlights the importance biotic agent use in this setting spans most common indication for the tran- of postprocedural follow-up and sur- the boundary between prophylaxis srectal approach is prostate biopsy (7). veillance to ensure timely and appro- and treatment. As abscesses are typi- Suggested prophylaxis for this proce- priate antibiotic therapy, in addition cally polymicrobial, broad-spectrum dure includes preprocedural gentami- to up-front prophylaxis in patients at antibiotic agents are warranted in the cin 80 mg intramuscularly (IM) 30 1622 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR minutes before the procedure, fol- Procedure classification: clean hives) (13). However, if a patient has lowed by a 5-day course of oral cipro- Organisms encountered: skin flora (S had an anaphylactic reaction to peni- floxacin 250 mg twice daily (7). Sieber epidermidis, S aureus, Corynebacte- cillins, they should never receive pen- et al (113) advocate the use of oral rium species) icillin. In addition, use of cephalospo- ciprofloxacin 500 mg twice daily for 4 Routine prophylaxis recommended: rins in these patients should be days, commencing the day before bi- yes approached with caution because of opsy. In their retrospective review of First-choice antibiotic agent:(i)1g the 15% cross-hypersensitivity with 4,439 biopsies performed with the use cefazolin IV these agents (9). However, it is advis- of this antibiotic regimen (113), infec- Alternate choices: (ii) if penicillin-al- able in this setting to query hospital tive complications (ie, urinary tract in- lergic, can use vancomycin or records to determine if the patient has fection) occurred in a minority of pa- clindamycin ever received a cephalosporin in the tients (n ϭ 5). Level of evidence:8 past without allergic reaction. If so, a Procedure classification: nontransrectal, cephalosporin can be used in the fu- clean; transrectal, contaminated ture. Although monobactams (eg, az- MANAGEMENT OF treonam) are structurally related to Organisms encountered: transrectal, ␤ Gram-negative bacteria, Entero- ANTIBIOTIC -lactam agents, they are unrelated in coccus species, E coli, B fragilis, HYPERSENSITIVITY terms of allergic potential (9). There is no cross-reactivity between monobac- other anaerobes Patients frequently state that they tam and ␤-lactam agents, and these Routine prophylaxis recommended: have a penicillin allergy (9). However, drugs may be used safely in patients yes (transrectal route); no (non- penicillin allergies are only uncom- with anaphylactic reactions to ␤-lac- transrectal route) monly manifested as an anaphylactic tam agents (9,120). However, it should First-choice antibiotic agent: transrec- reaction. Macpherson et al (118) stud- be noted that there is cross-reactivity tal, no consensus ied 1,260 patients attending a pread- between carbapenem (eg, imipenem, Common antibiotic choices: transrec- mission clinic before routine surgery, meropenem) and ␤-lactam agents. tal, (i) 80 mg gentamicin IM plus among whom 22% described an anti- 250 mg ciprofloxacin twice daily biotic allergic reaction. After review of orally for 5 days; (ii) 500 mg cip- the patient medical charts and on fur- rofloxacin twice daily orally for 4 MANAGEMENT OF ther follow-up questioning, the major- VALVULAR HEART DISEASE days, commencing the day be- ity of “allergies” were known side ef- fore biopsy fects of the drugs (118). In all patients Current AHA recommendations for Level of evidence:4 who had been administered an alter- antibiotic prophylaxis against infec- native agent as a result of a history of tive endocarditis (IE) are for those car- previous “penicillin allergy,” none diac conditions associated with the PERCUTANEOUS were found to have had a previous highest risk of adverse outcomes from VERTEBROPLASTY allergic reaction (13,118). Patients who endocarditis (121). These include cases Vertebroplasty is employed to treat have a questionable penicillin allergy, of prosthetic cardiac valve or pros- the pain associated with osteoporotic or have had only fever or rash, may be thetic material used for cardiac valve compression fractures and pathologic safely given ␤-lactam antibiotic agents repair; previous IE; unrepaired cya- vertebral compression (13). Postproce- without fear of anaphylaxis (9,13). If a notic congenital heart disease, includ- dure infection is rare in this setting, patient has a penicillin allergy, the cli- ing palliative shunts and conduits; and clinical data from randomized tri- nician should determine whether it is completely repaired congenital heart als are currently lacking. However, of an anaphylactic or nonanaphylactic defect with prosthetic material or de- when infection does occur in this set- variety (9). Two percent of the popu- vice, whether placed by surgery or by ting, surgical debridement of the in- lation has some degree of penicillin catheter intervention during the first 6 fected vertebral body can be techni- hypersensitivity, but most reactions to months after the procedure; repaired cally very difficult (13,115,116). The ␤-lactam agents are nonanaphylactic congenital heart disease with residual common pathogens are from skin and usually manifest clinically as a defects at the site or adjacent to the site flora. Prophylactic antibiotics are mild maculopapular rash or fever (9). of a prosthetic patch or prosthetic de- therefore commonly administered to Such considerations are not insignifi- vice (which inhibit endothelialization); these patients, with some investigators cant, as the cost of prescribing alterna- and cardiac transplantation recipients adding antibiotics to the polymethyl tive agents to a patient with a docu- who develop cardiac valvulopathy methacrylate before intraosseous in- mented penicillin allergy can be more (121). Notably, except for these condi- jection, although the efficacy of the lat- than double the cost of an appropriate tions, IE prophylaxis is not recom- ter has not been proven (13,117). An penicillin and generally results in the mended for any other form of congen- appropriate prophylactic regimen use of more broad-spectrum agents, ital heart disease, including mitral against the involved pathogens in- which are more likely to confer antibi- valve prolapse (121). volves 1 g cefazolin IV 30 minutes be- otic resistance (13,119). Penicillin aller- For patients at high risk who un- fore the procedure (13). If the patient