The American Journal of Emergency Medicine

Volume 25, Issue 1, Pages.1-131 (January 2007)

1. Masthead Page A2

2. Editorial Board Page A4

Table of Contents 3. Pages A6-A8

4. Manuscript Submission and Editorial Review Policy Pages A10-A11

Original Contributions

5. Antithrombotic therapy for prevention of stroke in clinical practice in the ED Pages 1-5 Asia Kogan, Reuma Shapira, Ada Tamir and Gad Rennert

6. Bilevel positive airway pressure in the treatment of status asthmaticus in pediatrics Pages 6-9 Sara L. Beers, Thomas J. Abramo, Andrea Bracken and Robert A. Wiebe

7. Is there a link between hyperbilirubinemia and elevated urine nitrite Pages 10-14 Susan Watts, David Bryan and Keith Marill

8. Hospital-based healthcare provider (nurse and physician) integration into an emergency medical services–managed mass-gathering event Pages 15-22 Christian Martin-Gill, William J. Brady, Kevin Barlotta, Anthony Yoder, Allen Williamson, Benjamin Sojka, Dayton Haugh, Marcus L. Martin, Marge Sidebottom and Leonard Sandridge 9. How can we identify low- and high-risk patients among unselected patients with possible acute coronary syndrome? Pages 23-31 Kirsten Melgaard Nielsen, Ole Faergeman, Mogens Lytken Larsen and Anders Foldspang

10. Prevalence of elevated blood pressure in 563 704 adult patients with stroke presenting to the ED in the United States Pages 32-38 Adnan I. Qureshi, Mustapha A. Ezzeddine, Abu Nasar, M. Fareed K. Suri, Jawad F. Kirmani, Haitham M. Hussein, Afshin A. Divani and Alluru S. Reddi

11. Impact of a negative prior stress test on emergency physician disposition decision in ED patients with chest pain syndromes Pages 39-44 Rebecca H. Nerenberg, Frances S. Shofer, Jennifer L. Robey, Aaron M. Brown and Judd E. Hollander

Brief Reports

12. Discrepancies in interpretation of ED body computed tomographic scans by radiology residents Pages 45-48 Nelson Tieng, Diana Grinberg and Siu Fai Li

13. Factors associated with myocardial infarction after emergency endoscopy for upper gastrointestinal bleeding in high-risk patients: a prospective observational study Pages 49-52 Ching-Tai Lee, Shih-Pei Huang, Tsu-Yao Cheng, Tsung-Hsien Chiang, Chi-Ming Tai, Wei-Chih Su, Chien-Hua Huang, Jaw-Town Lin and Hsiu-Po Wang

14. Endotracheal intracuff pressures in the ED and prehospital setting: is there a problem? Pages 53-56 James E. Svenson, M. Bruce Lindsay and Jill E. O'Connor

15. Onset of symptoms after methadone overdose Pages 57-59 Frank LoVecchio, Anthony Pizon, Brad Riley, Azadeh Sami and Carmella D'Incognito

16. Sumatriptan for the treatment of undifferentiated primary headaches in the ED Pages 60-64 James R. Miner, Stephen W. Smith, Johanna Moore and Michelle Biros

Diagnostics

17. Prognostic implications of myocardial necrosis triad markers' concentration measured at admission in patients with suspected acute coronary syndrome Pages 65-68 Filip M. Szymański, Marcin Grabowski, Krzysztof J. Filipiak, Grzegorz Karpiński, Anna Hrynkiewicz, Przemysław Stolarz, Artur Oręziak, Robert Rudowski and Grzegorz Opolski

18. Delayed traumatic thoracic spinal epidural hematoma: a case report and literature review Pages 69-71 Cheng-Ta Hsieh, Yung-Hsiao Chiang, Chi-Tun Tang, Jui-Ming Sun and Da-Tong Ju

19. Hypertrophic cardiomyopathy: electrocardiographic manifestations and other important considerations for the emergency physician Pages 72-79 Brian S. Kelly, Amal Mattu and William J. Brady

Clinical Notes

20. Ultrasound-guided hip arthrocentesis in the ED Pages 80-86 Kalev Freeman, Andreas Dewitz and William E. Baker

Case Reports

21. Unusual presentation of herpes simplex virus encephalitis: bilateral thalamic involvement and normal imaging of early stage of the disease Pages 87-89 Hakan Gümüş, Sefer Kumandaş, Hüseyin Per, Fulya Tahan, Esad Köklü and Musa Karakükçü

22. Subcortical stroke presenting as micrographia Pages 89-90 Mark A. Marinella

23. Acute paraplegia caused by an accidental ingestion of hydrogen peroxide Pages 90-92 Te-Ming Liu, Kuo-Chin Wu, Ko-Chi Niu and Hung-Jung Lin

24. Takotsubo cardiomyopathy: an unusual syndrome mimicking an ST-elevation myocardial infarction Pages 92-95 Tom E. Kolkebeck, Casey L. Cotant and Richard A. Krasuski

25. Anaphylaxis to topical bacitracin ointment Pages 95-96 Karen Greenberg, James Espinosa and Victor Scali

Laceration of the popliteal artery and compartment syndrome resulting from 26. stingray envenomation Pages 96-97 Charlotte Derr, Barbara J. O'Connor and Sandra L. MacLeod

27. An unusual primary blast injury: Traumatic brain injury due to primary blast injury Pages 97-98 Serkan Yılmaz and Murat Pekdemir

28. Gastric outlet obstruction in an infant: lactobezoar Pages 98-99 Robert J. Graham and Phil Stein

29. An unusual case of pediatric abdominal distension Pages 99-101 Christopher T.Y. Hsu, Maria Carmen G. Diaz and David Rappaport

30. Massive tissue emphysema after cardiopulmonary resuscitation Pages 101-102 Raffi Bekeredjian, F. Joachim Meyer, Norbert Frey and Boris Radeleff

31. Succinylcholine-induced masseter muscle rigidity associated with rapid sequence intubation Pages 102-104 Christopher S. Roman and Alex Rosin

32. Pontine hemorrhage and isolated abducens nerve palsy Pages 104-105 Scott C. Sherman and Babak Saadatmand

33. ED treatment of brainstem anesthesia after retrobulbar block Pages 105-106 James M. Dahle and Kenneth V. Iserson

34. Fatal afebrile streptococcal meningitis in a chronic alcoholic patient Pages 106-108 Ramsis F. Ghaly, William P. Gibbons and Timothy P. Plackett

35. Atrial flutter with cardiac tamponade as initial presentation of tuberculosis pericarditis Pages 108-110 I-Chuan Chen, Te-Fa Chiu and Jih-Chang Chen

36. Importance of bedside echocardiography for detection of unsuspected isolated right ventricular infarction as a cause of cardiovascular collapse Pages 110-114 Mitja Lainscak and Andrej Pernat

37. Acute myocardial infarction diagnosed early by multidetector computed tomography Pages 114-116 Ju-Sing Fan, Chorng-Kuang How, Chii-Hwa Chern, Lee-Min Wang, Chun-I Huang and Jen-Dar Chen

38. Life-threatening epistaxis from a rare treatable cause Pages 116-118 Mohammed Iqbal Syed, Andy Chin and Louise Clark

39. Spontaneous intracranial hemorrhage caused by sustained hypertension in a child Pages 118-120 Jainn-Jim Lin, Chang-Teng Wu, Shao-Hsuan Hsia, Kuang-Lin Lin and Huei- Shyong Wang

40. A rare diagnosis in ED: cerebral pyogenic ventriculitis due to infective endocarditis Pages 120-122 Selahattin Kiyan, Ersin Aksay, Murat Ersel and Sedat Yanturali

41. Should early venous oximetry be indicated in suspected cases of fulminant myocarditis? Pages 122-123 Arvind K. Baronia, Afzal Azim, Gagan Narula, Mohan Gurjar, Banani Poddar, Sudeep Kumar, Sanjay Gambhir and Sukanta Barai

Correspondence

42. Thiamine for the treatment of acute decompensated heart failure Pages 124-126 Howard A. Smithline

43. Early diagnosis of gas-forming pyometra in an aged patient can prevent mortality Pages 126-127 Chia-Chun Tai, Wan-Ching Lien, Hsiu-Po Wang and Kao-Lang Liu

44. Image pitfall of computed tomography in diagnosis of aortic dissection Pages 127-129 Wen-Chu Chiang, Pei-Chieh Kao, Chan-Ping Su and Juan Hs

45. Other causes of unilateral pulmonary edema Pages 129-131 Ritesh Agarwal, Ashutosh N. Aggarwal and Dheeraj Gupta

The American Journal of Emergency Medicine

THE AMERICAN JOURNAL OF EMERGENCY MEDICINE (ISSN any means now or hereafter known, electronic or mechanical, 0735-6757) is published seven times a year by Elsevier Inc., including photocopy, recording, or any information storage and 360 Park Avenue South, New York, NY 10010-1710. Months retrieval system, without permission in writing from the Pub- of issue are: January, March, May, July, September, October lisher. Printed in the United States of America. and November. Business and Editorial Offices: 1600 John PERMISSIONS: Permissions may be sought directly from F. Kennedy Blvd., Ste. 1800, Philadelphia, PA 19103-2899. Elsevier’s Rights Department in Philadelphia, PA, USA: phone Customer Service Office: 6277 Sea Harbor Drive, Orlando, 215-239-3600, e-mail [email protected]. Requests FL 32887-4800. Periodicals postage paid at New York, NY and may also be completed on-line via the Elsevier homepage (www. additional mailing offices. elsevier.com/locate/permissions). POSTMASTER: Send address changes to THE AMERICAN JOUR- REPRINTS: For 100 or more copies of an article in this NAL OF EMERGENCY MEDICINE, Elsevier Customer Service, 6277 Sea publication, please contact the Commercial Reprints Depart- Harbor Drive, Orlando, FL 32887-4800. ment, Elsevier Science Inc., 360 Park Avenue South, New Manuscripts, correspondence, and editorial material should York, New York 10010-1710. Tel. (212) 633-3813 Fax: be sent to The Editor, THE AMERICAN JOURNAL OF EMERGENCY (212) 633-3820 e-mail: [email protected] MEDICINE, P.O. Box 1494, West Bethesda, MD 20827-1494. ADVERTISING: Display Advertising and related corres- For enquiries relating to the submission of articles (in- pondence should be addressed to Inez Herrero, 360 Park Avenue, cluding electronic submission where available) please visit South, New York, NY 10010. Tel: 212-633-3122 / Fax: 212-633- Elsevier’s Author Gateway at http://authors.elsevier.com. The 3820; E-mail: [email protected]. Classified & Recruitment Author Gateway also provides the facility to track accepted Advertising and related correspondence should be addressed to articles and set up e-mail alerts to inform you of when Simone Imbert, 360 Park Avenue, South, New York, NY 10010. Tel: an article’s status has changed, as well as detailed artwork 212-462-1908 / Fax: 212-633-3820; E-mail: [email protected] guidelines, copyright information, frequently asked questions The ideas and opinions expressed in THE AMERICAN JOURNAL OF and more. EMERGENCY MEDICINE do not necessarily reflect those of the Editor Contact details for questions arising after acceptance of an or the Publisher. Publication of an advertisement or other product article, especially those relating to proofs, are provided after mention in THE AMERICAN JOURNAL OF EMERGENCY MEDICINE should registration of an article for publication. not be construed as an endorsement of the product or the manufac- YEARLY SUBSCRIPTION RATES: United States and pos- turer’s claims. Readers are encouraged to contact the manufacturer sessions: individual, $262.00; institution, $384.00; student and resi- with any questions about the features or limitations of the products dent, $123.00. All other countries: individual, $378.00; institution, mentioned. The Publisher does not assume any responsibility for $501.00; student and resident, $189.00. For all areas outside the any injury and/or damage to persons or property arising out of or United States and possessions, there is no additional charge for related to any use of the material contained in this periodical. The surface delivery. To receive student/ resident rate, orders must reader is advised to check the appropriate medical literature and the be accompanied by name of affiliated institution, date of term, product information currently provided by the manufacturer of and the signature of program/residency coordinator on institution each drug to be administered to verify the dosage, the method and letterhead. Orders will be billed at individual rate until proof of duration of administration, or contraindications. It is the responsi- status is received. bility of the treating physician or other health care professional, Prices are subject to change without notice. Current prices relying on independent experience and knowledge of the patient, to are in effect for back volumes and back issues. Single issues, determine drug dosages and the best treatment for the patient. both current and back, exist in limited quantities and are offered The appearance of the code at the bottom of the first page for sale subject to availability. Back issues sold in conjunction of an article in this journal indicates the copyright owner’s with a subscription are on a prorated basis. consent that copies of the article may be made for personal or Checks should be made payable to Elsevier and sent to internal use, or for the personal or internal use of specific THE AMERICAN JOURNAL OF EMERGENCY MEDICINE, W.B. Saun- clients, for those registered with the Copyright Clearance Cen- ders, Periodicals Department, PO Box 628239, Orlando, FL ter, Inc (222 Rosewood Drive, Danvers, MA 01923; (978) 32862-8239. 750-8400; www.copyright.com). This consent is given on Correspondence regarding subscriptions or change of address the condition that the copier pay the stated per-copy fee for should be directed to THE AMERICAN JOURNAL OF EMERGENCY that article through the Copyright Clearance Center, Inc for MEDICINE, Elsevier, Periodicals Department, 6277 Sea Harbor Dr, copying beyond that permitted by Sections 107 or 108 of the Orlando, FL 32887-4800. Telephone number, (800) 654-2452; out- US Copyright Law. This consent does not extend to other side the United States and Canada, (407) 345-4000. Changes of kinds of copying, such as copying for general distribution, for address should be sent preferably 60 days before the new address advertising or promotional purposes, for creating new collec- becomes effective. Missing issues will be replaced free of charge tive works, or for resale. Absence of the code indicates that the if the Publisher is notified at the above address within 2 months of material may not be processed through the Copyright publication of the issue for US and Canadian subscribers and within Clearance Center, Inc. 4 months for subscribers from all other countries. THE AMERICAN JOURNAL OF EMERGENCY MEDICINE is indexed © 2007 Elsevier Inc. All rights reserved. No part of this and abstracted in Index Medicus, EMBASE/Excerpta Medica, publication may be reproduced or transmitted in any form or by Current Concepts/Clinical Medicine, ISI/BIOMED, and BIOSIS.

Elsevier, John F. Kennedy Blvd, Philadelphia, PA 19103-2899. Director, Journals Production Production Editor Michael T. Miller Karen Stover Executive Publisher Publisher Christine Rullo Theresa Monturano The American Journal of Emergency Medicine

EDITOR J. Douglas White, MD, MPH, MBA, Medical College of Virginia/VCU, Richmond

EDITORIAL BOARD William J. Brady, MD, University of Virginia, Richard M. Nowak, MD, MBA, Henry Ford Hospital, Charlottesville Detroit Neal E. Flomenbaum, MD, Cornell Medical Center, New York Jonathan Olshaker, MD, Boston University, Boston Glenn C. Hamilton, MD, Wright State University, Joseph P. Ornato, MD, Medical College of Virginia, Dayton Richmond Gabor D. Kelen, MD, Johns Hopkins University, Baltimore Norman A. Paradis, MD, University of Colorado, Denver Toby L. Litovitz, MD, Georgetown University, Howard A. Werman, MD, Ohio State University, Columbus Washington, DC Charles J. McCabe, MD, Massachusetts General Loren Yamamoto, MD, MPH, MBA, University of Hawaii, Hospital, Boston Honolulu

EDITORIAL CONSULTANTS 2006 Neal Abarbanell, MD Michael Heller, MD Patrick Ray, MD David Amponsah, MD Johan Herlitz, MD, PhD Philip Rice, MD Joel Bartfield, MD C. James Holliman, MD Alfred Sacchetti, MD Steven L. Bernstein, MD James F. Holmes, MD, MPH Philip Salen, MD Paul Biddinger, MD Stephen Huff, MD David M. Schreck, MD, MS Polly Bijur, MD Fredric Husty, MD William Scruggs, MD Michael Blaivas, MD Ken Iserson, MD Donna Seeger, MD Judith Brillman, MD Jeanne Jacoby, MD Philip Shayne, MD Sean Bush, MD Gary Josephsen, MD Ronald Sing, MD Christopher R. Carpenter, MD Marshall Kapp, MD Adam Singer, MD Jeffrey Caterino, MD Lawrence Edward Kass, MD David E. Slattery, MD William Chiang, MD Ijaz Khan, MD Corey M. Slovis, MD Richard Christensen, MD, MA Bruce Klein, MD Richard Sobel, MD William Cordell, MD Wendy Klein-Schwartz, PharmD Matthew Spencer, MD Frank Counselman, MD Michael Kontos, MD Tom Stair, MD Cameron Crandall, MD John G. Laffey, MD LG Stead, MD Sandra J. Cunningham, MD Jerroid Leikin, MD Milton Tenenbein, MD Daniel Davis, MD Philip D. Levy, MD Kevin M. Terrell, DO, MS Robert Derlet, MD Siu Fai Li, MD Stephen Thomas, MD Deborah Dierecks, MD Joseph Losek, MD Joseph Varon, MD Charles Emerman, MD Frank Lovecchio, MD Arvind VenKat, MD Amy Ernst, MD Michael Lyons, MD Gary Michael Vilke, MD Lorrie Garces, MD Scott Melanson, MD Rade Vukmir, MD, JD Leslie A. Geddes, ME, PhD James R. Miner, MD Terry Walman, MD Nina Gentile, MD Antonio Muniz, MD Daniel Walsh, MD Louis Graff, MD Kristen E. Nordenholz, MD Richard Wersman, MD Colin Graham, MD David Overton, MD Howard Werman, MD Steven Green, MD Manish Patel, MD Michael D. Witting, MD Eric A. Gross, MD W. Frank Peacock, MD Tim Wolfe, MD Blaine Hannafin, MD Andrew Perron, MD Allan B. Wolfson, MD Raymond G. Hart, MD, MPH Michael Phelan, MD Keith Wrenn, MD Mark Hauswald, MD Jesse M. Pines, MD, MBA Huiyun Xiang, MD Kennon Heard, MD Rumen D. Powers, MD Leslie Zun, MD The American Journal of Emergency Medicine Contents VOLUME 25 • NUMBER 1 • JANUARY 2007

ORIGINAL CONTRIBUTIONS Antithrombotic Therapy for Prevention of Stroke in Clinical Practice in the ED 1 Asia Kogan, Reuma Shapira, Ada Tamir, and Gad Rennert Bilevel Positive Airway Pressure in the Treatment of Status Asthmaticus in Pediatrics 6 Sara L. Beers, Thomas J. Abramo, Andrea Bracken, and Robert A. Wiebe Is There a Link Between Hyperbilirubinemia and Elevated Urine Nitrite 10 Susan Watts, David Bryan, and Keith Marill Hospital-Based Healthcare Provider (Nurse and Physician) Integration Into an Emergency 15 Medical Services–Managed Mass-Gathering Event Christian Martin-Gill, William J. Brady, Kevin Barlotta, Anthony Yoder, Allen Williamson, Benjamin Sojka, Dayton Haugh, Marcus L. Martin, Marge Sidebottom, and Leonard Sandridge How Can We Identify Low- and High-Risk Patients Among Unselected Patients With Possible 23 Acute Coronary Syndrome? Kirsten Melgaard Nielsen, Ole Faergeman, Mogens Lytken Larsen, and Anders Foldspang Prevalence of Elevated Blood Pressure in 563 704 Adult Patients With Stroke Presenting 32 to the ED in the United States Adnan I. Qureshi, Mustapha A. Ezzeddine, Abu Nasar, M. Fareed K. Suri, Jawad F. Kirmani, Haitham M. Hussein, Afshin A. Divani, and Alluru S. Reddi Impact of a Negative Prior Stress Test on Emergency Physician Disposition Decision in ED 39 Patients With Chest Pain Syndromes Rebecca H. Nerenberg, Frances S. Shofer, Jennifer L. Robey, Aaron M. Brown, and Judd E. Hollander BRIEF REPORTS Discrepancies in Interpretation of ED Body Computed Tomographic Scans by Radiology Residents 45 Nelson Tieng, Diana Grinberg, and Siu Fai Li Factors Associated With Myocardial Infarction After Emergency Endoscopy for Upper 49 Gastrointestinal Bleeding in High-Risk Patients: A Prospective Observational Study Ching-Tai Lee, Shih-Pei Huang, Tsu-Yao Cheng, Tsung-Hsien Chiang, Chi-Ming Tai, Wei-Chih Su, Chien-Hua Huang, Jaw-Town Lin, and Hsiu-Po Wang Endotracheal Intracuff Pressures in the ED and Prehospital Setting: Is There a Problem? 53 James E. Svenson, M. Bruce Lindsay, and Jill E. O’Connor Onset of Symptoms After Methadone Overdose Frank LoVecchio, Anthony Pizon, 57 Brad Riley, Azadeh Sami, and Carmella D’Incognito Sumatriptan for the Treatment of Undifferentiated Primary Headaches in the ED 60 James R. Miner, Stephen W. Smith, Johanna Moore, and Michelle Biros DIAGNOSTICS Prognostic Implications of Myocardial Necrosis Triad Markers’ Concentration Measured at 65 Admission in Patients With Suspected Acute Coronary Syndrome Filip M. Szyman´ski, Marcin Grabowski, Krzysztof J. Filipiak, Grzegorz Karpin´ski, Anna Hrynkiewicz, Przemyslaw Stolarz, Artur Ore˛ziak, Robert Rudowski, and Grzegorz Opolski

(Continued on next page) Contents continued Delayed Traumatic Thoracic Spinal Epidural Hematoma: A Case Report and Literature Review 69 Cheng-Ta Hsieh, Yung-Hsiao Chiang, Chi-Tun Tang, Jui-Ming Sun, and Da-Tong Ju Hypertrophic Cardiomyopathy: Electrocardiographic Manifestations and Other Important 72 Considerations for the Emergency Physician Brian S. Kelly, Amal Mattu, and William J. Brady CLINICAL NOTES Ultrasound-Guided Hip Arthrocentesis in the ED Kalev Freeman, Andreas Dewitz, 80 and William E. Baker CASE REPORTS Unusual Presentation of Herpes Simplex Virus Encephalitis: Bilateral Thalamic Involvement 87 and Normal Imaging of Early Stage of the Disease Hakan Gümüs¸, Sefer Kumandas¸, Hüseyin Per, Fulya Tahan, Esad Köklü, and Musa Karakükçü Subcortical Stroke Presenting as Micrographia Mark A. Marinella 89 Acute Paraplegia Caused by an Accidental Ingestion of Hydrogen Peroxide 90 Te-Ming Liu, Kuo-Chin Wu, Ko-Chi Niu, and Hung-Jung Lin Takotsubo Cardiomyopathy: An Unusual Syndrome Mimicking an ST-Elevation 92 Myocardial Infarction Tom E. Kolkebeck, Casey L. Cotant, and Richard A. Krasuski Anaphylaxis to Topical Bacitracin Ointment Karen Greenberg, James Espinosa, 95 and Victor Scali Laceration of the Popliteal Artery and Compartment Syndrome Resulting From 96 Stingray Envenomation Charlotte Derr, Barbara J. O’Connor, and Sandra L. MacLeod An Unusual Primary Blast Injury: Traumatic Brain Injury Due to Primary Blast Injury 97 Serkan Yılmaz and Murat Pekdemir Gastric Outlet Obstruction in an Infant: Lactobezoar Robert J. Graham and Phil Stein 98 An Unusual Case of Pediatric Abdominal Distension Christopher T.Y. Hsu, 99 Maria Carmen G. Diaz, and David Rappaport Massive Tissue Emphysema After Cardiopulmonary Resuscitation Raffi Bekeredjian, 101 F. Joachim Meyer, Boris Radeleff, and Norbert Frey Succinylcholine-Induced Masseter Muscle Rigidity Associated With Rapid Sequence Intubation 102 Christopher S. Roman and Alex Rosin Pontine Hemorrhage and Isolated Abducens Nerve Palsy Scott C. Sherman 104 and Babak Saadatmand ED Treatment of Brainstem Anesthesia After Retrobulbar Block James M. Dahle 105 and Kenneth V. Iserson Fatal Afebrile Streptococcal Meningitis in a Chronic Alcoholic Patient Ramsis F. Ghaly, 106 William P. Gibbons, and Timothy P. Plackett Atrial Flutter With Cardiac Tamponade as Initial Presentation of Tuberculosis Pericarditis 108 I-Chuan Chen, Te-Fa Chiu, and Jih-Chang Chen Importance of Bedside Echocardiography for Detection of Unsuspected Isolated Right Ventricular 110 Infarction as a Cause of Cardiovascular Collapse Mitja Lainscak and Andrej Pernat Acute Myocardial Infarction Diagnosed Early by Multidetector Computed Tomography 114 Ju-Sing Fan, Chorng-Kuang How, Chii-Hwa Chern, Lee-Min Wang, Chun-I Huang, and Jen-Dar Chen Life-Threatening Epistaxis From a Rare Treatable Cause Mohammed Iqbal Syed, 116 Andy Chin, and Louise Clark Spontaneous Intracranial Hemorrhage Caused by Sustained Hypertension in a Child 118 Jainn-Jim Lin, Chang-Teng Wu, Shao-Hsuan Hsia, Kuang-Lin Lin, and Huei-Shyong Wang

(Continued on next page) Contents continued A Rare Diagnosis in ED: Cerebral Pyogenic Ventriculitis Due to Infective Endocarditis 120 Selahattin Kiyan, Ersin Aksay, Murat Ersel, and Sedat Yanturali Should Early Venous Oximetry Be Indicated in Suspected Cases of Fulminant Myocarditis? 122 Arvind K. Baronia, Afzal Azim, Gagan Narula, Mohan Gurjar, Banani Poddar, Sudeep Kumar, Sanjay Gambhir, and Sukanta Barai CORRESPONDENCE Thiamine for the Treatment of Acute Decompensated Heart Failure Howard A. Smithline 124 Early Diagnosis of Gas-Forming Pyometra in an Aged Patient Can Prevent Mortality 126 Chia-Chun Tai, Wan-Ching Lien, Hsiu-Po Wang, and Kao-Lang Liu Image Pitfall of Computed Tomography in Diagnosis of Aortic Dissection 127 Wen-Chu Chiang, Pei-Chieh Kao, Chan-Ping Su, and Juan Hsu Other Causes of Unilateral Pulmonary Edema Ritesh Agarwal, Ashutosh N. Aggarwal, 129 and Dheeraj Gupta Manuscript Submission and Editorial Review Policy

The scope of The American Journal of Emergency Medicine is as Cover Letter broad as the definition of emergency medicine itself, encompassing all The cover letter accompanying all submitted manuscripts must (1) activities concerned with acute medical care. AJEM invites the submission be signed by all authors, and (2) contain the following language: ‘‘The of original research, reports, correspondence, and opinion relating to manuscript, as submitted or its essence in another version, is not under acute adult and pediatric medicine and surgery and the related fields consideration for publication elsewhere, and will not be published of trauma, toxicology, critical care, resuscitation, emergency medical elsewhere while under consideration by AJEM. The authors have no services, behavioral emergencies, and environmental medicine. commercial associations or sources of support that might pose a Original contributions will be accepted on the basis of significance, conflict of interest. All authors have made substantive contributions to validity, and clarity. Authors will be expected to justify conclusions by the study, and all authors endorse the data and conclusions.’’ Authors the data presented, maintain a lucid prose style, and describe method- with a potential conflict of interest should cite it in the cover letter. ology in sufficient detail for readers to evaluate results accurately. AJEM, in turn, is committed to a confidential, expeditious, and professional Author Responsibility editorial process. Reviews will be objective, rigorous, and responsible. Submissions are reviewed for possible publication with the under- Articles published in AJEM are indexed and abstracted in Index standing that they are original and not simultaneously under consid- Medicus, Excerpta Medica, Current Contents/Clinical Medicine, eration by another journal. Accepted manuscripts become the property ISI/BIOMED, and BIOSIS. of AJEM and may not be published elsewhere without the written per- For the convenience of prospective authors, AJEM is a participating mission of AJEM. Any material previously published elsewhere must journal in the International Committee of Medical Journal Editors’ be accompanied by written consent of its author and publisher when ‘‘Uniform Requirements for Manuscripts Submitted to Biomedical submitted to AJEM. Photographs of an identifiable subject should be Journals’’ (N Engl J Med 1997;336:309-315). This agreement provides accompanied by a release signed by the subject or responsible party for a standardized manuscript format, allowing authors to submit authorizing publication. If required, institutional clearance to publish articles to any one of over 500 scholarly medical publications without should be submitted with the manuscript. AJEM is not responsible for revision simply to accommodate the vagaries of any individual jour- statements made by any contributor. Authors should keep copies of nal’s technical and stylistic requirements. all submitted materials. Photographs and figures are not returned, REVIEW POLICY even if a manuscript is not accepted. All original contributions, investigations, and reports will be subjected Repetitive Publication to multiple-peer review. To protect the integrity and anonymity of the review process, all reviews will be conducted in double-blinded fashion. Authors submitting papers to this journal must confirm that the paper, To promote quality composition and investigation, legible comments as submitted or its essence in another version, has not been published from all referees will accompany returned manuscripts. To encourage elsewhere, is not under consideration for publication elsewhere, and will criticism, correspondence, and open discussion of controversial issues, not be published elsewhere while under consideration by AJEM. Prior letters to the editor will be printed promptly. As a courtesy to contributors publication of some content of the paper may not preclude the paper’s and to ensure the timeliness of AJEM’s content, authors will routinely publication in AJEM. Authors must provide full information in the be notified of the action taken upon their manuscripts within 10 weeks cover letter sent with the submitted manuscript on any possibly of submission. repetitive publication of content, including: (1) reworked data already reported; (2) cases or subjects in a study cohort already described in a GUIDE FOR AUTHORS published report; (3) previously reported single or multiple cases; (4) We invite submissions on clinical and laboratory research and topics content already published or to be published in another format such as the pertinent to adult and pediatric emergency medicine including emergency proceedings of a meeting or symposium, a chapter in a book, or a letter to medical and health services, trauma, toxicology, resuscitation, behavioral the editor; (5) content published in a language other than English. emergencies, critical care, and environmental medicine. In general AJEM does not publish surveys, papers that focus on patient satis- Conflict of Interest faction, quality assurance, or didactics. The following are journal Authors are expected to disclose any commercial associations or features for which we invite submissions: sources of support that might pose a conflict of interest in connection with ORIGINAL CONTRIBUTIONS: Reports of new clinical and laboratory the submitted article. All funding sources supporting the work must be investigations and research. acknowledged in a footnote on the title page. All affiliations with or BRIEF REPORTS: Short papers, series of cases, and preliminary reports financial involvement in any organization on entity with a direct financial of work in progress; studies with small numbers pointing to the need for interest in the subject matter or materials of the research discussed (eg, further investigation. Brief reports should be limited to 2,000 words of text employment, consultancies, stock ownership or other equity interest, (exclusive of tables, references, and figure legends). patent-licensing arrangements) should be cited in the cover letter. RESEARCH SEMINARS: Discussions of the history, methodology, and Human Research and Informed Consent future of a particular area or subject in emergency medicine research. REVIEWS: Definitive, in-depth, state-of-the-art reviews of clinical and When appropriate, manuscripts reporting the results of experimental research subjects. Unsolicited reviews are not generally published in investigations on human subjects should include a statement indicating AJEM. Before submitting any unsolicited reviews, please forward an approval by the institution’s Human Research Committee. outline to the Editor for consideration. Author Approval THERAPEUTICS: Detailed reviews of important devices and drugs All accepted manuscripts are subject to copyediting. Authors will used in the practice of emergency medicine. receive page proof of their article before publication. DIAGNOSTICS: Concise articles guiding clinical practice, with references to additional, authoritative sources. Manuscript Submission CONTROVERSIES: Editorial viewpoints on current controversies. All manuscripts (including figures) must be submitted to AJEM CLINICAL NOTES: Descriptions of new techniques and procedures in through our Web site (http://ees.elsevier.com/ajem/). Submission emergency medicine practice and investigation. items should include separate files for a cover letter, title page, CORRESPONDENCE: Letters to the editor are limited to 800 words abstract, manuscript text, references, legends for table/figure, tables, of text (exclusive of references, tables, and figure legends). These and figures. Revised manuscripts should also be accompanied by a submissions should not contain an abstract. unique file (separate from the covering letter) with responses to CASE REPORTS: Case reports should describe a case unique to the reviewers’ comments. emergency medicine literature, and are limited to 800 words of text The preferred order of files for electronics submission is as follows: and an abstract < 250 words. Accepted case reports in their entirety cover letter, response to reviews (revised manuscripts only), title page, will be published digitally at our web site (www.ajemjournal.com), manuscript file(s), table(s), figure(s). Files should be labelled with while the abstract will be published in each printed issue. appropriate and descriptive file names (e.g., SmithText.doc, Fig1.eps, Table3.doc). Upload text, tables and graphics as separate files. Do not Acknowledgments import figures or tables into the text document; submit them as separate Acknowledge only people who have made substantive contributions files. Complete instructions for electronic artwork submission can be to the study, and specify the contributions. Authors are responsible for found on the Author Gateway, accessible through the journal home page. obtaining written permission from everyone acknowledged by name All manuscripts must be submitted double-spaced in English. Please visit because readers may infer their endorsement of the data and conclusions. http://ees.elsevier.com/ajem to submit your manuscript electronically. The website guides authors stepwise through the creation and upload- Abbreviations, Symbols, and Nomenclature ing of the various files. Note that original source files, not PDF files, are Usage should conform to that recommended in Council of Biology required. Once the submission files are uploaded the system automati- Editors Style Manual (5th ed., 1983) available from the American cally generates electronic (PDF) proof, which is then used for review- Institute of Biological Sciences, 1950 Rockville Pike, Bethesda, MD ing. All correspondence, including the Editor’s decision and request for 20814. Avoid abbreviations. Do not abbreviate names of organizations, revisions, will be by e-mail. institutions, symptoms, diseases, or anatomic characteristics. A list of acceptable abbreviations is included in ‘‘Uniform Requirements for Copyright Manuscripts Submitted to Biomedical Journals’’ (see below). Generic A copyright transfer agreement will be sent to corresponding authors names of drugs are preferred; a brand name may be given of each manuscript accepted for publication. Authors are responsible only with the first use of generic name. When the brand name of a for applying for permission for both print and electronic rights for all product or pharmaceutical is used, supply the manufacturer’s borrowed materials and are responsible for paying any fees related to name and location (city and state). the application of these provisions. Units of Measurement Title Page Use SI units for linear dimensions, weight, clinical chemistry, and On the title page include (1) the title; (2) a short running head of hematology. Use the Celsius scale for all temperatures. The use of fewer than 50 characters/spaces placed at the foot of the page; (3) other SI units is encouraged. author(s) names, highest degrees, department(s) and institution(s); (4) name and address of author to whom reprint requests should be sent; References (5) source(s) of support in the form of equipment, drugs, or grants Cite references consecutively in the text. Do not cite review (including grant numbers); (6) the name of organization and date of articles. Use the same number each time the reference appears in assembly if the article has been presented; and (7) ‘‘Key Words,’’ a list the text. At the conclusion of the article, list references in numerical of three to ten important words or phrases for indexing. Whenever pos- order, typed double spaced. Abbreviate journal titles according to sible, use terms from the medical subject heading of Index Medius. Index Medicus style. Please provide inclusive pagination Punctuation To ensure blinded, impartial review, do not indicate the authors of the is shown below. article on any other page. Journal articles: List all authors when three or fewer; when four or Abstract more, list first three and add et al. On the second page include a structured abstract of fewer than Abraham E, Baraff LJ: Oral versus parenteral therapy of pye- 150 words stating the objective, methods, results, and conclusion. Be lonephritis. Curr Ther Res 1982;31:536-542 concise yet detailed. Books: Capitalize all important words in title. Ludwig S, Fleisher GR, Henretig FM, et al (eds): Pediatric Text Emergency Medicine. Baltimore, MD, Williams & Wilkins, When appropriate, divide the text into Introduction, Methods, Re- 1983, pp 203-209. sults, and Discussion. Chapter in a book: List editors of book. Introduction: Clearly state the purpose of the article, summarize the Eliastam M: Cardiac emergencies. In Eliastam M, Sternbach rationale for the study or observation, give only strictly pertinent GL, Bresler MJ (eds): Manual of Emergency Medicine. Chi- references, and do not review the subject extensively. cago, IL, Yearbook, 1983, pp 1-28 Methods: Identify the methods, apparatus, and procedures in sufficient detail to allow other workers to reproduce the results. Give references to References to unpublished information should be included parenthet- established methods, including statistical method; provide references and cally in the text. Do not cite review articles. brief descriptions of methods that have been published but may not be Tables well known; describe new or substantially modified methods, giving Type tables double-spaced on separate sheets with number and title. reasons for using them and evaluating their limitations. Results: Present your results in logical sequence in the text, tables, and Do not submit tables as photographs. Omit internal horizontal and illustrations. Do not repeat in the text all of the data in the tables and/or vertical rules. Cite each table in the text in consecutive order. illustrations; emphasize or summarize only important observations. Figures Discussion: Emphasize the new and important aspects of the study and Submit figures electronically as separate files. Complete instructions conclusions that follow from them. Do not repeat in detail data given in for electronic artwork submission can be found on the Author Gateway, the Results sections. Include in the Discussion the implications of the accessible through the journal home page, http://ees.elsevier.com/ajem/. findings and their limitations and relate the observations to other relevant studies. Link the conclusions with the goals of the study, but avoid Editorial Inquiries unqualified statements and conclusions not completely supported by your Authors are strongly encouraged to use the Journal gateway for all data. Avoid claiming priority and alluding to work that has not been inquiries regarding submitted manuscripts for the fastest response and most completed. State new hypotheses when warranted, but clearly label them current information/status. Other inquiries (i.e., unrelated to a submitted as such. Recommendations, when appropriate, may be included. manuscript) can be directed to [email protected]. American Journal of Emergency Medicine (2007) 25,1–5

www.elsevier.com/locate/ajem

Original Contribution Antithrombotic therapy for prevention of stroke in clinical practice in the ED

Asia Kogan MDa,*, Reuma Shapira MDa, Ada Tamir DScb, Gad Rennert MD, PhDb aEmergency Department, Carmel Medical Center, Haifa 34362, Israel bCommunity Medicine and Epidemiology Department, Carmel Medical Center, Haifa 34362, Israel

Received 6 February 2006; revised 20 March 2006; accepted 20 March 2006

Abstract Objective: The objective of this study was to analyze the emergency medicine department’s practice of recommending anticoagulation medication for stroke prevention, its compliance with clinical guidelines, and the role of the emergency physician in recommending anticoagulation medication. We also determined the occurrence of thromboembolic events in patients with atrial fibrillation (AF) during a follow-up period of up to 8 months after their discharge from the ED. Materials and Methods: Over a 6-month period, patients presenting to the ED with AF were registered using a predesigned 2-part questionnaire. The first part considered the management of the patients with AF at the ED; the second part evaluated data on patients who returned to the ED and their treatment follow-up, based on the drug supply registry. Results: This study included 102 ED visits by patients with AF, of whom 38 were hospitalized and 64 were discharged. Thirty-six of the discharged patients required anticoagulation according to American College of Cardiology/American Heart Association Task Force on Practice Guidelines/ European Society of Cardiology guidelines. Of these patients, 28 (78%) were recommended anticoagulation medication and 8 (25%) were not: 5 because of contraindications and 3 because of unknown reasons. No patient returned to the ED with a thromboembolic event during the 8-month follow-up period. Conclusion: The prevalence of anticoagulation recommendation for stroke prevention and compliance with clinical guidelines were found to be greater than previously reported. Our data show that most of the patients with AF and risk factors for stroke were recommended anticoagulation medication. This study illustrates the importance of applying clinical guidelines in daily practice and integrating them into patients’ medical files in the ED. D 2007 Elsevier Inc. All rights reserved.