is gies are only uncommonly caused by dergo an invasive respiratory tract allergic to penicillin or cephalosporins, an anaphylactic reaction (eg, broncho- procedure to treat an established in- vancomycin may be administered (13). spasm, laryngospasm, hypotension, or fection, such as drainage of an ab Volume 21 Number 11 Venkatesan et al • 1623 scess or empyema, the antibiotic reg- at high risk who have an established domized controlled data to help for- imen administered to these patients GI or GU tract infection or for those mulate the prophylactic regimens for should contain an agent active who receive antibiotic therapy to pre- interventional procedures. Clinical against S viridans (121). Suggested vent wound infection or sepsis associ- practice is largely informed, at regimens include 2 g amoxicillin ated with a GI or GU tract procedure is present, via existing surgical data orally, 2 g ampicillin IM or IV, or 1 g antienterococcal coverage considered and interventional radiology cohort cefazolin IM or IV or 1 g ceftriaxone reasonable, such as with penicillin, studies and clinical series. Given the IM or IV (121). In patients allergic to ampicillin, piperacillin, or vancomycin ability of antibiotic resistance to ob- penicillin or ampicillin and able to (121). However, as noted by the cur- viate historically effective regimens, take oral medications, 2 g cephalexin rent AHA guideline authors (121), no an ongoing review of current pri- orally or 600 mg clindamycin orally published studies demonstrate that mary surgical and IR literature on or 500 mg azithromycin orally or 500 such therapy would prevent entero- this topic by interventional radiolo- mg clarithromycin orally may be ad- coccal IE. For patients at high risk gists remains crucial. It is equally im- ministered (121). In patients allergic scheduled for an elective ureteral stent portant to remain abreast of current to penicillins or ampicillin and un- placement or other urinary tract ma- hospital and/or area practice. These able to take oral medication, 1g cefa- nipulation who have an enterococcal measures underscore a timely and zolin IM or IV or 1 g ceftriaxone IM urinary tract infection or colonization, evidence-based approach to antibi- or IV or 600 mg clindamycin IM or IV antibiotic therapy to eradicate Entero- otic prophylaxis for interventional may be administered (121). If the in- coccus species from the urine before procedures. fection is known or suspected to be the procedure is considered reason- caused by S aureus, the regimen should able (121). Amoxicillin or ampicillin is Acknowledgments: contain an agent active against S aureus, the preferred agent for enterococcal Dr. Aradhana M. Venkatesan authored the first draft of such as an antistaphylococcal penicillin coverage for these patients. Vancomy- this document and served as topic leader or cephalosporin, or vancomycin in pa- cin may be administered to patients during the revisions of the draft. Dr. San- tients unable to tolerate a ␤-lactam agent unable to tolerate ampicillin (121). joy Kundu is chair of the SIR Standards of (121). Vancomycin should be adminis- For patients at high risk who un- Practice Committee. Dr. John F. Cardella tered if the infection is known or sus- dergo an invasive procedure that in- is Councilor of the SIR Standards Divi- pected to be caused by a methicillin re- volves infected skin, skin structure, sion. All other authors are listed alpha- sistant strain of S aureus (121). or musculoskeletal tissue, the thera- betically. Other members of the Stan- In a departure from previous rec- peutic regimen administered for dards of Practice Committee and SIR who ommendations, the AHA indicated treatment of the infection should participated in the development of this clinical practice guideline are as follows in their 2007 guidelines (121) that ad- contain an agent active against Staph- ␤ (listed alphabetically): John “Fritz” An- ministration of prophylactic antibi- ylococcus species and -hemolytic gle, MD, Ganesh Annamalai, MD, Ste- otic agents solely to prevent endocar- Streptococcus species, such as an an- phen Balter, PhD, Daniel B. Brown, MD, ditis should not be recommended for tistaphylococcal penicillin or a ceph- Danny Chan, MD, Christine P. Chao, patients who undergo GU or GI tract alosporin (121). Vancomycin or clin- MD, Horacio D’Agostino, MD, Sean R. procedures. Reasons cited for this damycin may be administered to Dariushia, MD, Brian D. Davison, MD, departure include the lack of studies patients unable to tolerate a ␤-lactam Debra Ann Gervais, MD, S. Nahum Gold- showing a possible association between agent or who are known or sus- berg, MD, Alan T. Hirsch, MD, Eric J. GI or GU tract procedures and IE pected to have an infection caused Hohenwalter, MD, Sanjeeva P. Kalva, (122,123). As indicated by Wilson et al by a methicillin-resistant strain of MD, Arshad Ahmed Khan, MD, Venkat- aramu Krishnamurthy, MD, Ashish Ma- (121), the cases of IE temporally associ- Staphylococcus (121). hajan, MD, Patrick C. Malloy, MD, Gloria ated with a GI or GU tract procedure M. Martinez-Salazar, MD, J. Kevin have been largely anecdotal, with a sin- CONCLUSIONS McGraw, MD, Philip M. Meyers, MD, gle or very small number of cases re- Steven F. Millward, MD, Charles A. ported. No published data demon- Effective antibiotic prophylaxis Owens, MD, Darren Postoak, MD, Anne strate a conclusive link between for vascular and IR requires a thor- C. Roberts, MD, Wael A. Saad, MD, Ta- procedures of the GI or GU tract and ough knowledge of likely pathogens, run Sabharwal, MD, Marc R. Sapoval, the development of IE (121). Also cited procedure-specific infection risks, MD, Cindy Kaiser Saiter, NP, Marc S. is the increase in Enterococcus species and appropriate antibiotic coverage Schwartzberg, MD, Samir S. Shah, MD, Nasir H. Siddiqi, MD, Raymond H. resistant to penicillins, vancomycin, (7). Choice of medication should also Thornton, MD, Patricia E. Thorpe, MD, T. and aminoglycosides (122–124). These take into account patient-specific fac- Gregory Walker, MD, Ronald R. Ward- were antibiotics recommended for IE tors, including renal and hepatic rope, RN, Darryl A. Zuckerman, MD, prophylaxis in previous AHA guide- function and antibiotic allergy his- Curtis W. Bakal, MD, Curtis A. Lewis, lines (125). Therefore, only in patients tory. There continues to be few ran- MD, MBA, JD, Kenneth S. Rholl, MD.