1. Introduction During the last decades, AF has become the most frequent arrhythmia seen in the ED and the cause of more Atrial fibrillation (AF) is a growing public health hospital admissions than any other rhythm disturbance. problem associated with significant morbidity and mor- Along with these developments, a profound change has tality [1]. taken place in the treatment of patients with AF [2,3]. A change has also been seen in patients’ behavior: increasing numbers of patients tend to abandon convention- * Corresponding author. al health care facilities and go directly to the ED.

Randomized controlled trials with warfarin have conclu- the ED up until August 2005 and their diagnosis upon sively demonstrated that long-term anticoagulation therapy returning. The data were obtained from OFEK (DB Motion, can reduce the risk of stroke by approximately 68% per year Omer, Israel), a designated patient information management in patients with valvular heart disease and AF as well as in system. This system includes hospitalization records of all patients without valvular disease [4-7]. Despite these hospitals in Israel. Because there is no private internal conclusive evidence of the medication’s efficacy, several medicine hospital in Israel, this system enabled us to studies have shown suboptimal use of warfarin worldwide: electronically follow all participants in all the area hospitals. it is recommended for only 15% to 44% of patients with AF All medical decisions in the ED were made according to who have no contraindication [2,8]. guidelines developed in our hospital as a cooperative project Initial therapy of AF in the ED is often directed toward of the ED with the internal medicine and cardiology achieving rate control as well as obtaining and maintaining departments in accordance with ACC/AHA/ESC guidelines. sinus rhythm, in addition to the goal of reaching an In this part, we also evaluated adherence to anticoagu- international normalized ratio (INR) of anticoagulation lation by referring to the drug supply registry. between 2 and 3. Anticoagulation should be continued All discharged patients who were recommended anti- indefinitely in groups of high-risk patients with AF coagulation (also those continuing anticoagulation) were (according to American College of Cardiology/American given a guidance paper explaining anticoagulation—the Heart Association Task Force on Practice Guidelines/ drugs, the correct way to use them, their side effects and European Society of Cardiology [ACC/AHA/ESC] recom- risks, and the way to maintain a therapeutic level—in the mendations from 2001 [9]) because most strokes occur in 3 most common languages in Israel (Hebrew, Arabic, patients who discontinued their anticoagulant medication, Russian). They were also given a discharge letter to inform those whose INR was subtherapeutic, or those without their primary physician of the need to provide anticoagu- anticoagulant therapy [5,6,9]. lation medication and follow and control INR. The importance of providing therapy protocols for practicing physicians in the ED is undisputed, and such a protocol is particularly significant in the treatment of 3. Results patients with AF. This study implemented the protocol’s guidelines not only within the boundaries of the ED but also During the study period (between December 2004 and in cooperation with the cardiology and internal medicine May 2005), 102 ED visits by patients with AF were departments in the hospital as well as with outpatient clinics recorded; 38 of these patients were hospitalized and 64 in the community. were discharged. Baseline characteristics are listed in Table 1. There were 50 men and 52 women in the study; 20 patients (19.6%) 2. Materials and methods were younger than 65 years, 33 (32.4%) were between 65 and 75 years old, and 49 (48%) were older than 75 years. The Carmel Medical Center has approximately 500 beds The most common complaints upon arrival at the ED (including cardiac care units) and serves a population of were palpitations (55%), dyspnea (32.8%), and chest pain 200000 people. Its ED has approximately 80000 visits a year. (19.6%), whereas 21.6% of patients had complaints Patients who presented to our ED with AF between unrelated to AF. December 2004 and May 2005 were registered and included Duration of AF was less than 48 hours upon arrival at the in this study. ED in 26.5% of the patients, more than 48 hours in 38.2%, Data regarding each patient were provided by the treating and unknown in 35.3%. ED physician using a 2-part questionnaire. Part 1, which The clinical characteristics of the 64 patients discharged was filled prospectively, included demographic variables, from the ED are listed in Table 2. Among the discharged medical history, details regarding AF duration (less or more patients, 55 (86%) had risk factors for TE, the most common than 48 hours), symptoms, risk factors for thromboembo- of which was hypertension (77.5%), then coronary heart lism (TE), INR in the ED, treatment before entering the ED disease (36.4%), heart surgery (23.6%), congestive heart and in the ED (including antiarrhythmic treatments, rate failure (20%), valvular heart disease (16.7%), cardiomegaly control, antiaggregants, and anticoagulants), medical or (3.6%), thromboembolic event in the past (3.6%), and electric cardioversion, complications, admission diagnosis, diabetes mellitus (1.8%). and recommendations upon discharge. We evaluated rea- Thirty-six (56%) of the discharged patients with risk sons for not recommending anticoagulation medication to factors for TE had not received anticoagulation medication patients for whom it was indicated according to ACC/AHA/ before. Twenty-eight of these patients were discharged with ESC guidelines [9]. At a later stage, this part was used a first-time recommendation for anticoagulation medication, routinely in our ED patients’ files. as were 2 additional patients who demonstrated no risk Part 2 of the questionnaire, which was filled retrospec- factor. At the 6-month follow-up, 19 of the patients for tively, included patients from the first part who returned to whom anticoagulation medication was recommended were Antithrombotic therapy in the ED 3

cation before their ED visit, and it was recommended that Table 1 Clinical and demographic characteristics of the patients (N = 102) they continue the regimen. Of the 55 patients with risk factors, 23 (41.8%) had 1 risk Variables n (%) factor, 12 (21.8%) had 2 risk factors, and 20 (36.4%) had Male 50 (49) 3 or more risk factors. Female 52 (51) Among the 38 (37.3%) patients who had been treated Age (y) with anticoagulation before their ED visit, the INR was b65 20 (19.6) lower than 2 (nontherapeutic) in 27.8% and greater than 65-75 33 (32.4) N75 49 (48) 3.5 (overtherapeutic) in 36.6%; adequate coagulation Complaint upon arrival measures were found in only 35.6%. Palpitation 57 (56.8) In the ED, 68.6% received rate control, 28.7% did Dyspnea 34 (33.8) medical cardioversion, 2.0% did electric cardioversion, and Chest pain 20 (19.8) 64.7% did anticoagulants (warfarin and/or enoxaparin). Unrelated to AF 23 (22.8) Thirty-five of the patients returned to sinus rhythm in the AF duration (h) ED (38.6%), 2 of them after electric cardioversion. N48 39 (38.8) Indications for hospitalization (38 patients) were arrhyth- b 48 27 (26.8) mias in 10 patients, TE in 1 patient, congestive heart failure Unknown 36 (35.8) in 7 patients, syncope in 1 patient, ischemic heart disease in Discharged 64 (63.8) 6 patients, combination of ischemic heart disease and Indications for admission Ischemic heart disease 6 syncope in 1 patient, syncope and arrhythmia in 1 patient, Arrhythmia 10 and other reasons in 11 patients. TE 1 Sixty percent of the discharged patients were recom- Congestive heart failure 7 mended a change in their medication regimen, including Syncope 1 the addition of anticoagulation and/or rate control Syncope + ischemic heart disease 1 and/or antiaggregants. Syncope + Arrhythmia 1 Other 11 Coumadin treatment 38 (37.8) 4. Discussion INR b2 11 (27.8) Atrial fibrillation is the most frequent arrhythmia seen in 2-3.5 13 (35.6) the ED [10] and a very common reason for hospitalization N 3.5 14 (36.8) [11,12]. Patients with AF compose a heterogenic group Treatment in the ED Rate control 69 (68.8) Medical cardioversion 29 (28.8) Electric cardioversion 2 (2) Table 2 Clinical characteristics of the discharged patients Anticoagulants 65 (64.8) (n= 64) Sinus rhythm achieved 39 (38.8) Variable n (%) Risk factors for TE 55 (86) Hypertension 43 (77.5) still adhering to their therapeutic anticoagulant regimen, Ischemic heart disease 20 (36.4) 6 had ceased taking the anticoagulation medication, and no Heart surgery 13 (23.6) information was found regarding 5 others. Congestive heart failure 11 (20) Nineteen of the discharged patients were hospitalized Cardiomegaly 2 (3.6) during a mean follow-up period of 6 months; none of these TE in the past 2 (3.6) hospitalizations was caused by a thromboembolic event Diabetes mellitus 1 (1.8) (10 for cardiac disease, 3 for pulmonary disease, and 6 for 1 risk factor 23 (41.8) 2 risk factors 12 (21.8) other reasons). z3 risk factors 20 (36.4) Eight (22%) of the discharged patients with risk factors Anticoagulant treatment before 20 (31.2) for TE were not recommended anticoagulation medication: No anticoagulation before 44 (68.8) 2 because of bleeding history, 5 were not compatible with Anticoagulation recommendation total 49 (76.5) anticoagulation (dementia, recurrent falls), and 3 for New anticoagulation recommendation 28 (77.7) unknown reasons. Nine of the discharged patients did not No recommendation despite risk factors 8 (22) have risk factors for TE, but 3 of them received a Reasons for no recommendation recommendation for anticoagulation medication nonetheless Bleeding 2 (1 had already taken coumadin before). Twenty of the Not compatible with treatment 5 discharged patients had been taking anticoagulation medi- Unknown 3 4 A. Kogan et al. regarding etiology, natural course, therapeutic approach, and Before this project, we had completed a joint quality complications—mostly thromboembolic (stroke). Because project with the cardiology and internal medicine depart- AF is not always symptomatic, we found that 22% of ments for patients with AF that helped finalize the guide- our patients arrived at the ED for other complaints, and lines for ED physicians treating patients with AF. In the AF was found coincidentally and treated according to current project, we evaluated the application of these our guidelines. guidelines in the ED, which were used in conjunction with The prevalence of AF starts to increase after age 40 years ACC/AHA/ESC guidelines. and rises rapidly after age 65 years, reaching a prevalence of We found that the guidelines were applied in the cases of approximately 10% in patients older than 80 years. This more than 75% of the discharged patients by recommending finding is in accordance with our finding that more than anticoagulation prophylaxis in the setting of the ED. They 80% of the study patients were older than 65 years. Along were also applied at the scene of the ED in more than 65% of with the increasing prevalence of AF, the rate of ischemic the patients when trying to convert the rhythm to sinus rate. stroke in patients with AF increases from 1.5% between the These guidelines were also implemented at our outpatient ages of 50 and 59 years to 23.5% between the ages of clinics in general and in the cardiology clinics in particular 80 and 89 years [3]. by maintaining closer contact with them regarding changes Risk factors, such as congestive heart failure, hyperten- in medication and rhythm during follow-up. All of these sion, cardiac disorders, diabetes, and previous embolism, steps, the guidelines for treating physicians, the guidance pose an increased risk for thromboembolic events in patients papers for patients, and the dissemination of both to clinics with AF [1,2,9]. The risk increases if more than 1 risk factor within the community, led to greater than 60% patient are found. In our study, more than 58% of the patients had adherence to the recommended anticoagulation medication more than 3 risk factors, placing them at greater risk for TE regimen and consequently to no thromboembolic event in if not properly treated. the discharged patients during the follow-up period. We The ED population represents a wide spectrum of ascribe this high rate of adherence to the steps described, to patients with AF and other risk factors, many of whom the discharging physician’s ability to provide clear instruc- are not hospitalized [10,13]. Finding that more than 67% of tions to the patient regarding medication administration, and our patients with AF were discharged led us to examine the to the cooperation of the outpatient clinics. Undoubtedly, therapeutic approach to patients with AF in general and the guidelines for clinical practice improve the outcome of recommendation of antithrombotic therapy for these patients health care. in particular, with a special focus on the patients with AF who were discharged from the ED and needed to receive the correct recommendations upon discharge. Although the References literature reports that only 15% to 44% of patients with AF who need antithrombotic therapy get a recommendation [1] The Stroke Prevention in Atrial Fibrillation Investigators. Predictors of thromboembolism in atrial fibrillation. Ann Intern Med 1992; for it upon discharge from the ED [8,14,15], we found that 116:1-12. antithrombotic (anticoagulant) treatment was recommended [2] Laguna P, Martin A, Del Arco C, et al. Risk factors for stroke and to 78% of our discharged patients. thromboprophylaxis in atrial fibrillation: what happens in daily The rate of ischemic stroke per year in patients with AF clinical practice? The GEFAUR-1 Study. Ann Emerg Med 2004; is 2 to 7 times greater than the 5% rate reported in 44:3-11. [3] Go AS, Hylek EM, Phyllips KA, et al. Prevalence of diagnosed atrial nonrheumatic patients [16,17]. Most thromboembolic events fibrillation in adults. National implications for rhythm management in this patient group occurred in patients who stopped their and stroke prevention: the Anticoagulation and Risk factors In Atrial anticoagulation therapy or had subtherapeutic INR. fibrillation (ATRIA) Study. JAMA 2001;285:2370-5. Although a high percentage (36.3%) of our study [4] Laupacis A, Albers GW, Dalen J, et al. Antithrombotic therapy in population had been receiving anticoagulation therapy atrial fibrillation. Chest 1998;114:579s-87s. [5] Albers GW, Dalen JE, Laupacis A, et al. Antithrombotic therapy in before their arrival, only 47% of them were well within the atrial fibrillation. Chest 2001;119:194s-216s. optimal INR range (29% had overcoagulation and 24% had [6] Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to undercoagulation). This demonstrated the need of patients— prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann those already receiving anticoagulation therapy and those Intern Med 1999;131:492-501. starting this therapeutic regimen for the first time—for [7] Gottlieb LK, Salem-Shatz S. Anticoagulation in atrial fibrillation. Does efficacy in clinical trials translate into effectiveness in practice? proper guidance regarding the administration of anticoagu- Arch Intern Med 1994;154:1945-53. lation therapy. To this end, we devised a guidance paper for [8] Bungard TJ, Ghali WA, Teo KK, et al. Why do patients with patients, as mentioned in the Materials and Methods section. atrial fibrillation not receive warfarin? Arch Intern Med 2000;160: The incidence of bleeding among patients receiving oral 41-5. anticoagulation therapy is between 0.5% and 1%; of these [9] Fuster V, Ryden LE, Asinger RW, et al. ACC/AHA/ESC guidelines for the management of patients with atrial fibrillation executive cases, 10% require treatment [12,13]. In the current study, summary: a report of the American College of Cardiology/American not even one case of bleeding was recorded during the Heart Association Task Force on Practice Guidelines/European follow-up period. Society of Cardiology Committee for Practice Guidelines and Antithrombotic therapy in the ED 5

Policy Conference (Committee to Develop Guidelines for the treatment strategy with low-molecular-weight heparin. Ann Emerg Management of Patients With Atrial Fibrillation). Circulation 2001; Med 2002;40(2):187-92. 104:2118-50. [14] Evans A, Kalra L. Are the results of randomized controlled trials on [10] Koenig BO, Ross MA, Jackson RE. An emergency department anticoagulation in patients with atrial fibrillation generalizable to observation unit protocol for acute-onset atrial fibrillation is feasible. clinical practice? Arch Intern Med 2001;161:1442-6. Ann Emerg Med 2002;39(4):374-81. [15] Thomson R, McElroy H, Sudlow M. Guidelines on anticoagulant [11] Bialy D, Lehmann MH, Schumacher DN, et al. Hospitalization for treatment in atrial fibrillation in Great Britain: variation in content and arrhythmias in the United States: importance of atrial fibrillation implications for treatment. BMJ 1998;316:509-14. [abstr]. J Am Coll Cardiol 1992;19:41A. [16] The AFFIRM Investigators. A comparison of rate control and rhythm [12] Mulcahy B, Cotes WC, Henneman PL, et al. New-onset atrial control in patients with atrial fibrillation. N Engl J Med 2002; fibrillation: when is admission medically justified? Acad Emerg Med 347:1825-33. 1996;3(2):114-9. [17] Van Gelder IC, Hagens VE, Bosker HA, et al. for the Rate Control vs. [13] Kim MH, Morady F, Conlon B, et al. A prospective, randomized, Electrical Cardioversion for Persistent Atrial Fibrillation Study Group. controlled trial of emergency department–based atrial fibrillation N Engl J Med 2002;347:1835-40. American Journal of Emergency Medicine (2007) 25,6–9

www.elsevier.com/locate/ajem

Original Contribution Bilevel positive airway pressure in the treatment of status asthmaticus in pediatrics

Sara L. Beers MDa,b,*, Thomas J. Abramo MDa,b, Andrea Bracken RTTa,b, Robert A. Wiebe MDa,b aUniversity of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390, USA bPediatric Emergency Services, Children’s Medical Center Dallas, Dallas, TX 75235, USA

Received 2 September 2005; revised 27 June 2006; accepted 2 July 2006

Abstract Objectives: The aim of this study was to examine the safety, patient tolerance, and possible benefit of bilevel positive airway pressure (BiPAP) in conjunction with b-2 agonist therapy in the treatment of pediatric patients with status asthmaticus who were refractory to conventional medical therapy. Methods: This descriptive retrospective chart review examined all patients with the diagnosis of acute asthma treated with BiPAP in an urban academic pediatric emergency department (ED) from April 1, 2003, to August 31, 2004. Results: Eighty-three patients with status asthmaticus refractory to conventional pharmacological treatment were placed on BiPAP with b-2 agonist nebulization in the ED. The number of subjects tolerating BiPAP was 73 (88%) of 83 patients. All patients placed on BiPAP in the ED were initially designated for admission to the pediatric intensive care unit (PICU). However, only 78% (57/73) were actually admitted to the PICU. Sixteen patients on BiPAP were admitted to a ward service; of these patients, none were subsequently transferred to the PICU. In addition, there was an immediate improvement in subjects’ clinical status upon initiation of BiPAP, with 77% showing a decrease in respiratory rate, averaging 23.6% (range, 4%-50%), and 88% showing an improved oxygen saturation, averaging 6.6 percentage points (1-28 percentage points). There were no adverse events due to the use of BiPAP. Conclusions: These results suggest that the addition of BiPAP in treating pediatric status asthmaticus is safe and well tolerated. This intervention shows promise as a beneficial adjunct to conventional medical treatments. However, further prospective investigation is warranted to confirm these findings. D 2007 Elsevier Inc. All rights reserved.

1. Introduction

The National Institutes of Health guidelines recommend * Corresponding author. University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75235, USA. Tel.: +214 456 2014; fax: +214 use of continuous or intermittent nebulized b-2 agonist 456 8132. medications and systemic corticosteroids to treat the E-mail address: [email protected] (S.L. Beers). symptoms of acute asthma [1]. However, approximately

5% to 10% of patients with acute asthma fail to respond to medical management. The routine treatment for managing this conventional therapy and become candidates for other patients with acute asthma in our ED includes up to treatment options [2]. When first-line treatment fails, 3 treatments of inhaled albuterol (2.5 mg for children clinicians often add other therapeutic modalities such as b10 kg; 5 mg for children N10 kg) and ipratroprium bromide aerosolized anticholinergics or epinephrine, and intravenous (250 lg for children b10 kg; 500 lg for children N10 kg) magnesium, theophylline, or terbutaline. Mechanical airway and early administration of oral or intravenous cortico- support such as bilevel positive airway pressure (BiPAP) steroids (2 mg/kg, maximum 80 mg). If wheezing persists, and, ultimately, intubation with mechanical ventilation can an hour-long inhalation is provided (albuterol, 15-20 mg). be used if all other medical management has failed. Depending on the severity of respiratory distress, presence Using BiPAP to treat status asthmaticus is a concept that of hypoxemia, and clinical concern for respiratory failure, has its origin from intermittent positive pressure breathing additional therapies used may have included intravenous (IPPB). Both IPPB and BiPAP are similar in that they both magnesium sulfate, subcutaneous or intravenous terbutaline, provide noninvasive positive pressure support. However, and/or aerosolized epinephrine. It is this population of BiPAP expands on this concept with continuous flow patients who failed our routine treatment and qualified for alternating between a higher inspiratory pressure and a BiPAP therapy. lower end-expiratory pressure. Previous studies have looked In our ED we used nasal BiPAP with a mask that covers at the use of IPPB in patients with an acute asthma only the nose. The patients in this study received BiPAP exacerbation receiving nebulized b-agonist therapy [3-12]. from one of 2 machines, both made by Repironics Most of these studies found no significant benefit with the (Carlsbad, CA): the S/T-D30 Ventilatory Support System use of IPPB [6-12]. The problem with these early studies or the BiPAP Vision Ventilatory Support System. All was that the range in disease severity was varied and the patients who received BiPAP also concurrently received studies were not controlled. However, 2 studies concluded continuous inhaled b-2 agonists. there is evidence that in severe asthma failing to respond to Patients were excluded from the study population if other methods of b agonist inhalation, there is benefit of they had any comorbid illnesses including obstructive sleep IPPB [3,5]. apnea, congenital heart disease, sickle cell disease, To date, the information available on the use of BiPAP to tracheostomy, diabetic ketoacidosis, cystic fibrosis, bron- treat status asthmaticus in the pediatric population is a chopulmonary dysplasia, or any chronic lung disease other limited to a single brief report describing its use in 3 patients. than asthma. This report in the intensive care setting expressed a positive From the patient’s chart, the change in respiratory rate outcome [13]. It has been our experience that pediatric and oxygen saturation, the need for PICU admission, patients with acute asthma who do not respond to endotracheal intubation, patients admitted to the wards with conventional medical management benefit clinically from later transfer to the PICU, duration of BiPAP treatment, the implementation of BiPAP. Thus, the purpose of this tolerance for BiPAP therapy, and complications related to investigation was to examine the safety, patient tolerance, BiPAP use (aspiration, epistaxis, pneumomediastinum, or and possible benefit of BiPAP in conjunction with b-2 pneumothorax) were recorded. agonist therapy in the treatment of pediatric patients with All of the charts were individually reviewed by the status asthmaticus who were refractory to conventional primary investigator. asthma therapy. A descriptive analysis of the continuous and categorical data was performed using proportions, frequency distributions, and medians (25th and 75th quartiles). No statistical 2. Methods analysis was performed in this descriptive study.

The study was approved by the institutional review board of the University of Texas Southwestern Medical 3. Results Center. The study population was drawn from patients seen in the emergency department (ED) of a large urban Eighty-three patients with acute asthma were placed on pediatric specialty hospital. Charts reviewed were those BiPAP during the designated period. The median age was of all patients with the diagnosis of acute asthma treated 8 years (25th quartile, 5 years; 75th quartile, 11 years) with with BiPAP from April 1, 2003, to August 31, 2004. These a range of 2 to 17 years. Sixty-four percent were male. charts were retrospectively identified by reviewing the Average length of time on BiPAP was 5.8 hours. Subjects charts of all patients who were billed for BiPAP. Charts toleratingBiPAPwere73(88%)of83.Forpatients were then reviewed for subjects meeting the inclusion and not tolerating BiPAP, intolerance was established within exclusion criteria. 10 minutes of the nasal mask application. Of the 10 patients BiPAP was implemented for the treatment of children not tolerating BiPAP, there were no documented complica- with status asthmaticus who were in need of an escalation in tions. The median age of patients not tolerating BiPAP was treatment because they were refractory to conventional 4.5 years (25th quartile, 4 years; 25th quartile, 10.25 years) 8 S.L. Beers et al. with a range of 2 to 11 years. In general, it was the younger It is thought that the positive pressure decreases the subjects who did not tolerate BiPAP. workload of fatigued muscle groups used for inspiration, Seventy-seven percent showed a decrease in respiratory the diaphragm and accessory muscles [14]. The positive rate, with an average decrease of 23.6% (range, 4%-50%). pressure of BiPAP relieves the patient’s need for auto- The remaining 23% had no change in their respiratory rate positive end expiratory pressure (PEEP), thereby reducing while on BiPAP. No patients had an increased respiratory the patient’s energy expenditure for exhalation [14].In rate while on BiPAP. Eighty-eight percent showed improved addition, positive pressure has a direct bronchodilator effect oxygen saturation, averaging 6.6 percentage points (1-28 and recruits smaller airways and collapsed alveoli, which percentage points). These patients with improvement in improves ventilation perfusion mismatch [14,15]. Positive their oxygen saturation had oxygen saturations before pressure may also improve the delivery of bronchodilators BiPAP therapy of 76% to 99%, averaging 93%. The to the smaller airways of the lung [16]. remaining 12% had no change in their oxygen saturation The need for endotracheal intubation and mechanical while on BiPAP. Patients with no change in the oxygen ventilation in this BiPAP cohort was very low, with only 2 saturation had oxygen saturations before starting BiPAP of of the 83 patients intubated. Complications associated with 94% to 100%, averaging 98%. No patients had a decrease in intubating a patient with acute severe asthma include sudden oxygen saturation while on BiPAP. and/or prolonged severe bronchospasm barotrauma, pneu- Seventy-eight percent (57/73) were admitted to the mothorax, ventilator-induced lung injury, cardiovascular PICU. Two required intubation. Sixteen patients (22%) instability, and nosocomial infections [17-19].BiPAP were weaned off BiPAP in the ED and qualified for ward therapy shows promise as an adjunct to bridge the gap admission. None of these ward admissions were subse- between maximum medical therapy and mechanical venti- quently transferred to the PICU. There were no adverse lation in patients with acute status asthmaticus who are in outcomes such as aspiration, pneumothorax, pneumome- need of ventilatory support. This noninvasive means of diastinum, or epistaxis. There were no deaths. airway support may decrease the need for intubation and mechanical ventilation, thus decreasing the incidence of associated complications. 4. Discussion There are several limitations to this study. This is a retrospective chart review with no control group. In Traditionally, there has been a fear of using BiPAP, addition, the medical management before the initiation of particularly in the pediatric population, because of the BiPAP varied among study subjects. Although all subjects potential complications. However, our series had no adverse before starting BiPAP received routine care, some addi- outcomes or deaths. In addition, we showed in our study tionally received intravenous magnesium sulfate and/or that BiPAP is very well tolerated with an almost 90% intravenous terbutaline. Thus, it is difficult to discern success rate, even in children as young as 2 years. The between the effect of the BiPAP and the effect of the other positive benefit of BiPAP can be attributed to the significant interventions. Furthermore, the clinical decision to initiate improvement in the child’s work of breathing, which then BiPAP therapy may have differed among the attending reduces the level of patient anxiety. physicians, as no clear indications for starting BiPAP The patients in this study received BiPAP from one of was determined prospectively. Lastly, no clinical asthma 2 machines, both made by Repironics. The respiratory score was documented to objectively document the sever- therapists started with an inspired positive airway pressure ity of the clinical symptoms. However, the authors feel, as of 10 cm H2O and an expired positive airway pressure of a physician group, that BiPAP therapy was only reserved 5cmH2O. The rise time and inspiratory time are adjusted for those patients with whom medical treatment was near for patient comfort. The inspiratory pressure is then adjusted its maximum. to achieve an exhaled tidal volume of 6 to 9 mL/kg. We use In conclusion, our investigation is the largest series of a nasal mask and not a full face mask to reduce the risk of pediatric patients that examines the use of BiPAP in the gastric insufflation, vomiting, and aspiration. For the full treatment of status asthmaticus. Our findings suggest the benefit of nasal BiPAP to be achieved, the child needs to use of BiPAP in conjunction with continuous nebulized b-2 close his or her mouth to maintain the positive pressure. We agonist therapy is safe and well tolerated. In addition, it did not find this to be a problem because the children may be beneficial in treating acute status asthmaticus in the generally followed the simple instruction of closing their pediatric population because the subjects generally tolerat- mouths while wearing the BiPAP mask. The in-line ed the BiPAP, had no complications, had improved aerosolized b-2 agonists are introduced into the circuit oxygenation, and had a decrease in respiratory rate between the whisper valve and nasal mask. This is a good while receiving BiPAP. Because these results show location both to minimize the loss of medicine out of the promise, further scientific validation in the form of a circuit and to minimize adherence to the circuit tubing. prospective study is warranted to determine the true There are many theories suggesting the mechanism by efficacy and safety of BiPAP in the treatment of pediatric which BiPAP works to improve patients with acute asthma. status asthmaticus. BiPAP in the treatment of status asthmaticus in pediatrics 9

References [10] Campbell I, Hill A, Middleton H, et al. Intermittent positive pressure breathing. Br Med J 1978;1:1186. [1] National Institutes of Health. National Heart Lung and Blood [11] Gupta S, Poddar S, Nath S. Aerosolized bronchodilator therapy with Institute. Expert panel report 2: guidelines for the diagnosis and without IPPB in bronchial asthma—a comparative evaluation. and management of asthma 1997. Indian J Chest Dis Allied Sci 1978;20:118. [2] Busse WW, Banks-Schlegel S, Wenzel SE. Pathophysiology of severe [12] Weber R, Petty W, Nelson H. Aerosolized terbutaline in asthmatics— asthma. J Allergy Clin Immunol 2000;106:1033-42. comparison of dosage strength and method of administration. [3] Choo-Kang Y, Grant I. Comparison of two methods of administering J Allergy Clin Immunol 1979;63:116. bronchodilator aerosol to asthmatic patients. Br Med J 1975;2:119. [13] Akingbola O, Simakajornboon N, Hadley E, et al. Noninvasive [4] Cayton R, Webber B, Paterson J, et al. A comparison of salbutamol positive-pressure ventilation in pediatric status asthmaticus. Pediatr given by pressure packed aerosol or nebulization via IPPB in acute Crit Care Med 2002;3:181-4. asthma. Br J Dis Chest 1978;72:222. [14] Trill P, McGuire J, Baden H. Noninvasive positive-pressure ventila- [5] Webber B, Collins J, Branthwaite M. Severe acute asthma: a tion in children with lower airway obstruction. Pediatr Crit Care Med comparison of three methods of inhaling salbutamol. Br J Dis Chest 2004;5:337-42. 1982;76:69. [15] Soroksky A, Stav D, Shpirer I. A pilot prospective, randomized, [6] Chang N, Levinson H. The effect of a nebulized bronchodilator placebo-controlled trial of bilevel positive airway pressure in acute administered with or without intermittent positive pressure breathing asthmatic attack. Chest 2003;123:1018-25. on ventilatory function in children with cystic fibrosis and asthma. [16] Dolovich M, Killian D, Wolff R, et al. Pulmonary aerosol deposition Am Rev Respir Dis 1972;106:867. in chronic bronchitis: intermittent positive pressure breathing versus [7] Webber B, Shenfield G, Paterson J. A comparison of three different quiet breathing. Am Rev Respir Dis 1977;115:397. techniques for giving nebulized albuterol to asthmatic patients. [17] Roberts. Acute severe asthma: differences in therapies and outcomes Am Rev Respir Dis 1974;109:293. among pediatric intensive care units. Crit Care Med 2002;30:581-5. [8] Shenfield G, Evans M, Paterson J. The effect of different nebulizers [18] Tuxen D. Detrimental effects of positive end-expiratory pressure with and without intermittent positive pressure breathing on during controlled mechanical ventilation of patients with severe the absorption and metabolism of salbutamol. Br J Clin Pharmacol airflow obstruction. Am Rev Respir Dis 1987;136:872-9. 1974;1:295. [19] Tuxen D, Lane S. The effects of ventilatory pattern on hyper- [9] Loren M, Chai H, Midlich D, et al. Comparison between simple inflation, airway pressures, and circulation in mechanical ventilation nebulization and intermittent positive pressure breathing in asthmatic of patients with severe air-flow obstruction. Am Rev Respir Dis children with severe bronchospasm. Chest 1977;72:145. 1989;140:5-9. American Journal of Emergency Medicine (2007) 25,10–14

www.elsevier.com/locate/ajem

Original Contribution Is there a link between hyperbilirubinemia and elevated urine nitriteB

Susan Watts PhDa,*, David Bryan MDa, Keith Marill MDb aDepartment of Emergency Medicine, Texas Tech University Health Sciences Center, El Paso, TX 79905, USA bMassachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA

Received 5 December 2005; revised 27 June 2006; accepted 2 July 2006

Abstract Objective: The aim of this study was to determine whether hyperbilirubinemia affects the association between a positive urine nitrite test and a positive urine culture. Methods: We conducted an institutional review board–approved, retrospective review of 12 months of patient data, compiling information for patients having urinalysis, urine culture, and total serum bilirubin. Patients were divided into 3 groups according to total serum bilirubin: less than 1.5 mg/dL, 1.5 to 3.0 mg/dL, and greater than 3.0 mg/dL. The point estimates and 95% confidence intervals of the sensitivity, specificity, false-positive proportion (proportion of false positive to all positive tests), and other test characteristics of urine nitrite as an indicator of urinary tract infection were calculated and tested for trend as a function of the 3 total serum bilirubin ranges. Results: Three thousand one hundred seventy-four patients met our study criteria. Specificity of the nitrite test decreased as a function of increasing total serum bilirubin (0.974, 0.966, and 0.855 for the 3 total bilirubin levels, respectively) with a significant trend ( P b .0001). There was no significant trend in comparable sensitivity values (0.380, 0.417, and 0.241, respectively) with P = .55. The false-positive proportion also increased as a function of total serum bilirubin (17.5%, 17.3%, and 72.0%) with P b .0001. Thus, if a patient’s total serum bilirubin was elevated to the point of jaundice (N3.0 mg/dL), it was approximately 4 times more likely that a positive urine nitrite test would be a bfalse positiveQ (ie, nitrite-positive/culture-negative) compared with those with normal serum bilirubin levels. Conclusions: Specificity of the urine nitrite test for urinary tract infection decreases as a function of increasing serum bilirubin. Most patients with hyperbilirubinemia and a positive nitrite test in our sample did not have an associated urinary tract infection. D 2007 Elsevier Inc. All rights reserved.

1. Introduction 85% to 99% [1,2]. In spite of this, many physicians in our emergency department (ED) have independently noticed an Published literature says that the specificity of urine apparent association between elevated bilirubin and false- nitrite for urinary tract infection (UTI) is high, ranging from positive urine nitrite tests. A typical example is a young man with signs and symptoms of viral hepatitis, who has a B positive urine nitrite test with no evidence of a UTI (no This research has not been previously presented at a meeting or published in any form. There was no external funding for this project. white blood cells or bacteria in his urine and a negative * Corresponding author. Tel.: +1 915 545 7333; fax: +1 915 545 7338. urine culture). Such cases are not rare events in our border E-mail address: [email protected] (S. Watts). city ED because the seropositivity rate for hepatitis A in

Table 1 Test characteristics of urine nitrite for urinary tract infection as a function of 3 serum bilirubin ranges n TP FP FN TN FPP SENS SPEC PPV NPV (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) All patients 10204 964 246 1595 7399 0.203 0.377 0.968 0.797 0.823 with UA/UC (0.181-0.227) (0.358-0.395) (0.964-0.972) (0.774-0.819) (0.815-0.831) Study population 3174 295 79 481 2319 0.211 0.380 0.967 0.789 0.828 (0.172-0.257) (0.346-0.414) (0.960-0.974) (0.747-0.830) (0.814-0.842) Total serum bilirubin b1.5 mg/dL 2656 245 52 399 1960 0.175 0.380 0.974 0.825 0.831 (0.135-0.224) (0.343-0.418) (0.967-0.981) (0.782-0.868) (0.816-0.846) z1.5 mg/dL 365 43 9 60 253 0.173 0.417 0.966 0.827 0.808 and V3.0 mg/dL (0.087-0.308) (0.322-0.513) (0.944-0.988) (0.724-0.930) (0.765-0.852) N3.0 mg/dL 153 7 18 22 106 0.720 0.241 0.855 0.280 0.828 (0.504-0.871) (0.086-0.397) (0.793-0.917) (0.104-0.456) (0.763-0.893) TP, true positive; FP, false positive; FN, false negative; TN, true negative; FPP, false positive proportion (FP/FP + TP); SENS, sensitivity; SPEC, specificity. counties along the Texas-Mexico border is 2 to 3 times Healthcare LLC, Tarrytown, NY) or semiautomated Clinitek greater than the rest of the country [3]. In addition, the 200+ (Bayer). prevalence of chronic hepatitis C in our county has been Positive urine cultures were those that had more than estimated to be greater than 2% (N10000 cases), whereas 100000 colony-forming units per milliliter. Cultures grow- most Texas counties have fewer than 1000 cases and a ing out multiple organisms resembling contamination were prevalence less than 1.75% [4]. considered negative cultures (based on the report from the We undertook this study to determine if, in fact, the microbiology laboratory). anecdotal observations correlating hyperbilirubinemia with false-positive urine nitrite tests were correct. We tested this 2.4. Data analysis hypothesis by determining whether the specificity and Using standard formulas, we calculated the specificity, positive predictive value of urine nitrite for UTI differed for sensitivity, positive predictive value (PPV), and negative patients with normal and elevated total serum bilirubin levels. predictive value (NPV) of urine nitrite for UTI for patients whose total serum bilirubin was less than 1.5 mg/dL, 1.5 mg/dL or greater but 3.0 mg/dL or less, or greater than 2. Methods 3.0 mg/dL. Similarly, we determined the proportion of 2.1. Study design This project was an institutional review board–approved, retrospective review of patient records. Patient consent was waived. 2.2. Study setting and population This study was conducted at a 327-bed urban teaching hospital and its associated outpatient clinics. In 2004, there were more than 60000 visits to the 33-bed ED, about 81000 clinic visits, and approximately 19000 hospital admissions. 2.3. Study protocol We reviewed 12 months of electronic records (Jan 2004 to Dec 2004) and compiled laboratory results into an Excel 2003 (Microsoft Corp, Redmond, Wash) spreadsheet for those patients having a workup that included urinalysis (UA), urine culture (UC), and total serum bilirubin (from complete metabolic panel or liver function tests). Specimens from UA and UC were collected on the same day, and serum Fig. 1 Receiver operating characteristic curve for the accuracy tests were measured within 3 days of the urine studies. All of the urine nitrite test for urinary tract infection as a function of urines were tested using the automated Clinitek Atlas (Bayer total serum bilirubin. 12 S. Watts et al.