APPENDIX A: CONSENSUS METHODOLOGY Reported complication-specific rates in some cases reflect the aggregate of major and minor complications. Thresholds are derived from critical evaluation of the literature, evaluation of empirical data from Standards of Practice Committee members’ practices, and, when available, the SIR HI-IQ System national database. 1624 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

Consensus on statements in this document was obtained with a modified Delphi technique (1–3).

APPENDIX B

Levels of Evidence and Classes of Recommendation to Adhere to Definitions Established by the American College of Cardiology/American Heart Association Guidelines Task Force (4,5) Level of Evidence Definition 1 Randomized clinical trials or metaanalyses of multiple clinical trials with substantial treatment effects 2 Randomized clinical trials with smaller or less significant treatment effects 3 Prospective, controlled, nonrandomized, cohort studies 4 Historic, nonrandomized, cohort, or case-control studies 5 Case series: patients compiled in serial fashion, lacking a control group 6 Animal studies or mechanical model studies 7 Extrapolations from existing data collected for other purposes, theoretical analyses 8 Rational conjecture (common sense); common practices accepted before evidence-based guidelines

APPENDIX C: SIR STANDARDS OF PRACTICE COMMITTEE CLASSIFICATION OF COMPLICATIONS BY OUTCOME Minor Complications A. No therapy, no consequence B. Nominal therapy, no consequence; includes overnight admission (up to 23 hours) for observation only. Major Complications C. Require therapy, minor hospitalization (Ͼ or ϭ to 24 hrs, but Ͻ 48 hours) D. Require major therapy, unplanned increase in level of care, prolonged hospitalization (Ͼ 48 hours). E. Permanent adverse sequelae F. Death Volume 21 Number 11 Venkatesan et al • 1625

Practice Guidelines for Antibiotic Prophylaxis During Vascular and Interventional Radiology Procedures Potential Routine Organisms Prophylaxis First-choice Procedure Encountered Recommended Antibiotic Angiography, angioplasty, thrombolysis, S aureus, S epidermidis No None arterial closure device placement, stent placement Endograft placement S aureus, S epidermidis Yes Cefazolin1gIV Superficial venous insufficiency treatment S aureus, S epidermidis No None IVC filter placement S aureus, S epidermidis No None Tunneled central venous access S aureus, S epidermidis No consensus None

Embolization and chemoembolization (if S aureus, S epidermidis, Streptococcus Yes No consensus intent to create infarction or high spp, Corynebacterium spp, and/or likelihood of infarction) enteric flora (if prior sphincter of Oddi manipulation or bilioenteric surgery)

UAE S aureus, S epidermidis, Streptococcus Yes No consensus spp, and/or E coli

TIPS creation S aureus, S epidermidis, Yes No consensus Corynebacterium spp, biliary pathogens, enteric Gram- negative rods, anaerobes, Enterococcus spp Fluoroscopically guided gastrostomy and S aureus, S epidermidis, No consensus (if 1g cefazolin IV (pull gastrojejunostomy tube placement Corynebacterium spp introducer gastrostomy) “push” technique); Yes (if pull gastrostomy)

Liver and biliary interventions Enterococcus spp, Streptococcus spp, Yes No consensus aerobic Gram-negative organisms (E coli, Klebsiella spp, etc) Clostridium spp, Candida spp, and anaerobes GU procedures E coli, Proteus, Klebsiella, Yes No consensus Enterococcus

Tumor ablation S aureus, S epidermidis, Streptococcus No consensus No consensus spp, and/or E coli

Percutaneous abscess drainage S aureus, S epidermidis, Yes No consensus Corynebacterium spp; aerobic Gram-negative bacteria and anaerobes

Percutaneous biopsy Transrectal: bowel flora, mostly No Nontransrectal, none; anaerobes and aerobic Gram- (nontransrectal); transrectal, 80 mg negative, Streptococcus spp yes (transrectal) gentamicin IV/IM plus 250 mg ciprofloxacin twice daily orally for 5 d Percutaneous vertebroplasty S aureus, S epidermidis, Yes 1 g cefazolin IV Corynebacterium spp 1626 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

Practice Guidelines for Antibiotic Prophylaxis During Vascular and Interventional Radiology Procedures

Level of Common Antibiotic Choices Comments Evidence 1 g cefazolin IV (if high risk stent infection); if penicillin- Procedure classification: clean 5, 8 allergic, can use vancomycin or clindamycin

If penicillin-allergic, can use vancomycin or clindamycin Procedure classification: clean 5, 8 None Procedure classification: clean 8 None Procedure classification: clean 5, 6 1g cefazolin IV (eg, immunocompromised patients before Procedure classification: clean (nontunneled catheter: 8 chemotherapy; history of catheter infection); if no prophylaxis) penicillin-allergic, can use vancomycin or clindamycin 1.5–3 g ampicillin/sulbactam IV (hepatic Special considerations: if patient without intact 4, 7, 8 chemoembolization); 1 g cefazolin and 500 mg sphincter of Oddi, consider tazobactam/piperacillin metronidazole IV (hepatic chemoembolization); 2 g and bowel preparation; procedure classification: ampicillin IV and 1.5 mg/kg gentamicin (hepatic clean-contaminated; contaminated (bilioenteric chemoembolization); 1 g ceftriaxone IV (hepatic surgery) chemoembolization or renal, splenic embolization); if penicillin-allergic, use vancomycin or clindamycin and aminoglycoside (i) 1 g cefazolin IV; (ii) 900 mg clindamycin IV plus 1.5 If history of hydrosalpinx, doxycycline 100 mg twice 4, 5, 8 mg/kg gentamicin IV; (iii) 2 g ampicillin IV; (iv) 1.5–3 daily for 7 d; procedure classification, clean; clean- g ampicillin/sulbactam IV; if penicillin-allergic, can contaminated use vancomycin or clindamycin (i) 1 g ceftriaxone IV; (ii) 1.5-3 g ampicillin/sulbactam IV; Procedure classification: clean; clean-contaminated 2, 4, 5 if penicillin-allergic, can use vancomycin and aminoglycoside