Table 2 Distribution of organisms isolated from urine cultures stratified by three total serum bilirubin ranges Organism Frequency of isolation (%) Total T bili b1.5 mg/dL T bili z1.5 and V3.0 mg/dL T bili N3.0 mg/dL Escherichia coli 410 (65%) 66 (59%) 18 (58%) 494 (64%) Klebsiella sp 78 (12%) 6 (5%) 2 (6%) 86 (11%) Proteus sp 15 (2%) 6 (5%) 2 (6%) 23 (3%) Other Enterobacteriaceaea 27 (4%) 4 (4%) 2 (6%) 33 (4%) Pseudomonas sp 17 (3%) 2 (2%) 0 (0%) 19 (2%) Enterococcus fecalis 38 (6%) 11 (10%) 1 (3%) 50 (6%) Staphylococcus sp 19 (3%) 4 (4%) 3 (10%) 26 (3%) Streptococcus sp 18 (3%) 4 (4%) 0 (0%) 22 (3%) Candida sp and other yeasts 5 (1%) 4 (4%) 1 (3%) 10 (1%) No identificationb 6 (1%) 5 (4%) 2 (6%) 13 (2%) Totalc 633 (100%) 112 (101%) 31 (98%) 776 (99%) T bili, total serum bilirubin. a Includes Burkholderia, Citrobacter, Enterobacter, Morganella, Serratia, and Kluyvera. b Includes patients whose laboratory report had a blank field for organism ID or which stated bno workup done.Q c May not equal 100% because of rounding. false-positive urine nitrite tests (ie, the proportion of nitrite- respectively; PPV was 0.825 and 0.827, respectively; and the positive and culture-negative samples divided by all nitrite- proportion of false positives was 17.5% and 17.3%, positive samples) for patients with normal and elevated respectively (Table 1). However, when patients’ bilirubin serum bilirubin levels. 95% Confidence intervals (CIs) reached the level of jaundice [5] (N3.0 mg/dL), specificity were calculated using the Bayesian Calculator on the Web dropped to 0.855, PPV decreased to 0.280, and the site of the University of British Columbia, Department of proportion of false positives increased to 72% (Table 1). Healthcare and Epidemiology (http://www.healthcare. For patients with bilirubin greater than 3.0 mg/dL, there was ubc.ca/calc/bayes.html) and the Clinical Calculator on only an approximately 1-in-4 chance that a positive urine the VassarStats Web site of Vassar College (http://faculty. nitrite test was associated with a positive urine culture. vassar.edu/lowry/clin1.html). Hypothesis testing was per- Testing for trend as a function of the 3 serum bilirubin ranges formed using the Cochran-Armitage trend test with exact demonstrated highly significant relationships ( P b.0001) for computations (Statxact 3, Version 3.0.2, Cytel Software the proportion of false-positive, specificity, and PPV test Corporation, Cambridge, Mass). characteristics. Sensitivity did not vary significantly as a The frequency of isolation for each identified organism function of serum bilirubin with trend test ( P = .55). Because was compiled for each range of serum bilirubin and percent- of a decrease in specificity, the accuracy of the urine nitrite ages calculated. A difference in distribution of organisms test for urinary tract infection decreased as a function of between patients with the 3 different serum bilirubin ranges serum bilirubin (Fig. 1). was tested with the asymptotic Kruskal-Wallis test. The distribution of isolated organisms was similar among patients with the 3 serum bilirubin ranges ( P =.08) (Table 2). Cultures with no growth or with multiple isolates 3. Results suggestive of contamination were classified as negative cultures by the laboratory, and no additional information Our study population consisted of 3174 patients. In this about isolates was available. group, the prevalence of UTI was 24.4%, and the specificity of the urine nitrite test for a positive culture was 0.967, similar to that of the total population of patients with only 4. Discussion UA and UC results (Table 1). When we stratified the study population by serum Nitrite is not found in bnormalQ urine samples, and its bilirubin level, we found that the proportion of false-positive presence is presumed to indicate a urinary tract infection urine nitrite results was almost 4 times higher (72% vs 17%) caused by Gram-negative coliform bacteria that reduce in patients with serum bilirubin greater than 3.0 mg/dL nitrate (NO3) to nitrite (NO2) as part of their normal compared with those with normal or moderately elevated metabolic activity [6]. In our study population, about 21% bilirubin (Table 1). Similarly, specificity and PPV of nitrite of positive urine nitrite tests appeared to be false positives for a positive urine culture decreased as serum bilirubin because they were not associated with UTI. We also found increased. If total serum bilirubin was within normal limits that this false-positive percentage was markedly increased if (b1.5 mg/dL) or moderately elevated (z1.5 but V3.0 mg/dL), serum bilirubin was greater than 3.0 mg/dL, the level at the specificity of the urine nitrite test was 0.974 and 0.966, which jaundice is likely to occur. Hyperbilirubinemia and urine nitrite 13

In such cases of hyperbilirubinemia, we had assumed that nitrite levels for people with liver disease, but one study urine bilirubin causes a false-positive nitrite test via [35] reported that patients with cirrhosis had a nitrite chemical interference or color-masking. However, a review clearance rate that was 4 times greater than that of the of laboratory literature does not support this assumption. controls (18 vs 4.2 mL/min, respectively). Thus, it is The chemical reactions on urine dipsticks are highly reasonable to expect that a person with elevated plasma specific, with little to no likelihood of cross-reactivity, and nitrite could have elevated urine nitrite levels, even if there quality control materials include specific testing to detect is renal impairment, as is often seen with cirrhosis. interferences between analytes [7]. The test for nitrite on urine dipsticks is a modified Griess reaction, which is highly 4.1. Limitations specific for nitrite and has been a standard method for The major limitation of the current study is that our quantitation of nitrite for more than 100 years. There are conclusions are based on composite data that were not reports of false positives that mimic the pink color of a separated by patient treatment location (inpatient vs positive nitrite test on a urine dipstick, caused by outpatient), disease severity, comorbidities, age, or sex. medications that color urine red or which become red under Any one of these demographic features could affect acidic conditions (eg, phenozopyridine) [8]. It is unlikely prevalence, so generalizing the results of this study to that the yellow-brown color of bilirubin would cause this specific populations should be done with care. Similarly, type of reaction. because of the higher prevalence of hepatitis in our region, An alternate explanation for the presence of nitrite in these results may not be generalizable to other geographic apparently sterile urine is that urine nitrite may be derived areas. In particular, false-positive proportion, PPV, and NPV from a source other than microbial metabolism, namely, the will be affected by the prevalence of UTI overall and within renal clearance of elevated plasma nitrite, a stable metab- each serum bilirubin stratum, whereas sensitivity and olite of nitric oxide produced in response to disease and specificity will generally not be affected. In addition, the infection. Plasma nitrite and nitrate levels are used as a possibility of confounding by altered renal function, or marker for plasma nitric oxide because it is a very short- current or recent antibiotic treatment was not measured lived molecule, quickly oxidized to nitrite in the presence of or assessed. oxygen. Nitric oxide is involved in immune responses mediated by macrophages and neutrophils along with 4.2. Conclusions inflammatory processes initiated by endothelial cells and the nervous system in response to disease and infection. Our data strongly suggest that in patients with hyper- Nitric oxide’s role in inflammatory processes is well bilirubinemia, the presence of nitrite in urine is much less documented [9,10], and previously published studies likely to predict a UTI, compared with patients whose [9-27] indicate that it is reasonable to expect elevated bilirubin is within normal limits. The results of our study serum and urine nitrites in patients with various systemic also suggest a need for research that would assess the diseases. Elevated levels of nitrite or nitrate in the blood relationship between plasma nitrite levels and urine nitrite have been reported for Helicobacter pylori infection levels in patients with different types of systemic disease. [11,12], sepsis [13,14], gastrointestinal infection [15], and for various types of liver disease [16], including hepatitis [17],cirrhosis[18-23], and hepatocellular carcinoma References [24-26]. Additional studies have documented elevated nitrite levels in urine for disease states, such as acute [1] Rahn DD, Boreham MK, Allen KE, Nihira MA, Schaffer JI. pancreatitis [27], multiple sclerosis [28], and inflammatory Predicting bacteriuria in urogynecology patients. Am J Obstet bowel disease [29]. Gynecol 2005;192(5):1376-8. There are several published studies that provide measured [2] Deville WL, Yzermans JC, van Duijn NP, Bezemer PD, van der Windt DA, Bouter LM. The urine dipstick test useful to rule out levels of serum nitrite for various types of liver disease. The infections. A meta-analysis of the accuracy. BMC Urol 2004;4(4):1-14. levels vary widely from study to study for both patients and [3] Doyle TJ, Bryan RT. Infectious disease morbidity in the US region controls depending on whether the patients were on bordering Mexico, 1990-1998. J Infect Dis 2000;182(5):1503-10. nitrogen-restricted diets or not. However, the reported [4] Yalamanchili K, Saadeh S, Lepe R, Davis GL. The prevalence of average or median concentration of NO in patients with hepatitis C virus infection in Texas: implications for future health care. 2 Proc (Bayl Univ Med Cent) 2005;18(1):3-6. liver disease was at least 1.7 times, up to 2.9 times, greater [5] Roche SP, Kobos R. Jaundice in the adult patient. Am Fam Physician than their respective controls in all studies [17,21-23,25,26]. 2004;69(2):299-304. Similarly, a couple of studies have examined the relation- [6] Talaro KP, Talaro A. Foundations in microbiology. 4th ed. New York7 ship between blood and urine nitrite levels and have McGraw-Hill; 2002. b Q reported correlations between them [30], although excretion [7] Peele Jr J, Gadsden R, Crews R. Evaluation of Ames’ Clini-Tek . Clin Chem 1977;23(12):2238-41. of nitrite and the final urine concentration are affected by [8] Henry JB, editor. Basic examination of urine. Clinical diagnosis renal function and renal disease [31-34]. There are no and management by laboratory methods. Philadelphia, WB7 Saunders; studies that have reported simultaneous blood and urine 2001. p. 367-401. 14 S. Watts et al.

[9] Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. [22] Campillo B, Bories PN, Benvenuti C, Dupeyron C. Serum and urinary Molecular biology of the cell. 4th ed. New York: Garland Science; nitrate levels in liver cirrhosis: endotoxemia, renal function and 2002. hyperdynamic circulation. J Hepatol 1996;25(5):707-14. [10] Lamas S, Michel T, Endothelium-derived nitric oxide and control of [23] Oekonomaki E, Notas G, Mouzas IA, Valatas V, Skordilis P, Xidakis vascular tone. In: Jameson JL, editor. Principles of molecular C, et al. Binge drinking and nitric oxide metabolites in chronic liver medicine. Totowa (NJ): Humana Press; 1998. disease. Alcohol Alcohol 2004;39(2):106-9. [11] Kodama K, Sumii K, Kawano M, Kido T, Nojima K, Sumii M, et al. [24] Moriyama A, Masumoto A, Nanri H, Tabaru A, Unoki H, Imoto I, Helicobacter pylori infection increases serum nitrate and nitrite more et al. High plasma concentrations of nitrite/nitrate in patients with prominently than serum pepsinogens. Helicobacter 2002;7(1):9-13. hepatocellular carcinoma. Am J Gastroenterol 1997;92(9):1520-3. [12] Tari A, Kodama K, Kurihara K, Fujihara M, Sumii K, Kajiyama G. [25] Notas G, Xidakis C, Valatas V, Kouroumalis A, Kouroumalis E. Does serum nitrite concentration reflect gastric carcinogenesis in Levels of circulating endothelin-1 and nitrates/nitrites in patients with Japanese Helicobacter pylori–infected patients? Dig Dis Sci 2002; virus-related hepatocellular carcinoma. J Viral Hepat 2001;8(1):63-9. 47(1):100-6. [26] Moriyama A, Tabaru A, Unoki H, Abe S, Masumoto A, Otsuki M. [13] Wong HR, Carcillo JA, Burckart G, Shah N, Janosky JE. Increased Plasma nitrite/nitrate concentrations as a tumor marker for hepatocel- serum nitrite and nitrate concentrations in children with the sepsis lular carcinoma. Clin Chim Acta 2000;296(1-2):181-91. syndrome. Crit Care Med 1995;23(5):835-42. [27] Rahman SH, Ammori BJ, Larvin M, McMahon MJ. Increased nitric [14] Nakae H, Endo S, Kikuchi M, Yamada Y, Shibata M, Ishikura H, et al. oxide excretion in patients with severe acute pancreatitis: evidence of an Nitrite/nitrate (NOx) levels and hemodynamics during septic shock. endotoxin mediated inflammatory response? Gut 2003;52(2):270-4. Surg Today 2000;30(8):683-8. [28] Giovannoni G, Silver NC, O’Riordan J, Miller RF, Heales SJ, Land [15] Charmandari E, Meadows N, Patel M, Johnston A, Benjamin N. JM, et al. Increased urinary nitric oxide metabolites in patients with Plasma nitrate concentrations in children with infectious and multiple sclerosis correlates with early and relapsing disease. Mult noninfectious diarrhea. J Pediatr Gastroenterol Nutr 2001;32(4): Scler 1999;5(5):335-41. 423-7. [29] Goggins MG, Shah SA, Goh J, Cherukuri A, Weir DG, Kelleher D, [16] Geller DA, Di Silvio M, Nussler AK, Wang SC, Shapiro RA, et al. Increased urinary nitrite, a marker of nitric oxide, in active Simmons RL, et al. Nitric oxide synthase expression is induced in inflammatory bowel disease. Mediators Inflamm 2001;10(2):69-73. hepatocytes in vivo during hepatic inflammation. J Surg Res 1993; [30] Rejdak K, Petzold A, Sharpe MA, Smith M, Keir G, Stelmasiak Z, 55(4):427-32. et al. Serum and urine nitrate and nitrite are not reliable indicators of [17] Tankurt E, Kirkali G, Ozcan MA, Mersin N, Ellidokuz E, Akpinar intrathecal nitric oxide production in acute brain injury. J Neurol Sci HA. Increased serum nitrite and nitrate concentrations in chronic 2003;208(1-2):1-7. hepatitis. J Hepatol 1998;29(3):512-3. [31] Kurioka S, Koshimura K, Sugitani M, Murakami Y, Nishiki M, Kato [18] Heller J, Kristeleit H, Brensing KA, Woitas RP, Spengler U, Y. Analysis of urinary nitric oxide metabolites in healthy subjects. Sauerbruch T. Nitrite and nitrate levels in patients with cirrhosis of Endocr J 1999;46(3):421-8. the liver: influence of kidney function and fasting state. Scand J [32] Baylis C, Vallance P. Measurement of nitrite and nitrate levels in Gastroenterol 1999;34(3):297-302. plasma and urine—what does this measure tell us about the activity of [19] Hori N, Okanoue T, Mori T, Kashima K, Nishimura M, Nanbu A, et al. the endogenous nitric oxide system? Curr Opin Nephrol Hypertens Endogenous nitric oxide production is augmented as the severity 1998;7(1):59-62. advances in patients with liver cirrhosis. Clin Exp Pharmacol Physiol [33] Rabbani GH, Islam S, Chowdhury AK, Mitra AK, Miller MJ, Fuchs 1996;23(1):30-5. G. Increased nitrite and nitrate concentrations in sera and urine of [20] Arkenau HT, Stichtenoth DO, Frolich JC, Manns MP, Boker KH. patients with cholera or shigellosis. Am J Gastroenterol 2001;96(2): Elevated nitric oxide levels in patients with chronic liver disease and 467-72. cirrhosis correlate with disease stage and parameters of hyperdynamic [34] Dubey NK, Dey PK, Saxena S, Batra S, Khapekar P, Gupta A. Serum circulation. Z Gastroenterol 2002;40(11):907-13. nitrite and urinary nitrite excretion in nephrotic syndrome. Indian [21] Hokari A, Zeniya M, Esumi H, Kawabe T, Gershwin ME, Toda G. Pediatr 2001;38(9):1025-9. Detection of serum nitrite and nitrate in primary biliary cirrhosis: [35] Campillo B, Bories PN, Benvenuti C, Dupeyron C. Serum and urinary possible role of nitric oxide in bile duct injury. J Gastroenterol Hepatol nitrate levels in liver cirrhosis: endotoxemia, renal function and 2002;17(3):308-15. hyperdynamic circulation. J Hepatol 1996;25(5):707-14. American Journal of Emergency Medicine (2007) 25,15–22

www.elsevier.com/locate/ajem

Original Contribution Hospital-based healthcare provider (nurse and physician) integration into an emergency medical services–managed mass-gathering event

Christian Martin-Gill MDa, William J. Brady MDb,e,*, Kevin Barlotta MDb, Anthony Yoder EMT-Pe, Allen Williamson RNb, Benjamin Sojka EMT-Pe, Dayton Haugh EMT-Ce, Marcus L. Martin MDb, Marge Sidebottom BSc, Leonard Sandridge BS, MSd aAffiliated Residency in Emergency Medicine, University of Pittsburgh, Pittsburgh, PA 15260, USA bDepartment of Emergency Medicine, University of Virginia, Charlottesville, VA 22908, USA cOffice of Emergency Preparedness, University of Virginia Health System, Charlottesville, VA 22908, USA dVice President’s Office, University of Virginia, Charlottesville, VA 22908, USA eCharlottesville-Albemarle Rescue Squad, Charlottesville, VA 22908, USA

Received 12 October 2005; revised 26 May 2006; accepted 2 July 2006

Abstract Introduction: This report describes not only the implementation of a coordinated emergency medical services–hospital–based healthcare team but also investigates the integration of nurse-physician teams at a mass gathering medical care event. Methods: A review of resource utilization, patient encounters, and local ED census was performed during this period at a college football stadium. Results: During this 4-year period, 1681 patients presented for medical care during 26 events with a total attendance of 1544244 (1.09 patients per thousand attendees [PT]). The majority of patient contacts were for minor complaints (1451, 87.6%), whereas 205 (12.4%) received full evaluations (focused history and physical examination most often with pulse oximetric and electrocardiographic monitoring). A total of 109 patients were transported (4.19 PT), representing 6.48% of all patients. Patient census for the event medical deployment increased from 0.44 PT in 2001 to 1.75 PT in 2004. The number and percent of patients transported also increased between 2001 (0.02 PT, 4.48%) and 2004 (0.12 PT, 6.67%). However, 118 (57.6%) patients who received full evaluations were able to be discharged by a physician, avoiding transport. Chief complaints and management of patients receiving full evaluations were consistent across this period, with altered mental status (52.7%) and chest pain (12.7%) as the most common complaints. Average ED census during this period was found to be significantly higher on event days (176.2) than nonevent days (161.2) (t = 8.04, P b .001), although this produced only a minor impact on the emergent care system. Conclusion: This study describes one potential deployment plan for a mass gathering medical event and suggests that the incorporation of physicians into a mass gathering setting may be associated with an

* Corresponding author. Department of Emergency Medicine, University of Virginia, Charlottesville, VA, USA. E-mail address: [email protected] (W.J. Brady).

0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ajem.2006.07.003 16 C. Martin-Gill et al. absolute increase in patient census and transports, while decreasing the percent of patients transported. The impact on local emergency medical services and ED resources, although not specifically investigated in this study, was likely minimal. D 2007 Elsevier Inc. All rights reserved.

1. Introduction presence of the physician itself may affect both the magnitude of demand as well as the nature of the medical Mass gatherings of people, that is, those events number- need at these events. These issues are an important aspect to ing greater than 1000 participants, are a common occurrence consider as EMS personnel, physicians, and event coor- and include scenarios as diverse as family reunions, dinators consider the issues surrounding an enhanced university graduations, musical concerts, college football clinical presence at mass-gathering events. Finally, we have games, automobile races, and papal masses. During these found limited information on the impact that these events events, patients present to medical staff with a variety of may have on local EDs. complaints, from simple headache to cardiac arrest [1-3]. This report describes not only the implementation of a The majority of research on mass gathering medical care coordinated EMS-hospital–based healthcare team at a large (MGMC) has focused on describing variables that influence collegiate sporting event but also investigates the impact of the number and types of patients seen at mass gatherings in physician presence at a mass gathering. Medical resources an effort to anticipate medical care needs at these events including personnel, equipment, and operational issues are [4,5]. Several groups have expressed a need for uniformity reviewed. A review of resource utilization, patient encoun- or standardization of MGMC, and several guidelines have ters, and local ED census was performed during this period been written including recommendations for physician and at a college football stadium. nonphysician staffing levels based on crowd size and other variables such as nature of the event, local environmental conditions, and the presence of alcohol [6-10]. 2. Methods Yet, the need for physician involvement at mass gathering events continues to be debated; in fact, several This investigation is a retrospective study of patient studies have aimed to describe the arguments for and encounters, resource utilization, and ED use during the first against physician staffing [11,12]. Arguments in favor of a 4 years that physicians were involved in medical care at a physician presence include enhancement of overall care, university football stadium. The study period was summer reduced liability, and allowing a safe return to the event 2001 to fall 2004. The stadium, located at the University of without transport to a medical facility [13,14]. For example, Virginia, has a capacity for 60000 spectators and is located in their study of a large motor sports event, Grange et al in the city of Charlottesville, Va. The urban-suburban [11] found that having physicians on-site decreased population of the Charlottesville-Albemarle County region emergency medical services (EMS) transports by 89%. is approximately 120000 people and is served by a mixed They also found that 48% of patients required care outside volunteer–career EMS system. Spectators at these events the scope of prehospital personnel. Having physicians on- include people of all ages, ranging from children of local site may furthermore decrease the cost of care, improve residents, local residents, and University students to customer satisfaction, decrease local hospital impact, spectators from outside of the area. On game days, the improve disaster response, and be beneficial for media population inside the urban ring (40000 persons) of the image concerns [15,16]. region markedly increases to greater than 60000 persons. Several investigations, however, have described that the In the year 2000, the stadium was expanded in capacity majority of patients seen at these events present with only from approximately 45000 to 60000 spectators. With this minor injuries or ailments, suggesting that on-site physi- expansion came the addition of 4 first-aid stations. Before cians may not be mandatory for proper medical function of 2001, all medical care at the stadium was provided by the occasion [2,3,15-21]. Critical illness has specifically prehospital EMS providers working under local protocols been found to be uncommon at mass gathering events, for all event participants. Starting with the 2001 season, the although rare events such as cardiac arrests do occur [1,22]. athletic department requested the addition of the hospital- Furthermore, mass gathering events may not have a based healthcare providers (physicians and nurses) of the significant impact on ED volume; in addition, even events university’s ED to be involved in medical care at the associated with a large influx of patients to the ED may not stadium. This request was influenced by the collapse of a significantly impact the emergent care system [23,24]. balcony at the University of Virginia graduation in 1997 that In reviewing the literature, we found no studies that have lead to 1 fatality and 19 injuries, which raised awareness of analyzed the time when physicians were first integrated into the need for emergency preparedness at the university’s a mass-gathering environment. It is thus unclear how the mass-gathering events. The addition of physicians at the Medical services-managed mass-gathering event 17 stadium was meant to supplement and not replace the This is the standard of medical practice provided by the preexisting EMS-based medical care system as well as to EMS agencies in this area. All patient care at first-aid fully implement the new first-aid rooms. stations is documented by physicians, nurses, and/or Personnel staffing of the stadium includes EMS person- emergency medical technicians; these are maintained by nel, nurses, and physicians (Fig. 1). Staffing by prehospital the stadium medical program coordinator. For patients providers from 2 local EMS agencies has remained similar evaluated only by EMS providers or transported by an since the incorporation of physicians at the stadium. EMS unit, separate documentation is created using standard Prehospital providers are distributed in teams of 2, including prehospital care forms and maintained by the EMS agency, a combination of advanced life support (ALS) and basic life similar to other prehospital encounters. support providers. There is typically 1 team stationed at Incident command (a senior EMS provider) dispatches each of the 4 corners of the stadium, with an ALS team at 2 teams to locations within the complex from its location near opposite corners, and a separate ALS team on the field. Two the press box, overlooking the stadium. The incident bicycle EMS teams with ALS capability are also present and commander has full operational control over all medical circulate throughout the complex, focusing primarily on the personnel present at the event. Calls for assistance are parking areas and other external regions of the event, received from event staff or through the 911 system and including the 4 streets encircling the stadium. Since 2001, 2 routed to incident command, who dispatches medical teams first-aid rooms have each been staffed with a physician and based on location, access, and nature of the call. Typically, 1 or 2 nurses. Starting in 2003, 2 additional first-aid rooms prehospital provider teams either take patients to the first- have each been staffed with paramedic personnel. An aid stations for further evaluation for locations within the additional physician and nurse are available and float stadium or transport them directly to the hospital for calls between these locations during events, overseeing care at originating outside of the stadium. An ambulance, staffed by all first-aid rooms. During the hour before and after the hospital-based paramedics, is located adjacent to the first- event, EMS providers additionally staff an ALS-equipped aid rooms and transports the majority of patients from the mobile first-aid room located adjacent to the parking areas. stadium. Three ambulances staffed by EMS personnel are Physicians and nurses who provide care at the stadium available to transport patients; a fourth ambulance staffed by are ED staff members at the local university hospital. This EMS personnel is stationed adjacent to the field. Patients staffing includes 1 attending physician, 2 resident physi- transported from the stadium are primarily taken to a nearby cians, and 3 emergency nurses per game. There are no 600-bed university hospital with an annual ED census charges for patient care at the stadium, except if intravenous of 60000. lines or ALS medications are administered, and these are For the purposes of this study, descriptive information billed by the hospital where the patient is being transported. related to patient care was gathered from data obtained and

Fig. 1 University of Virginia football stadium with first-aid station, minimal EMS staffing/deployment, and adjacent surface roads. 18 C. Martin-Gill et al. stored during the 4-year study period since physicians have week immediately after an event day. Series of event days been present on-site (2001-2004). These data included the were matched with the following series of nonevent days, of total number of patient encounters, types of encounters, and the same day of the week. number of transports to the hospital. Patient encounters were All data were entered and frequency analyses were separated into those requiring a full evaluation (focused performed using a commercially available spreadsheet history and physical) and those requiring only a minor (Microsoft Excel 2000, Microsoft Corporation, Redmond, contact. Complaints requiring full evaluations included Wash). A paired-samples t test for means, comparing ED chest pain, shortness of breath, and altered level of census on event and nonevent days, was performed using consciousness, and all patients who are transported from SPSS 12.0 for Windows (SPSS, Inc, Chicago, Ill). first-aid stations received full evaluations. Minor contacts included requests for minor wound care, acetaminophen, or an ice pack. Patient care records for patients who required a 3. Results full evaluation between 2001 and 2004 were reviewed, and information gathered included age, sex, chief complaint, and During the 4-year period since the addition of the nurse- type of treatment rendered. Patient transport information physician team, 26 football games were held with an from 2001 to 2004 was also obtained from the EMS agency average game attendance of 59394 and a total attendance regarding transports to the ED. All patients who were of 1544244 (Table 1). A total of 1681 patients were evaluated at the stadium or the streets immediately adjacent evaluated, of whom 1656 were evaluated at first-aid stations to the stadium on event days near the time of the event were and 25 were evaluated by prehospital teams only. These selected. Finally, ED census information for event days and 25 patients included 8 patients found in the areas matching nonevent days was also obtained for 2001 to immediately adjacent to the stadium, who were seen by 2004. Nonevent days were selected as the same day of the EMS personnel only and were all transported to a hospital.

Table 1 Summary of event information 2001 2002 2003 2004 2001-2004 Event information No. of events 776626 Total attendance 424008 394998 356275 368963 1544244 Average attendance 60573 56428 59379 61494 59394

Patients presenting to first-aid station No. of patients 181 320 511 644 1656 Minor contacts, n (%) 160 (88.4) 269 (84.1) 476 (93.2) 546 (84.8) 1451 (87.6) Full evaluations, n (%) 21 (11.3) 51 (15.9) 35 (6.8) 98 (15.2) 205 (12.4) Average patients per game 25.9 45.7 85.2 107.3 63.7 Patients per 1000 spectators 0.43 0.81 1.43 1.75 1.07 No. of transports 4 14 27 42 87 Average transports per game 0.57 2.00 4.50 7.00 3.35 Transports per 1000 spectators 0.01 0.04 0.08 0.11 0.06 Percent of patients transported (%) 2.21 4.38 5.28 6.52 5.25

Patients presenting to EMS team only No. of patients 5 10 9 1 25 Patients per 1000 spectators 0.01 0.03 0.03 0.003 0.02 No. of transports 588122 ALS transports 12115 BLS transports 467017

Total patients No. of patients 186 330 520 645 1681 Average patients per game 26.6 47.1 86.7 107.5 64.7 Patients per 1000 spectators 0.44 0.84 1.46 1.75 1.09 No. of transports 9 22 35 43 109 Average transports per game 1.29 3.14 5.83 7.17 4.19 Transports per 1000 spectators 0.02 0.06 0.10 0.12 0.07 Percent of patients transported (%) 4.84 6.67 6.73 6.67 6.48 BLS indicates basic life support. Medical services-managed mass-gathering event 19

Fig. 2 Patient census and transports over time. Fig. 3 Emergency department census on event days vs nonevent days. This census averaged 64.7 patients per game and 1.09 patients per 1000 participants. Of these patients, 109 Of all individuals seeking medical attention during 2001 were transported to the hospital for further care, averaging to 2004, 205 (12.2%, 0.13 per 1000 participants) received 4.19 transports per game and 0.07 transports per 1000 par- full evaluations at a first-aid station (Table 2). Approxi- ticipants. Patients who were transported represented 6.48% mately half of these patients presented with altered mental of all patients encountered at the event. status, whereas others presented with chest pain (n = 26, Over the 4-year period of 2001 to 2004, there was a 12.7%) and weakness (n = 22, 10.7%) as their chief notable increase in the number of patients presenting each complaint. The most common intervention performed was year, from 0.44 per 1000 participants in 2001 to 1.75 per intravenous line placement (n = 114, 55.6%), while 1000 participants in 2004 (Fig. 2). During this time, the electrocardiogram (single lead), oxygen, or medication number of transports per game also steadily increased, from was each administered to approximately 20%. 0.02 per 1000 participants in 2001 to 0.12 per 1000 partic- Over the period of 2001 to 2004, 24 of the 26 event days ipants in 2004. The percent of patients being transported had a higher ED census than their matched nonevent days. increased from 2001 (4.84%) to 2002 (6.67%) and remained The average ED census on event days (176.2) was 15.0 visits at approximately 6.7% from 2002 to 2004. (9.3%) higher than on nonevent days (161.2) (Fig. 3). This

Table 2 Patient care information for patients receiving full evaluations 2001 2002 2003 2004 2001-2004 n%n%n%n%n% Demographics No. of patients 21 51 35 98 205 Patients per 1000 0.05 0.13 0.10 0.27 0.13 Average age 38.7 31.2 30.9 33.6 31.1 Males 13 61.9 37 72.5 21 60.0 56 57.1 127 62.0 Females 8 38.1 14 27.5 14 40.0 42 42.9 78 38.0

Chief complaint Chest pain 6 28.6 6 11.8 4 11.4 10 10.2 26 12.7 Shortness of breath 2 9.5 4 7.8 2 5.7 7 7.1 15 7.3 Abdominal pain 0 0.0 1 2.0 0 0.0 0 0.0 1 0.5 Altered mental status 10 47.6 22 43.1 24 68.6 52 53.1 108 52.7 Weakness 1 4.8 3 5.9 3 8.6 15 15.3 22 10.7 Trauma 1 4.8 13 25.5 0 0.0 0 0.0 14 6.8 Other 1 4.8 2 3.9 2 5.7 14 14.3 19 9.3