— Procedure classification: clean-contaminated; special 3, 5 considerations: if head and neck cancer, consider (i) a second-generation cephalosporin; (ii) second- generation cephalosporin followed by oral course of first-generation cephalosporin, or (iii) clindamycin or vancomycin in penicillin-allergic patients (i) 1 g ceftriaxone IV; (ii) 1.5–3 g ampicillin/sulbactam Procedure classification: clean-contaminated, 5 IV; (iii) 1 g cefotetan IV and 4 g mezlocillin IV; (iv) contaminated 2 g ampicillin IV and 1.5 mg/kg gentamicin IV; (v) if penicillin-allergic, can use vancomycin or clindamycin and aminoglycoside (i) 1 g cefazolin IV; (ii) 1 g ceftriaxone IV; (iii) 1.5–3 g Procedure classification: clean-contaminated; 4, 5 ampicillin/sulbactam IV; (iv) 2 g ampicillin IV and contaminated; cover any organisms already found in 1.5 mg/kg gentamicin IV; (v) if penicillin-allergic, can urine use vancomycin or clindamycin and an aminoglycoside (i) 1.5–3 g ampicillin/sulbactam IV (liver); (ii) 1 g Procedure classification: (site-dependent) clean; clean- 8 ceftriaxone IV (renal); (iii) 1 g ceftriaxone IV (bone); contaminated (eg, bilioenteric anastomosis/bypass) (iv) if penicillin-allergic, can use vancomycin or clindamycin for Gram-positive coverage and aminoglycoside for Gram-negative coverage (i) 1–2 g cefoxitin IV every 6 h (ii) 1–2 g cefotetan IV Procedure classification: dirty; antibiotics should cover 8 every 12 h; (iii) 3 g ampicillin/sulbactam IV every anticipated organisms for empiric treatment and 6 h; (iv) 3.375 g piperacillin/tazobactam IV; if then adjusted for final culture results penicillin-allergic, can use vancomycin or clindamycin for Gram-positive coverage; aminoglycoside for Gram-negative coverage with or without metronidazole for anaerobic coverage 500 mg ciprofloxacin twice daily for 4 d commencing the Procedure classification: nontransrectal, clean; 4 day before biopsy transrectal, contaminated

If penicillin-allergic, can use clindamycin or vancomycin Procedure classification: clean 8 Volume 21 Number 11 Venkatesan et al • 1627