Management Intravenous line 18 85.7 37 72.5 18 51.4 41 41.8 114 55.6 Oxygen administration 10 47.6 7 13.7 7 20.0 16 16.3 40 19.5 Electrocardiogram 14 66.7 8 15.7 7 20.0 14 14.3 43 21.0 Medication administration 8 38.1 6 11.8 7 20.0 17 17.3 38 18.5 20 C. Martin-Gill et al. represented a significantly higher number of patients on integration of physicians into a mass-gathering setting with event days (t = 8.04, P b .001). a preexisting medical care team composed of prehospital personnel only. Our intent was both to describe the resources that were used and to analyze any changes in 4. Discussion patient volume, complaints, and transports during the first 4 years, which may have been influenced by the physician Mass gathering events, defined as planned public presence. We further attempted to describe whether there gatherings attended by over 1000 persons [25], have the was a significant difference in local ED census on event potential for large numbers of patients seeking medical days vs nonevent days. attention. A great variability exists both in the number and During the first 4 years that hospital-based healthcare type of medical presentations, as well as in the services providers have been present at the stadium, we found the provided at these events [4,10]. In 1990, the American average patient census (1.09 per 1000 participants) to be College of Emergency Physicians published an official consistent with published reports from other mass gather- guide to aid in the planning and delivery of emergency ings [4,25]. However, the patient census increased signif- healthcare at these settings [26]. Although this publication icantly during this period (Fig. 2). The number of patients was a tremendous step forward in addressing the need for evaluated per game nearly doubled from 2001 to 2002, organized guidance in planning for emergency medical care increased by 74% from 2002 to 2003, and then increased at mass gathering events, no national standard has been another 20% in 2004. Such a large and consistent increase in established. Instead, policy and operative procedures used patient census cannot be explained solely by an increase in by medical directors at these gatherings have been based pathology. In addition, the average game attendance and largely on case reports, precedent, and the occasional local average age and sex of patients were consistent during this government regulation [5,8,10,25,27]. period (Tables 1 and 2). Instead, there may have been an Inherent in the dilemma of preparing an emergency increased interest in seeking medical care while at the event response plan for a large event is the unpredictable number because of the increased resources that became available at of people who may require medical attention. Underscoring the stadium. Anecdotal examples from the authors’ experi- the variable nature of mass gathering medicine, reported ence at these events include various examples of event patient presentation rates (patients presenting per 1000 participants who would have ordinarily not sought medical participants) have ranged in value from 0.14 to 90.0 [25]. attention for complaints such as chest pain or traumatic Although this is one of the most perplexing variables to injury until after the event, who after talking to event staff account for, many authors have attempted to analyze the and finding that hospital-based healthcare providers were relationship between environmental factors, event type, and crowd size with patient presentation rates [25,28]. Arbon available at the stadium went to a first-aid station for et al [28] developed regression models to predict patients medical attention. Thus, a progressive increase in awareness per 1000 participants as well as transportation-to-a-hospital of these new resources by spectators and event personnel rates attempting to establish a predictive model. However, may explain the increasing patient census each year. these predictive models still relied on historical data for The majority of patient encounters during this 4-year accuracy and further support the importance of past period were for minor contacts, including requests for Band- experience in preparing for a mass gathering event. Aids and acetaminophen (Table 1). This is consistent with Adding complexity to the issue of predicting patients per previous reports and has been cited as a reason that 1000 participants is the difficulty in predicting the level and physicians may not be needed on-site at mass gatherings amount of emergency response personnel required at mass- [2,3,15-21]. The number and percent of patients transported gathering events. Opinions vary considerably from those over this 4-year period (0.07 per 1000 participants and that maintain that on-site facilities are not required at all 6.48%, respectively) were also consistent with previous events, to those that advocate the involvement of on-site reports [4,25]. However, although the patient census physicians to maximize patient care [3,10-17,19,25]. Al- significantly increased from 2001 to 2004, the percent of though the role of EMS personnel as basic first aid, ALS, full evaluations and minor contacts remained consistent, and transportation providers has been well established, the meaning that minor contacts did not represent an increasing number of personnel and the role of physicians have been majority of patient encounters or solely account for the largely based on past experience with a specific event rather increased patient census. Furthermore, the total and percent than on the application of a predictive model [28,29]. of patients transported did increase during this period. This A large number of mass-gathering venues do incorporate suggests that the increased patient census was not all due to physicians as part of the medical team, including 75% of low acuity complaints, but instead reflects an increase in major league baseball stadiums [30]. Most studies that have medical complaints of similar or slightly higher potential for discussed physician involvement on-site have described acuity. If the increased patient census is indeed attributed to established MGMC systems with or without a physician the hospital-based healthcare providers’ presence, there may presence. This study provides the first description of the be a direct benefit to patients with significant medical Medical services-managed mass-gathering event 21 problems, as more of these patients are likely to present for diogram (single lead), or medication stayed consistent after care when physicians are on-site. 2001. These differences might be explained by the small Several authors have commented that having physicians sample size of each of these interventions or a delay in on-site at mass gatherings is likely to decrease the number initiating care until en route to the hospital. of transports to the hospital and may decrease the burden on Limited information is known about the effects of mass- local EMS systems [11,13,14]. Before 2001, local EMS gathering events on local EDs. Cooke et al [23] analyzed protocols dictated that all patients seeking medical care at ED census at 14 hospitals over a 9-week period surrounding the stadium required transport to the hospital unless the a 3-week international soccer venue in the United Kingdom. patient refused care. Although the authors’ experience They found that although there was a significant increase in suggests that patients now labeled as bminor contactsQ attendance during the tournament at 2 hospitals, there was may or may not have been transported to a hospital, all no significant increase in ED census on match days. In patients who since 2001 received full evaluations would another study, Simon et al [24] analyzed medical care at 2 have needed transport to the hospital according to previous major children’s hospitals near the 1996 Olympic Games in protocols. Of the 230 patients who either received full Atlanta, Ga. They found that although there was a large evaluations or presented only to EMS providers, 109 influx of people presenting to the hospitals, this resulted in a (47.4%) were transported. This group included 118 relatively minor impact on emergent medical care for (57.6%) patients over 4-years who received full evaluations pediatric patients [24]. In our study, we found a significant but were not transported to the hospital. Having hospital- increase in ED census on event days when compared to based healthcare providers on-site allowed a large percent- matched nonevent days (Fig. 3). The average increase of age of patients who were evaluated to be discharged at the 15.0 patients per day represented 9.3% of the average stadium and avoid transport, a finding consistent with that nonevent day census. Considering that there has been no published by Grange et al [11]. However, considering that increase in staff on event days and no reported problems, it during these 4 years the total patients per 1000 participants appears that this increase in patients was easily accommo- increased by almost 300%, the number of full evaluations dated with established resources. Thus, the increase in ED per 1000 participants increased by more than 400%, and the census on event days, although significant, appeared to number of transports per 1000 participants increased by produce only a minor impact on the emergent care system, 500%, the effects of having hospital-based healthcare supporting the observations of Simon et al. We do not know, providers on-site may be mixed. Having physicians on-site however, whether not having physicians at the stadium who may indeed decrease the percent of patients who are can discharge patients and avoid certain transports would transported, but their presence at these events may be affect the impact of these mass gatherings on the ED. associated with an absolute increase in number of patients Several authors have argued that although cardiac arrest and number of transports. events are rare at mass-gathering events, the most important These mixed effects may each provide a benefit and may aspect of MGMC is the rapid response to these events by support the presence of physicians at these events. Having personnel with ALS training and equipment [2,19,31]. an increased number of patients who need medical Survey studies of professional stadiums found that all assistance seek care at events where such care is available National Football League stadiums have advanced cardiac is a direct benefit to spectators, as well as to event sponsors life support (ACLS) medication and equipment, whereas all who benefit through improved customer satisfaction. major league baseball stadiums have at least 1 ACLS Physicians may also be able to keep this increased patient provider and 96% have ACLS equipment [30,32]. At our census from resulting in an increased burden for the football stadium, rapid-response bicycle teams equipped transport system, including event-based or local EMS with defibrillators and other ACLS equipment, as well as ambulances, as well as decreasing the burden on local separate ALS teams at opposite ends of the stadium and on EDs. Thus, physician involvement in MGMC may provide a the field, provide this level of care. Our review of 26 events large public health benefit, while minimizing the impact on over the past 4 years revealed that no cardiac arrests EMS and ED systems. occurred during this time. Still, we agree that, although rare, The review of patients who received full evaluations these events are of utmost importance and should be revealed that the number and types of complaints in this considered a fundamental aspect of medical care at mass group were consistent during the 4-year period (Table 2). gatherings. Furthermore, we believe this to be independent Small variations in the number of each complaint can be of physician staffing at events, as prehospital personnel with explained by the small sample size of each complaint and proper ACLS equipment and training can provide this differences in coding from year to year. For example, an standard of care in cardiac arrest situations and provide intoxicated patient who fell may have had a chief complaint rapid transport to a higher care facility. of either baltered mental statusQ or bfall,Q which could have The major limitation of this study is that it was unable to affected the yearly numbers. The percent of patients compare the periods before and after physicians became receiving an intravenous line did decrease across all 4 years, involved in medical care at the stadium. Although this study but the percent of patients receiving oxygen, an electrocar- suggests several effects of having on-site physicians, we 22 C. Martin-Gill et al. encourage future studies that directly compare such periods [10] Sanders AB, Criss E, Steckl P, et al. An analysis of medical care at to fully identify and explain these effects. Another limitation mass gatherings. Ann Emerg Med 1986;15(5):515-9. of this study was that there was likely a lack of [11] Grange JT, Baumann GW, Vaezazizi R. On-site physicians reduce documentation by prehospital personnel for patients who ambulance transports at mass gatherings. Prehosp Emerg Care 2003; were not transported to a first-aid station or hospital. 7(3):322-6. [12] Parrillo SJ. Medical care at mass gatherings: considerations for Considering that 88% of patients determined to have physician involvement. Prehospital Disaster Med 1995;10(4):273-5. presented only to EMS teams were transported to the [13] Boyle MF, De Lorenzo RA, Garrison R. Physician integration into hospital, it is likely that minor contacts, such as requests for mass gathering medical care: the United States Air Show. Prehospital a Band-Aid, made directly to EMS staff were not well Disaster Med 1993;8(2):165-8. documented. A third limitation is that the data on ED census [14] Brunko M. Emergency physicians and special events. J Emerg Med were only obtained from the university medical center that 1989;7:405-6. [15] Baker WM, Simone BM, Niemann JT, Daly A. Special event medical receives the vast majority of transports from the stadium. It care: the 1984 Los Angeles Summer Olympics experience. Ann is possible that some event participants may have presented Emerg Med 1986;15(2):185-90. to one other local, private hospital in the city. Considering [16] Thompson JM, Savoia G, Powell G, Challis EB, Law P. Level of that the majority of people who attend the stadium are from medical care required for mass gatherings: the XV Winter Olympic out of the area and that the university hospital is closest to Games in Calgary, Canada. Ann Emerg Med 1991;20(4):385-90. [17] McDonald CC, Koenigsberg MD, Ward S. Medical control of mass the stadium, it is unlikely that this represented a large gatherings: can paramedics perform without physicians on-site? number of patients. Prehospital Disaster Med 1993;8(4):327-31. This study suggests that the incorporation of physicians [18] Ounanian LL, Salinas C, Shear CL, Rodney WM. Medical care at the and nurses, as well as the availability of first-aid stations, 1982 US festival. Ann Emerg Med 1986;15(5):520-7. may result in a higher number of patients presenting with [19] Varon J, Fromm RE, Chanin K, Filbin M, Vutpakdi K. Critical illness medical complaints at mass-gathering events. This increase at mass gatherings is uncommon. J Emerg Med 2003;25(4):409-13. [20] Weaver WD, Sutherland K, Wirkus MJ, Bachman R. Emergency in patient census represents an increase in complaints with medical care requirements for large public assemblies and a new both a low and high potential for acuity. Physicians are strategy for managing cardiac arrest in this setting. Ann Emerg Med likely to decrease the percent of patients who are trans- 1989;8(2):155-60. ported, but may increase the absolute number of transports. [21] Zeitz KM, Schneider DP, Jarrett D, Zeitz CJ. Mass gathering events: Finally, although there was a significantly higher ED census retrospective analysis of patient presentations over seven years. Prehospital Disaster Med 2002;17(3):147-50. on event days compared to nonevent days, this appeared to [22] Pons PT, Holland B, Alfrey E, Markovchick V, Rosen P, Dinerman N. produce only a minor impact on the local ED. An advanced emergency medical care system at National Football League games. Ann Emerg Med 1980;9(4):203-6. [23] Cooke MW, Allan TF, Wilson S. A major sporting event does not References necessarily mean an increased workload for accident and emergency departments. Br J Sports Med 1999;33(5):333-5. [1] Bock HC, Cordell WH, Hawk AC, Bowdish GE. Demographics of [24] Simon HK, Stegelman M, Button T. A prospective evaluation of emergency medical care at the Indianapolis 500 mile race (1983- pediatric emergency care during the 1996 Summer Olympic Games in 1990). Ann Emerg Med 1992;21(10):1204-7. Atlanta, Georgia. Pediatr Emerg Care 1998;14(1):1-3. [2] Shelton S, Haire S, Gerard B. Medical care for mass gatherings at [25] De Lorenzo RA. Mass gathering medicine: a review. Prehospital collegiate football games. South Med J 1997;90(11):1081-3. Disaster Med 1997;12(1):68-72. [3] Spaite DW, Criss EA, Valenzuela TD, Meislin HW, Smith R, Nelson [26] Leonard RB, Nuji EK, Petrilli R, Calabro JJ. Provision of emergency A. A new model for providing prehospital medical care in large medical care for crowds. Dallas (Tex)7 ACEP; 1990. p. 1-25. stadiums. Ann Emerg Med 1988;17(8):825-8. [27] Jaslow D, Drake M, Lewis J. Characteristics of state legislation [4] Michael JA, Barbera JA. Mass gathering medical care: a twenty-five governing medical care at mass gatherings. Prehosp Emerg Care year review. Prehospital Disaster Med 1997;12(4):305-12. 1999;3(4):316-20. [5] Milsten AM, Maguire BJ, Bissell RA, Seaman KG. Mass-gathering [28] Arbon P, Bridgewater FH, Smith C. Mass gathering medicine: a medical care: a review of the literature. Prehospital Disaster Med predictive model for patient presentation and transport rates. 2002;17(3):151-62. Prehospital Disaster Med 2001;16(3):150-8. [6] Butler II WC, Gesner DE. Developing a mass gathering EMS plan. [29] Franaszek J. Medical care at mass gatherings. Ann Emerg Med JEMS 1999;24(11):66. 1986;15(5):600-1. [7] Grange JT. Planning for large events. Curr Sports Med Rep [30] Ma OJ, Pirrallo RG, Rubin JM. Survey of medical services at 2002;1(3):156-61. major league baseball stadiums. Prehospital Disaster Med 1995; [8] Jaslow D, Yancy II A, Milsten A. Mass gathering medical care: 10(4):268-71. National Association of EMS Physicians Standards and Clinical [31] Hodgetts TJ, Cooke MW. The largest mass gathering. BMJ 1999; Practice Committee. Prehosp Emerg Care 2000;4(4):359-60. 318:957-8. [9] Leonard RB. Medical support for mass gatherings. Emerg Med Clin [32] Roberts DM, Blackwell TH, Marx JA. Prehosp Emerg Care 1997; North Am 1996;14(2):383-97. 1(3):149-55. American Journal of Emergency Medicine (2007) 25,23–31

www.elsevier.com/locate/ajem

Original Contribution How can we identify low- and high-risk patients among unselected patients with possible acute coronary syndrome?

Kirsten Melgaard Nielsen MD, PhDa,*, Ole Faergeman MD, DMSca, Mogens Lytken Larsen MD, DMSca, Anders Foldspang MD, PhD, DMScb aDepartment of Internal Medicine and Cardiology, Aarhus Sygehus University Hospital, Tage Hansens Gade 2, 8000 Aarhus C, Denmark bDepartment of Health Services Research, Institute of Public Health, University of Aarhus, 8000 Aarhus C, Denmark

Received 4 April 2006; revised 19 June 2006; accepted 21 June 2006

Abstract Objective: Prognosis among patients admitted with possible acute coronary syndrome (ACS) may differ from that of patients with definite ACS. The aim of this study was to identify risk factors for mortality among unselected patients and to use the statistical model to identify patients at low or high mortality risk. Methods: From April 1, 2000, to March 31, 2002, we identified all consecutive patients aged 30 to 69 years admitted to the 2 coronary care units covering the municipality of Aarhus, Denmark (population, 138290). ACS was considered a possible diagnosis if the physician at admission (1) had noted the presence or absence of chest pain, (2) performed a 12-lead electrocardiogram, and (3) measured markers of myocardial necrosis. In 1576 consecutive patients these criteria were fulfilled. Results: By logistic regression, predictors of mortality were age 60 and older, ST elevation, right bundle-branch block, arrhythmia, elevated markers of myocardial necrosis, and the diagnosis of ACS. The predictive validity of the model, as indicated by receiver operating characteristic curve area, was 85.7%, 87.8%, and 80.1% for 7-, 30-, and 365-day mortality, respectively. Conclusions: Mortality may be predicted with high precision based on a statistical model. Identification of survivors by the use of a statistical model was superior as compared to simply ruling out the clinical diagnosis of ACS. D 2007 Elsevier Inc. All rights reserved.

The study was performed during tenures of grants from Pfizer, 1. Introduction Denmark, the Danish Heart Foundation, Foundation of Laegekredsfor- eningen of Aarhus, Research Initiative of Aarhus University Hospital, Confirming or ruling out acute coronary syndrome Foundation of Kong Christian d. 10, Foundation of Jacob Madsens, (ACS) consisting of unstable angina pectoris (UAP) or Institute of Epidemiology and Social Medicine, University of Aarhus, and myocardial infarction (MI) constitutes a major challenge of the Foundation of Laegernes Forsikringsforening. * Corresponding author. Tel.: +45 86525665; fax: +45 8949 7619. a chest pain clinic. It is, however, also important to ascertain E-mail address: [email protected] (K.M. Nielsen). risk in patients not fulfilling the diagnostic criteria of ACS.

Among patients with possible ACS, the prognosis is poorer Symptoms, history of ACS, 12-lead ECG, and measure- in patients with confirmed ACS than in patients in whom ments of MMN were recorded by a study nurse or by one of ACS can be excluded [1], although mortality is also rather the authors. Patients were considered to have possible ACS high in the latter type of patients [2,3]. if the physician who examined the patient at admission had Although the plasma concentration of troponin is a wanted to rule out MI by (1) querying the patient about sensitive and fairly specific marker of minimal myocardial presence or absence of chest pain, (2) performing a 12-lead injury, it does not seem to predict mortality very well [4]. ECG, and (3) measuring MMN. A diagnosis of ACS was Combining electrocardiographic (ECG) data with measure- made if there was UAP with either significant 1-mm ST ments of markers of myocardial necrosis (MMN) in a depression or T-wave inversion (FRISC [Fast Revascular statistical model improves prediction of adverse outcomes during Instability in Coronary disease] criteria) [14] or acute [5], as does inclusion of age [6-8] and a history of ischemic myocardial infarction (AMI) according to international heart disease [9-12]. criteria [15]. A specialist in cardiovascular medicine Most prognostic studies have, however, been performed performed diagnostic evaluation of the ACS diagnosis in in clinical trials with a quite select group of patients the first 120 patients with a study diagnosis of ACS. As characterized by well-defined symptoms or ECG changes there was full agreement in the diagnostic evaluation, we did [5,6,8,13]. Our aim was to develop a prognostic prediction not use an end point committee in the evaluation of the model based on a cohort of consecutive patients with diagnosis of ACS. possible ACS. We thus wanted to avoid the preliminary exclusion of any possible patients with ACS because of 2.3. ECG and biomedical data at admission predefined inclusion criteria, ie, specific symptoms or ECG Information concerning history of ACS and the presence findings. From simple clinical findings at admittance, we or absence of chest pain was obtained from the patients’ wanted to estimate the initial acute, subacute, and long-term charts.CodingofECGwasperformedbasedonthe risk of dying and to compare the precision of prediction admission 12-lead ECG. ST-T changes were measured in afforded by the clinical diagnosis of ACS as such with the millimeters at 60 milliseconds after the J point. For changes prediction afforded by a statistical model incorporating to be considered abnormal, they had to appear in at least other information as well. Thus, we hypothesized that a 2 continuous leads, and ST elevation had to be at least 2 mm simple statistical model was better in identifying patients at in leads V -V or 1 mm in any other lead. ST depression, high and low risk of dying compared to, as traditionally, 1 2 T-wave inversion, and Q-waves were present if they were at including or excluding a clinical diagnosis of ACS. least 1 mm in any lead. Clinically significant arrhythmia was noted (Minnesota code category 8), and atrial fibrillation and atrial flutter were given a separate category. We also 2. Methods categorized right or left bundle-branch blocks (RBBB and 2.1. Design and patients LBBB) separately, and we consigned other clinically significant abnormalities such as other conduction defects The present design was a cohort study. The cohort or left ventricular hypertrophy to a separate category of consisted of all persons aged 30 to 69 years residing in bother abnormalities.Q The principle of the hierarchy of the the municipality of Aarhus, Denmark, as of April 1, 2001 Minnesota coding [16] was used for coding the ECG, but (ie, the midpoint population), in total 138,290 persons, we found the above simplified categories to fit standards who were followed up concerning episodes of possible used in most studies nowadays [14,15]. ACS during April 1, 2000, to March 31, 2002. We Serial measurements of MMN were performed at identified 1576 patients with possible ACS at the 2 coronary admission, after 6 to 8 hours, and after 12 to 16 hours as care units serving the municipality. Patients with possible part of the hospital routine. The maximum values of both ACS were not admitted elsewhere (if primarily admitted at troponin T (TnT) and creatine kinase MB (CKMB) mass other departments they were transferred to one of the were noted, and levels of TnT of 0.1 lg/L or higher and 2 coronary care units). Status as dead or alive 1 year later CKMB 10.0 lg/L or higher were considered abnormal. was available in all of these 1576 patients with use of Identical biochemical assays were used for analyzing TnT Danish registers. and CKMB in the 2 coronary care units. Patients were observed in hospital for at least 6 to 8 hours before ACS 2.2. Case finding could be ruled out. The type of ACS (MI or UAP) was decided as well after viewing the results of 2 measurements Consecutive patients with possible ACS surviving until of TnT and CKMB (6-8 hours). admission were identified from the 2 coronary care units based on daily visits by the research team to the units in the 2.4. Sociodemographic data inclusion period. All patients belonging to the cohort and admitted alive to the coronary care units, irrespective of the Data on age, sex, and death were obtained from the admission diagnosis, were screened for possible ACS. Central Office of Civil Registration for each patient, based Possible acute coronary syndrome: who dies? 25

Table 1 Mortality by chest pain, ECG, MNN, and diagnosis at admission in 1576 consecutive patients with possible ACS (Aarhus, Denmark, April 2000 to March 2002) Symptom/findings/diagnosis Patients 7-d Mortality 30-d Mortality 365-d Mortality No. %a %b ORc %b ORc %b ORc Symptom Chest pain 1281 81.3 1.0 0.8 1.5 0.4* 3.8 0.4** Previous ACSc 288 18.3 1.4 0.9 2.0 0.8 5.4 1.0 ECGd ST elevation 123 7.8 4.9 5.2** 4.9 2.2 8.1 1.4 ST depression 160 10.2 1.3 0.9 3.8 1.7 6.3 1.1 T-wave inversion 200 12.7 2.5 2.4 4.5 2.4* 8.5 1.8 Q wave 177 11.2 1.7 1.5 3.4 1.7 7.3 1.5 LBBB 49 3.1 2.0 1.5 2.0 0.8 8.2 1.4 RBBB 48 3.0 4.2 4.6* 8.3 5.5** 8.3 2.0 Arrhythmia 235 14.9 3.8 5.7*** 7.7 7.4*** 12.3 3.5*** Any abnormality 717 45.5 1.8 3.2* 3.6 5.3** 7.1 2.2** MMNe TnT z0.1 382 24.2 2.6 3.6* 5.2 4.7*** 9.2 2.3** CKMB z10.0 319 20.2 2.5 2.9** 5.0 3.6*** 8.2 1.8* Diagnosis ACS 591 37.5 2.4 6.2*** 4.4 7.1*** 8.6 2.9*** UAP 187 11.9 1.6 1.4 2.1 1.0 7.0 1.4 AMI 404 25.6 2.7 4.3** 5.4 5.9*** 9.4 2.5*** ACS excluded 985 62.5 0.3 0.2*** 0.5 0.1*** 2.4 0.3*** a Percent of total. b Percent of group with characteristic. c Odds ratio (OR) controlled for age. Persons without the characteristic constitute the reference group. d ECG at admission. Arrhythmia consists of all types of arrhythmias except atrial fibrillation/flutter. e MMN maximum value. One marker elevated if either TnT level is 0.1 lg/L or higher or CKMB is 10.0 lg/L or higher. * P b .05. ** P b .01. *** P b .001. on the unique 10-digit individual identification number v2 test or Fisher exact test for the 2 Â 2 table as well as assigned to all Danish citizens. multiple logistic regression. Variables included in the multivariate model, including age and sex, are shown in 2.5. Statistical analysis Table 1. We chose not to include many variables concerning comorbidity, smoking, and others, as we wanted a simple The 7-, 30-, and 365-day mortality rates constituted the model with use of the 3 diagnostic components: chest pain, dependent variables. Statistical analyses included Pearson ECG changes, and measurements of MMN.

Table 2 Mortality by age and clinical predictors (chest pain, ECG changes, MNN, and diagnosis) at admission in 1576 consecutive patients with possible ACS in age group 30 to 69 years (Aarhus, Denmark, April 2000 to March 2002) Predictor 7-d Mortality 30-d Mortality 365-d Mortality OR 95% CI OR 95% CI OR 95% CI Age 60-69 1.1** 1.0-1.2 1.1*** 1.0-1.2 1.1*** 1.1-1.1 Symptom Chest pain – – 2.7 0.3-22.2 0.9 0.4-2.1 ECGa Arrhythmia 7.6*** 2.2-27.0 6.1*** 2.4-15.4 2.6*** 1.5-4.5 ST elevation 5.3*** 1.5-18.3 – – – – RBBB 10.0** 1.6-58.3 8.2*** 2.1-31.8 – – MMNb Elevated – – 26.2*** 3.1-220.6 3.5** 1.2-10.3 Diagnosis ACS – – – – 2.2*** 1.1-4.4 Interactionsc Chest pain Â MMN – – 0.1** 0.0-0.9 0.2*** 0.1-0.7 Based on multiple logistic regression. Model fits: P = .1, P = .3, and P = .3 for 7-, 30-, and 365-day mortality, respectively. Controlled for sex. a ECG at admission. Arrhythmia, not including atrial fibrillation/flutter. b MMN elevated if either TnT level is 0.1 lg/L or higher or CKMB mass is 10.0 lg/L or higher. c Interactions: chest pain Â MMN: chest pain and elevated MMN, ie, a patient with chest pain and elevated MMN has an OR of dying in 30 days of 2.7 Â 26.2 Â 0.1 = 7.1, as compared to a patient with elevated MMN and no chest pain, who will have an OR of 26.2. * Pb .01. ** Pb .001. 26 K.M. Nielsen et al.

Fig. 1 ROC curves based on estimated probabilities* of dying in 7, 30, and 365 days among 1576 patients with possible ACS. Aarhus, Denmark, April 2000 to March 2002.

Regression models were reduced by forward selection of chest pain, ischemic ECG, and elevated MMN were tested candidate variables by use of the v2 distributed À2ln(likeli- with baseline variables (as shown in Table 1), as were 2- and likelihood ratio) and the Wald v2 as significance tests. 3-factor interactions of chest pain, ST elevation, and Regression model goodness of fit was estimated by the elevated MMN. P less than .05 was applied as the general Hosmer-Lemeshow test. Two- and 3-factor interactions of level of significance. Individual probabilities of death were Possible acute coronary syndrome: who dies? 27

Table 3 Estimated (based on the regression model of Table 2) and observed mortality and estimated probability predictive validity parameters in 1576 patients with possible ACS (Aarhus, Denmark, April 2000 to March 2002) Estimated probability of death (%) Observed mortality and survival Predictive validity parameters at cut point No. No. Total Test-positive Sensitivity Specificity PPV NPV deceased survived (%) (%) (%) (%) (%) 7-d Mortality b.0 9 1498 1507 z.0 8 61 69 Total 17 1559 1576 4.4 47.1 96.1 11.6 99.4 30-d Mortality b10.0 16 1486 1502 z10.0 15 59 74 Total 31 1545 1576 4.7 48.4 96.2 20.3 98.9 365-d Mortality b15.0 44 1423 1467 z15.0 31 78 109 Total 75 1501 1576 6.9 41.3 94.8 28.4 97.0 estimated based on regression coefficients and applied as 2.6. Ethics continuous predictive tests with estimation of criterion validity parameters (sensitivity, specificity, predictive value The regional Committee of Ethics in Medical Science of positive and negative test results). Relationships between and the Danish Data Protection Agency approved the study predictive validity parameters were illustrated by receiver and its database. operating characteristic (ROC) curves. After statistical analysis, patients were considered to be at relatively low mortality risk if (1) a diagnosis of ACS had been ruled out 3. Results or if (2) a low risk had been estimated based on the logistic 3.1. Patients and mortality regression model. Disagreement between these 2 methods of identifying patients at low risk was tested by the McNemar From the background population of 138290 persons v2 test. The log-rank test was used to test the difference in (median age, 46.0 years; 49.9% men), 1576 patients survival between groups, as illustrated by Kaplan-Meier (median age, 57.0 years; 66.4% men) were admitted alive survival plots [17]. with possible ACS. A history of ACS was noted in 18.3%,

Table 4 Comparison of the ability of a clinical diagnosis of ACS with that of a regression model to identify patients who would die or survive within 7, 30, and 365 days among 1576 patients with possible ACS (Aarhus, Denmark, April 2000 to March 2002) Prediction of mortality Observed 7-d mortality Observed 30-d mortality Observed 365-d mortality Clinical model Regression model No. of No. of No. of No. of No. of No. of risk statusa risk statusb deceased survivors deceased survivors deceased survivors Low Low 3 963 4 960 20 950 High Low 3 462 6 454 19 443 Low High 0 19 1 20 4 11 High High 11 115 20 111 32 97 Total 17 1559 31 1545 75 1501 v2c 3.0 408.0* 3.6 397.4* 9.8** 411.1* a Clinical model risk status: ACS is excluded = low risk. ACS is confirmed = high risk. b High risk of death as estimated by multiple regression based on Table 2 by selected cut points of estimated probability at 5.0%, 10.0%, and 15.0% for 7-, 30-, and 365-day mortality, respectively. c Comparison of patients with disagreement (in bold) by use of McNemar v2 test for paired observations. A clinical diagnosis of ACS/not ACS was not better than the statistical model in identifying deceased within 7 and 30 days (not significantly different between 3 and 0 or between 6 and 1). The clinical diagnosis of ACS/not ACS was better at identifying deceased within 365 days (19 is just significantly different from 4). On the contrary, the statistical model is much better in identifying survivors as compared to the clinical diagnosis of not having ACS (462 compared to 19, 454 compared to 20, and 443 compared to 11). * Pb .001. ** Pb .05. 28 K.M. Nielsen et al. and chest pain in 81.3%. The admission ECG indicated 3.2. Predictors of mortality myocardial ischemia (either ST depression or T-wave inversion) in 18.9% of patients, and MMN was above the After controlling for age, we found mortality to be accepted cut point in 25.9% of patients. ACS was confirmed positively associated with abnormal ECG (any abnormality, in 591 (37.5%) patients, of whom 404 had AMI and 187 had ST elevation, T-wave inversion, RBBB, or arrhythmia), UAP (Table 1). elevated MMN (either TnT or CKMB), and diagnosis of Of all patients with possible ACS, 1.1%, 2.0%, and 4.8% either ACS or AMI. Mortality was negatively associated haddiedby7,30,and365days,respectively.The with chest pain (Table 1). corresponding figures for the subgroup of patients in whom In the final regression model, predictors with positive ACS was confirmed were 2.4%, 4.4%, and 8.6%, and in association to 7-day mortality were age 60 years and older, patients in whom ACS was excluded were 0.3%, 0.5%, and arrhythmia (other than atrial fibrillation/flutter), ST eleva- 2.4% (Table 1). tion, and RBBB. Positive predictors of 30-day mortality

Fig. 2 Kaplan-Meier survival curves among 1576 consecutive patients stratified by age, diagnosis of ACS, and estimated mortality probability.* Aarhus, Denmark, April 2000 to March 2002. Possible acute coronary syndrome: who dies? 29 were age 60 years and older, arrhythmias (other than atrial for possible ACS, and most of them have included only fibrillation/flutter), RBBB and elevated MMN. Interaction patients with chest pain or other subgroups with specific between chest pain and MMN was negatively associated ECG changes. In a chest pain clinic or in a coronary care with 30-day mortality. Thus, the presence of chest pain was unit, it is, however, good clinical practice also to consider a characterized by a relatively low risk in patients with diagnosis of ACS in patients with atypical symptoms (eg, elevated MMN, who otherwise were at high risk. Positive arrhythmias or dyspnea), even in the absence of chest pain. predictors of 365-days mortality were age 60 years and We therefore assigned a patient to the category of possible older, arrhythmia (other than atrial fibrillation/flutter), ACS if the physician who examined the patient seemed to MMN, and ACS diagnosis. Interaction between chest pain want to rule out MI by querying the patient about possible and MMN was also negatively associated with 365-day chest pain and by measuring MMN as well as performing a mortality (Table 2). 12-lead ECG. Thus, we included patients who could have been at high risk due to heart disease apart from ACS or 3.3. Precision of prediction of mortality from concurrent diseases. As there was full agreement The predictive validity of the regression-based estimated between the study diagnosis of ACS (in the first 120 patients probabilities was high, as indicated by areas under the ROC with ACS) and the diagnosis evaluated by a specialist in curves of 85.7%, 87.8%, and 80.1% for the 7-, 30-, and cardiovascular medicine, we believe that few patients were 365-day mortality, respectively (Fig. 1). The estimated diagnosed incorrectly. probabilities were especially precise in identifying patients Second, ACS is currently the object of intense diagnostic at relatively low risk of dying (high specificity, high and therapeutic attention, and the risk of dying from ACS has negative predictive value [NPV]) but not as accurate in probably been reduced in routine clinical practice as well as identifying high-risk patients (low sensitivity; low positive in scientific clinical studies [18], especially in patients with predictive value [PPV]). For example, nearly all (98.9%) only chest pain and ischemic ECG changes as opposed to patients with an estimated risk of less than 10% survived for those with both biochemical, ECG and/or hemodynamic 30 days, whereas only about half (48.4%) of patients who evidence that myocardial damage has taken place. actually died within 30 days were found among the 4.7% of Chest pain did not differentiate well between patients patients with an estimated mortality risk of 10% or lower with a favorable or a poor prognosis, maybe because it was (corresponding PPV, 20.3%) (Table 3). present in more than 80% of the patients in the study. Based on the regression models, we arbitrarily defined Patients admitted to a coronary care unit or being evaluated low risk as an estimated probability below 5% of dying in a chest pain clinic without typical chest pain may within 7 days (10.0% and 15.0% within 30 and 365 days, obviously show some other serious signs of disease, such as respectively). By this method, we identified survivors syncope or shortness of breath. significantly better than by defining them as patients RBBB was a significant predictor of mortality in the final without a clinical diagnosis of ACS. In contrast, the ability regression model (Table 2). This finding could be due to small of the regression model to identify patients who died was numbers. Thus, of 48 patients with RBBB, 4 patients died, not significantly better than that of the clinical ACS 3 of them in the acute phase. Nevertheless, this finding is diagnosis (Table 4). consistent with those of other studies, which have indicated Age-stratified survival rates of patients in whom a that patients with bundle-branch block are often treated diagnosis of ACS had been confirmed or ruled out were suboptimally [19]. Moreover, these patients have a poor compared by use of Kaplan-Meier plots, as were patients at prognosis if they also have MI [20]. On the contrary, it is not relatively high and low mortality risk based on the estimated our opinion that RBBB in general is more dangerous than predictive probabilities (Fig. 2). In both cases, actual LBBB. Atypical results are sometimes the manifestations of survival differed significantly between the compared risk multivariate models, and for this study, risk of acute and b Q groups, but differences were much more outspoken when subacute deaths seems to be accumulated in RBBB. based on regression-based estimated probabilities. The ability of the regression model to initially identify patients who would later die within the observation period was not significantly different from that of the diagnosis of 4. Discussion ACS (Table 4). Neither risk assessment system adequately identified patients at high, long-term risk. In contrast, the The results of this study are apparently surprising in ability of the regression model to identify patients destined to several ways. Although MNN identified patients at high survive appeared to be substantially better than that of the risk of dying, ST depression and chest pain did not. We diagnosis of ACS, as it consistently showed higher believe, however, that these data are understandable in the specificity and higher NPVs. These findings are likely to context of the study itself and in the context of a modern be due to the high quality of treatment accorded to patients university hospital. with ACS in hospitals such as ours because the character- First, most previous studies have been conducted with istics of this regression model, but not the diagnosis of ACS, the aim of identifying patients at high risk after admission resulted in part from the structure of the study database. 30 K.M. Nielsen et al.

Thus, the results of our study are consistent with the current References practice of, first of all, ruling out AMI or UAP. Others have found that although the ACS diagnosis was missed at [1] Luscher MS, Ravkilde J, Thygesen K. Clinical application of two discharge in about 2% of patients [21], it seemed safe to novel rapid bedside tests for the detection of cardiac troponin T and discharge patients without evidence of UAP or AMI [22].By creatine kinase-MB mass/myoglobin in whole blood in acute myocardial infarction. Cardiology 1998;89:222-8. use of supplementary risk stratification based on a statistical [2] Hillis GS, Taggart P, Hillis L, Zhao N, Dalsey WC, Mangione A. model such as the one employed in the present study, low- Biochemical and clinical predictors of long-term outcome in patients risk patients may be identified and reassured with even with nonspecific chest pain and nondiagnostic electrocardiograms. higher accuracy, especially those in the youngest age group. Am Heart J 2003;145:88-94. Improved estimates of mortality risk could enable us to [3] Hillis GS, Oliner C, O’Neil BJ, et al. Coronary artery disease in patients with chest pain who have low-risk clinical characteristics and improve the targeting of therapeutic interventions to those negative cardiac troponin I. Am J Emerg Med 2001;19:118-21. at highest risk and to discharge somewhat earlier those at [4] Peacock IV WF, Emerman CL, McErlean ES, et al. Prediction of lower risk, and thus also to focus on hospital resource short- and long-term outcomes by troponin T levels in low-risk consumption. The heterogeneity of the patient group in the patients evaluated for acute coronary syndromes. Ann Emerg Med present study allows us, in principle, to generalize our 2000;35:213-20. [5] Holmvang L, Luscher MS, Clemmensen P, Thygesen K, Grande P. conclusions to a broader group of patients admitted to Very early risk stratification using combined ECG and biochemical hospital in everyday practice, but we recognize the assessment in patients with unstable coronary artery disease limitation due to differences in diagnostic practices between (a thrombin inhibition in myocardial ischemia [TRIM] substudy). countries and between hospital units. Prognostic studies The TRIM Study Group. Circulation 1998;98:2004-9. like the present one should thus be repeated in different [6] Kaul P, Newby LK, Fu Y, et al. Troponin T and quantitative ST-segment depression offer complementary prognostic information populations and different health service contexts, and in the risk stratification of acute coronary syndrome patients. J Am the regression model should be validated as well. As future Coll Cardiol 2003;41:371-80. risk groups may be identified with increasing accuracy, and [7] Porela P, Pulkki K, Helenius H, et al. Prediction of short-term outcome as therapeutic interventions improve, risk patterns and, in patients with suspected myocardial infarction. Ann Emerg Med consequently, estimated predictive probabilities will 2000;35:413-20. [8] Bjorklund E, Lindahl B, Johanson P, et al. Admission troponin T and change, which itself underscores the need for continuous measurement of ST-segment resolution at 60 min improve early risk risk monitoring. stratification in ST-elevation myocardial infarction. Eur Heart J 2004;25: 113 - 20. [9] Koukkunen H, Penttila K, Kemppainen A, et al. Ruling out myocardial 5. Conclusions infarction with troponin T and creatine kinase MB mass: diagnostic and prognostic aspects. Scand Cardiovasc J 2001;35:302-6. The main lesson from this study seems to be that among [10] Savonitto S, Ardissino D, Granger CB, et al. Prognostic value of the admission electrocardiogram in acute coronary syndromes. JAMA unselected patients with possible ACS, such as they present 1999;281:707-13. themselves in everyday clinical practice, those at high or [11] Launbjerg J, Fruergaard P, Madsen JK, Mortensen LS, Hansen JF. low risk of dying can be identified already at admission on Ten-year mortality of patients admitted to coronary care units with and the basis of information on chest pain, ECG, and MMN, without myocardial infarction. Risk factors from medical history and which are all readily available to the clinician in the diagnosis at discharge. DAVIT-Study Group. Danish Verapamil Infarction Trial. Cardiology 1994;85:259-66. coronary care unit or in the chest pain clinic. In this patient [12] Savonitto S, Fusco R, Granger CB, et al. Clinical, electrocardio- group, the strongest predictors of mortality seem to have graphic, and biochemical data for immediate risk stratification in acute been arrhythmia, ST elevation, elevated MMN, and RBBB. coronary syndromes. Ann Noninvasive Electrocardiol 2001;6:64-77. Among patients with possible ACS, as much attention [13] Jacobsen MD, Wagner GS, Holmvang L, et al. Clinical significance of should be directed at those presenting with atypical abnormal T waves in patients with non–ST-segment elevation acute coronary syndromes. Am J Cardiol 2001;88:1225-9. symptoms or atypical changes in the ECG as to patients [14] Wallentin L, Lagerqvist B, Husted S, Kontny F, Stahle E, Swahn E. with definite ACS. Statistical prediction of prognosis at Outcome at 1 year after an invasive compared with a non-invasive strategy admittance should be further developed in other populations in unstable coronary-artery disease: the FRISC II invasive randomised and other clinical contexts so that regression models could trial. FRISC II Investigators. Fast Revascularisation during Instability in become tools to support the identification of low-risk Coronary artery disease. Lancet 2000;356:9-16. [15] Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction patients and to sustain early reassurance and early hospital redefined—a consensus document of the Joint European Society of discharge, and, thus, to focus clinical intervention on those Cardiology/American College of Cardiology Committee for the redefi- with the highest need. nition of myocardial infarction. J Am Coll Cardiol 2000;36:959-69. [16] Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Acknowledgments Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation 1994;90:583-612. We thank Birgitte Gustafson and Vivian Ellerup for able [17] Armitage P, Berry G, Matthews JNS. Statistical methods in medical technical assistance in collecting and processing data. research. Oxford7 Blackwell Science LTd; 2002. Possible acute coronary syndrome: who dies? 31

[18] Abildstrom SZ, Rasmussen S, Madsen M. Significant decline in case National Registry of Myocardial Infarction 2 Investigators. Ann fatality after acute myocardial infarction in Denmark—a population- Intern Med 1998;129:690-7. based study from 1994 to 2001. Scand Cardiovasc J 2002;36:287-91. [21] Pope JH, Aufderheide TP, Ruthazer R, et al. Missed diagnoses of [19] Gunnarsson G, Eriksson P, Dellborg M. Bundle branch block and acute cardiac ischemia in the emergency department. N Engl J Med acute myocardial infarction. Treatment and outcome. Scand Cardio- 2000;342:1163-70. vasc J 2000;34:575-9. [22] Koukkunen H, Pyorala K, Halinen O. Low-risk patients with chest [20] Go AS, Barron HV, Rundle AC, Ornato JP, Avins AL. Bundle-branch pain and without evidence of myocardial infarction may be safely block and in-hospital mortality in acute myocardial infarction. discharged from emergency department. Eur Heart J 2004;25:329-34. American Journal of Emergency Medicine (2007) 25,32–38

www.elsevier.com/locate/ajem

Original Contribution Prevalence of elevated blood pressure in 563704 adult patients with stroke presenting to the ED in the United States

Adnan I. Qureshi MDa,*, Mustapha A. Ezzeddine MDa, Abu Nasar MSa, M. Fareed K. Suri MDa, Jawad F. Kirmani MDa, Haitham M. Hussein MDa, Afshin A. Divani PhDa, Alluru S. Reddi MDb aEpidemiological and Outcomes Research Division, Zeenat Qureshi Stroke Research Center, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, NJ 07103, USA bDepartment of Medicine, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, NJ 07103, USA

Received 14 March 2006; revised 5 July 2006; accepted 10 July 2006

Abstract Purpose: The aim of this study was to estimate the prevalence of elevated blood pressure in adult patients with acute stroke in the United States (US). Methods: Patients with stroke were classified by initial systolic blood pressure (SBP) into 4 categories using demographic, clinical, and treatment data from the National Hospital Ambulatory Medical Care Survey, the largest study of use and provision of emergency department (ED) services in the United States. We also compared the age-, sex-, and ethnicity-adjusted rates of elevated blood pressure strata, comparable with stages 1 and 2 hypertension in the US population. Results: Of the 563704 patients with stroke evaluated, initial SBP was below 140 mm Hg in 173120 patients (31%), 140 to 184 mm Hg in 315207 (56%), 185 to 219 mm Hg in 74586 (13%), and 220 mm Hg or higher in 791 (0.1%). The mean time interval between presentation and evaluation was 40 F 55, 33 F 39, 25 F 27, and 5 F 1 minutes for increasing SBP strata ( P = .009). A 3- and 8-fold higher rate of elevated blood pressure strata was observed in acute stroke than the existing rates of stages 1 and 2 hypertension in the US population. Labetalol and hydralazine were used in 6126 (1%) and 2262 (0.4%) patients, respectively. Thrombolytics were used in 1283 patients (0.4%), but only in those with SBP of 140 to 184 mm Hg. Conclusions: In a nationally representative large data set, elevated blood pressure was observed in over 60% of the patients presenting with stroke to the ED. Elevated blood pressure was associated with an earlier evaluation; however, the use of thrombolytics was restricted to patients with ischemic stroke with SBP below 185 mm Hg. D 2007 Elsevier Inc. All rights reserved.

* Corresponding author. Tel.: +1 973 972 7852; fax: +1 973 972 9960. E-mail address: [email protected] (A.I. Qureshi).

1. Introduction using a national probability sample of visits in noninstitutional general and short-stay hospitals in the 50 states and the District In 1981, Wallace and Levy [1] reported that blood of Columbia. A total of 663 hospitals were selected for the pressure was elevated in 84% of the 334 consecutive NHAMCS sample. Within the ED, 100 patient visits were admissions for acute stroke on the day of admission. There systematically selected over a 4-week reporting period. The was spontaneous reduction of blood pressure (an average of hospital staff collected data following an in-service by 20 mm Hg systolic [SBP] and 10 mm Hg diastolic [DBP]) specially trained interviewers from the US Bureau of the within 10 days following the acute event without any Census. Data were collected on various aspects of patient specific antihypertensive therapy, with only one third of the visits, including patient, hospital, and visit characteristics. cases remaining hypertensive on the 10th day of hospital- Among the items collected were patient’s age, sex, race, and ization. Subsequently, several other studies [2-8] have also ethnicity; patient’s expressed reason for visit; physician’s described elevation of blood pressure in the acute period of diagnoses; diagnostic services ordered or provided; procedures stroke. In a systematic review of 18 studies [9], 52% of the provided; medications; providers seen; visit disposition; im- patients with stroke had elevated blood pressure at the time mediacy with which patient should be seen; and expected of admission. Further studies have evaluated the prognostic source of payment. Items collected that are specific to the ED significance of the initial elevated blood pressure observed include mode of arrival, waiting time, duration of time in the in patients with stroke [2-8]. Either lower or higher blood ED, initial vital signs, and cause of injury. All data were pressure after ischemic stroke and higher blood pressure submitted to and coded centrally by Constella Group, Inc, after intracerebral hemorrhage were found to be associated Durham, NC, and subjected to quality control procedures. The with poor outcomes [5,6]. Furthermore, elevated blood error rate was less than 2%, and nonresponse rates were 5% or pressure among patients with intracerebral hemorrhage may less for NHAMCS data items. National estimates were increase the risk of hematoma expansion with subsequent determined using (1) inflation by reciprocals of the sampling neurologic deterioration [10,11]. selection probabilities; (2) adjustment for nonresponse; and (3) Recently, there has been renewed interest in the treatment a population weighting ratio adjustment [10,11]. of elevated blood pressure in acute stroke. Among patients with ischemic stroke, use of intravenous or intraarterial 2.2. Identification of patients with stroke thrombolysis within 6 hours of symptom onset reduces in the ED death and disability at 3 to 6 months [12,13]. Also, acutely elevated blood pressure increases the risk of thrombolytic- We selected all the ED visits for adult patients (20 years related intracranial hemorrhages in ischemic stroke and may or older) with any stroke and stroke subtypes from require concomitant antihypertensive treatment [14-16]. NHAMCS data. We identified the patients with stroke and However, optimal management strategies of elevated blood stroke subtypes using International Classification of Dis- pressure in patients with acute ischemic stroke are unclear. ease, 9th Revision, Clinical Modification (ICD-9-CM) Before initiating further studies examining antihyperten- primary diagnosis codes [20]. Diagnostic code fields were sive treatment strategies, it is necessary to define the screened for specific codes to identify patients with stroke magnitude of the problem. Most of the existing data using ICD-9 codes 430, 431, 432, 433,434, 436, and 437 as addressing this issue are derived either from single-center primary diagnoses; ischemic stroke was identified by 433, studies or post hoc analysis of multicenter studies, which 434, 436, 437.0, and 437.1; intracerebral hemorrhage, by evaluated novel neuroprotective agents [2-8].Insuch 431 and 432; and subarachnoid hemorrhage, by 430. studies, the magnitude of the problem could not be 2.3. Categorization of initial blood pressure evaluated because of variability in patient selection, study in the ED design, referral patterns, and the definition of elevated blood pressure. We therefore performed the present study to In the first analysis, we determined the prevalence of high determine the national prevalence of elevated blood pressure SBP (defined by value z140 mm Hg), high DBP (defined by in adult patients with stroke, using a nationally representa- value z90 mm Hg), and high mean arterial pressure (defined tive sample of the United States (US) population. by value z107 mm Hg). The SBP and DBP criteria were derived from the definition of hypertension by the Joint National Committee on Prevention, Detection, Evaluation, 2. Methods and Treatment of High Blood Pressure (JNC 7) [21]. In the 2.1. National Hospital Ambulatory Medical second analysis, we determined the proportion of patients in Care Survey 4 elevated blood pressure categories defined by SBP and DBP: (1) b120/b80 mm Hg; (2) 120-139/80-89 mm Hg; (3). We used the data from the National Hospital Ambulatory 140-159/90-99 mm Hg; and (4) z160/z100 mm Hg. These Medical Care Survey (NHAMCS). The NHAMCS is designed categories are comparable with prehypertension, stage to collect data on the use and provision of ambulatory care 1 hypertension, and stage 2 hypertension categories defined services in hospital emergency departments (EDs) [17-19] by JNC 7. 34 A.I. Qureshi et al.

Table 1 Prevalence of various categories of elevated blood pressure based on initial measurement among adult patients with stroke (National Hospital Ambulatory Medical Care Survey 2003) Elevated blood pressure strata All strokes Ischemic stroke Intracerebral Subarachnoid (n = 563704) (n = 276734) hemorrhage hemorrhage (n = 45330) (n = 4245) SBP z140 mm Hg 390584 (69.3%) 211713 (76.5%) 33992 (75.0%) 4245 (100.0%) DBP z90 mm Hg 172186 (30.5%) 89315 (32.3%) 10707 (23.6%) 1756 (41.4%) MAP z107 mm Hg 235843 (41.8%) 125227 (45.3%) 14938 (33.6%) 4245 (100.0%) SBP b120 and DBP b80 mm Hg 43959 (7.8%) 14313 (5.2%) 3662 (8.1%) 0 (0.0%) SBP 120-139 mm Hg/DBP 80-89 mm Hg 107807 (19.1%) 47988 (17.3%) 5960 (13.1%) 0 (0.0%) SBP 140-159 mm Hg/DBP 90-99 mm Hg 183152 (32.5%) 83559 (30.2%) 20699 (45.7%) 0 (0.0%) SBPz160 mm Hg/DBP z100 mm Hg 228786 (40.6%) 130874(47.3%) 15009 (33.1%) 4245 (100%) Age-, sex-, and race/ethnicity–adjusted ratesa SBPb120 and DBPb80 mm Hg 39766 (7.1%) 24963 (9.0%) 3278 (7.2%) Not estimatedb SBP 120-139mm Hg/DBP 80-89 mm Hg 147286 (26.1%) 16207 (5.9%) 14460 (31.9%) Not estimatedb SBP 140-159mm Hg/DBP 90-99 mm Hg 199559 (35.4%) 104305 (37.7%) 15243 (33.6%) Not estimatedb SBPz160 mm Hg/ DBPz100 mm Hg 177099 (31.4%) 131190 (47.4%) 12349 (27.2%) Not estimatedb MAP, mean arterial pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure. a Adjusted to United States population. b Not estimated due to small sample size.

2.4. Statistical analyses All values are presented as frequencies with percentages (for categorical values) and means with SD (for continuous We calculated age-, sex-, and ethnicity-adjusted rates for values). We used the Cochran-Mantel-Haenszel test and elevated blood pressure categories. To calculate the adjusted analysis of variance for categorical and continuous data, rates, NHAMCS provided weights were adjusted to the US respectively, using SUDAAN software (Release 9.0.1, census population by dividing into 42 strata (2 sex types, Research Triangle Institute, Research Triangle Park, NC). 3 race types, and 7 age decades) using the previously Bonferroni correction procedure was used to adjust for described method [13]: Wnew = Wij Â (Ci/RjWij) Â RWij, multiple comparisons. where Wnew is the weight adjusted for age, sex, and race to the US census population; Wij is the original NHAMCS provided weight for each individual j in stratum i; Ci is the 3. Results proportion of population within each stratum i, defined by sex race and age decade; RjWij is the summation of Of the 563704 adult patients evaluated with stroke, SBP NHAMCS-provided weights within each stratum i defined of 140 mm Hg or higher was observed in 63%, DBP of by proportion of population; and RWij is the sum of all original NHAMCS-provided weights. The above method was used to separately calculate Wnew for any strokes, ischemic strokes, and intracerebral hemorrhage. Because high SBP was the most prevalent, further analysis was performed using SBP defined strata. Variables pertaining to the ED visit were compared across 4 SBP groups: below 140 mm Hg, hypertension 140 to 184 mm Hg, severe hypertension 185 to 219 mm Hg, and 220 mm Hg or higher. These groups were chosen to differentiate between patients with no significant elevated blood pressure (b140 mm Hg), elevated blood pressure (140-184 mm Hg), severe elevated blood pressure precluding thrombolytic therapy (185-219 mm Hg), and severe elevated blood pressure requiring emergent treatment (N220 mm Hg) [14]. The variables compared were age, sex, ethnicity, mode of arrival, time interval between presentation and physician contact, type of primary provider, stroke subtype, medication Fig. 1 Strata of SBP according to stroke and stroke subtypes used including thrombolytics and antihypertensive medica- (National Hospital Ambulatory Medical Care Survey 2003). SBP, tion, and disposition after ED evaluation. The prevalence of systolic blood pressure; ICH, intracerebral hemorrhage; SAH, SBP strata was also presented for each stroke subtype. subarachnoid hemorrhage. Prevalence of elevated blood pressure in acute stroke 35

90 mm Hg or higher in 28%, and mean arterial pressure of rates for elevated blood pressure categories defined by both 107 mm Hg or higher in 38% of the patients. The proportion SBP and DBP are also presented in Table 1. of patients with SBP 140 mm Hg or higher according to An analysis of rates of categories defined by initial SBP stroke subtypes was as follows: ischemic stroke (67%), was as follows: below 140 mm Hg (n = 173120 [31%]), 140 intracerebral hemorrhage (75%), and subarachnoid hemor- to 184 mm Hg (n = 315207 [56%]), 185 to 219 mm Hg (n = rhage (100%) (see Table 1). The proportion of patients with 74586 [13%]), and 220 mm Hg or higher (n = 791 [0.1%]). DBP of 90 mm Hg or higher or mean arterial pressure of Fig. 1 demonstrates the distribution of the different strata 107 mm Hg or higher according to stroke subtypes is defined by SBP according to all strokes and subtypes of presented in Table 1. The age-, sex-, and ethnicity-adjusted stroke. There was no significant relationship between age

Table 2 Demographic and clinical characteristics according to strata defined by initial SBP among adult stroke patients (National Hospital Ambulatory Medical Care Survey 2003) b140 mm Hg 140-184 mm Hg 185-219 mm Hg z220 mm Hg (n = 173120) (n = 315207) (n = 74586) (n = 791) Mean age (FSD) 67.6 F 18.2 69.7 F 13.1 68.6 F 10.0 68.9 F 0.5 Age group (y) 20-29 4078 (2.4%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 30-39 1600 (0.9%) 4578 (1.5%) 0 (0.0%) 0 (0.0%) 40-49 29949 (17.3%) 17097 (5.4%) 5237 (7.0%) 0 (0.0%) 50-59 22343 (12.9%) 55744 (17.7%) 6323 (8.5%) 0 (0.0%) 60-69 23243 (13.4%) 58155 (18.4%) 26048 (34.9%) 791 (100.0%) 70-79 36891 (21.3%) 113210 (35.9%) 28205 (37.8%) 0 (0.0%) z80 55016 (31.8%) 66423 (21.1%) 8773 (11.8%) 0 (0.0%) Sex Men 59618 (34.4%) 128847 (40.9%) 30814 (41.3%) 26 (3.3%) Women 113502 (65.6%) 186360 (59.1%) 43772 (58.7%) 765 (96.7%) Race/ethnicity White 128018 (73.9%) 277421 (88.0%) 67514 (90.5%) 26 (3.3%) Black 35789 (20.7%) 30697 (9.7%) 0 (0.0%) 765 (96.7%) Other 9313 (5.4%) 7089 (2.2%) 7072 (9.5%) 0 (0.0%) Waiting time to see physician (min) 40.1 F 55.0 33.5 F 38.7 25.0 F 27.2 5.2 F 0.9a Strata of time interval between arrival and physician evaluation Less than 15 min 86719 (50.1%) 135026 (42.8%) 34605 (46.4%) 765 (96.7%) 15-60 min 64545 (37.3%) 67502 (21.4%) 24610 (33.0%) 26 (3.3%) N1 h or unknown/no triage 21856 (12.6%) 112679 (35.7%) 15371 (20.6%) 0 (0.0%) Mode of arrival Ambulance 94108 (54.4%) 178566 (56.7%) 29976 (40.2%) 765 (96.7%) Walk-in 76493 (44.2%) 126723 (40.2%) 38852 (52.1%) 26 (3.3%) Unknown 2519 (1.5%) 9918 (3.1%) 5758 (7.7%) 0 (0.0%) Length of visit (h) 40.3 F 65.9 44.1 F 69.3 50.8 F 73.2 7.5 F 29.4a Disposition Admit to hospital 121379 (70.1%) 213205 (67.6%) 54591 (73.2%) 791 (100.0%) Admit to ICU 7651 (4.4%) 17399 (5.5%) 0 (0.0%) 0 (0.0%) Died in ED 3662 (2.1%) 0 (0.0%) 0 (0.0%) 0 (0.0%) Refer to other physician/clinic for follow-up 34124 (19.7%) 44847 (14.2%) 13233 (17.7%) 0 (0.0%) Left against medical advice 3584 (2.1%) 0 (0.0%) 0 (0.0%) 0 (0.0%) Admit for 23 hour observation 0 (0.0%) 11052 (3.5%) 0 (0.0%) 0 (0.0%) Transfer to other hospital 779 (0.4%) 26984 (8.3%) 326 (0.4%) 0 (0.0%) Treatment Thrombolytic therapy 0 (0%) 1283 (0.4%) 0 (0.0%) 0 (0.0%) Labetalol 0 (0.0%) 0 (0.0%) 5361 (7.2%) 765 (96.7%) Hydralazine 0 (0.0%) 2262 (0.7%) 0 (0.0%) 0 (0.0%) Nicardipine/nitroprusside/ 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) enalapril/nitroglycerin/nitrates ICU, intensive care unit; LPN, licensed practice nurse; EMT, emergency medical technician. a Significant difference derived from comparison between values from SBP strata after adjusting for multiple comparisons. 36 A.I. Qureshi et al. strata and SBP strata among patients with stroke (see [26]. However, spontaneous reduction in the initial blood Table 2). The mean time interval between presentation and pressure over the next few days described previously in evaluation was 40 F 55, 33 F 39, 25 F 27, and 5 F most patients is inconsistent with chronic hypertension 1 minutes for increasing SBP strata ( P = .009). Labetalol [1-4]. Other mechanisms such as stress response in acute and hydralazine were used in 6126 (1%) and 2262 (0.4%) stroke leading to abnormal sympathetic activity, altered patients, respectively. None of the patients received intra- parasympathetic activity, raised levels of circulating cat- venous nicardipine, nitroprusside, enalaprate, nitroglycerin, echolamines [18] and brain natriuretic peptide [19], have or nitrates. Among patients with ischemic stroke, thrombo- been suggested to contribute to the acute rise in blood lytics were used in 1283 patients (0.4%) with SBP between pressure. Thus, the observed elevated blood pressure in 140 and 184 mm Hg and not used in any of the patients in acute stroke probably has multifactorial etiology and the higher SBP strata. therefore requires further studies. A high prevalence of elevated blood pressure appears to be associated with all stroke subtypes (see Fig. 1). However, 4. Discussion differences in underlying pathophysiology among stroke subtypes mandates different management strategies. Ische- The present study, which is one of the largest to date, mic stroke results from occlusion of an artery with demonstrates that acutely elevated blood pressure was subsequent reduction in regional cerebral blood flow within observed in over 60% of the patients presenting with stroke the affected region [27]. However, the reduction in regional to the ED. The results are derived from settings that are cerebral blood flow is not homogenous but is demarcated representative of the nationwide admissions. Therefore, the into regions of severe reduction (core) and moderate study provides more meaningful data compared with single reduction (penumbra). The penumbra remains viable for center studies or post hoc analysis of randomized trials. We hours sustained through collateral supply; however, wors- found that high SBP was most prevalent in the acute period. ening of ischemic injury in the penumbra region with Admission SBP in patients with stroke has been linked to systemic blood pressure reduction is possible. In practice, adverse outcomes, including poor clinical outcomes [2], neurologic deterioration and worse outcome associated with hematoma expansion [10], and cardiovascular stress [22]. blood pressure reduction has been demonstrated in a subset The relationship with high mean arterial pressure or DBP is of patients with ischemic stroke [28]. Based on theoretical less consistently described. We found that the age-, sex-, and and clinical considerations, aggressive blood pressure ethnicity-adjusted rates of elevated blood pressure catego- reduction is not recommended in patients with ischemic ries comparable with prehypertension, stage 1 hypertension, stroke in the acute phase [29]. This may have been the and stage 2 hypertension were 19%, 31%, and 30% among possible reason for less aggressive treatment of elevated patients with acute stroke. These rates were several-fold blood pressure in the current study. However, it should be higher than age-, sex-, and ethnicity-adjusted rates observed noted that a lower systemic blood pressure is desirable for the US population in 1999 to 2000 (prehypertension, whenever thrombolytics are considered in order to reduce 37%; stage 1 hypertension, 12%; and stage 2 hypertension, the risk of postthrombolytic intracranial hemorrhage [29]. 4%] [23]. Although a direct comparison is not possible The management of elevated blood pressure in acute because elevated blood pressure is not synonymous with intracerebral hemorrhage is based on 2 conflicting patho- hypertension, the indirect comparison provides some physiologic processes. There is potential benefit of reducing estimate of expected blood pressure ranges in general rates of hematoma expansion with systemic blood pressure population. The recognition of elevated blood pressure led reduction vs the potential provocation of ischemia in the to early evaluation of the patient in the ED. perihematoma region [24,30,31]. Recent data from preclin- It has been proposed that cerebral ischemia invokes a ical and clinical studies have supported the relative safety of protective response by increasing systemic blood pressure blood pressure reduction in acute intracerebral hemorrhage, to improve cerebral perfusion [1]. Increase in systemic although the benefit remains unclear [32,33]. Patients with blood pressure has also been described in patients with subarachnoid hemorrhage remain at risk for second rupture increased intracranial pressure, particularly in the presence acutely following an initial rupture of intracranial aneurysm of brainstem compression [24,25]. This pathophysiologic [34]. The benefit of reduction of blood pressure has been process has particular relevance for elevated blood pressure linked to reduction of hemodynamic stress on aneurysm wall observed in association with intracerebral and subarachnoid and prevention of second rupture. There has been only hemorrhages. The mechanisms that cause elevated blood indirect evidence available to support this concept [35]. pressure in the acute period of stroke are not clearly Some issues need to be considered prior to interpretation understood. A high prevalence of chronic hypertension is of our results. We used the data from the NHAMCS, a large observed among patients with stroke. It is therefore size,datasetwithstandardizeddesigntoprovidea reasonable to assume that in at least a proportion of these representative estimate of the total US experience [18,19]. patients, the elevated blood pressure is merely a reflection Since the study is based on cross sectional survey, the data of inadequately treated or undetected chronic hypertension can only be used to determine the prevalence rather than Prevalence of elevated blood pressure in acute stroke 37 incidence of diagnoses. Furthermore, the data set provides [7] Aslanyan S, Fazekas F, Weir CJ, et al. Effect of blood pressure during minimal clinical details on time interval between symptom the acute period of ischemic stroke on stroke outcome: a tertiary onset and evaluation, severity of neurological deficits, and analysis of the GAIN International Trial. Stroke 2003;34(10):2420-5. diagnostic study results. Thus, there is an undefined [8] Leonardi-Bee J, Bath PM, Phillips SJ, Sandercock PA. Blood pressure variation in patients’ time interval between symptom onset and clinical outcomes in the International Stroke Trial. Stroke 2002; 33(5):1315-20. and initial blood pressure recording. The actual prevalence [9] Willmot M, Leonardi-Bee J, Bath PM. High blood pressure in acute of elevated blood pressure may have been higher if all stroke and subsequent outcome: a systematic review. Hypertension blood pressure recordings were measured consistently early 2004;43(1):18-24. after symptom onset. We used primary ICD-9-CM codes to [10] Kazui S, Minematsu K, Yamamoto H, et al. Predisposing factors to identify patients with stroke and stroke subtypes admitted enlargement of spontaneous intracerebral hematoma. Stroke 1997;28(12):2370-5. to the hospital. A previous study [36] reported that if both [11] Ohwaki K, Yano E, Nagashima H, et al. Blood pressure management primary and secondary ICD-9-CM codes were used, 97% in acute intracerebral hemorrhage: relationship between elevated blood of all strokes and transient ischemic attacks could be pressure and hematoma enlargement. Stroke 2004;35(6):1364-7. detected, but the true-positive rate for stroke was 72%. [12] The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic Using only primary discharge diagnoses of ICD-9-CM stroke. N Engl J Med 1995;333(24):1581-7. codes, the true positive rate increased to 83% but decreases [13] Furlan A, Higashida R, Wechsler L, et al. Intra-arterial prourokinase the yield of detection of stroke and transient ischemic for acute ischemic stroke. The PROACT II study: a randomized attacks to 84%. Therefore, we may have underestimated the controlled trial. Prolyse in Acute Cerebral Thromboembolism. JAMA total of number of stroke evaluations. Further underesti- 1999;282(21):2003-11. [14] Bowes MP, Zivin JA, Thomas GR, et al. Acute hypertension, but not mation may have occurred because of multiple ED visits by thrombolysis, increases the incidence and severity of hemorrhagic the same patient because patients are identified by visits transformation following experimental stroke in rabbits. Exp Neurol rather than as individuals in the NHAMCS data set. 1996;141(1):40-6. Although the prevalence of elevated blood pressure in [15] Huynh T, Cox JL, Massel D, et al. Predictors of intracranial patients with acute stroke was determined, it was difficult hemorrhage with fibrinolytic therapy in unselected community patients: a report from the FASTRAK II project. Am Heart J 2004; to define whether or not hypertension was acute or chronic 148(1):86-91. because of unavailability of supporting data. Despite these [16] Aylward PE, Wilcox RG, Horgan JH, et al. Relation of increased concerns, the objective of the study, however, was to arterial blood pressure to mortality and stroke in the context of determine the prevalence of elevated blood pressure contemporary thrombolytic therapy for acute myocardial infarction. A randomized trial. GUSTO-I Investigators. Ann Intern Med associated with stroke admissions, using a methodology 1996;125(11):891-900. that has a high specificity. [17] McCaig LF, Burt CW. National Hospital Ambulatory Medical Care In summary, we provide estimates of elevated blood Survey: 2002 Emergency Department Summary. Advance data from pressure observed among patients with stroke evaluated in vital and health statistics. No. 340, Hyattsville, MD7 National Center the ED. Elevated blood pressure is highly prevalent across for Health Statistics; 2004. [18] Selim AJ, Berlowitz DR, Fincke G, et al. Risk-adjusted mortality rates different stroke subtypes. These data provide further support as a potential outcome indicator for outpatient quality assessments. for planning clinical studies that address the management of Med Care 2002;40(3):237-45. elevated blood pressure and clinical outcomes in patients [19] Oster A, Bindman AB. Emergency department visits for ambulatory with acute stroke [37-39]. care sensitive conditions: insights into preventable hospitalizations. Med Care 2003;41(2):198-207. [20] International Classification of Diseases, 9th Revision, Clinical Modification. Washington (DC)7 National Center for Health Statistics, References US Public Health Service; 1980. [21] Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the [1] Wallace JD, Levy LL. Blood pressure after stroke. JAMA Joint National Committee on Prevention, Detection, Evaluation, and 1981;246(19):2177-80. Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003; [2] Okumura K, Ohya Y, Maehara A, et al. Effects of blood pressure 289(19):2560-72. levels on case fatality after acute stroke. J Hypertens 2005;23(6): [22] Hirashima Y, Takashima S, Matsumura N, et al. Right sylvian fissure 1217-23. subarachnoid hemorrhage has electrocardiographic consequences. [3] Rodriguez-Garcia JL, Botia E, de La Sierra A, et al. Significance of Stroke 2001;32(10):2278-81. elevated blood pressure and its management on the short-term [23] Qureshi AI, Suri MF, Kirmani JF, Divani AA. Prevalence and trends outcome of patients with acute ischemic stroke. Am J Hypertens of prehypertension and hypertension in United States: National Health 2005;18(3):379-84. and Nutrition Examination Surveys 1976 to 2000. Med Sci Monit [4] Vemmos KN, Tsivgoulis G, Spengos K, et al. Blood pressure course in 2005;11(9):CR403-9. acute ischaemic stroke in relation to stroke subtype. Blood Press [24] Qureshi AI, Tuhrim S, Broderick JP. Spontaneous intracerebral Monit 2004;9(3):107-14. hemorrhage. N Engl J Med 2001;344(19):1450-60. [5] Vemmos KN, Tsivgoulis G, Spengos K, et al. U-shaped relationship [25] Qureshi AI, Geocadin RG, Suarez JI, Ulatowski JA. Long-term between mortality and admission blood pressure in patients with acute outcome after medical reversal of transtentorial herniation in patients stroke. J Intern Med 2004;255(2):257-65. with supratentorial mass lesions. Crit Care Med 2000;28(5):1556-64. [6] Castillo J, Leira R, Garcia MM, et al. Blood pressure decrease during [26] Arboix A, Roig H, Rossich R, et al. Differences between hyperten- the acute phase of ischemic stroke is associated with brain injury and sive and non-hypertensive ischemic stroke. Eur J Neurol 2004; poor stroke outcome. Stroke 2004;35(2):520-6. 11(10):687-92. 38 A.I. Qureshi et al.

[27] Janardhan V, Qureshi AI. Mechanisms of ischemic brain injury. Curr [33] Qureshi AI, Mohammad YM, Yahia AM, et al. A prospective Cardiol Rep 2004;6(2):117-23. multicenter study to evaluate the feasibility and safety of aggressive [28] Ahmed N, Nasman P, Wahlgren NG. Effect of intravenous nimodipine antihypertensive treatment in patients with acute intracrebral hemor- on blood pressure and outcome after acute stroke. Stroke 2000; rhage. J Intensive Care Med 2005;20(1):34-42. 31(6):1250-5. [34] Hillman J, von Essen C, Leszniewski W, Johansson I. Significance of [29] Adams Jr HP, Adams RJ, Brott T, et al. Guidelines for the early bultra-earlyQ rebleeding in subarachnoid hemorrhage. J Neurosurg management of patients with ischemic stroke: a scientific statement 1988;68(6):901-7. from the Stroke Council of the American Stroke Association. Stroke [35] Wijdicks EF, Vermeulen M, Murray GD, et al. The effects of treating 2003;34(4):1056-83. hypertension following aneurysmal subarachnoid hemorrhage. Clin [30] Qureshi AI, Wilson DA, Hanley DF, Traystman RJ. Pharmacologic Neurol Neurosurg 1990;92(2):111-7. reduction of mean arterial pressure does not adversely affect regional [36] Broderick J, Brott T, Kothari R, et al. The Greater Cincinnati/Northern cerebral blood flow and intracranial pressure in experimental Kentucky Stroke Study: preliminary first-ever and total incidence intracerebral hemorrhage. Crit Care Med 1999;27(5):965-71. rates of stroke among blacks. Stroke 1998;29(2):415-21. [31] Qureshi AI, Bliwise DL, Bliwise NG, et al. Rate of 24-hour blood [37] Potter J, Robinson T, Ford G, et al. CHHIPS (Controlling Hyperten- pressure decline and mortality after spontaneous intracerebral hemor- sion and Hypotension Immediately Post-Stroke) pilot trial: rationale rhage: a retrospective analysis with a random effects regression model. and design. J Hypertens 2005;23(3):649-55. Crit Care Med 1999;27(3):480-5. [38] Qureshi AI. Antihypertensive treatment of acute cerebral hemorrhage. [32] Powers WJ, Zazulia AR, Videen TO, et al. Autoregulation of cerebral The 30th International Stroke Conference, New Orleans, 2005. blood flow surrounding acute (6 to 22 hours) intracerebral hemor- [39] Kirmani JF. Hypertension in acute stroke treatment. The 30th rhage. Neurology 2001;57(1):18-24. International Stroke Conference, New Orleans 2005. American Journal of Emergency Medicine (2007) 25,39–44

www.elsevier.com/locate/ajem

Original Contribution Impact of a negative prior stress test on emergency physician disposition decision in ED patients with chest pain syndromes

Rebecca H. Nerenberg BS, Frances S. Shofer PhD, Jennifer L. Robey RN, BSN, Aaron M. Brown BS, Judd E. Hollander MD*

Department of Emergency Medicine, Hospital of the University of Pennsylvania, Philadelphia, PA 19104, USA

Received 24 April 2006; revised 26 May 2006; accepted 28 May 2006

Abstract Objective: Many emergency department (ED) patients with potential acute coronary syndromes (ACS) have prior visits and prior cardiac testing; however, the effect of knowledge of prior testing on the emergency physician disposition decision making is not known. We studied the impact of prior noninvasive testing (ie, stress testing) for myocardial ischemia on disposition decision making in ED patients with potential ACS. Methods: We performed a prospective cohort study of ED patients with chest pain who received an electrocardiogram for potential ACS. Data included demographics, medical history, stress test history, and TIMI risk score. Patients were followed in-house; 30-day telephone interviews were performed for follow-up. Main outcomes were ED disposition (admit/discharge) and a composite of 30-day death, acute myocardial infarction, and revascularization stratified on the basis of prior stress testing known at the time of presentation. Standard statistical techniques were used with 95% confidence intervals (CI). Results: There were 1853 patients enrolled and 97% had follow-up. Patients had a mean age of 53 F 14 years; 60% were women, 67% were black. There were 1491 (79%) patients without a prior stress test, 291 (16%) had a normal prior stress test result, and 89 (5%) had an abnormal prior stress test result. Admission rates were 92% (95% CI, 87%-98%) for patients with a prior abnormal stress test, 73% (95% CI, 67%-78%) for patients with a normal prior stress test, and 70% (95% CI, 67%-72%) for patients without a prior stress test. Adverse outcomes were the highest among patients with prior abnormal stress test but did not differ significantly between patients with no prior stress test and patients with prior normal stress test (10.1% [95% CI, 3.6-16.7%] vs 5.2% [95% CI, 4.1-6.4%] vs 4.8% [95% CI, 2.4-7.3%]). Conclusion: Patients without prior stress tests and patients with prior normal stress tests were admitted for potential ACS at the same rate and had the same 30-day cardiovascular event rates. This suggests that prior stress testing does not affect subsequent disposition decisions. Perhaps cardiac catheterization or computed tomography coronary angiograms would have more of an impact on subsequent visits, making them potentially more cost-effective in the low-risk patient. D 2007 Elsevier Inc. All rights reserved.

Presented at the SAEM 2006 annual meeting, San Francisco, CA, May 2006. * Corresponding author. Tel.: +1 267 250 7167; fax: +1 215 662 3953.

1. Introduction 2.3. Participants

Each year, approximately 5.8 million people present to Consecutive patients aged 30 years or older who presented the ED with complaints of chest pain, representing 5% of all to the ED between July 13, 2003, and May 31, 2005, with ED visits [1]. The decision to admit patients for potential chest pain who received an electrocardiogram were enrolled. acute coronary syndrome (ACS) is not always clear, Patients who were younger than 30 years, admitted cocaine resulting in hundreds of thousands of unnecessary admis- use, and those with ST segment elevation myocardial sions, whereas approximately 2% of patient with acute infarction (STEMI) were excluded, as these cohorts would myocardial infarction (AMI) are inadvertently discharged be unlikely to have heavy reliance on a prior stress test. from the ED [2]. Many strategies have been implemented During the study period, trained research assistants were over the years, both to decrease unnecessary admissions and present in the ED from 8 am to midnight 7 days per week to to avoid missing patients with ACS [3-10]. identify and enroll appropriate patients. Written consent for One diagnostic modality commonly used to evaluate study participation was obtained. Patients were treated by patients with potential ACS is an exercise or pharmacologic board-certified emergency physicians working in conjunc- stress test. The result of the stress test helps brule inQ or brule tion with house staff. Patient treatment was at the discretion outQ ACS during any given admission. But what impact do of the treating physician and was not influenced by study these stress tests have on the patient’s next ED visit? Are enrollment. Data collection was approved by the institu- they used by physicians to assist with subsequent disposi- tional review board. tion decision making? Does a previous negative stress test have any prognostic value for a future ED visit? 2.4. Measurements Chan et al [11] evaluated patients admitted to monitored Patients received a structured history and physical telemetry beds who received inpatient vs outpatient vs no stress testing and found no difference in 30-day outcome examination as well as an electrocardiogram, with all measures. Tatum et al [7] looked at the utility of ED information collected prospectively in the ED. The struc- myocardial perfusion imaging in low-risk patients for ACS, tured history included questions about the onset of chest and Lai et al [12] studied the necessity for inpatient stress pain, the number of previous episodes of chest pain, and the testing for low-risk patients. Smith et al examined the character of the pain. Patients were queried about family incidence of AMI in patients with a documented negative history of MI before 55 years of age, medical history of stress test within the previous 3 years and found that 4.8% hypercholesterolemia, hypertension, coronary artery disease of them were diagnosed with AMI [13]. This implies that a (CAD), diabetes mellitus, angina, and congestive heart recent negative stress test does not conclusively rule out failure as well as prior stress testing, coronary artery bypass AMI when a patient has a new episode of symptoms. grafting (CABG), and cardiac catheterization. A TIMI risk Given that stress testing is often used in the evaluation of score was calculated for each patient by the treating patients with chest pain, it would be expected that prior physician. Patient care including disposition was determined stress test results may impact ED physician’s decision by the treating physician and recorded as part of the data set. making during subsequent visits. There are no data, Physicians were blinded to study hypothesis. Data were however, about what impact, if any, physician knowledge collected in accordance with the standardized reporting of a previous stress test has on the disposition decision guidelines [14]. Admitted patients were followed up daily by making and therefore on 30-day outcome. The goal of this one of the investigators. Thirty-day cardiovascular outcomes study was to determine whether ED physician knowledge of were obtained via follow-up telephone calls to patients or a previous stress test impacted disposition decision making their contacts. The national death index was searched for for patients with potential ACS. patients that could not be contacted for 30-day follow-up.