References 14. Cantwell C, Perumpillichira JJ, Maher 27. Kadir S. Cholangiography. In: Kadir 1. Sacks D, McClenny TE, Cardella JF, MM, et al. Antibiotic prophylaxis S, ed. Diagnostic angiography. Phila- Lewis CA. Society of Interventional for percutaneous radiologic gastros- delphia: Saunders, 1986; 642–678. Radiology clinical practice guidelines. tomy and gastrojejunostomy insertion 28. Haley RW, Culver DH, Morgan WM, J Vasc Interv Radiol 2003; 14(suppl): in outpatients with head and neck White JW, Emori TG, Hooton TM. S199–S202. cancer. J Vasc Interv Radiol 2008; 19: Identifying patients at high risk of 2. Fink A, Kosecoff J, Chassin M, Brook 571–575. surgical wound infection: a simple RH. Consensus methods: character- 15. Koc M, Zulfikaroglu B, Kece C, Ozalp multivariate index of patient suscep- istics and guidelines for use. Am J N. A prospective randomized study tibility and wound contamination. Public Health 1984; 74:979–983. of prophylactic antibiotics in elective Am J Epidemiol 1985; 12 I:206–I215. 3. Leape LL, Hilborne LH, Park RE, et al. laparoscopic cholecystectomy. Surg 29. Joint Commission on Hospital Ac- The appropriateness of use of coro- Endosc 2003; 17:1716–1718. creditation. 2009 National Patient nary artery bypass graft surgery in 16. Sanchez-Manuel FJ, Lozano-Garcia J, Safety Goals. Available at http://www. New York State. JAMA 1993; 269:753– Seco-Gil JL. Antibiotic prophylaxis jointcommission.org/GeneralPublic/ 760. for hernia repair. Cochrane Database NPSG/09_npsgs.htm. Accessed June 4. Gibbons RJ, Smith SC Jr, Antman E, Syst Rev 2003; 2:CD0033769. 24, 2010. et al. American College of Cardi- 17. Leone M, Albanese J, Tod M, et al. 30. Bratzler DW, Houck PM; Surgical In- ology/American Heart Association Ceftriaxone (1 g intravenously) pene- fection Prevention Guideline Writers clinical practice guidelines, part 1: tration into abdominal tissues when Workgrou. Antimicrobial prophy- where do they come from? Circula- laxis for surgery: an advisory state- tion 2003; 107:2979–2986. administered as antibiotic prophy- 5. Gibbons RJ, Smith SC Jr, Antman E, laxis during nephrectomy. J Che- ment from the National Surgical In- et al. American College of Cardi- mother 2003; 15:139–142. fection Prevention Project. Am J Surg ology/American Heart Association 18. Ad Hoc Committee of the Committee 2005; 189:395–404. clinical practice guidelines, part II: on Trauma, Division of Medical Sci- 31. Harbarth S, Samore MH, Lichtenberg evolutionary changes in a continuous ences, National Academy of Sciences, D, Carmeli Y. Prolonged antibiotic quality improvement project. Circula- National Research Council. Post- prophylaxis after cardiovascular sur- tion 2003; 107:3101–3107. operative wound infections: the influ- gery and its effect on surgical site in- 6. Spies JB, Rosen R, Lebowitz AS. ence of ultraviolet irradiation of the fections and antimicrobial resistance. Antibiotic prophylaxis in vascular operating room and various other fac- Circulation 2000; 101:2916–2921. and interventional radiology: a ratio- tors. Ann Surg 1964;160:1–192. 32. Jones RN, Pfaller MA. Bacterial re- nal approach. Radiology 1988; 166: 19. Zarrinpar A, Kerlan R. A guide to sistance: a worldwide problem. Diag 301–387. antibiotics for the interventional radi- Microbiol Infect Dis 1998; 31:379–388. 7. McDermott VG, Schuster MG, Smith ologist. Semin Interv Radiol 2005; 22: 33. Goldmann DA, Weinstein RA, Wen- T. Antibiotic prophylaxis in vascu- 69–79. zel RP, et al, Strategies to prevent lar and interventional radiology. AJR 20. Burke JF. The effective period of and control the emergence and spread Am J Roentgenol 1997; 169:31–38. preventive antibiotic action in experi- of antimicrobial-resistant organisms 8. Dravid VS, Gupta A, Zegel HG, Mo- mental incisions and dermal lesions. in hospitals: a challenge to hospital rales AV, Rabinowitz B, Freiman DB. Surgery 1961; 50:161–168. leadershi. JAMA 1996; 275:234–240. Investigation of antibiotic prophylaxis 21. Classen DC, Evans RS, Pestotnik SL, 34. Shawker TH, Kluge RM, Ayella RJ. usage for vascular and nonvascular Horn SD, Menlove RL, Burke J. The Bacteremia associated with angiogra- interventional procedures. J Vasc In- timing of prophylactic administration phy. JAMA 1974; 229:1090–1092. terv Radiol 1998; 9:401–406. of antibiotics and the risk of surgical 35. Sande MA, Levinson ME, Lukas DS, 9. Ryan JM, Ryan BM, Smith T. wound infection. N Engl J Med 1992; Kaye D. Bacteremia associated with Antibiotic prophylaxis in interven- 326:281–286. cardiac catheterization. N Engl J Med tional radiology. J Vasc Interv Radiol 22. Stone HH, Haney BB, Kolb LD, Gehe- 1969; 281:1104–1106. 2004; 15:547–556. ber CE, Dawkins EJ. Antibiotic pro- 36. Frazee BW, Flaherty J. Septic endar- 10. Patel S, Tuite CM, Mondschein JI, phylaxis in gastric, biliary and colonic teritis of the femoral artery following Soulen MC. Effectiveness of an ag- surgery. Ann Surg 1976; 184:443–452. angioplasty. Rev Infect Dis 1990; 13: gressive antibiotic regimen for che- 23. Stone HH, Haney BB, Kolb LD, Gehe- 620–623. moembolization in patients with pre- ber CE, Hooper CA. Prophylactic 37. Franco J, Motaganahalli R, Habeeb M, vious biliary intervention. J Vasc and preventive antibiotic therapy, Wittgen C, Peterson G. Risk factors Interv Radiol 2006; 17:1931–1934. for infectious complications with an- 11. Kim W, Clark TW, Baum RA, Soulen timing, duration and economics. Ann MC. Risk factors for liver abscess af- Surg 1979; 189:691–699. gio-seal percutaneous vascular clo- ter hepatic chemoembolization. J Vasc 24. Nelson CL, Green TG, Porter RA, sure devices. Vascular 2009; 17:218– Interv Radiol 2001; 12:965–968. Warren RD. One day versus seven 221. 12. Geschwind JF, Kaushik S, Ramsey days of preventive antibiotic therapy 38. Whitton Hollis H Jr, Rehring TF. DE, Choti MA, Fishman EK, Kobeiter in orthopedic surgery. Clin Orthop Femoral endarteritis associated with H. Influence of a new prophylactic Relat Res 1983; 176:258–263. percutaneous suture closure: new antibiotic therapy on the incidence of 25. DiPiro JT, Cheung RP, Bowden TA Jr, technology, challenging complica- liver abscesses after chemoemboliza- Mansberger JA. Single dose sys- tions. J Vasc Surg 2003; 38:83–87. tion treatment of liver tumors. J Vasc temic antibiotic prophylaxis of surgi- 39. Park Y, Roh HG, Choo, SW, et al. Interv Radiol 2002; 13:1163–1166. cal wound infections. Am J Surg 1986; Prospective comparison of collagen 13. Beddy P, Ryan BM. Antibiotic pro- 152:552–559. plug (Angio-seal) and suture-medi- phylaxis in interventional radiology- 26. Thadepalli H, Mandal AK. Anti- ated (the Closer S) closure devices at anything new? Tech Vasc Interv Ra- biotic prophylaxis in the surgical pa- femoral access sites. Korean J Radiol diol 2006; 9:69–76. tient. Infect Med 1998; 6:71–80. 2005; 6:248–255. 1628 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

40. Dieter RS. Coronary artery stent terventional Radiology, Cardiovascular noma. Am J Gastroenterology 1997; infection. Clin Cardiol 2000; 23:808– Interventional Radiological Society of 92:2257–2259. 810. Europe, American College of Phlebology, 64. de Baere T, Roche A, Amenabar JM, et 41. Chalmers N, Eadington DW, Gandan- and Canadian Interventional Radiology al. Liver abscess formation after lo- hamo D, Gillespie IN, Ruckley CV. Association. J Vasc Interv Radiol 2010; cal treatment of liver tumors. Hepa- Case report: infected false aneurysm 21:14–31. tology 1996; 23:1436–1440. at the site of an iliac stent. Br J Radiol 53. Millward SF, Peterson RA, Moher D, 65. Song SY,Chung JW, Han JK, et al. 1993; 66:946–948. et al. LGM (Vena Tech) vena caval Liver abscess after transcatheter oily 42. Liu P, Dravid V, Freiman D, Zegel H, filter: experience at a single institu- chemoembolization for hepatic tu- Weinberg D. Persistent iliac endar- tion. J Vasc Interv Radiol 1994; 5:351– mors: incidence, predisposing factors, teritis with pseudoaneurysm forma- 356. and clinical outcome. J Vasc Interv tion following balloon-expandable 54. Peyton JW, Hylemon MB, Greenfield Radiol 2001; 12:312–320. stent placement. Cardiovasc Intervent LJ, Crute SL, Sugerman H, Quereshi 66. Spigos DG, Jonasson O, Mozes M, Ca- Radiol 1995; 18:39–42. GD. Comparison of Greenfield and pek V. Partial splenic embolization 43. Deiparine MK, Ballard JK, Taylor FC, vena caval ligation for experimental in the treatment of hypersplenism. Chase DR. Endovascular stent infec- septic thromboembolism. Surgery AJR Am J Roentgenol 1979; 132:777– tion. J Vasc Surg 1996; 23:529–533. 1983; 93:533–537. 782. 44. Bunt TJ, Gill HK, Smith DC, Taylor 55. Miller DL, O’Grady N Guidelines 67. Brown DB, Cardella JF, Sacks D, et al. FC. Infection of a chronically im- for the prevention of intravascular Quality improvement guidelines for planted iliac artery stent. Ann Vasc catheter-related infections: recom- transhepatic arterial chemoemboliza- Surg 1997; 11:529–532. mendations relevant to interventional tion, embolization, and chemothera- 45. Malek AM, Higashida RT, Phatouros radiology. J Vasc Interv Radiol 2003; peutic infusion for hepatic malig- CC, et al. Endovascular manage- 14:133–136. nancy. J Vasc Interv Radiol 2009; 20 ment of extracranial carotid artery 56. van de Wetering MD, van Woensel JB. (suppl):S219–S226. dissection achieved using stent angio- Prophylactic antibiotics for prevent- 68. Atassi B, Bangash AK, Bahrani A, et plasty. Cardiovasc Intervent Radiol ing early central venous catheter al. Multimodality imaging follow- 2000; 23:57–60. Gram positive infections in oncology ing 90Y radioembolization: a compre- 46. Leroy O, Martin E, Prat A, et al. patients. Cochrane Database Syst Rev hensive review and pictorial essay. Fatal infection of coronary stent im- 2007; 1:CD003295. Radiographics 2008; 28:81–99. plantation. Cathet Cardiovasc Diagn 57. Philipneri M, Al-Aly Z, Amin K, 69. Mascarenhas NB, Mulcahy MF, Le- 1996; 39:168–170. Gellens ME, Bastani B. Routine re- wandowski RJ, Salem R, Ryu RK. 47. Deitch JS, Hansen KJ, Regan JD, placement of tunneled, cuffed, hemo- Hepatic abscess after yttrium-90 ra- Burkhart JM, Ligush J Jr. Infected dialysis catheters eliminates paraspi- dioembolization for islet-cell tumor renal artery pseudoaneurysm and nal/vertebral infections in patients hepatic metastasis. Cardiovasc Inter- mycotic aortic aneurysm after percu- with catheter-associated bacteremia. vent Radiol 2010; 33:650–653. taneous transluminal renal artery Am J Nephrol 2003; 23:202–207. 70. Ravina JH, Herbreteau D, Ciraru- angioplasty and stent placement in a 58. Brodwater BK, Silbers JS, Smith TP, et Vigneron N, et al. Arterial embolisa- patient with a solitary kidney. J Vasc al. Conversion of indwelling chest tion to treat uterine myomata. Lancet Surg 1998; 28:340–344. port catheters to tunneled central ve- 1995; 346:671–672. 48. McCready RA, Siderys H, Pittman JN, nous catheters. J Vasc Interv Radiol 71. Walker WJ, Pelage J. Uterine artery et al. Septic complications after car- 2000; 11:1137–1142. embolization in symptomatic fibroids: diac catheterization and percutaneous 59. Loo S, vanHeerden PV, Gollege CL, clinical results in 400 women with im- transluminal coronary angioplasty. J Roberts BL, Power BM. Infection in aging follow up. Br J Obstet Gynaecol Vasc Surg 1991; 14:170–174. central lines: antiseptic-impregnanted 2002; 109:1262–1272. 49. Becker GJ, Kovacs M, Mathison MN, vs. standard non-impregnated cathe- 72. Pron G, Mocarski E, Cohen M, et al. et al. Risk stratification and out- ters. Anaesth Intens Care 1997; 25: Hysterectomy for complications after comes of transluminal endografting 637–639. uterine artery embolization for for abdominal aortic aneurysm: 60. Meyer P, Reizine D, Aymard A, leiomyoma: results of a Canadian 7-year experience and long-term fol- Guerin JM, Merland JJ, Habib Y. multicenter clinical trial. J Am Assoc low-u. J Vasc Interv Radiol 2001; 12: Septic complications in interventional Gynecol Laparosc 2003; 10:99–106. 1033–1046. radiology: evaluation of risk and pre- 73. de Blok S, de Vries C, Prinssen HM, 50. Katzen BT, Becker GJ,Mascioli CA, et ventive measures: preliminary stud- Blaauwgeers HL, Jorna-Meijer LB. al. Creation of a modified angiogra- ies. J Intervent Radiol 1988; 3:73–75. Fatal sepsis after uterine artery embo- phy (endovascular) suite for translumi- 61. Reed RA, Teitelbaum GP, Daniels JR, lization with microspheres. J Vasc In- nal endograft placement and combined Pentecost MJ, Katz MD. Prevalence terv Radiol 2003; 14:779–783. interventional-surgical procedures. J of infection following hepatic chemo- 74. Vashisht A, Studd J, Carey A, Burn P. Vasc Interv Radiol 1996; 7:161–167. embolization with cross-linked colla- Fatal septicemia after fibroid emboli- 51. Heikkinen L, Valtonen M, Lepantalo gen with administration of prophylac- zation. Lancet 1999; 354:307–308. M, Saimanen E, Jarvinen A. Infra- tic antibiotics. J Vasc Interv Radiol 75. Pelage JP, Jacob D, LeDref O, La- renal endoluminal bifurcated stent 1994; 5:367–371. combe P, Laurent A. Fatal sepsis af- graft infected with Listeria monocyto- 62. Hemingway AP, Allison D. Com- ter uterine artery embolization with genes. J Vasc Surg 1999; 29:554–556. plications of embolization: analysis of microspheres. J Vasc Interv Radiol 52. Khilnani NM, Grassi CJ, Kundu S, et 410 procedures. Radiology 1988; 166: 2004; 15:405–406. al. Multi-society consensus quality 669–672. 76. Walker WJ, Pelage JP, Sutton C. improvement guidelines for the treat- 63. Chen C, Chen PJ, Yang PM, et al. Fibroid embolisation. Clin Radiol ment of lower-extremity superficial ve- Clinical and microbiological features 2002; 57:325–331. nous insufficiency with endovenous of liver abscess after transarterial em- 77. Pron G, Bennett J, Common A, et al. thermal ablation from the Society of In- bolization for hepatocellular carci- Technical results and effects of oper- Volume 21 Number 11 Venkatesan et al • 1629