2.5. Noninvasive (stress) testing 2. Methods Patients who reported prior stress tests from our 2.1. Study design institution generally received an exercise test with either echocardiography or sestamibi, or a dobutamine or persan- We performed a prospective cohort study to determine tine stress test with sestamibi. Exercise treadmill testing what impact, if any, knowledge of a previous stress test has without imaging is uncommon at our institution. upon emergency physician disposition decision making During the hospitalization, when patients received stress during a subsequent ED visit for potential ACS. testing, patients were injected with 10 mCi technetium Tc 2.2. Setting 99m sestamibi before getting resting tomographic images of the heart using single-photon tomographic techniques. This study was conducted in the ED of a university- Approximately 2 hours later, they underwent the stress affiliated, urban tertiary care center with an ED census of portion of the test. Exercise stress tests were performed on a more than 55000 visits per year. treadmill test using a Bruce or modified Bruce protocol. Prior stress testing and EP decision making 41

Alternatively, patients received an infusion of 0.142 mg/kg Table 2 Disposition by prior stress test per minute of dipyramidole over 4 minutes, or 5 lg/kg per minute of dobutamine increasing to 10, 20, and Disposition Prior stress test 30 lg/kg per minute over 3-minute intervals. Approximate- Abnormal None Normal ly 30 minutes after the exercise or infusions, tomographic Admitted (95% CI) (%) 92 (87-98) 67 (67-72) 72 (67-78) imaging of the heart was again obtained. A board-certified/ Discharged (%) 8 30 27 eligible cardiologist interpreted the test. For purposes of this Left against 022 study, the stress test was considered positive if there were medical advice (%) signs of reversible ischemia. Nondiagnostic tests were those that did not achieve a heart rate 80% of predicted maximum posite of 30-day cardiovascular events including death, and those that had equivocal images (eg, an inability to nonfatal AMI, percutaneous coronary intervention, or distinguish breast attenuation from inferior ischemia). CABG. AMI was defined as a troponin I (TnI) level greater 2.6. Outcome measures than 2.0 or creatine kinase-MB (CK-MB) level greater than 10, in accordance with institutional norms. Self (or proxy)- The primary outcome was the emergency physician reporting of AMI, percutaneous coronary intervention, disposition decision. The secondary outcome was a com- CABG, or death was recorded at 30-day follow-up. Acute coronary syndrome included AMI and unstable Table 1 Patient characteristics angina. Unstable angina was objectively defined in accordance with the standardized reporting guidelines as those Prior stress test P patients with consistent symptoms and proven underlying Abnormal None Normal coronary disease including a positive exercise stress test Patient characteristics (N1.5 mm ST-segment depression), reversible ischemic Age (mean, y) 60 F 10 52 F 14 54 F 13 b.05 changes on stress echocardiogram or sestamibi scan, 70% Sex or greater blockage of at least one coronary artery on Female 39 (43) 900 (62) 164 (56) b.05 angiography, or elevation of cardiac TnI or CK-MB Male 51 (57) 552 (38) 131 (44) level above normal but less than the limits set for AMI Race (0.3 ng/mL V cardiac TnI b 2 ng/mL; or 5 ng/mL V CK-MB African American 58 (64) 974 (67) 192 (65) .57 b 10 ng/mL) [14]. White 26 (29) 425 (29) 93 (32) Other 6 (7) 52 (4) 10 (3) 2.7. Primary data analysis Cardiac risk factors Hypertension 76 (85) 740 (52) 178 (61) b.05 Data are presented as means F SD for continuous Diabetes mellitus 32 (36) 285 (20) 61 (21) b.05 variables and frequency and percent for categorical varia- b Hypercholesterolemia 48 (54) 307 (22) 104 (36) .05 bles. To determine whether patients differed by stress Family history of 19 (21) 208 (15) 43 (15) .23 test category (abnormal, normal, or none) and baseline MI at age b55 y b characteristics and outcomes (AMI, ACS, or 30-day events), Tobacco use 48 (54) 513 (36) 109 (37) .05 2 Undiagnosed chest pain 10 (11) 127 (9) 60 (21) b.05 analysis of variance and the v test was used for continuous Prior cardiac events and categorical data, respectively. Assuming a 70% admis- Congestive heart failure 21 (24) 149 (10) 36 (12) b.05 sion rate in patients without a prior stress test, this study had CAD 68 (76) 224 (16) 66 (23) b.05 90% power to detect a 10% absolute decrease in the Angina 50 (56) 171 (12) 51 (17) b.05 b MI 38 (43) 162 (11) 44 (15) .05 Table 3 Admission by location and prior stress test Associated symptoms Location Prior stress test Shortness of breath 52 (58) 763 (53) 144 (49) .28 Diaphoresis 24 (27) 281 (19) 61 (21) .23 Abnormal None Normal Nausea 24 (27) 290 (20) 65 (22) .25 Cardiac catheterization 2 (2) 4 (.28) 0 (0) Vomiting 8 (9) 89 (6) 15 (5) .42 laboratory Lightheadedness 21 (23) 252 (17) 44 (15) .18 ICU (includes CCU, 21 (23.6) 107 (7.5) 20 (6.87) Syncope 0 (0) 36 (2) 2 (1) .05 CICU, MICU, SICU) Palpitations 11 (12) 181 (12) 37 (13) 1.0 Telemetry bed 54 (60.7) 771 (54.3) 73 (59.5) Chest pain at rest 78 (87) 1196 (82) 247 (84) 1.44 Nonmonitored bed 5 (5.6) 70 (4.93) 17 (5.84) Multiple episodes 49 (54) 690 (48) 148 (50) .33 Discharged 7 (7.87) 431 (30.3) 79 (27.2) of chest pain Other 0 (0) 38 (2.67) 2 (0.69) Chest pain characteristics (median, min) Values are presented as number (%). Chest pain onset 180 240 240 .19 ICU, indicates intensive care unit; CCU cardiac care unit; CICU, Chest pain duration 120 120 180 .70 cardiac intermediate care unit; MICU, medical intensive care unit; Values are presented as number (%) unless otherwise indicated. SICU, surgical intensive care unit. 42 R.H. Nerenberg et al.

Table 4 Discharge diagnoses by prior stress test Final diagnosis Prior stress test Abnormal, No. (%) 95% CI None, No. (%) 95% CI Normal, No. (%) 95% CI ACS 22 (25.0) 13.0-34.0 130 (8.7) 7.3-10.1 31 (10.8) 7.2-14.4 AMI 9 (10.2) 3.9-16.5 62 (4.2) 3.2-5.2 12 (4.2) 1.9-6.5 CP NOS 39 (44.3) 33.9-54.7 988 (66.4) 64.0-68.8 190 (66.0) 60.5-71.5 Nonischemic 18 (20.5) 12.1-29.3 204 (13.7) 12.0-15.4 53 (18.4) 13.9-22.9 Definitive other diagnosis 0 (0) 0-3.4 17 (1.1) 0.6-1.6 2 (0.7) 0.1-4.0 CP NOS, chest pain not otherwise specified. admission rate for patients with a prior negative stress test received a stress test before discharge and that testing is (PASS 2002, Jerry Hintze, Number Cruncher Statistical relatively low yield [11,18]. Amsterdam et al found that System, Kaysville, UT). All comparisons were performed only 13% of 1000 low-risk patients receiving immediate using SAS statistical software (Version 9.1, SAS Institute, exercise testing in the ED had a positive exercise test result Cary, NC). [17]. They found no deaths and only one AMI by 30 days in the 64% of patients with a negative test. The rates of positive test results are lower in chest pain centers with 3. Results mandatory predischarge stress testing. Mikhail et al [18] noted that 420 patients underwent stress testing, 32 were During the 21 months of the study, 1853 eligible admitted for positive results, and less than 2% received participants were enrolled. The average patient age was revascularization. Chan et al [11] found a revascularization F 53 14 years; 60% were women, and 67% were Black. rate of only 1%. Lai et al evaluated 344 patients admitted to Patient characteristics are presented in Table 1. a chest pain observation unit who did not receive stress Of the 1853 enrolled patients, 295 (16%) had a testing before discharge and found a 60-day death or MI rate documented normal stress test before presentation, of 0.6% [12]. They concluded that low-risk patients who 90 (4.9%) had a prior documented abnormal stress test. rule out for MI in a chest pain observation unit do not One thousand four hundred ninety-one (79%) patients require inpatient stress testing and that the safety of had no prior stress test. Disposition by prior stress test, discharging these patients from the chest pain observation admission location by prior stress test, and discharge unit is not significantly different than providing immediate diagnoses are presented in Tables 2-4. stress testing [12]. This is further supported by the evidence Thirty-day follow-up was available on 1801 (97%) from Mikhail et al [18] who found no diagnoses of AMI patients. Adverse outcomes (including AMI and death) within 5 months in patients discharged form the chest pain were significantly higher for patients with a prior abnormal observation unit. Chan et al [11] found no difference in stress test compared with either prior normal or no prior 30-day outcomes regardless of whether the patients received b stress test ( P .0001; Table 5). However, there was no predischarge testing, outpatient testing, or no provocative difference in adverse outcome rate between patients with testing within the 30-day period. Although testing may help a prior normal stress test and patients without a prior risk-stratify patients, there is no evidence that predischarge stress test. testing offers advantages over outpatient testing with regard to patient outcomes [11]. Shaver et al [19] took a different approach to studying 4. Discussion this problem. They evaluated the impact of a negative evaluation of underlying CAD upon cardiac resource Most low-risk patients with chest pain are admitted to utilization over the ensuing year. They found that there the hospital or observed in chest pain units to rule out was no association between a negative evaluation for an ACS [2,4,5,7,9,11,15]. Typically, this cohort of patients underlying CAD and subsequent cardiac-related ED visits, receives predischarge provocative testing to risk-stratify admissions, or cardiac resource test utilization over the year patients in terms of prognosis and to assist further after the index visit [19]. Thus, it appears that negative management. Previous studies have evaluated the usefulness and cost-effectiveness of stress testing in the ED, the chest pain observation unit, as an inpatient and as an Table 5 Adverse outcome by prior stress test outpatient, and the impact stress testing has on repeat ED Prior stress testing Adverse 30-day outcome, No. (%) 95% CI visits [12,13,15-19]. Abnormal 9/89 (10.1) 3.6-16.7 Several studies have shown that, in actual clinical None 74/1421 (5.2) 4.1-6.4 practice, the minority of patients (21%-22%) admitted to a Normal 14/291 (4.8) 2.4-7.3 non–intensive care telemetry bed to rule out an ACS Prior stress testing and EP decision making 43 tests do not provide a sufficient level of comfort to different than the patients who were enrolled. However, patients or providers when patients have persistent or re- because the large majority of the patients (N85%) seen in the current symptoms. ED arrive during the time the research assistants are present, We extended this previous research to investigate this represents only a small cohort of patients. whether knowledge of prior negative stress test impacts In addition, a patient with a prior stress test at any point emergency physician clinical decision making in the ED and in time was classified as having a prior test. Therefore, a found that knowledge of prior negative stress test did not patient who had a normal stress test 5 weeks ago was placed impact the likelihood of admission to the hospital. Admis- in the same category as a patient who had a normal stress sion rates in patients without a prior negative test were the test several years ago. It is possible that the timing of the same as admission rates in patients that never had a test, prior normal stress test could have an impact on disposition suggesting that prior negative tests do not impact emergency decision making. For purposes of this analysis, all prior physician decision making for subsequent visits. stress tests were considered the same because patients often Emergency physicians do not appear to be influenced by do not know the type of test they received. It would require a prior normal stress test, but do admit patients with a prior a larger number of patients with prior stress tests at different abnormal stress test more frequently. We also found that points in time to determine if timing or type of test affects patients with a prior normal stress test are at the same risk of physician disposition decision making. adverse cardiovascular events as patients who have not Cardiac risk factors differed between patients who had previously undergone stress testing. Thus, knowledge of prior abnormal, normal, or no prior stress testing. Patients previously normal stress did not and should not impact with an abnormal prior stress test were significantly more clinical decision making in the ED. From a pathophysiology likely to have hypertension, diabetes mellitus, congestive perspective, this makes sense. Stable angina is caused by a heart failure, previous MI, and to use tobacco than patients fixed obstruction to flow; however, unstable angina, non ST with a normal prior stress test or no prior stress test. There segment elevation myocardial infarction (NSTEMI) and were significant differences between all 3 groups of patients STEMI are caused by plaque rupture and thrombus in history of hypercholesterolemia, known CAD, and formation in a lesion that may or may not have been angina. Because we used revascularization as part of our significant enough to result in angina (or a positive stress secondary composite outcome, it should be noted that this test) before the acute process. Stress testing assesses outcome might be affected by workup bias and the clinical whether a fixed obstruction to flow is present and cannot judgment of the treating physician. It is possible, then that predict subsequent plaque rupture resulting in ischemia. the difference between disposition and 30-day cardiovascu- Alternative methods of risk stratification may prove to be lar outcomes may be partially explained by the differences more successful in terms of both risk stratification and in baseline patient characteristics as well as workup bias. impact on future symptoms. deFillipi et al [20] have shown Although we were able to follow-up with 97% of that in low-risk patients, coronary angiography is more our patients, we are missing data on 30-day outcome sensitive and more cost-effective than exercise treadmill for 3% of our study population. Because this is a small testing. Patients with negative coronary angiography had fraction, however, it is not likely to greatly influence the fewer repeat ED visits, fewer hospitalizations, and had overall results. higher satisfaction rates and better understanding of their Lastly, our study population was derived from a single, disease than patients who had negative stress tests [20]. urban, university-affiliated center, where decision making Thus, using coronary angiography in place of stress testing may be different than in other locations. We cannot be sure low-risk patients may not only aid in patients’ understand- that our results generalize to other practitioners or settings. ing of their disease, but may also decrease ED visits and hospitalizations and may aid in physician disposition decision making. Alternatively, computed tomography 5. Conclusions coronary angiography correlates very well with cardiac In conclusion, physician knowledge of a prior stress test catheterization, potentially offering a noninvasive assess- significantly impacts disposition decision making only in ment of coronary anatomy with the same benefits on future the case of an abnormal stress test. Emergency physicians clinical decision making and patient understanding of their do not appear to make disposition decisions based on a prior disease as cardiac catheterization. Future studies should normal stress test. assess this possibility. This study has several potential limitations, the first of which is selection bias. We attempted to avoid selection bias by having trained research assistants present in the ED to References screen patients for study enrollment. However, these [1] McCaig LF, Burt CW. National hospital ambulatory medical care research assistants were only available 16 hours per day, survey: 2003 emergency department summary. Advance data from so it is possible that the patients who presented during the vital and health statistics. Hyattsville (Md)7 National Center for Health 8 hours the research assistants were not available were Statistics; 2005. 44 R.H. Nerenberg et al.

[2] Pope JH, Aufderheide TP, Ruthazer R, et al. Missed diagnoses of chest pain admitted to floor telemetry beds. Am J Emerg Med 2003; acute cardiac ischemia in the emergency department. N Engl J Med 21:282-7. 2000;342:1163-70. [12] Lai C, Noeller TP, Schmidt K, et al. Short-term risk after initial [3] Dagnone E, Collier C, Pickett W, et al. Chest pain with nondiagnostic observation for chest pain. J Emerg Med 2003;25:357-62. electrocardiogram in the emergency department: a randomized con- [13] Smith SW, Jackson EA, Bart BA, et al. Incidence of myocardial trolled trial of two cardiac marker regimens. CMAJ 2000;162:1561-6. infarction in emergency department chest pain patients with a recent [4] Selker HP, Beshansky JR, Griffith JL, et al. Use of the acute cardiac negative stress imaging test. Acad Emerg Med 2005;12:51 [abstract]. ischemia time-insensitive predictive instrument (ACI-TIPI) to assist [14] Hollander JE, Blomkalns AL, Brogran GX, et al. Standardized with triage of patients with chest pain or other symptoms suggestive of reporting guidelines for studies evaluating risk stratification of acute cardiac ischemia. A multicenter, controlled clinical trial. Ann emergency department patients with potential acute coronary syn- Intern Med 1998;129:845-55. dromes. Ann Emerg Med 2004;44:589-98. [5] Limkakeng Jr A, Gibler WB, Pollack C, et al. Combination of [15] Diercks DB, Gibler WB, Liu T, et al. Identification of patients at risk Goldman risk and initial cardiac troponin I for emergency depart- by graded exercise testing in an emergency department chest pain ment chest pain patient risk stratification. Acad Emerg Med 2001; center. Am J Cardiol 2000;86(3):289-92. 8:696-702. [16] Polanczyk CA, Johnson PA, Hartley LH, et al. Clinical correlates and [6] Lloyd-Jones DM, Camargo Jr CA, Lapuerta P, et al. Electrocardio- prognostic significance of early negative exercise tolerance test in graphic and clinical predictors of acute myocardial infarction in patients with acute chest pain seen in the hospital emergency patients with unstable angina pectoris. Am J Cardiol 1998;81:1182-6. department. Am J Cardiol 1998;81:288-92. [7] Tatum JL, Jesse RL, Kontos MC, et al. Comprehensive strategy for [17] Amsterdam EA, Kirk JD, Diercks DB, Lewis WR, Turnipseed S. the evaluation and triage of the chest pain patient. Ann Emerg Med Immediate exercise testing to evaluate low risk patients presenting to 1997;29:116-25. the emergency department with chest pain. J Am Coll Cardiol 2002; [8] Reilly BM, Evans AT, Schaider JJ, et al. Impact of a clinical decision 40:251-6. rule on hospital triage of patients with suspected acute cardiac [18] Mikhail MG, Smith FA, Gray M, et al. Cost-effectiveness of ischemia in the emergency department. JAMA 2002;288:342-50. mandatory stress testing in chest pain center patients. Ann Emerg [9] Pollack Jr CV, Sites FD, Shofer FS, et al. Application of the TIMI risk Med 1997;29:88-98. score for unstable angina and non-ST elevation acute coronary [19] Shaver KJ, Marsan RJ, Sease KL, et al. Impact of a negative syndrome to an unselected emergency department chest pain evaluation for underlying coronary artery disease on one-year resource population. Acad Emerg Med 2006;13:13-8. utilization for patients admitted with potential acute coronary [10] Antman EM, Cohen M, Bernink PJ, et al. The TIMI risk score for syndromes. Acad Emerg Med 2004;11:1272-7. unstable angina/non-ST elevation MI: a method for prognostication [20] deFilippi CR, Rosanio S, Tocchi M, et al. Randomization of a strategy and therapeutic decision making. JAMA 2000;284:835-42. of predischarge coronary angiography versus exercise testing in low- [11] Chan GW, Sites FD, Shofer FS, Hollander JE. Impact of stress risk patients in a chest pain unit: in-hospital and long-term outcomes. testing on 30-day cardiovascular outcomes for low-risk patients with J Am Coll Cardiol 2001;37:2042-9. American Journal of Emergency Medicine (2007) 25,45–48

www.elsevier.com/locate/ajem

Brief Reports Discrepancies in interpretation of ED body computed tomographic scans by radiology residents

Nelson Tieng MD, Diana Grinberg MD, Siu Fai Li MD*

Department of Emergency Medicine, Jacobi Medical Center, Bronx, NY 10461, USA

Received 5 March 2006; revised 18 April 2006; accepted 19 April 2006

Abstract Objective: In academic institutions, radiology residents are often relied on for providing preliminary reports of imaging studies done in the ED. We examined the prevalence of discrepant interpretations of body computed tomographic (CT) scans in our institution. Methods: We conducted a retrospective study on a consecutive series of body CT scans at an urban ED. We compared the preliminary interpretation by radiology residents with the final interpretation by radiology attending physicians. An interpretation was characterized as having no discrepancy, minor discrepancy, or major discrepancy. A major discrepancy was defined as a discrepancy that resulted in a change in diagnosis, treatment, or disposition. Results: Two hundred three body CT scans were identified during the study period. Of these CT scans, 20 had major discrepancies (10%), 40 had minor discrepancies (20%), and 143 had no discrepancy (70%). Major discrepancies included missed appendicitis, normal appendix, missed bowel obstruction, and missed colon cancer. Computed tomographic scans with abnormal findings were more likely to contain major discrepancies (relative risk = 6.0; 95% confidence interval = 1.8-20). Conclusion: Discrepancies between radiology residents and radiology attending physicians were common at our institution. Emergency department physicians should exercise caution when relying on residents’ interpretation of body CT scans. D 2007 Elsevier Inc. All rights reserved.

1. Introduction concern because emergency medicine physicians are re- liant on radiologists for interpretations, including those Some medical centers offer round-the-clock in-house of computed tomographic (CT) scans done on critically attending coverage in the radiology department, whereas ill patients. others rely on radiology residents during off-hours to Studies have reported discrepancy rates in the range of provide preliminary readings. In most teaching hospitals, 2% to 13% [1-9] in residents’ interpretations of CT scans. including our institution, the latter setup is the case. The The discrepancy rates vary according to the site (eg, head, residents are student physicians and therefore less experi- appendix) and level of experience of the resident physician. enced and perhaps more subject to error. This is a particular Few studies have examined the discrepancy rate of interpretations of body CT scans [5,7]. Given the conflicting * Corresponding author. Tel.: +1 718 918 5820; fax: +1 718 918 7549. outcomes of previous studies, particularly when it came to E-mail address: [email protected] (S.F. Li). interpretation of the body CT scans, we wished to determine

0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ajem.2006.04.008 46 N. Tieng et al. the prevalence and quality (major vs minor) of discrepant by the reviewer. A minor discrepancy was defined a interpretations of body CT scans in our own institution. discrepancy that would not have an impact in the ED course. Each report was independently reviewed by 2 emergency medicine physicians. In cases of disagreement, a third 2. Materials and methods physician adjudicated the interpretation. The primary outcome was the proportion of major We conducted a retrospective study on a consecutive discrepancies between radiology residents’ and attending series of body CT scans ordered by the emergency physicians’ interpretations of body CT scans. Secondary department (ED) at an inner-city level I trauma center from outcomes included the proportion of minor discrepancies, February to March 2005. The series of CT scans included the proportion of major discrepancies stratified by CT trauma and nontrauma cases. Computed tomographic scans findings (ie, CT scans with positive findings vs CT scans that were initially interpreted by a radiology attending were with negative findings), and the types of discrepancies excluded from the study. The interpretations of radiology encountered. We estimated that 196 CT scans were needed in residents and attending physicians were obtained from the the sample based on an estimated discrepancy rate of 15%, a hospital electronic database. At our institution, off-hour confidence interval (CI) width of F0.05, and a 2-tailed a of studies are primarily read by PGY-3 or PGY-4 radiology .05. We estimated that 40 days were needed to collect 196 residents. Computed tomographic scans were obtained using CT scans using an estimate of 5 CT scans per day. Interrater a General Electric LightSpeed multislice scanner. agreement was determined using simple agreement. The The preliminary interpretation by residents was compared study was approved by our Institutional Review Board. with the final interpretation by attending physicians, as stated in dictated report, by the physician chart reviewers. The interpretations were characterized as having no discrepancy, 3. Results minor discrepancies, or major discrepancies. A major discrepancy was defined as a discrepancy that resulted in a Two hundred three body CT scans were identified during change in diagnosis, treatment, or disposition, as determined the study period. There were 180 abdominal and pelvic CT

Table 1 Major discrepancies in body CT interpretations Finding Patient outcome Type of scan Indication Lung mass Disposition unchanged; referred for outpatient workup Chest/abd/pelvic Trauma Appendicitis Patient recalled, rescanned; follow-up scan negative Abd/pelvic Abd pain Lung metastases Disposition unchanged; additional workup as inpatient Chest PE Partial/early SBO Patient recalled, did not return Abd/pelvic Abd pain Cecal mass Patient contacted, referred for outpatient follow-up Abd/pelvic Abd pain Colonic thickening Disposition unchanged; additional workup as inpatient Abd/pelvic Abd pain Sigmoid irregularity Patient recalled and admitted Abd/pelvic Abd pain Adnexal masses (?TOA) Patient recalled, referred for outpatient Abd/pelvic Abd pain follow-up after further workup Appendicitis Patient recalled, rescanned; follow-up Abd/pelvic Renal colic CT showed improvement of appendiceal changes/periappendiceal inflammation; patient discharged Obstructive kidney stone ED contacted; working diagnosis was Abd/pelvic Abd pain appendicitis; management changed Left adnexal mass Patient contacted, referred for outpatient follow-up Abd/pelvic Abd pain Pancreatic hypodensity Patient contacted, referred for outpatient follow-up Abd/pelvic Aorta dissection Appendicitis Patient recalled and admitted; Abd/pelvic Abd pain eventual diagnosis was colon cancer No PID No change in patient management Abd/pelvic Abd pain Diverticulitis Disposition unchanged; inpatient management altered Abd/pelvic Abd pain No renal calculi No change in patient management Abd/pelvic Renal colic Normal appendix Patient discharged Abd/pelvic Abd pain Diaphragmatic hernia Disposition unchanged; inpatient Abd/pelvic Abd pain containing large and management unchanged small bowel (patient considered to be at poor surgical risk) Sigmoid thickening Disposition unchanged; inpatient workup/management altered Abd/pelvic Abd pain Renal calculus No change in patient management Abd/pelvic Renal colic abd indicates abdominal; PID, pelvic inflammatory disease; PE, pulmonary embolism; SBO, small bowel obstruction; TOA, tubo-ovarian abcess. Discrepancies in interpretation of ED body CT scans 47 scans, 17 chest CT scans, and 6 CT scans of the chest, With the exception of one case in which the outcome abdomen, and pelvis. There were 57 scans without contrast, was not known (the patient was recalled but did not return 7 with oral contrast only, 29 with intravenous contrast only, to the hospital for further evaluation), patient outcomes did and 110 with both oral and intravenous contrast. Most CT not appear to be adversely affected despite the major scans were done to evaluate abdominal pain (n = 108), discrepancies in body CT scan interpretations. For instance, followed by evaluations for nephrolithiasis (n = 52), trauma there were 3 patients in whom the diagnosis of appendicitis (n = 16), pulmonary embolism (n = 15), and others (n = 12; was appended by the radiology attending. All were initially eg, aortic dissection). discharged from the ED and recalled. Subsequent follow-up Of the CT scans, 20 had major discrepancies (10%; 95% and repeat CT scans were negative, and the patients did not CI = 0.06-0.15), 40 had minor discrepancies (20%; 95% need to undergo surgery for acute appendicitis (or, worse CI = 0.14-0.26), and 143 had no discrepancy (70%; 95% CI = yet, perforated appendicitis). In some cases, such as the 0.64-0.77). Major discrepancies included missed diverticu- patient with diverticulitis, the patients were admitted to the litis, missed appendicitis, normal appendix, missed bowel hospital and the amended CT report was communicated to obstruction, missed colon cancer, missed lung metastases, the inpatient physicians; in other cases, such as the patients and missed adnexal masses (Table 1). Interrater agreement with possible pelvic inflammatory disease or renal colic, of major discrepancies was 0.91 (95% CI = 0.86-0.95), the treatment and disposition of the patients were not whereas overall interrater agreement was 0.75 (95% CI = affected at all. 0.69-0.81). There were several limitations to our study. Perhaps There were abnormal findings on 99 CT scans (49%). most important is the inherent selection bias of the study There were 17 major discrepancies among the CT scans (ie, the findings may be institution-dependent and not with abnormal findings (17%) and 3 major discrepancies generalizable). Second, there is clustering bias (ie, our among the CT scans with normal findings (3%). Computed study followed a consecutive series of patients during a tomographic scans with abnormal findings were more likely constrained period during which preliminary interpretations to contain major discrepancies (relative risk [RR] = 6.0, were made by a limited number of radiology residents in 95% CI = 1.8-20). Most major discrepancies (18/20, 90%) their rotation). Third, our primary outcome was discrepant occurred in the abdomen/pelvis portion of the CT scans, interpretations between radiology residents and radiology whereas only 2 (10%) occurred in the chest portion of the attending physicians, not accuracy. Thus, in cases in which CT scans. However, abdominal/pelvic CT scans were not a discrepancy arises, the resident, not the attending, may be more likely to result in major discrepancies as a whole (RR = correct. This was true in one of the cases in which there 1.1; 95% CI = 0.28-4.5). Similarly, most major discrep- was a discrepancy in the interpretation of a body CT scan ancies occurred in CT scans with oral and intravenous for appendicitis. In addition, we do not know if the contrast (13/20, 65%), but such scans were not statistically radiologists received input from other physicians (eg, more likely to result in major discrepancies (RR = 1.6, 95% emergency medicine physicians or surgeons) before or CI = 0.65-3.8). during their interpretations of the CT scans. Finally, we did not examine if any adverse outcome occurred as a result of discrepant readings. 4. Discussion Discrepancies between residents’ and attending physicians’ interpretations of body CT scans were relatively At our institution, discrepant interpretations of body common in our institution (10%) and have the potential to CT scans between radiology residents and radiology result in major changes in diagnosis, treatment, or disposi- attending physicians are relatively common. Overall, tion. Emergency medicine physicians should exercise there were discrepancies in 30% of body CT scans and caution when relying on residents’ interpretations of body 10% of body scans contained major discrepancies that CT scans. may have a significant impact in the ED management of the patient, such as missed appendicitis and missed bowel obstruction. Several studies have assessed the practice of residents’ References interpretations of CT scans, with mixed results. The rate of discrepancies appears to be lowest with neuroimaging (head [1] Wysoki MG, Nassar CJ, Koenigsberg RA. Head trauma: CT scan CT), with discrepancy rates of 2% or less and even lower interpretation by radiology residents versus staff radiologists. Radiol- rates of adverse clinical outcomes [1-4]. The discrepancy ogy 1998;208:125-8. rates were more variable for body CT scans: 7% in a study on [2] Roszler MH, McCarroll KA, Rashid T. Resident interpretation of appendicitis [6], 11% in a series of trauma patients [9], and emergency computed tomography scans. Invest Radiol 1991;26(4): 374-6. between 1% and 13% in other series of body CT scans [5,7]. [3] Lal NR, Murray UM, Eldevik OP. Clinical consequences of misinter- Compared with other studies, the prevalence of discrepant CT pretations of neuroradiologic CT scans by on-call radiology residents. interpretations was slightly higher in our institution. AJNR Am J Neuroradiol 2000;21(1):124-9. 48 N. Tieng et al.

[4] Erly WK, Berger WG, Krupinski E. Radiology resident evaluation of [7] Wechsler RJ, Spettell CM, Kurtz AB. Effects of training and experience head CT scan orders in the emergency department. AJNR Am J in interpretation of emergency body CT scans. Radiology 1996;199: Neuroradiol 2002;23(1):103-7. 717-20. [5] Carney E, Kempf J, DeCarvalho V. Preliminary interpretations of after- [8] van Rossum AB, van Erkel AR, van Persijn van Meerten EL. hours CT and sonography by radiology residents versus final Accuracy of helical CT for acute pulmonary embolism: ROC analysis interpretations by body imaging radiologists at a level I trauma center. of observer performance related to clinical experience. Eur Radiol AJR Am J Roentgenol 2003;181(2):367-73. 1998;8(7):1160-4. [6] Albano MC, Ross GW, Ditchek JJ. Resident interpretation of [9] Velmahos GC, Fill C, Vassiliu P. Around-the-clock attending radiology emergency CT scans in the evaluation of acute appendicitis. Acad coverage is essential to avoid mistakes in the care of trauma patients. Radiol 2001;8:915-8. Am Surg 2001;67(12):1175-7. American Journal of Emergency Medicine (2007) 25,49–52

www.elsevier.com/locate/ajem

Brief Reports Factors associated with myocardial infarction after emergency endoscopy for upper gastrointestinal bleeding in high-risk patients: a prospective observational study

Ching-Tai Lee MDa, Shih-Pei Huang MD, PhDb, Tsu-Yao Cheng MDc, Tsung-Hsien Chiang MDa, Chi-Ming Tai MDa, Wei-Chih Su MDe, Chien-Hua Huang MDd, Jaw-Town Lin MD, PhDb, Hsiu-Po Wang MDd,* aDepartment of Internal Medicine, E-Da Hospital/I-Shou University, Kaohsiung, Taiwan bDepartment of Internal Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan cDepartment of Laboratory Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan dDepartment of Emergency Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan eDepartment of Gastroenterology, Buddist Tzu Chi General Hospital Taipei Branch, Taipei, Taiwan

Received 29 March 2006; revised 30 April 2006; accepted 30 April 2006

Abstract Background: Because myocardial infarction (MI) after emergency endoscopy for upper gastrointestinal bleeding carries high mortality, we investigated factors associated with procedure-related MI in high- risk patients. Methods: Consecutive patients with coronary artery disease or age-based risk for coronary artery disease (men, age N45 years; women, N55 years) who underwent emergency endoscopy were enrolled at a single ED. Demographic, laboratory, and outcome data were recorded. Patients fit 1 of 3 groups: MI before endoscopy (pre-panendoscopy [PES] MI), MI after endoscopy (post-PES MI), or non-MI. Results: We enrolled 108 high-risk patients, including 5 (4.6%) with MI diagnosed preendoscopy. Five patients (4.6%) had MIs postendoscopy. Compared with non-MI patients, significantly more post-PES MI patients had heart disease (60.0% vs 12.2%; P = .021), lower systolic pressure on arrival (86.2 F 16.6 vs 128.0 F 27.2 mm Hg; P = .002), lower diastolic pressure on arrival (50.0 F 6.3 vs 69.5 F 15.8 mm Hg; P = .003), lower hemoglobin on arrival (6.7 F 1.1 vs 9.1 F 2.4 g/dL; P = .021), and more persistent shock status preendoscopy (80.0% vs 13.3%; P = .002). There was no significant difference in factors including duration of procedure and rates of recurrent bleeding, postprocedure complication, and mortality. Conclusions: Heart disease, lower blood pressure or hemoglobin level on arrival, and persistent shock before endoscopy are associated with increased risk for procedure-related MI. D 2007 Elsevier Inc. All rights reserved.

* Corresponding address. Hsiu-Po Wang, MD, Department of Emergency Medicine, National Taiwan University Hospital, 7 Chung-Shan South Road, Taipei, Taiwan. Tel.: +886 2 23123456x5695; fax: +886 2 23947899. E-mail address: [email protected] (H.-P. Wang).

1. Introduction within 24 hours, intolerance to procedures, pregnancy, and recent MI (within 6 weeks of ED admission). Upper gastrointestinal (UGI) bleeding is an emergency frequently encountered at EDs, and coronary artery disease 2.3. Data collection (CAD) is common in older patients admitted with UGI bleeding. Several studies have shown that myocardial Clinical data that were recorded included age, sex, infarction (MI) with significant upper or lower gastrointes- duration of UGI symptoms, history of CAD, and comorbid tinal (GI) bleeding occurs in 30% to 49% of patients conditions. Chronic obstructive lung disease, diabetes admitted to intensive care units (ICUs), with an overall mellitus, old cerebral vascular accident, chronic renal mortality rate of 5% to 10% [1,2]. However, predisposing insufficiency, liver cirrhosis, malignancy, hypertension, factors for acute MI after emergency endoscopy for upper hyperlipidemia, and heart disease, including valvular heart or lower GI hemorrhage in patients in the ED setting disease, congestive heart failure, and arrhythmia, were remain uncertain. recorded as comorbid conditions. Earlier studies involving the subset of patients sufficient- Blood pressure, heart rate, oxygen saturation, hemoglo- ly ill to be admitted to ICUs demonstrated that factors bin level, cardiac enzyme levels, and electrocardiogram associated with MI in the setting of GI hemorrhage include were checked on arrival and followed 4 hours after the advanced age, cardiac comorbidities, anemia, hypotension, endoscopic procedure. During the procedure, maximal heart and elevated Acute Physiology and Chronic Health Evalu- rate, lowest oxygen saturation, duration of the procedure, ation II score at the time of ICU admission [1-4]. However, and method of hemostasis were recorded. Persistence of none of the previous studies evaluated the influence of shock, usage of inotropic agents, and amount of blood endoscopy and therapeutic agents on subsequent cardiac transfusion (if any) before endoscopy were also recorded. morbidity and mortality. We designed the current, prospec- Finally, data were collected regarding procedure-related tive observational study to investigate clinical and endo- complications, recurrence of bleeding, total duration of scopic factors associated with MI after endoscopy done in hospitalization, and mortality. the ED for UGI bleeding. 2.4. Data analysis Patients were categorized into 1 of 3 groups: MI before 2. Patients and methods endoscopy (pre-PES MI), MI after endoscopy (post-PES MI), and non-MI. Factors including demographic data, 2.1. Inclusion criteria and general endoscopic course of treatment, and outcomes were compared between technique post-PES MI and non-MI groups. Categorical data were We defined high-risk patients as individuals with at least analyzed using Fisher exact test, and continuous data were one of the following: men older than 45 years, women older compared using Mann-Whitney U test. A P value less than than 55 years, or patients with documented CAD. High-risk .05 was considered statistically significant. All analyses patients who visited our ED between May 2004 and March were performed by SPSS 11.0 version for Windows. 2005 with a diagnosis of UGI bleeding were enrolled. Upper gastrointestinal bleeding was defined with symptoms of tarry stool, melena, coffee-ground vomitus, hematemesis, or 3. Results anemia suspected to be related to UGI bleeding. Myocardial infarction was defined by the presence of at least 2 of the 3.1. Demographic data following as verified by a cardiologist: elevation of cardiac During the study period, 108 high-risk patients meeting enzyme (eg, troponin I) or diagnosis-appropriate chest pain the study criteria were enrolled. Of them, 5 patients (4.6%) or electrocardiographic change. were diagnosed with acute MI before endoscopy. An All patients underwent endoscopic examination within additional 5 patients (4.6%) were diagnosed with acute MI 24 hours after admission to the ED. Endoscopic intervention after the procedure. In the pre-PES MI group, 1 of the was performed by endoscopists who had a similar training 5 patients had an ST-elevated MI. All 5 patients in the post- background. Oxygen via nasal cannula was routinely PES MI group had non–ST-elevated MIs. None of the supplied during the endoscopic procedure. No conscious- 10 patients had typical symptoms such as chest pain, cold ness sedation was applied. sweating, or dyspnea. 2.2. Exclusion criteria When post-PES patients were compared with non-MI patients, mean age was similar (68.0 F 10.4 vs 70.2 F Exclusion criteria included decompensated liver cirrhosis 9.0 years, post-PES and non-MI, respectively; P = .884). with obvious bleeding tendency (prothrombin time [PT] Sex and history of CAD were also similar. The post-PES MI prolong more than 3 seconds), underlying advanced group included significantly more patients with underlying malignancy, failure to perform endoscopic examination heart diseases (60.0% vs 12.2%; P = .021); there was a trend MI after emergency endoscopy for upper gastrointestinal bleeding 51

Table 1 Demographic data of patients with and without MI after endoscopy Total (N = 108) Pre-PES MI (n = 5) Post-PES MI P Present (n = 5) No (n = 98) Age, mean F SD, y 70.3 F 9.1 73.6 F 4.0 68.0 F 10.4 70.2 F 9.0 .884 Sex Male/Female 73/35 4/1 4/1 65/33 1.000 CAD Present, n (%) 32 (29.6) 3 (60) 2 (40) 27 (27.6) .619 Comorbidity, mean F SD 1.9 F 1.3 2.8 F 1.2 2.8 F 1.2 1.9 F 1.3 .108 COPD 7 0 1 6 .302 DM 37 3 2 32 1.000 CVA 21 2 1 18 1.000 CRF 20 1 2 17 .229 Cirrhosis 12 1 0 11 1.000 Cancer 16 1 0 15 1.000 HTN 62 5 3 54 1.000 Hyperlipid 19 0 2 17 .229 Heart disease 16 1 3 12 .021 CAD indicates coronary artery disease; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; CRF, chronic renal failure; HTN, hypertension. toward more comorbidities, although the difference com- group also had significantly more patients who had pared with the non-MI group was not significant (2.8 F persistent shock status before their endoscopic procedure 1.2 vs 1.9 F 1.3; P = .108). Demographic data are (80.0% vs 13.3%; P = .002). No statistically significant summarized in Table 1. difference existed in heart rate, oxygen saturation, follow-up blood pressure, or follow-up hemoglobin level (Table 2). 3.2. Clinical course Procedure duration was longer for the post-PES MI The post-PES MI group had a shorter duration of group, but the difference was not statistically significant F F symptoms than the non-MI group, although the difference (625.2 412.6 vs 575.7 366.0 seconds; P = .747). did not reach significance (1.6 F 1.1 vs 3.3 F 1.1 days; A trend toward larger volume of blood transfusion was F F P = .606). Post-PES MI patients had significantly lower found for the post-PES MI group (2.2 1.6 vs 1.3 systolic blood pressure on arrival (86.2 F 16.6 vs 128.0 F 1.6 unit; P = .163) (Table 2). 27.2 mm Hg; P = .002), lower diastolic blood pressure on 3.3. Outcome arrival (50.0 F 6.3 vs 69.5 F 15.8 mm Hg; P = .003), and lower hemoglobin level on arrival (6.7 F 1.1 vs 9.1 F Hospitalization stay (in days) was similar between post- 2.4 g/dL; P = .021) than non-MI patients. The post-PES MI PES MI and non-MI patients (6.0 F 4.1 vs 6.2 F 9.8 days;

Table 2 Comparisons of data for study patients with and without MI complicating UGI bleeding Total (N = 108) Pre-PES MI (n = 5) Post-PES MI P Present (n = 5) No (n = 98) Symptom (d) 3.2 F 6.9 2.0 F 1.1 1.6 F 1.1 3.3 F 1.1 .606 SBP (mm Hg) at ED 126.6 F 28.3 139.2 F 24.9 86.2 F 16.6 128.0 F 27.2 .002 DBP (mm Hg) at ED 68.8 F 16.0 74.8 F 14.8 50.0 F 6.3 69.5 F 15.8 .003 HR (per min) at ED 93.6 F 18.7 90.2 F 8.1 80.0 F 24.5 94.5 F 18.4 .145 Spo2 (%) at ED 98.3 F 9.7 98.8 F 1.6 96.3 F 38.9 98.4 F 1.8 .758 Hb (g/dL) at ED 8.9 F 2.4 8.8 F 2.2 6.7 F 1.1 9.1 F 2.4 .021 Shock presented, n (%) 17 (15.7) 0 (0) 4 (80) 13 (13.3) .002 PES duration (s) 578.1 F 368.6 a 625.2 F 412.6 575.7 F 366.0 .747 BT (unit) 1.3 F 1.6 1.4 F 1.2 2.2 F 1.6 1.3 F 1.6 .163 Hospital (d) 6.2 F 9.4 6.8 F 3.2 6.0 F 4.1 6.2 F 9.8 .536 Death, n (%) 4 (3.7) 0 (0) 1 (20) 3 (3.1) .183

SBP indicates systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; Spo2, oxygen saturation; Hb, hemoglobin; BT, blood transfusion. Values are mean F SD unless otherwise indicated. a Endoscopy was not performed. 52 C.-T. Lee et al.

P = .536). There were 4 patients (3.7%) who had bleeding- with procedure-related MI, whereas that of underlying CAD related mortality, 1 from the post-PES MI group and 3 from alone was not. Comparisons of clinical and laboratory data the non-MI group. The difference was not significant in our study suggest that lower blood pressure on arrival, (20.0% vs 3.1%; P = .183) (Table 2). lower hemoglobin level on arrival, and persistent shock status before endoscopy are significant factors associated with MI after emergency endoscopy. There is also a trend 4. Discussion indicating that the need for a greater volume of blood transfusion may increase risk for MI. These findings suggest Myocardial infarction is a known complication of GI that low myocardial perfusion and oxygen delivery before bleeding. Previous studies have shown that patients with GI endoscopy may play important roles in the pathogenesis of hemorrhage admitted to ICUs, who developed an MI had MI rather than coronary stenosis alone. an overall mortality rate of 5% to 10% [1,2]. In the study In general, GI endoscopy as well as a decreased blood by Cappell [5], 0.94% of patients from the general volume may produce significant cardiovascular stress. There population had a simultaneous UGI bleed and acute MI. may be an association between longer durations of However, factors associated with MI after emergency endoscopic examination and acute MI. A study of larger therapeutic endoscopy have not been studied [1-5]. With scale may be warranted to clarify the issue. our series of 108 high-risk patients in the ED, 10 (9.2%) patients had an MI before or after their emergency endoscopy. This difference in incidence of MI between 5. Conclusion reports may result from differences between the studied populations. Our study population included high-risk Myocardial infarction after emergency endoscopy is not patients who presented with UGI bleeding in an ED, uncommon in high-risk patients with UGI bleeding. Our whereas previous studies focused on critical patients who study disclosed that underlying heart disease, lower blood were admitted to ICUs. Our findings indicate that pressure on arrival in the ED, lower hemoglobin level on simultaneous MI and UGI bleeding should be considered arrival, and persistent shock status before endoscopic and ruled out during clinical practice in the ED. examination are all associated with higher risk of MI after Certainly, GI bleeding deleteriously affects myocardial emergency endoscopy. These findings indicate that the function. Massive acute blood loss causes hypovolemia, establishment of stable hemodynamics and oxygenation hypoperfusion, and decreased oxygenation of the myocar- before emergency endoscopy might reduce the risk of dium, which may cause compensatory tachycardia to procedure-related MI, especially in patients with known maintain adequate blood pressure and tissue perfusion. This heart disease. may increases myocardial workload and oxygen demand to the point that myocardial injury results [4,6]. In our study, MI, as a complication of UGI bleeding, was overwhelm- References ingly of non–ST segment elevation type (9/10, 90%). This pattern suggests that plaque rupture or thrombosis may have [1] Emenike E, Srivastava S, Amoateng-Adjepong Y, et al. Myocadial a lesser role in pathogenesis in this setting. Excessive infarction complicating gastrointestinal hemorrhage. Mayo Clin Proc myocardial oxygen demand is a more likely explanation. 1999;74:235-41. [2] Bhatti N, Amoateng-Adjepong Y, Qamar A, et al. Myocardial Interestingly, none of the 10 patients had typical symptoms infarction in critically ill patients presenting with gastrointestinal of an acute MI, such as chest pain or cold sweats. This may hemorrhage: retrospective analysis of risks and outcomes. Chest be because the overt symptoms of GI bleeding mask the 1998;114:1137-42. typical symptoms of evolving MI. Based on this clinical [3] Inyat N, Amoateng-Adjepong Y, Upadya A, et al. Risks for develop- finding, we suggest physicians send cardiac enzymes in ing critical illness with gastrointestinal hemorrhage. Chest 2000;118: 473-8. high-risk patients with GI bleeding even when the patient [4] Cappell MS. Gastrointestinal bleeding associated with myocardial has no chest pain and a normal electrocardiogram to avoid a infarction. Gastroenterol Clin North Am 2000;29:423-44. delay in diagnosis of myocardial ischemia or acute MI. [5] Cappell MS. A study of the syndrome of simultaneous acute upper Our study also suggests that risk for MI after emergency gastrointestinal bleeding and myocardial infarction in 36 patients. Am J endoscopic procedures may rise with number of comorbid Gastroenterol 1995;90:1444-9. [6] Kondo Y, Nomura M, Nada T, et al. Ischemic electrocardiographic conditions. In our series, the presence of underlying heart changes after massive blood transfusion: finding based on myocardial disease such as valvular abnormalities, history of arrhyth- scintigraphy using 99mTc-MIBI and 123I-MIBG. Acta Cardiol 1998; mia, or presence of heart failure was significantly associated 53:279-83. American Journal of Emergency Medicine (2007) 25,53–56

www.elsevier.com/locate/ajem

Brief Reports Endotracheal intracuff pressures in the ED and prehospital setting: is there a problem?

James E. Svenson MD, MSa,*, M. Bruce Lindsay MD, MAa, Jill E. O’Connor RNb aSection of Emergency Medicine, University of Wisconsin, Madison, WI 53792, USA bMed Flight, University of Wisconsin, Madison, WI 53792, USA

Received 12 March 2006; revised 31 August 2006; accepted 5 September 2006

Abstract Introduction: Cuffed endotracheal tubes are used to prevent gas leak and also pulmonary aspiration in ventilated patients. The pressure exerted on the tracheal wall is similar to intracuff pressure. The perfusion pressure for the tracheal mucosa is 40 cm H2O. Cuff pressures greater than 40 cm H2O may cause various ischemic changes and complications. High cuff pressures have also been implicated in postoperative sore throat and nonischemic complications. Postintubation endotracheal tube cuff pressures are not routinely measured in the ED or prehospital setting. The time spent in these settings may be long enough for pressure-induced tracheal mucosal injury to occur. The purpose of this study is to assess cuff pressures in intubated patients before aeromedical transport. Methods: All intubated patients transported by an aeromedical transport program during a 3-month period were included in this study. Patients were intubated either by helicopter physicians or before helicopter arrival at the referring hospital or by ambulance personnel. Cuff pressure was measured using a manometer (Cuffpressure, Posey Co, USA). This measurement was recorded, and correction was performed, if necessary, to achieve a cuff pressure of 14 to 27 cm H2O while preventing an air leak. Data were analyzed for the distribution of intracuff pressures and incidence of elevated pressure on first measurement and the need for correction. Results: There were 62 patients in this study. The mean first recorded pressure was 63 F 34 cm H2O. Initial cuff pressures were greater than 40 cm H2O in 36 (58%) patients and required correction. Conclusions: In this study, most cuff pressures exceeded safe pressure and required correction. Measurement of intracuff pressure is a simple and inexpensive procedure and should be done whenever a patient is intubated, in either the prehospital or hospital setting, because this may reduce long- term morbidity. D 2007 Elsevier Inc. All rights reserved.

1. Introduction

High compliance endotracheal tube cuffs are used to Presented at the 11th International Conference on Emergency Medicine, prevent gas leak and also pulmonary aspiration in intubated June 2006, Halifax, Nova Scotia. patients. The pressure exerted on the tracheal wall by the * Corresponding author. Tel.: +1 608 265 5808; fax: +1 608 262 2641. balloon cuff is similar to the intracuff pressure [1,2]. E-mail address: [email protected] (J.E. Svenson). Ischemic changes in the tracheal mucosa may occur if the

0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ajem.2006.09.001 54 J.E. Svenson et al. pressure exerted by the cuff of the tube exceeds the capillary perfusion pressure, typically about 30 cm H2O [3,4].High cuff pressure has been implicated in postoperative sore throat and nonischemic complications [5]. The incidence of possible tracheal ischemia can be minimized by keeping the intracuff pressure lower than the perfusion pressure [6].To allow effective ventilation, the cuff should be inflated until it just prevents an air leak [7]. Endotracheal tube cuff pressures are not routinely measured in the prehospital setting or the ED [8]. It has been demonstrated that palpation of the pilot balloon is insufficient to detect high cuff pressures [9]. Many patients Fig. 1 Distribution of endotracheal tube cuff pressures. intubated in the ED or prehospital setting do not have range on first measurement. The number of patients intracuff pressure measured and adjusted. Mucosal damage intubated by program physicians during the study period has been shown to occur after 15 minutes in an animal was small and so data were analyzed as one group. Because model, longer than the time spent in these settings [6]. the data seemed to indicate convincingly the need for The purpose of this study is to assess endotracheal tube pressure measurement and monitoring after a short period, cuff pressures in patients intubated before transport by a formal data collection was stopped, and mandatory mea- critical care aeromedical service. surement and adjustment of intracuff pressures were adopted as a formal policy. 2. Methods 3. Results This was a retrospective study. As part of a quality improvement program, intracuff pressures were measured During the study period, there were 345 patients trans- and adjusted on intubated patients transported by a critical ported. Of these patients, 69 (20%) were younger than care aeromedical program. This program is a regional air 18 years and so were not included. Of the remaining ambulance service. All patients are transported by helicop- 276 patients, 54 (19%) were transported from a scene, ter. The helicopter provides critical care transport for whereas 222 (81%) were transported from another hospital. interfacility transfers (within approximately 250 miles) and There were 62 (22%) patients who were intubated. All scene flights (within approximately 75 miles). The program patients intubated during the study period were included. Of transports approximately 1300 patients per year. The the scene patients, 6 (11%) were intubated, 4 by ambulance medical crew consists of an attending flight physician and personnel and 2 by flight physicians. Of the 222 patients a flight nurse. transported from another facility, 56 (25%) were intubated. All transported patients older than 18 years who Fifty (89%) patients were intubated before helicopter arrival required tracheal intubation were eligible for this study. (either in the ED or by ambulance personnel before arrival The 3-month study period was from April through June in the ED). The primary intubator in these cases was not 2005. Intubations were performed either by the program recorded. The mean first recorded intracuff pressure was physician or by ambulance personnel or at the referring 63 F 34 cm H2O (range, 4-120) (Fig. 1). There were 4 (6%) hospital before helicopter arrival. Inflation of endotracheal patients who had pressures less than 20 cm H2O, 9 (15%) tubes by flight physicians has routinely been done by patients had pressures between 20 and 30 cm H2O, and injecting enough air to prevent an air leak or by estimating 49 (79%) had pressures greater than 30 cm H2O. The pressure using the pilot balloon. How other intubators presence or absence of an air leak at lower pressures was not (paramedics, ED physicians) in this study determine the recorded. The pressure was greater than 40 cm H2Oin amount of air injected is unknown. Cuff pressures were not 36 (58%) patients. Forty-five (72%) patients had pressure measured at outside hospitals before transport. All flight corrections of greater than 10 cm. physicians were familiar with the use of the manometer. Regardless of the place of intubation, after loading into the helicopter, the cuff pressure was measured just before 4. Limitations takeoff using a manometer (Posey Co, USA). One measurement was made and recorded, and correction was This study relied upon measurement of tube pressure by performed if necessary to achieve a cuff pressure 20 to numerous physicians that staff the program. All flight 30 cm H2O without an air leak [6,15]. The Institutional physicians were familiar with the intracuff manometer and Review Board approved this study. its use. However, there was no attempt to validate Data were analyzed for the distribution of intracuff independent measurements. The individual and interobserv- pressures and incidence of pressures outside the optimal er reliability were not measured. Endotracheal intracuff pressures in the ED and prehospital setting 55

For those intubated before helicopter arrival, there was microaspirations of secretions [14]. It may be difficult to no attempt to identify the skill level of the provider who ensure a tracheal seal especially in the noisy prehospital intubated the patient. A number of patients transported from environment. Although previous studies have shown differ- other EDs were intubated before arrival in the ED by ences in fluid leaks for various tube types at pressures less ambulance personnel, and these were not recorded. Thus, than 20 cm H2O, there is little difference or leak at higher the actual number of patients intubated in the prehospital pressures [15]. Thus, attention to pressures may also scene setting and those in an outside ED are not known. minimize problems of tubes that have underinflated cuffs. Given the small number of patients that had optimal A small number of patients in our study had underinflated pressures recorded, it appears that the skill level of the cuffs requiring adjustment. intubator does not matter. These data suggest that endotracheal cuff pressures are In retrospect, it would have been valuable to collect the elevated in the prehospital or ED setting. The Cuffpressure same data on patients in whom a combitube had been manometer is used routinely in our institution to measure placed. The size of the endotracheal tube and patient weight and adjust intracuff pressures. The precision of these could have been collected to further evaluate their relation- measurements has been shown to be 0.7 F 1.9 cm H2O, ship with measured pressure and whether they are related to but it is slightly less accurate at higher pressures [16].A excess pressures. more formal study with rigorous attention to accurate, reproducible measurements would be required to validate our data. In summary, endotracheal tube cuff pressures may 5. Discussion be significantly elevated in the prehospital or ED setting. Measurement of intracuff pressure is a simple and inexpen- In this study, most of the patients intubated in the sive procedure and should be done whenever a patient is prehospital or ED setting experienced elevated endotracheal intubated, in either the prehospital or hospital setting, intracuff pressures that require correction. because this may reduce long-term morbidity. The first-generation endotracheal tubes required high pressures to seal the lumen. Since the end of the 1970s, high-volume, low-pressure tubes have replaced these tubes. References These tubes have a wider support surface on the tracheal mucosa, and therefore the pressure exerted is lower. Despite [1] Carroll R, Hedden M, Safar P. Intratracheal cuffs: performance characteristics of tracheal cuffs. Int Anesthesiol Clin 1974;12:111-41. this, these low-pressure cuffs can exert pressures exceeding [2] Carroll RG, McGinnis GE, Grenvik A. Performance characteristics of the tracheal capillary pressure. High-compliance endotra- tracheal cuffs. Int Anesthesiol Clin 1974;12:111-41. cheal tube pressure increases slowly between the pressures [3] Mehta S, Mickiewicz M. Pressure in large-volume, low-pressure cuffs: its significance, measurement and regulation. Intensive Care of 10 and 20 cm H2O, after which the addition of small volumes increases the cuff pressure substantially [10]. Med 1985;11:267-72. [4] Seegoblin RD, Van G, Hasselt L. Endotracheal cuff pressure and Typically, tube pressure is estimated by palpation of the tracheal mucosal blood flow: endoscopic study of the effects of four inflation port balloon, but this has been shown to be large volume cuffs. Br Med J 1984;288:965-8. unreliable [9]. [5] McHardy FE, Chung F. Postoperative sore throat: cause, prevention In an animal model, cuff pressure greater than 30 cm and treatment. Anaesthesia 1999;54:444-53. H O for 15 minutes was sufficient to induce tracheal [6] Nordin U. The trachea and cuff induced tracheal injury: 2 an experimental study on causative factors and prevention. Acta mucosal lesions, which did not worsen with time beyond Otolaryngol 1976;345(Suppl 345):1-7. 15 minutes [6]. This time frame is longer than most [7] Guyton DC, Barlow MR, Besselievre TR. Influence of airway prehospital transports. Tracheal mucosal lesions and post- pressure on minimum occlusive endotracheal tube cuff pressure. Crit intubation sore throat were shown to be correlated with cuff Care Med 1997;25:91-4. overpressure [11]. Postoperative sore throat was significant- [8] Byrd RA, Mascia MF. What is the endotracheal tube cuff pressure in a cross-section of intubated patients? Anesthesiology 1996;85 ly decreased when low cuff pressure was maintained during (Suppl 3A):982. anesthesia [12]. [9] Fernandez R, Blanch L, Mancebo J, Bonsoms N. Endotracheal tube Prevention of overpressure in endotracheal tube cuffs can cuff pressure assessment: pitfalls of finger estimation and need for be achieved by several means. Frequent measurement and objective measurement. Crit Care Med 1990;18:1423-6. adjustment of cuff pressures is simple, requires little time, [10] Braz JRC, Navarro LHC, Takata IH, Nascimento P. Endotracheal tube cuff pressure: need for precise measurement. Sao Paulo Med J 1999; and is most amenable to the prehospital setting. Endotra- 117:243-7. cheal tubes with a pressure-regulated cuff system are [11] Combes X, Schauviliege F, Peyrouset O, Motamed C, Kirov K, effective but have a prohibitive cost [13]. Filling the cuff Dhonneur G, et al. Intracuff pressure and tracheal morbidity: influence with saline has been shown to provide constant cuff of filling cuff with saline during nitrous oxide anesthesia. Anesthe- pressures [11]. However, endotracheal tube cuffs are not siology 2001;95:1120-4. [12] Susuki N, Kooguchi K, Mizobz T, Hirose M, Takano Y. Postoperative designed to be filled with saline. hoarseness and sore throat after tracheal intubation: effect of low Problems attributable to insufficient cuff inflation have intracuff pressure of endotracheal tube and the usefulness of cuff been reported, including leaking of the tidal volume and pressure indicator. Masui 1999;48:1091-5. 56 J.E. Svenson et al.

[13] Brandt L. Prevention of nitrous oxide induced increase in endotracheal [15] Asai T, Shingu K. Leakage of fluid around high-volume, low- tube cuff pressure. Anesth Analg 1991;72:262-70. pressure cuffs: a comparison of four tracheal tubes. Anaesthesia 2001; [14] Metha S, Mickiewicz M. Work practices relating to intubation and 56:38-42. associated procedures in intensive care units in Sweden. Acta [16] Blanch PB. Laboratory evaluation of 4 brands of endotracheal tube Anaesthesiol Scand 1986;30:637-40. cuff inflator. Respir Care 2004;49:166-73. American Journal of Emergency Medicine (2007) 25,57–59

www.elsevier.com/locate/ajem

Brief Reports Onset of symptoms after methadone overdose

Frank LoVecchio DO, MPHa,b,c,*, Anthony Pizon MDa, Brad Riley MDa, Azadeh Samic, Carmella D’Incognitoc aBannerGood Samaritan Regional Poison Center, Phoenix, AZ, USA bDepartment of Emergency Medicine, Maricopa Medical Center, Phoenix, AZ, USA cAZ College of Osteopathic Medicine, Phoenix, AZ, USA

Received 26 January 2006; accepted 2 July 2006

Abstract Background: Methadone ingestion may cause delayed coma and require naloxone infusion. Few studies exist regarding the time development of symptoms following methadone overdose in adults. Methods: After a brief training period, reviewers who were blinded to the purpose of the study completed a standardized data collection sheet. Two consecutive years of poison center patient encounters were reviewed. Age, outcomes, coingestions, vital signs, clinical manifestations, hospital admissions, and mortality were abstracted. Data were analyzed using descriptive statistics. The first reviewer was designated to extract the data. The second reviewer conducted a review of 20% of all the charts for a j value to be calculated. Results: In total, 44 cases of isolated methadone overdose in patients older than 18 years were identified. A mean age of 32.5 (18-58) years and a mean presumed ingestion of 106 mg of methadone was calculated. Of the 44 patients, 32 received naloxone for symptoms consistent with opiate toxicity. All symptoms occurred within 9 hours of methadone ingestion, with a mean symptom onset of 3.2 hours. All patients had resolution of symptoms within 24 hours. No deaths were recorded. The j score for interreviewer reliability was 0.69, with a 95% confidence interval of 0.58 to 0.73. Limitations: This was a retrospective study that was limited by patient history. Conclusion: Acute methadone toxicity typically results in symptoms within 9 hours of ingestion. D 2007 Elsevier Inc. All rights reserved.

1. Introduction (CNS) and the organs of smooth muscle. Quantitatively and qualitatively, its actions are similar to those of Methadone is a long-acting synthetic opiate therapeuti- morphine [1]. cally used for analgesia and opioid addiction detoxification An overdose of methadone can result in serious life- and maintenance. As a l-receptor agonist, methadone has threatening effects. The classic triad consists of CNS a multitude of actions involving the central nervous system depression, respiratory depression, and myopic pupils. The treatment course for methadone overdose can extend for days * Corresponding author. Phoenix, AZ 85006, USA. Tel.: +1 602 239 because of the prolonged duration of its actions [2]. Currently, 2358; fax: +1 602 239 4138. there are few studies that examine the time relationship E-mail address: [email protected] (F. LoVecchio). between methadone overdose and symptom development.

The purpose of this study was to establish a time frame 8 and 59 hours. The slow release of methadone from the for symptom development seen in patients with isolated liver and other tissues enables the prolonged duration of its methadone overdose. action despite low plasma concentrations. Although the analgesic actions may be evident for only 4 to 8 hours, the elimination rate is substantially longer. This facilitates 2. Methods the potential for treatment difficulties and delayed overdose symptoms [4]. Our study of methadone overdose ingestions consisted of An overdose of methadone may present with signs and a retrospective chart review of 2 consecutive years of patient symptoms of respiratory depression, CNS depression, encounters. The reviewers underwent a brief training period constricted pupils, muscle flaccidity, bradycardia, and of systematic chart review. Blinded to the purpose of the hypotension. In severe overdose, apnea, cardiac arrest and study, they completed a standardized data collection sheet. death may occur. The onset of a coma represents the effect Information concerning age, outcomes, coingestions, vital of the opioid blocking the brain acetylcholine receptors in signs, clinical manifestations, hospital admissions, and any the ascending reticular activating system. Even at low doses, mortality were gathered from the records. methadone may cause somnolence and coma in elderly or The data were analyzed using descriptive statistics. In debilitated patients because of the exaggerated plasma level addition, a second person was tasked to review 20% of all seen in these patients [2,4]. This particular symptom was the charts to calculate a j value for the establishment of an examined as a concern for delayed expression. interreviewer reliability value. A few studies have been conducted to review the observational guidelines that have been adapted for the opioid-toxic patients. In most cases of opioid related deaths, 3. Results respiratory depression is the main cause. One examiner’s review reported that the development of noncardiogenic An extensive chart review identified 44 cases of isolated pulmonary edema was recorded to have occurred within methadone overdose in adult patients. The patient ages 2 hours of emergency department (ED) presentation. A ranged from 18 to 58 years, giving a mean age of 32.5 years. review of 61 of 64 heroin-related NCPE cases revealed less The mean dose of presumed ingestion was calculated to be than 1% developed delayed symptom presentation. In 106 mg of methadone. addition, a retrospective study of 124 patients with opioid All observed symptoms consistent with methadone overdose displayed noncardiac pulmonary edema (NCPE) overdose occurred within a 9-hour time frame. The mean symptoms within 20 minutes of ED presentation. Another for symptom onset was as early as 3.2 hours. Of the 44 cases retrospective study examined the recurrence of opioid reviewed, 32 patients received naloxone treatment for toxicity in 84 patients treated with naloxone. Methadone symptoms of opiate toxicity. All patients experienced was significantly more likely to cause a recurrence of resolution of symptoms within 24 hours. In addition, no toxicity. Interestingly, all recurrences were evident within deaths were recorded. 2 hours of symptom presentation [5]. The j score for interreviewer reliability was calculated to The treatment of a methadone overdose can be highly be 0.69. There was a 95% confidence interval between 0.58 effective when begun promptly. Airway and respiratory and 0.73. support are always a priority when needed. Monitoring vitals, pulse oximetry, and arterial blood gases are useful. Toxicology screens, on the other hand, are generally not 4. Discussion very useful because of their low sensitivity. The urine drug screens can confirm the presence of an opioid. However, Methadone has been used as a successful pharmaco- they cannot distinguish between a true positive and the logic intervention for the treatment of heroin dependence ingestion of poppy seeds [1,5]. and acute and chronic pain. Patients on methadone Often, patients require the use of an opiate antagonist maintenance therapy may become physically dependent, such as naloxone. Patients who receive infusions should but they should be able to refrain from abusing other drugs remain for observation for at least 4 hours after discontinu- [3]. However, methadone overdosing has become a ing the naloxone. This precaution is necessary because the growing phenomenon because of the increased availability short half-life of 1 to 3 hours for the opiate antagonists of this drug [2]. leaves the potential for recurrences of toxicity. Patients who Methadone differs from other opioid agonists in several were asymptomatic and did not require a treatment with important ways. Primarily, the pharmacokinetic properties naloxone need only to be observed for 6 hours. Additional of absorption, metabolism, and analgesic potency have a literature states the ideal treatment course for methadone high variability between patients. Metabolism occurs via the overdose will involve a lengthy stay for up to 72 hours. cytochrome P450 enzymes of CYP3A4 and CYP2D6. The These are the recommendations designated for the treatment terminal half-life of methadone has extensive range between of methadone overdose [1]. Onset of symptoms after methadone overdose 59

This study attempts to address the standard of care 5. Conclusion surrounding the time frame established for methadone toxicity. Although the symptoms of overdose can be quite Acute isolated methadone toxicity typically results in concerning, our study establishes a shortened time period symptom presentation within 9 hours of ingestion. This data for development of these symptoms. All effects were brings into question the admission guideline of up to evident within 9 hours of ingestion, with the mean period 72 hours of observation for methadone overdose patients. being 3.2 hours. Although methadone has an extended elimination rate, the time required for symptom development may not be as lengthy as the half-life of the drug. References Perhaps, a better standard can be established to observe [1] Micromedex (online database). Poisindex management: methadone and the patient for a shortened period after the last treatment related agents. Thomson Micromedex, USA. course is completed. Further investigation is warranted to [2] Wolff K. Characterization of methadone overdose: clinical consider- discover what treatment course may best benefit the ations and scientific evidence. Ther Drug Monit 2002;24(4):457-70. patient’s health, time, and resources. This retrospective [3] Up to date (online database): Michael Weaver. Heroin and other study was limited by patient history and the amount of opioids. [4] Nissen D, editor. Monogram: Methadonesin. Mosby’s drug consult. cases seen. Future studies could involve a larger data 15th ed. St. Louis (Mo)7 Mosby; 2005. p. 326-8. collection and a focused examination of specific symptoms [5] Ford MD. Clinical toxicology. 1st ed. W.B. Saunders Company; 2001. and their presentation. p. 628-35. American Journal of Emergency Medicine (2007) 25,60–64

www.elsevier.com/locate/ajem

Brief Reports Sumatriptan for the treatment of undifferentiated primary headaches in the EDB

James R. Miner MD*, Stephen W. Smith MD, Johanna Moore, Michelle Biros MS, MD

Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, MN 55415, USA

Received 19 January 2006; revised 12 June 2006; accepted 26 June 2006

Abstract Objective: In this study, we determine the effectiveness and adverse effects of sumatriptan when used in the emergency department (ED) as a first-line treatment for benign undifferentiated headaches, and determine if the International Headache Society (IHS) classification of migraine, probable migraine, or tension-type headache has any effect on the effectiveness of the treatment. We hypothesize that there is no difference in the effectiveness of pain relief or frequency and severity of adverse effects between patients with migraine, probable migraine, or tension-type headaches when treated with sumatriptan. Methods: This was a prospective observational study of adult ED patients undergoing treatment for primary headaches (ie, patients in whom head trauma, vascular disorders, infection, or disorders of facial or cranial structures have been clinically excluded). Other exclusions were renal impairment, hepatic impairment, and risk factors for coronary artery disease. Consenting patients then were asked to complete a 100-mm visual analog scale (VAS) representing their perceived pain, after which they were interviewed by a research assistant who completed a headache diagnosis worksheet, which differentiates the headache by IHS criteria. The patient repeated the VAS score at 30 and 60 minutes. Data were analyzed using descriptive statistics and v2 tests. Results: One hundred forty-seven patients were enrolled: 84 (57.1%) patients with migraine headache, 45 (30.7%) with a probable migraine headache, and 18 (12.2%) with a tension headache. A 50% reduction in VAS scores 60 minutes postdose was seen in 87 (59%) of 147 patients; 50 (60%) of 84 of migraine patients, 25 (56%) of 45 of probable migraine patients, and 12 (67%) of 18 tension patients ( P = .72). There were no serious adverse events reported. Forty-seven patients (32%) received rescue medications after the 60-minute VAS score: 29 (34.5%) patients in the migraine group, 15 (33.3%) patients in the probable migraine group, and 3 (15.8%) patients in the tension-type headache group ( P = .26). Conclusions: Most of the patients presenting with primary headaches had migraine or probable migraine headaches. There was no difference in sumatriptan’s effectiveness based on the classification of the headache using IHS criteria. D 2007 Elsevier Inc. All rights reserved.

This work was presented in part at the Society for Academic Emergency Medicine Annual Meeting, May 2004, Orlando, Fla, and at the American Headache Society Research Assembly, June 2004, Vancouver, BC. B This work was supported in part by a grant from GlaxoSmithKline. * Corresponding author. Tel.: +1 612 873 8791. E-mail address: [email protected] (J.R. Miner).

1. Introduction differences in etiology and treatment of the primary headaches. The International Headache Society (IHS) has There are many taxonomies of headache. Primary head- established several definitions for classifications of headaches, also known as benign headache, are the headaches aches [1]. Although these remain useful in determining that are not bsecondaryQ to some identified discrete long-term treatment strategies, most ED patients present pathology (eg, meningitis, sinusitis, subarachnoid hemor- with undifferentiated headaches [2]. These include mi- rhage, and toothache). Headaches present a frequent graine, probable migraine, episodic tension-type, and cluster diagnostic and therapeutic challenge to emergency physi- headaches. The benign causes of headache (primary head- cians. In fact, it is estimated that patients with the complaint aches) have been reported to account for up to 90% of of headache account for 1% to 2% of all emergency ambulatory patients presenting with a complaint of head- department (ED) visits [1]. Very little is known about the ache [3].

Fig. 1 Headache diagnosis worksheet. 62 J.R. Miner et al.

Sumatriptan has been shown to be effective in treating diabetes, a family history of coronary artery disease, males the spectrum of primary headache subtypes in clinic patients older than 40 years, or postmenopausal females. [3]. Its adverse effects include coronary artery spasm and chest pain, but these occur rarely. Although it has been 2.3. Study protocol established that sumatriptan is an effective treatment for Consenting patients were asked to complete a visual migraine headaches, it is unknown whether it is effective in analog scale (VAS) representing their perceived pain. This ED patients for primary headaches of the subtypes other scale was a 100-mm line with the words bno painQ on 1 end than migraine. It is difficult to formally establish the and bmost pain imaginableQ on the other. Patients were then diagnosis of migraine headaches in ED patients because it treated with sumatriptan 6 mg subcutaneously and inter- generally requires 5 episodes for the diagnosis to be made. viewed by a trained research assistant who completed a The objective of this pilot study was to determine the headache diagnosis worksheet, which differentiates the effectiveness of sumatriptan when used in the ED as a first- headache by IHS criteria (Fig. 1). Patients were asked to line treatment for undifferentiated primary headaches and to repeat the VAS 30 and 60 minutes after receiving the drug. determine if the IHS classification as migraine, probable Pain relief was defined as a 50% decrease in the pain VAS migraine, or tension type has any effect on the outcome. score. The patient’s physician was asked to report any adverse effects of the medication. Patients were contacted by telephone 48 hours after enrollment and queried regarding 2. Methods continuing pain and the need for further treatment. 2.1. Study design 2.4. Data collection and analysis This was a prospective observational study of adult ED Data were collected by a designated research assistant patients undergoing treatment for primary headaches. The and was then entered into an EXCEL (Microsoft Corp, institutional review board of the Hennepin County Medical Redmond, Wash) database. All analysis and interpretation of Center, Minneapolis, Minn, approved the study. data were performed using STATA 6.0 (STATA Corp, 2.2. Study setting and population College Station, Tex) statistical software. Descriptive statistics were used when appropriate. The This study was performed at the Hennepin County rates of patients achieving a 50% reduction in their pain Medical Center, an urban county hospital with approxi- VAS score by headache classification and other nonpara- mately 93000 ED patient visits per year between January metric parameters of the study groups were compared using 30, 2003, and September 30, 2004. A convenience sample v2 tests. of adult (older than 18 years) ED patients who were going to To find an effect size of 0.3 in a v2 table comparing 50% be treated for a primary headache were included (ie, patients pain relief in the 3 groups, with an a of .10 and 90% power, in whom head trauma, vascular disorders, infection, or we estimated that it was necessary to enroll 146 patients in disorders of facial or cranial structures have been clinically the study. excluded). Exclusion criteria included prior use of triptans or ergots for this headache, renal or hepatic impairment, MAO-I use at present or in the past, a history of vascular 3. Results disease, pregnancy or breast-feeding, or having at least 1 of the following coronary artery disease risk factors: hyper- One hundred sixty-seven patients were enrolled (mean tension, hypercholesterolemia, tobacco abuse, obesity, age, 28.4 F 14.4; 74% female). Fifteen patients did not

Table 1 Outcomes of the study Migraine Probable migraine Tension Total Baseline VAS score (95% CI) 87.7 (83.1-92.3) 75.4 (66.9-83.8) 77.4 (60.7-94.1) 82.3 (78.5-86.9) Available for follow-up 58 (66.7%) 30 (65.2%) 7 (36.8%) 95 (62.5%) Minutes in the department 51.2 F 105.1 107.8 F 155.6 90.2 F 177.5 72.9 F 133.9 (range, 15-730) (range, 20-730) (range, 25-720) (range, 15-730) % decrease in VAS score 34.7 (26.7-42.8) 19.5 (À12.8 to 51.8) 38.3 (9.1-67.5) 30.5 (19.3-41.7) at 30 min (95% CI) % decrease in VAS score 56.5 (48.2-61.8) 42.1 (17.4-66.9) 60.4 (38.4-82.4) 52.6 (43.4-61.8) at 60 min (95% CI) 50% decrease in VAS at 30 min 41 (49%) 20 (44%) 12 (67%) 73 (50%) 50% decrease in VAS at 60 min 50 (60%) 25 (56%) 12 (67%) 87 (59%) CI indicates confidence interval. Sumatriptan for the treatment of undifferentiated headaches in the ED 63 consent to be in the study. Four patients who gave consent headaches. No further adverse effects from those noted left the department before being discharged and did not give before discharge were found in the follow-up calls. a 30-minute VAS score (1 tension, 1 probable migraine, and 2 migraines). One patient was given a rescue medication before the 30-minute VAS score and was excluded 4. Discussion (migraine group). One patient was eventually diagnosed with viral meningitis and excluded from the study. This left Sumatriptan has been well described in the treatment of 146 patients with data appropriate for analysis. acute migraine headaches outside of the ED [1,4,5]. It is not Using IHS criteria, 84 (57.1%) patients presented with a known, however, if patients presenting to the ED with migraine headache, 45 (30.7%) with a probable migraine primary headaches will respond to a medicine indicated headache, and 18 (12.2%) with a tension headache. There only for migraine headaches. Most of the patients enrolled was no difference in the age ( P = .64) or sex ( P = .85) of in this study had either a migraine or a probable migraine the 3 groups. Fifty-nine patients reported having a previous headache. Tension headaches represented only a small headache diagnosis from a physician; they included 24 pa- portion of the patients who qualified for our study. The tients with migraine, 3 with cluster, 7 with tension, 18 who distribution of headache subtypes identified here is similar had unspecified diagnoses, and 7 with posttraumatic to previous reports of primary headache distributions [3], conditions. There was no difference in the distribution of but the rate of migraine, migrainous, and tension headaches these previous diagnoses between the 3 groups ( P = .82). among patients presenting to the ED and requiring a pain The outcome parameters are presented in Table 1. Fifty- medication has not been well described, and we cannot four (65%) of 83 patients with a history of migraine comment on whether or not this is unique to our institution. headaches had a 50% decrease in their VAS score at 60 The adverse events noted were similar in nature and minutes compared with 39 (61%) of 64 without a history of frequency to previous reports for sumatriptan [6]. We used a migraines ( P = .61). There was no difference between large number of exclusion criteria, including patients with headache types in terms of the measured pain relief by the cardiac risk factors, which may have prevented possible VAS scales. adverse events. We assume that the rate of adverse events Thirty-six adverse events were noted. Seven patients from sumatriptan would be much higher among such reported increased drowsiness, 7 reported worsening head- patients with cardiac risk factors such as hypertension, ache, 6 reported nausea, and 2 reported shortness of breath. obesity, or tobacco use. Thirteen patients reported chest and neck burning or We found very few differences between patients with tightness, all for less than 60 minutes. All of these patients migraines, probable migraine, or tension-type headaches in had an electrocardiogram performed, and none of these this study. The differentiation of headaches among migraine, patients were noted to have abnormalities after experiencing probable migraine, and episodic tension type does not appear these symptoms. One patient developed a fever 45 minutes to be associated with the effectiveness of sumatriptan, and its after the administration of sumatriptan and went on to be effectiveness in all groups was similar to previous reports for diagnosed with viral meningitis. This person did not have migraine headache treatments in the ED [2,7-9]. It is possible relief of their headache pain from the sumatriptan. that the IHS criteria do not offer a useful differentiation for Forty-seven patients (32.0%) received rescue medica- the treatment of headaches in the ED. tions after the 60-minute VAS score: 30 (20.4%) patients received droperidol, 11 (7.5%) received oral hydrocodone/ acetaminophen, 5 (3.4%) received parenteral morphine, and 5. Limitations 1 (0.6%) received lorazepam. There was no difference in the type of rescue medications given between headache-type This study’s primary limitation is that it is not a groups ( P = .77). Rescue medications were given to randomized comparison of treatments. We felt that to 29 (34.5%) patients in the migraine group, 15 (33.3%) establish the importance of the differentiation of primary patients in the probable migraine group, and 3 (16.7%) headache subtypes on headache treatment in the ED, we patients in the tension-type headache group ( P = .26). should first compare the effectiveness of a known treatment Ninety-five (64.6%) patients were available for follow- for migraine headaches to other primary headache subtypes up at 48 hours. Sixty-three (66.3%) of these patients (probable migraine and episodic tension-type headaches) reported continuing headache symptoms: 40 (63.5%) because it is not yet established if different subtypes of patients in the migraine group, 19 (30.2%) in the probable primary headaches in the ED will respond similarly to migraine group, and 4 (6.3%) in the tension group ( P = .85). pain treatments. Thirty-one patients (49.2%) described the continuing head- A further limitation was that only patients with no aches as light, 6 (9.5%) as moderate, and 26 (41.3%) as cardiac risk factors were enrolled. This is not typical of a severe. There was no difference in the description of the sample of a population of ED patients and limits the severity of pain between the 3 groups ( P = .95). Fourteen extrapolation of the findings here to patients without cardiac patients (14.7%) had sought further treatment for their risk factors. 64 J.R. Miner et al.