ator experience on uterine artery blind clinical trial. Ann Intern Med 104. Cochran ST, Barbaric ZL, Lee, JJ, embolization for fibroids: the Ontario 1987; 107:824–828 Kashfian P. Nephrostomy tube Uterine Fibroid Embolization Trial. 91. Hull M, Beane A, Bowen J, Settle C. placement: an outpatient procedure? The Ontario UFE Collaborative Group. Methicillin-resistant Staphylococcus Radiology 1991; 179:843–847. J Vasc Interv Radiol 2003; 14:545–554. aureus infection of percutaneous en- 105. Yoder IC, Pfister RC, Lindfors KK, 78. Rajan DK, Beecroft JR, Clark TW, et al. doscopic gastrostomy sites. Aliment Newhouse JH. Pyonephroses: imag- Risk of intrauterine infectious compli- Pharmacol Ther 2001; 15:1883–1888. ing and intervention. AJR Am J cations after uterine artery emboliza- 92. Maetani I, Tada T, Ukita T, Inoue H, Roentgenol 1983:141:735–740. tion. J Vasc Interv Radiol 2004; 15: Sakai Y, Yoshikawa M. PEG with in- 106. Larsen EH, Gasser TC, Madsen PO. 1415–1421. troducer or pull method: a prospec- Antibiotic prophylaxis in urologic 79. Mehta H, Sandhu C, Matson M, Belli tive randomized comparison. Gastroi- surgery. Urol Clin North Am 1986; AM. Review of readmissions due to ntest Endosc 2003; 57:837–841. 13:591–604. complications from uterine fibroid 93. Maetani I, Yasuda M, Seike M, et al. 107. Goldberg SN, Solbiati L, Hahn PF, et embolization. Clin Radiol 2002; 57: Efficacy of an overtube for reducing al. Large-volume tissue ablation 1122–1124. the risk of peristomal infection after with radiofrequency by using a clus- 80. Walker WA, Green A, Sutton C. PEG placement: a prospective, ran- tered, internally cooled electrode Bilateral uterine artery embolization domized comparison study. Gastroin- technique: laboratory and clinical ex- for myomata: results, complications test Endosc 2005; 61:522–527. perience in liver metastases. Radiol- and failures. Min Invas Ther Allied 94. Horiuchi A, Nakayama Y, Kajiyama ogy 1998; 209:371–379. Technol 1999; 8:449–454. M, Fujii H, Tanaka N. Naso- 108. Solbiati L, Ierace T, Goldberg SN, et 81. Saxon RR. A new era for transjugu- pharyngeal decolonization of methi- al. Percutaneous US-guided radio- lar intrahepatic portosystemic shunts. cillin-resistant Staphylococcus aureus frequency tissue ablation of liver me- J Vasc Interv Radiol 2004; 15:217–219. can reduce PEG peristomal wound in- tastases: treatment and follow up in 82. Sanyal AJ, Reddy KR. Vegetative in- fection. Am J Gastroenterol 2006; 101: 16 patients. Radiology 1997; 202:195– fection of transjugular intrahepatic 274–277. 203. portosystemic shunts. Gastroenterol- 95. Wollman B, D’Agostino HB. Per- 109. Wood BJ, Abraham J, Hvizda JL, Al- ogy 1998; 115:110–115. cutaneous radiologic and endoscopic exander JR, Fojo T. Radiofrequency 83. Deibert P, Schwarz S, Olschewski M, gastrostomy: a 3-year institutional ablation of adrenal tumors and adre- Siegerstetter V, Blum HE, Rossle M. analysis of procedure performance. nocortical carcinoma metastases. Can- Risk factors and prevention of early AJR Am J Roentgenol 1997; 169:1551– cer 2002; 97:557–560. infection after implantation or revi- 1553. 110. Shibata T, Yamamoto Y, Yamamoto sion of transjugular intrahepatic por- N, et al. Cholangitis and liver ab- 96. Shastri YM, Hoepffner N, Tessmer A, tosystemic shunts: results of a ran- scess after percutaneous ablation ther- Ackermann H, Schroeder O, Stein J. domized study. Dig Dis Sci 1998; 43: apy for liver tumors: incidence and New introducer PEG gastropexy does 1708–1713. risk factors. J Vasc Interv Radiol 2003; not require prophylactic antibiotics: 84. DeSimone JA, Beavis KG, Eschelman 14:1535–1542. multicenter prospective randomized DJ, Henning KJ. Sustained bactere- 111. Choi D, Lim HK, Kim MJ, et al. double-blind placebo-controlled study. mia associated with transjugular intra- Liver abscess after percutaneous ra- Gastrointest Endosc 2008; 67:620–628. hepatic portosystemic shunt (TIPS). diofrequency ablation for hepatocel- Clin Infect Dis 2000; 30:384–386. 97. Yee AC, Ho CS. Complications of lular carcinoma: frequency and risk 85. Preshaw RM. A percutaneous percutaneous drainage: benign versus factors. AJR Am J Roentgenol 2005; method for inserting a feeding gas- malignant diseases. AJR Am J Roent- 184:1860–1867. trostomy tube. Surg Gynecol Obstet genol 1987; 148:1207–1209. 112. Dupuy DE, Goldberg SN. Image- 1981; 152:658–660. 98. Brody LA, Brown KT, Getrajdman GI, guided radiofrequency tumor abla- 86. Brown AS, Mueller PR, Ferrucci JT Jr. et al. Clinical factors associated with tion: challenges and opportunities— Controlled percutaneous gastrostomy positive bile cultures during primary part II. J Vasc Interv Radiol 2001; 12: nylon T-fastener for fixation of the an- percutaneous biliary drainage. J Vasc 1135–1148. terior gastric wall. Radiology 1986; 158: Interv Radiol 1998; 9:572–578. 113. Sieber PR, Rommel FM, Agusta VE, 543–545. 99. Condon RE, Wittmann DH. The use Breslin JA, Huffnagle HW, Harpster 87. Schapiro G, Edmundowicz SA. of antibiotics in general surgery. Curr LE. Antibiotic prophylaxis in ultra- Complications of percutaneous endo- Probl Surg 1991; 28:801–949. sound guided transrectal prostate bi- scopic gastrostomy. Gastrointest En- 100. Clark CD, Picus D, Dunagan WC. opsy. J Urol 1997; 157:2199–2000. dosc Clin N Am 1996; 6:409–422. Bloodstream infections after interven- 114. Lorber B, Swenson RM. The bacteri- 88. McClave SA, Chang WK. Com- tional procedures in the biliary tract. ology of intraabdominal infections. plications of enteral access. Gastroin- Radiology 1994; 191:495–499. Surg Clin North Am 1975; 55:1349– test Endosc Clin N Am 2003; 58:739– 101. Sheen-Chen S, Chen W, Eng H, et al. 1354. 751. Bacteriology and antimicrobial choice 115. Alfonso Olmos M, Silva Gonzalez A, 89. American Society for Gastrointestinal in hepatolithiasis. Am J Infect Control Duart Clemente J, Villas Tome C. Endoscopy. Infection control during 2000; 28:298–301. Infected vertebroplasty due to un- gastrointestinal endoscopy: guide- 102. Cronan JJ, Horn DL, Marcello A, et al. common bacteria solved surgically: a lines for clinical application. From the Antibiotics and nephrostomy tube rare and threatening life complication ASGE. American Society for Gastroin- care: preliminary observations. Part of a common procedure: report of a testinal Endoscopy. Gastrointest En- II: bacteremia. Radiology 1989; 172: case and a review of the literature. dosc 1999;49:836–841. 1043–1045. Spine 2006; 31:E770–E773. 90. Jain NK, Larson DE, Schroeder KW, et 103. Goodwin WE, Casey WC, Wolf W. 116. Walker DH, Mummaneni P, Rodts GE al. Antibiotic prophylaxis for percu- Percutaneous trocar (needle) nephros- Jr. Infected vertebroplasty: report of taneous endoscopic gastrostomy. A tomy in hydronephrosis. JAMA 1955; two cases and review of the literature. prospective, randomized, double- 157:891–894. Neurosurg Focus 2004; 17:E6. 1630 • Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR

117. Mathis JM, Wong W. Percutaneous ditis: guidelines from the American 123. Shay DK, Maloney SA, Montecalvo vertebroplasty: technical consider- Heart Association. A guideline from M, et al. Epidemiology and mortal- ation. J Vasc Interv Radiol 2003; 14: the American Heart Association ity risk of vancomycin-resistant en- 953–960. Rheumatic Fever, Endocarditis, and terococcal bloodstream infections. 118. MacPherson RD, Willcox C, Chow C, Kawasaki Disease Committee, J Infect Dis 1995; 172:993–1000. Wang A. Anaesthetist’s response to Council on Cardiovascular Disease 124. Lucas GM, Lechtzin N, Puryear DW, patients’ self-reported drug allergies. in the Young, and the Council on Yau LL, Flexner CW, Moore R. Br J Anaesth 2006; 97:634–639. Clinical Cardiology, Council on Vancomycin-resistant and vancomy- 119. Borsch JE, Andersen KE, Bindslev- Cardiovascular Surgery and Anes- cin-susceptible enterococcal bactere- Jensen C. The prevalence of sus- thesia, and the Quality of Care and mia: comparison of clinical features pected and challenge-verified penicil- Outcomes Research Interdiscipli- and outcomes. Clin Infect Dis 1998; lin allergy in a university hospital population. Basic Clin Pharmacol nary Working Group. Circulation 26:1127–1133. Toxicol 2006; 98:357–362. 2007; 116:1736–1754. 125. Dajani AS, Taubert KA, Wilson W, et 120. Cunha BA. Antimicrobial selection 122. Strom BL, Abrutyn E, Berlin JA, et al. al. Prevention of bacterial endocar- in the penicillin-allergic patient. Risk factors for infective endocarditis: ditis: recommendations by the Amer- Drugs Today 2001; 37:377–383. oral hygiene and nondental expo- ican Heart Association. JAMA 1997; 121. Wilson W, Taubert KA, Gewitz M, et sures. Circulation 2000; 102:2842– 277:1794–1801. al. Prevention of infective endocar- 2848.

SIR DISCLAIMER The clinical practice guidelines of the Society of Interventional Radiology attempt to define practice principles that generally should assist in producing high quality medical care. These guidelines are voluntary and are not rules. A physician may deviate from these guidelines, as necessitated by the individual patient and available resources. These practice guidelines should not be deemed inclusive of all proper methods of care or exclusive of other methods of care that are reasonably directed towards the same result. Other sources of information may be used in conjunction with these principles to produce a process leading to high quality medical care. The ultimate judgment regarding the conduct of any specific procedure or course of management must be made by the physician, who should consider all circumstances relevant to the individual clinical situation. Adherence to the SIR Quality Improvement Program will not assure a successful outcome in every situation. It is prudent to document the rationale for any deviation from the suggested practice guidelines in the department policies and procedure manual or in the patient’s medical record.