The distribution of headaches in our study was heavily References weighted toward migraine headaches, with only a small number of patients classified as having a tension headache. [1] IHC Committee. ICD-10 guide for headaches. Cephalalgia 1997; It is not clear if this is an anomaly of our study population 17(Suppl 19):1-82. or is indicative of general ED patients who require [2] Morgenstern LB, Huber JC, Luna-Gonzales H, et al. Headache in the emergency department. Headache 2001;41(6):537-41. pain medications for a primary headache. A large study [3] Lipton RB, Stewart WF, Cady R, et al. Sumatriptan for the range of of all patients presenting with primary headache will headaches in migraine sufferers: results of the spectrum study. be needed to determine the prevalence of each head- Headache 2000;40(10):783-91. ache subtype. [4] Dowson AJ, Lipscombe S, Sender J, Rees T, Watson D. New guidelines for the management of migraine in primary care. Curr Med Res Opin 2002;18(7):414-39. [5] Silberstein SD. Evaluation and emergency treatment of headache. 6. Conclusion Headache 1992;32(8):396-407. [6] Silberstein SD, Rosenberg J. Multispecialty consensus on diagnosis and Most primary headaches in this study were classified treatment of headache. Neurology 2000;54(8):1553. as migraine or probably migraine. There was no difference [7] Miner JR, Fish SJ, Smith SW, Biros MH. Droperidol vs. prochlorper- in sumatriptan’s effectiveness for primary headaches based azine for benign headaches in the emergency department. Acad Emerg on the classification of the headache using IHS criteria. Med 2001;8(9):873-9. [8] Richman PB, Reischel U, Ostrow A, et al. Droperidol for acute Sumatriptan’s effectiveness in this sample of undifferen- migraine headache. Am J Emerg Med 1999;17(4):398-400. tiated primary headaches in the ED was similar to pre- [9] Diamond ML. Emergency department management of the acute vious reports. headache. Clin Cornerstone 1999;1(6):45-54. American Journal of Emergency Medicine (2007) 25,65–68

www.elsevier.com/locate/ajem

Diagnostics Prognostic implications of myocardial necrosis triad markers’ concentration measured at admission in patients with suspected acute coronary syndrome

Filip M. Szyman´ski MDa,b,*, Marcin Grabowski MD, PhDa,b, Krzysztof J. Filipiak MD, PhDa, Grzegorz Karpin´ski MDa, Anna Hrynkiewicz MDa, Przemysyaw Stolarz MD, PhDa, Artur Ore˛ziak MD, PhDa, Robert Rudowski PhDb, Grzegorz Opolski MD, PhDa a1st Department of Cardiology, Medical University of Warsaw, 02 097 Warsaw, Poland bDepartment of Medical Informatics and Telemedicine, Medical University of Warsaw, 02 097 Warsaw, Poland

Received 11 February 2006; revised 15 July 2006; accepted 17 July 2006

Abstract The aim of the study was to analyze the prognostic implications of 3 myocardial necrosis markers measured at admission in short-term observation of patients with suspected acute coronary syndrome. The study group consisted of 336 consecutive patients whose concentration of cardiac troponin I, creatine kinase–MB fraction, and myoglobin were measured at admission. All patients referred due to chest pain and suspected acute coronary syndrome and were followed up for 30 days. The patients who died had statistically higher concentration of cardiac troponin I (8.7 F 17.2 vs 0.9 F 3.2 ng/mL; P = .0006), myoglobin (215.2 F 181.5 vs 109.7 F 151.5 ng/mL; P = .003), and creatine kinase–MB (21.9 F 30.7 vs 8.8 F 25.9 ng/mL; P = .005), compared to patients who stayed alive. There was statistically significant increase in 30-day all-cause mortality with increasing numbers of positive markers—0.6% for patients with nonpositive marker, 3.4% for patients with 1 positive marker, and 11.5% for patients with at least 2 positive markers ( P = .001 for trend). D 2007 Elsevier Inc. All rights reserved.

1. Introduction approximately one third of patients with diabetes mellitus and elderly patients develop nonspecific symptoms [1-3]. Early accurate diagnosis of acute coronary syndrome Although electrocardiography is a simple and highly useful (ACS) and rapid evaluation of its severity may influence diagnostic method, its sensitivity is approximately 50% the patient’s prognosis. Chest pain symptoms are the first [4-7]. Approximately 40% of patients with myocardial infar- diagnostic indicator of acute myocardial infarction (AMI), ction have a nonspecific electrocardiogram (ECG) without but many patients may complain of atypical symptoms, and ST elevation, and 10% of patients have a normal ECG [8]. Therefore, serum markers may have an important role in the * Corresponding author. Tel.: +48 603 957 993. earlier and more accurate diagnosis of AMI. Clinicians E-mail address: [email protected] (F.M. Szyman´ski). commonly use creatine kinase–MB fraction (CK-MB) and

0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ajem.2006.07.007 66 F.M. Szyman´ski et al. cardiac troponin I (cTnI) to diagnose AMI, but circulating presented to the emergency department between December levels of these markers do not increase until 3 to 6 hours 2003 and December 2005 with acute chest pain of possible after the onset of symptoms. Myoglobin (MYO) is a small coronary etiology. All patients fulfilled one of the following (17.8-kd) cytoplasmic heme protein found in all muscles. criteria: (a) an ECG that showed dynamic changes or (b)a Myoglobin increases within 1 to 3 hours in the setting of high probability of an ACS according to the criteria of the myocardial necrosis, usually peaks within 6 to 9 hours, and cardiologist on duty. Patients with renal failure (creatinine may become normal in 24 hours [9]. Myoglobin might be N1.5 mg/dL), hypothyroidism, myopathy, and traumatic increased when cTnI levels are still normal [10]. It has been injury and those requiring defibrillation were excluded from suggested that concomitant use of cTnI and MYO is the the study. Study protocol was approved by the Regional optimal combination [11]. However, using only MYO to Ethics Committee. Informed consent was obtained from diagnose AMI has significant limitations. Myoglobin has each patient before participation in this study. limited specificity for myocardial necrosis in patients who All-cause mortality was evaluated at 30 days of the have renal insufficiency and skeletal muscle trauma—it may follow-up period by telephone contact or outpatient clinic be elevated in many noncardiac situations as it is reported visit. No patient was lost to follow-up. for other newer necrosis markers [12,13]. The aim of the present study was the analyses of the 2.2. Biomarker analytic techniques prognostic implications of myocardial infarction markers Patients did complete the full marker panel. Creatine measured at admission in short-term observation of patients kinase–MB, cTnI, and MYO were measured with a point- with suspected ACS. of-care device (Dade Behring Stratus CS, Newtown, Conn). Reference interval for CK-MB, cTnI, and MYO were 0.6- 3.5, 0.0 to 0.1, and 20 to 82 ng/mL, respectively. 2. Methods 2.3. Statistical analysis 2.1. Population The data of this study were analyzed by MedCalc We enrolled 336 consecutive patients (180 men), 24 to (MedCalc Software, Mariakerke, Belgium). Continuous 98 years of age (mean age: 65.9 F 14.4 years), who data are presented as mean F SD and were compared using

Table 1 Baseline characteristics according to 30-day all-cause mortality Characteristics (mean F SD or %): Survivals Nonsurvivals P No. of patients 323 13 – Demographics Age (y) (FSD) 65.6 F 14.4 72.9 F 10.4 P = .064 Age N 65 (y) (n [%]) 185 (57.3%) 10 (76.9%) P = .445 Age N 75 (y) (n [%]) 91 (28.2%) 6 (46.1%) P = .580 Male sex (n [%]) 176 (54.5%) 4 (30.8%) P = .162 Medical history (n [%]) Hypertension (n [%]) 172 (53.2%) 5 (38.5%) P = .445 Prior stroke (n [%]) 12 (3.7%) 0 (%) P = .957 Diabetes mellitus (n [%]) 37 (11.5%) 4 (30.8%) P = .098 Hypercholesterolemia (n [%]) 84 (26.0%) 4 (30.8%) P = .951 Current smoker (n [%]) 69 (21.4%) 3 (23.1%) P = .844 Prior myocardial infarction (n [%]) 72 (22.3%) 5 (38.4%) P = .306 Prior CABG (n [%]) 1 (0.3%) 0 (0%) P = .997 Physical findings Mean heart rate (beats per min) 79 F 16.3 98.7 F 27.3 P = .0001 Heart rate N100 beats per min no (%) 29 (9.0%) 8 (61,5%) P = .959 Mean systolic blood pressure (mm Hg) 133.8 F 25.2 103.5 F 29.4 P = .0009 Systolic blood pressure b100 mm Hg 20 (6.2%) 7 (53.8 %) P = .374 Mean diastolic blood pressure (mm Hg) 79.8 F 14.1 63.1 F 15.5 P = .0004 ECG findings ST elevation (n [%]) 47 (14.6%) 11 (84.6%) P b .0001 Biochemical test results CK-MB (ng/mL) (mean F SD) 8.8 F 25.9 21.9 F 30.7 P = .005 cTnI (ng/mL) (mean F SD) 0.9 F 3.2 8.7 F 17.2 P = .0006 MYO (ng/mL) (mean F SD) 109.7 F 151.5 215.2 F 181.5 P = .003 CABG, coronary artery bypass grafting; NS, nonsignificant. Myocardial necrosis markers 67

Fig. 2 Receiver-operating characteristic curve for prognostics value for 30-day all-cause mortality of cTnI (dotted lines), MYO (solid lines), CK-MB fraction (dashed lines) on admission, with area under the curve of 0.764 (95% CI, 0.70-0.81), 0.785 (95% CI, 0.72-0.83), and 0.714 (95% CI, 0.65-0.77), respectively. 3. Results

During the initial 30 days, 13 patients died. The demographic data and medical history of the study population according to 30-day mortality are shown in Table 1.All the patients who died had had statistically faster heart beat (98.7 F 27.3 vs 79 F 16.3 beats per minute; P = .0001) and lower systolic (103.5 F 29.4 vs 133.8 F 25.2 mm Hg; P = .0009) and diastolic (63.1 F 15.5 vs 79.8 F 14.1 mm Hg; P = .0004) blood pressure at admission, compared to the patients who survived. The maximum levels detectible for CK-MB, cTnI, and MYO were 150, 40.4, and 900 ng/mL, respectively. The patients who died, compared to patients who stayed alive, had statistically higher concentration of all markers: cTnI 8.7 F 17.2 vs 0.9 F 3.2 ng/mL, P = .0006; CK-MB 21.9 F 30.7 vs 8.8 F 25.9 ng/mL, P = .005; and MYO

Fig. 1 Mean cTnI (A), CK-MB fraction (B), and MYO (C) levels at admission (FSD) according to 30-day mortality.

the Mann-Whitney U or Student t test. Categorical variables were compared using either v2 or Fisher exact tests. Results were included in a receiver operating characteristic (ROC) curve analysis. Using multivariate logistic regression, the relationship between markers (dichotomized) and outcome was adjusted for the effects of clinical features. The resulting sensitivity, specificity, and positive and negative predictive values were calculated. In all comparisons, P = Fig. 3 Thirty-day all cause mortality according to number of .05 was considered to be significant. positive markers. 68 F.M. Szyman´ski et al.

215.2 F 181.5 ng/mL vs 109.7 F 151.5 ng/mL, P = .003 References (Fig. 1A-C). The area under the ROC curve using cTnI, CK-MB, and [1] Maynard C, Litwin PE, Martin JS, et al. Gender differences in the MYO to detect mortality was as follows: 0.764 (95% treatment and outcome of acute myocardial infarction: results from the confidence interval [CI], 0.70-0.81), 0.714 (95% CI, 0.65- myocardial infarction triage and intervention registry. Arch Intern Med 1992;152:972-6. 0.77), and 0.785 (95% CI, 0.72-0.83), respectively. There [2] Weaver WD, Litwin PE, Martin JS, et al. Effect of age on use of was no statistically significant difference between areas thrombolytic therapy and mortality in acute myocardial infarction. under the ROC curves (CK-MB vs cTnI, P = .558; CK-MB J Am Coll Cardiol 1991;18:657-62. vs MYO, P = .506; cTnI vs MYO, P = .855) (Fig. 2). [3] Uretsky BF, Farquhar DS, Berezin AF, et al. Symptomatic myocardial There was statistically significant increase in 30-day all- infarction without chest pain: prevalence and clinical course. Am J Cardiol 1977;40:498-503. cause mortality with increasing numbers of positive markers [4] Rude RE, Poole WK, Muller JE, et al. Electrocardiographic (over upper normal range): 0.6% (1/171) for patients with and clinical criteria for recognition of acute myocardial infarc- nonpositive marker, 3.4% (3/87) for patients with 1 positive tion based on analysis of 3,697 patients. Am J Cardiol 1983; marker, and 11.5% mortality for patients with at least 52:936-42. 2 positive markers ( P = .001 for trend) (Fig. 3). [5] Weaver WD, Eisenberg MS, Martin JS, et al. Myocardial infarction triage and intervention projectphase I: patient characteristics and feasibility of prehospital initiation of thrombolytic therapy. J Am Coll Cardiol 1990;15:925-31. 4. Discussion [6] Karlson BW, Herlitz J, Wiklund O, et al. Early prediction of acute myocardial infarction from clinical history, examination and Creatine kinase–MB, cTnI, and MYO are the most widely electrocardiogram in the emergency room. Am J Cardiol 1991;68: used cardiac necrosis markers. Of these, cTnI are the markers 171-5. of choice because of their higher sensitivity and specificity in [7] Brush JE, Brand DA, Acampora D, et al. Use of the initial electrocardiogram to predict in-hospital complications of acute myocardial detecting myocardial necrosis [14]. We sought to determine infarction. N Engl J Med 1985;312:1137-41. the sensitivity of a change in MYO for AMI in patients who [8] Fisch C. The clinical electrocardiogram: sensitivity and specificity. had normal levels of cTnI at presentation. Myoglobin is the Am Coll Cardiol (Current Journal Rev) 1997;6:71-5. earliest marker, with a negative predictive value for myo- [9] Mair J, Artner-Dworzak E, Lechleitner P, et al. Early diagnosis of cardial infarction diagnosis greater than that provided by acute myocardial infarction by newly developed rapid immunoturbi- dimetric assay for myoglobin. Br Heart J 1992;68:462-8. cTnI in the first hours [12-15]. Myoglobin is generally [10] Alexander JH, Sparapani RA, Hahaffey KW, et al. Association acknowledged to increase before CK-MB in AMI. However, between minor elevations of creatine kinase–MB level and mortality because of its fast clearance, the levels start to diminish in patients with acute coronary syndrome without ST-segment earlier than other markers [16]. Several authorities have elevation. JAMA 2000;283:347-53. suggested cTnI and MYO to be the optimal combination. [11] Panteghini M, Apple FS, Christenson RH, et al. Proposals from IFCC committee on standardization of markers of cardiac damage (C- In this study, we observed that the approach of SMCD): recommendations on use of biochemical markers of cardiac myocardial necrosis triad markers might be superior to the damage in acute coronary syndromes. Scan J Clin Lab Invest 1999; estimation of the researched markers alone. The number of 59:103-12. elevated markers in triad was predictive for fatal outcomes [12] DeWinter RJ, Koster RW, Sturk A, et al. Value of myoglobin, troponin in 30-day follow-up. Thus, better accuracy in estimating T, and CK-MB mass in ruling out an AMI in the emergency room. Circulation 1995;92:3401-7. those patients’ prognosis could be provided by the [13] Mayek xA, Grabowski M, Szpotan´ska M, et al. Exercise h-FABP introduction of the marker triad approach. However, wider plasma concentration in healthy subjects. Arch Med Sci 2005; studies focusing especially on cost-effectiveness of this 1(4):226-9. multimarker method are needed. [14] The Joint European Society of Cardiology/American College of All 3 markers of myocardial infarction have important Cardiology Committee. Myocardial infarction redefined. A consensus document of the joint European Society of Cardiology/American prognostic implication in the case of short-term observation College of Cardiology Committee for the redefinition of myocardial of patients with chest pain and suspected ACS. Mortality at infarction. J Am Coll Cardiol 2000;36:959-69. 30 days was higher in patients with all 3 markers increased [15] Zimmerman J, Fromm R, Meyer D, et al. Diagnostic marker than in patients with other combinations. We would re- cooperative study for the diagnosis of myocardial infarction. commend the combination of all 3 markers to be used in the Circulation 1999;99:1671-7. [16] Adams JE, Abendschein DR, Jaffe AS. Biochemical markers of early evaluation of patients with chest pain and an ECG myocardial injury—is MB creatine kinase the choice for the 1990s? nondiagnostic of AMI. Circulation 1993;88:750-63. American Journal of Emergency Medicine (2007) 25,69–71

www.elsevier.com/locate/ajem

Diagnostics Delayed traumatic thoracic spinal epidural hematoma: a case report and literature review

Cheng-Ta Hsieh MD, Yung-Hsiao Chiang MD, PhD, Chi-Tun Tang MD, Jui-Ming Sun MD, Da-Tong Ju MD*

Department of Neurological Surgery, Tri-Service General Hospital, National Defense Medical Center, Taipei 114, Taiwan, Republic of China

Received 16 May 2006; accepted 17 May 2006

Abstract Spinal epidural hematoma is a relatively uncommon disease, but an important cause leading to cord compression. Posttraumatic spinal epidural hematoma is a rare entity and remains a challenge for clinical physicians. Magnetic resonance imaging is the best choice for early diagnosis, and urgent surgical decompression with evacuation of hematoma could improve some neurologic deficits, especially vital cord functions. We presented a 77-year-old woman who sustained back pain after a fall 1 month before admission, complaining of progressive weakness and sensory loss in bilateral lower extremities since 2 weeks before admission. Radiography of the thoracic spine revealed decreased body height at T10 and compression fracture. Magnetic resonance imaging of the thoracic spine revealed epidural hematomas at the level of T11 to T12. An urgent decompressive laminectomy with evacuation of hematoma was performed immediately. Postoperatively, her previous neurologic deficits improved except for an underactive neurogenic bladder and fecal incontinence. D 2007 Elsevier Inc. All rights reserved.

1. Introduction SEH after a compression fracture and present a review of the relevant literature. Spinal epidural hematoma (SEH) is an uncommon entity but an important cause leading to cord compression. Most SEHs are considered spontaneous without an identifiable 2. Case report cause in 40% to 50% of reported cases and are associated with concurrent coagulopathy, vascular malformations, A 77-year-old woman with a negative medical history neoplasm, blood dyscrasia, minor trauma, and pregnancy had severe back pain after an axial falling accident 1 month [1-3]. Traumatic SEH is relatively uncommon, and the before admission. Initially, she was not able to ambulate, but incidence represents less than 1% to 1.7% of all spinal her symptoms later improved without treatment. However, injuries [2]. Herein, we report a case of delayed traumatic 2 weeks before admission, she sustained progressive weakness in the bilateral lower limbs and had been treated with medication at an outpatient department. Three days * Corresponding author. Tel.: +886 2 8792 7177; fax: +886 2 8792 before admission, she gradually developed bilateral lower 7178. extremity paraplegia and sensory loss. A physical exami- E-mail address: [email protected] (D.-T. Ju). nation was significant for weakness in the bilateral lower

Fig. 1 A, T2-weighted sagittal reconstruction MR image of the thoracolumbar spine with a compression fracture at T10 (asterisk) and a high signal intensity collection dorsal to the spinal cord at the level of T11 to T12. B, T2-weighted axial MR image of the thoracic spine revealed a high signal intensity collection located in the posterior and left aspect of the epidural space, with deviation of the cord to the right side. extremities (muscle power scored as 1/5), hyperreflexia in Paget disease, and ankylosing spondylitis have been knee jerks and ankle clonus (4+), and positive Babinski considered as risk factors for posttraumatic SEH [2]. signs, along with urine and fecal incontinence. Radiography However, posttraumatic SEH is rarely associated with spine of the thoracic spine revealed decreased body height at T10 fracture, with the incidence ranging from 0.5% to 7.5% [6]. with deformity, conforming to the diagnosis of a compres- Most traumatic SEHs affect men more than women and are sion fracture. Magnetic resonance imaging (MRI) of the usually located in the dorsal epidural space of the cervical thoracic spine revealed an epidural mass extending from the spine [4]. Hematomas located in the thoracic spine are T11 to T12 level and compression of the left posterolateral relatively uncommon, with incidence ranging from 15% to aspect of the spinal canal (Fig. 1). The mass displaced the 18.8% in all spinal levels [4,7]. In addition, posttraumatic cord to the right of the midline and anterior, and had a SEHs that occur in the thoracic spine are thought to be mixed intensity on T1- and T2-weighted images, consistent symptomatic with smaller hematomas because of the with blood. SEH was highly suspected, and the patient anatomically narrower canal of the thoracic spine when underwent laminectomy from T10 to T12 and evacuation of compared to the cervical or lumbar spine [2]. the epidural hematoma. Intact laminas and zygapohyses of The pathophysiologic mechanism of SEH remains involved thoracic spine segments (T10-T12) were observed obscure. The hematoma formed by cancellous bone during the operation. In addition, transpedicular vertebro- bleeding from fracture of vertebrae may be located in the plasty with bone cement was performed at T10. Pathologic ventral site of the epidural space. However, close fibrous examination demonstrated hematoma without tumor or adherence of the posterior longitudinal ligament to the abnormal blood vessels. Postoperatively, her previous ventral surface of the canal may prevent the occurrence of neurologic deficits improved except for underreactive hematoma [2]. Although SEH has been thought to originate neurogenic bladder and fecal incontinence. from the arterial system, bleeding from rupture of valveless venous plexus in the epidural spaces, possibly resulting from an abrupt change in venous pressure after blunt 3. Discussion trauma, is the most commonly accepted source [1,4]. The clinical picture of posttraumatic SEH is severe SEH is a rare disease and remains a challenge for clinical painful episode and then progressive symptoms of acute physicians. Traumatic SEH is less common than spontane- cord compression with neurologic deficits including senso- ous lesion, and incidence is estimated at less than 1% to ry, motor, or sphincter dysfunction [5]. The duration 1.7% of all spinal injuries [4]. Traumatic causes include between onset of symptoms and inciting event varies from vertebral fractures, obstetrical birth trauma, lumbar punctu- minutes to months [2]. Progressive spinal cord compromise res, postsurgical bleeding, epidural anesthesia, and missile can lead to paraplegia, quadriplegia, or even death [3].In injuries [5]. Cervical spondylosis, rheumatoid arthritis, our case, the onset of progressive neurologic deficits from Delayed traumatic spinal epidural hematoma 71 paraparesis to paraplegia, hypoesthesia, and sphincter of clinical presentation also significantly influenced the dysfunction was 2 weeks after the inciting event. As well, outcome in patients with SEH [10]. the hematomas were located in the dorsal aspect of the In conclusion, posttraumatic SEH is a rare entity and epidural space below the level of the compression fracture, remains a challenge for clinical physicians. MRI is the best not in the ventral side of the epidural space at the level of the choice for early diagnosis. Urgent surgical decompression destructive bony edge. Because of the hemodynamic change with evacuation of hematoma could improve some neuro- resulting from the abnormal structure, we favor the rupture logic deficits, especially vital cord functions. of the venous plexus as the origin of the bleeding. In the differential diagnosis of SEH, one has to consider a herniated vertebral disk, acute bone compression, cord References edema, cord contusion, and subdural and subarachnoid hemorrhage [5]. With advanced radiographic techniques, [1] Chang FC, Lirng JF, Luo CB, et al. Evaluation of clinical and MR MRI is considered the best modality for early diagnosis of findings for the prognosis of spinal epidural haematomas. Clin Radiol SEH. The MRI appearance of epidural hematomas varies 2005;60:762-70. with its age and could provide the important clue about the [2] Cuenca PJ, Tulley EB, Devita D, et al. Delayed traumatic spinal duration of symptoms [8]. In a study of 20 patients with epidural hematoma with spontaneous resolution of symptoms. J Emerg Med 2004;27:37-41. SEH, Chang et al [1] reported most hematomas examined [3] Dinsmore AJ, Leonard RB, Manthey D. Spontaneous spinal epidural within 36 hours were homogeneously isointense relative to hematoma: a case report. J Emerg Med 2005;28:423-6. the spinal cord on T1-weighted images and homogeneously [4] Foo D, Rossier AB. Post-traumatic spinal epidural hematoma. hyperintense on T2-weighted images. Heterogeneously Neurosurgery 1982;11(1 Pt 1):25-32. isointense or hyperintense on T1-weighted images and [5] Lefranc F, David P, Brotchi J, et al. Traumatic epidural hematoma of the cervical spine: magnetic resonance imaging diagnosis and mixed high- and low-signal change on T2-weighted images spontaneous resolution: case report. Neurosurgery 1999;44:408-10. were found when MRI was performed more than 36 hours [6] Kessel G, Bocher-Schwarz HG, Ringel K, et al. The role of endoscopy after symptom onset. in the treatment of acute traumatic anterior epidural hematoma of the Although SEH in some patients with mild neurologic cervical spine: case report. Neurosurgery 1997;41:688-90. symptoms were successfully managed with conservative [7] Bennett DL, George MJ, Ohashi K, et al. Acute traumatic spinal epidural hematoma: imaging and neurologic outcome. Emerg Radiol treatment, early surgical decompression is still the best 2005;11:136-44. choice of treatment for SEH [3]. In a study of 30 patients [8] Segal DH, Lidov MW, Camins MB. Cervical epidural hematoma after with SEH, Lawton et al [9] reported neurologic deficits chiropractic manipulation in a healthy young woman: case report. improved in 26 (87%) patients who underwent surgical Neurosurgery 1996;39:1043-5. evacuation of hematomas, and early surgery within 12 hours [9] Lawton MT, Porter RW, Heiserman JE, et al. Surgical management of spinal epidural hematoma: relationship between surgical timing and correlated with better neurologic outcomes. Except for the neurological outcome. J Neurosurg 1995;83:1-7. timing of surgical decompression of the spinal cord, [10] Kreppel D, Antoniadis G, Seeling W. Spinal hematoma: a literature preoperative neurologic status and the speed of development survey with meta-analysis of 613 patients. Neurosurg Rev 26:1-49. American Journal of Emergency Medicine (2007) 25,72–79

www.elsevier.com/locate/ajem

Diagnostics Hypertrophic cardiomyopathy: electrocardiographic manifestations and other important considerations for the emergency physician

Brian S. Kelly MDa, Amal Mattu MDb, William J. Brady MDc,* aDepartment of Emergency Medicine, Mount Carmel Health System, Columbus, OH 43123, USA bDepartment of Emergency Medicine, University of Maryland School of Medicine, Baltimore, MD 21201, USA cDepartment of Emergency Medicine, University of Virginia, Charlottesville, VA 22911, USA

Received 16 April 2006; accepted 17 April 2006

Abstract Hypertrophic cardiomyopathy (HCM) is one of the most common inherited primary cardiac disorders and the most common cause of sudden cardiac death in young athletes. With advances in technology, it is now recognized that HCM affects individuals of all ages. Many patients with HCM will have a benign course with few symptoms. Some patients, however, possess risk factors that greatly increase the likelihood of sudden death if their disease remains undiagnosed. Therefore, it is imperative that emergency physicians be familiar with the symptoms and typical electrocardiogram manifestations of HCM. Three illustrative cases are presented with a review of the disease. D 2007 Elsevier Inc. All rights reserved.

1. Introduction thy accurately reflects the primary diagnostic feature of a thickened myocardium without ventricular dilation in In the mid-19th century, French pathologists first absence of conditions known to secondarily result in described the pathology of a cardiac disorder known today ventricular hypertrophy (eg, systemic hypertension and as hypertrophic cardiomyopathy (HCM). Because, in part, aortic stenosis). of the heterogeneity of this disease as well as technological Over the last 50 years, scientific interest has led to advances in genetic testing, HCM has become the preferred increased understanding of this extremely variable disease. name, replacing previous names such as hypertrophic Hypertrophic cardiomyopathy is currently recognized to be obstructive cardiomyopathy and idiopathic hypertrophic one of the most common inherited diseases of the heart, subaortic stenosis, which described the typical, but not occurring in as many as 1 in 500 individuals [1-4]. Early universal, features of left ventricular outflow tract (LVOT) studies suggested annual mortality rates as high as 3% to 6% obstruction and the predilection for asymmetrical hypertro- due largely to sudden cardiac death (SCD) in patients with phy of the ventricular septum. Hypertrophic cardiomyopa- HCM [1,3,4]. These early studies were primarily conducted at tertiary referral centers and included patients preferen- tially referred because of the severity of their symptoms [4]. * Corresponding author. E-mail addresses: [email protected] (A. Mattu)8 The annual mortality of unselected populations with HCM [email protected] (W.J. Brady). is estimated to be 1% to 2% [1,3,4].

2. Case presentations rotic disease or systemic drug toxicity. The presumed cause of death was HCM. 2.1. Case 1 2.2. Case 2 A 30-year-old man presented to the emergency department (ED) after an episode of severe lightheadedness and A 25-year-old man presented to the ED complaining of palpitations that began while he was running to catch a bus. severe palpitations and severe lightheadedness with near- The symptoms persisted for 20 minutes despite resting, then syncope while playing basketball. The symptoms lasted for resolved without any interventions. The patient had no only 5 to 10 minutes before resolving with rest. The patient medical history or history of drug use, no family history of denied any medical history, family history of medical early cardiac conditions, and no prior similar episodes. illness, drug use, or prior similar episodes. Upon arrival to Upon arrival to the ED, he was completely asymptomatic. the ED, he was asymptomatic. His vital signs and his His vital signs and physical examination results, including physical examination were normal. An ECG was obtained the cardiac examination results, were normal. An electro- (Fig. 2) and interpreted by both an emergency physician and cardiogram (ECG) was obtained (Fig. 1) and interpreted by a cardiologist as notable for left ventricular hypertrophy the emergency physician as bLVH, otherwise normal.Q The (LVH) and evidence of a prior lateral myocardial infarction. cardiology interpretation was bNormal sinus rhythm, left No previous ECGs were available for comparison (Fig. 3). ventricular hypertrophy, lateral infarct age undetermined.Q The patient was admitted to the hospital for further No previous ECGs were available for comparison. evaluation of the episode of near-syncope and the abnormal Laboratory studies including one set of cardiac markers ECG. All laboratory studies, including serial cardiac were ordered on the patient and were all normal. A chest enzymes, were normal. The chest radiograph was normal radiograph was also obtained and demonstrated a normal as well. Twenty-four hours of cardiac monitoring failed to heart size and no abnormalities. The patient demonstrated reveal any dysrhythmias. The patient was then referred for a no dysrhythmias while being monitored in the ED. He was Doppler echocardiogram. There was no evidence of wall then discharged, still asymptomatic, with instructions for motion abnormality that would normally be expected with a outpatient follow-up. Two days later, the patient collapsed prior myocardial infarction, and the patient’s ejection while once again running to catch a bus. Paramedics were fraction was normal. However, the study did reveal called immediately. When they arrived the patient was ventricular septal hypertrophy with some LVOT obstruction, unconscious, pulseless, and apneic. His cardiac rhythm was diagnostic of HCM. The patient was then prescribed ventricular fibrillation (Fig. 2). Standard advanced life b-blocker medications and discharged for outpatient fol- support measures used by the paramedics and subsequently low-up with a cardiologist. by ED staff were unsuccessful and the patient died. At 2.3. Case 3 autopsy, significant hypertrophy of the ventricular septum was noted. There was no evidence of a prior myocardial A 29-year-old man presented to the ED complaining of infarction, nor was there evidence of coronary atheroscle- palpitations and lightheadedness during sexual activity. The

Fig. 1 (Case 1) Normal sinus rhythm with large-amplitude QRS complexes consistent with LVH and nonspecific T-wave abnormality. Deep narrow Q waves are also present in the lateral leads I, aVL, V5, and V6. 74 B.S. Kelly et al.

Fig. 2 (Case 1) Ventricular fibrillation. symptoms persisted for 10 to 15 minutes before resolving referred the patient for immediate Doppler echocardiogra- without any intervention. The patient reported that he had phy, which confirmed the diagnosis of HCM. The patient presented to the ED on 2 prior occasions in the past month was then admitted to the cardiology service and has been for similar complaints: on the first visit, he experience a managed well on b-blocker medications. near-syncopal episode while mowing the lawn, and on the second visit he experienced a syncopal episode while watching television. Both of those episodes were preceded by palpitations. He was discharged from the ED after the 3. Discussion first visit after having an ECG and normal laboratory 3.1. Pathophysiology studies. He was admitted to the ED observation unit during the second presentation, whereupon he underwent serial The heterogeneity of HCM can largely be explained cardiac enzyme testing and cardiac monitoring. He was through examination of the genetic mutations responsible discharged after the negative workup and instructed to for disease production. Hypertrophic cardiomyopathy is follow up with a primary care physician, but he had not yet known to result from variable penetrance of the mendelian done so. The patient denied any medical history, family dominant inheritance of mutations in genes that encode history of early cardiac disease, or drug use. proteins of the cardiac sarcomere. Laboratory studies have On the current visit, the patient remained asymptomatic implicated more than 150 different mutations in multiple at the time of ED arrival. His vital signs and physical sarcomeric proteins with most the mutations occurring in the examination, including cardiac examination, were normal. following proteins: including b myosin heavy chain An ECG was obtained (Fig. 4). The ECG was interpreted by (~35%), troponin T (~15%), myosin-binding protein C the emergency physician as being highly suggestive of (~15%), and a-tropomyosin (b5%) [1,2,4]. Recently, HCM. Of note, there was no significant difference in the mutations of genes coding for mitochondrial enzymes have ECG when it was compared with the ECGs from the been linked to the HCM phenotype [5]. Studies performed patient’s 2 prior visits. Each of the past ECGs had been on pedigrees of different genotypes have led to the interpreted (by separate cardiologists) as bNormal sinus discovery that some mutations carry a benign prognosis, rhythm, left ventricular hypertrophy, lateral myocardial whereas other mutations result in early phenotypic disease infarction of indeterminate age.Q The emergency physician expression and worse prognosis.

Fig. 3 (Case 2) Normal sinus rhythm with LVH and deep narrow Q waves in the lateral leads I, aVL, V5, and V6. Hypertrophic cardiomyopathy 75

Fig. 4 (Case 3) Normal sinus rhythm with LVH and deep narrow Q waves in the lateral leads I and aVL.

Phenotypic manifestation of HCM is complicated further especially, sudden cardiac death in a young relative by factors such as modifier genes or environmental increases the likelihood of HCM. In addition, because of influences [2]. These complexities result in the different autosomal dominant inheritance, a family history of HCM phenotypic manifestations of HCM even within families greatly increases the risk for HCM. with the same genetic defect [3]. It therefore comes as no Patients with HCM may be completely asymptomatic, surprise that the clinical presentation and disease course is but will usually present with complaints related to LVOT highly variable. obstruction and diastolic dysfunction. Symptoms are These genetic mutations result in LVH, which cannot be nonspecific and include dyspnea, exercise intolerance, attributed to other commonly known etiologies. The dizziness, and near syncope or syncope. The severity of magnitude of hypertrophy is highly variable ranging from these symptoms may change throughout the day. Because mild (13-15 mm) to severe (N30 mm) [2]. The distribution of the dynamic nature of the degree of obstruction in HCM, of hypertrophy is also variable, but is most often asymmetric patients may be asymptomatic at rest, but may become and most pronounced in the anterior ventricular septum [2]. symptomatic even with minimal exertion. In addition, Other patterns of hypertrophy associated with HCM include symptoms may worsen after large meals or alcohol symmetric or concentric LVH (~20%) and apical (~10%). ingestion [3]. Spontaneous increase in ventricular hypertrophy is common Chest pain is more common in younger patients and may during periods of rapid growth and is most apparent in signal cardiac ischemia secondary to a mismatch in oxygen adolescence. Upon completion of adolescence, the magni- supply and demand to the thickened ventricle. In severe tude of hypertrophy often remains stable throughout adult hypertrophy, filling of the coronary arteries during diastole life but may paradoxically decrease. may become compromised. In addition, coronary arteriolar Postmortem histologic examination of myocardial sam- narrowing may limit blood flow reserve during periods of ples in HCM reveals loss of the normal parallel myocyte increased oxygen demand. architecture, a condition known as myocyte disarray. Other It is particularly important to inquire about near syncope findings include myocardial fibrosis, coronary artery abnor- or syncope in patients with suspected HCM. Lightheaded- malities including increased collagen deposition in the ness and syncope occur commonly in patients with HCM vessel walls, and narrowing of the arteriolar lumen. The and may be related to decreases in blood pressure due to features may serve as the substrate for arrhythmias in HCM. LVOT obstruction. Alternatively, syncope should alert the Other abnormalities associated with HCM include clinician to the possibility of episodic atrial or ventricular malformations of the mitral valve leaflets and severe arrhythmias, especially when it occurs in association with systolic anterior motion (SAM) of the mitral valve. These palpitations [3]. It is believed that children who present abnormalities can lead to abnormal leaflet coaptation, mitral with syncope secondary to suspected HCM represent a regurgitation, and additional LVOT obstruction. particularly high-risk group and should receive aggressive 3.2. Clinical presentation initial evaluation. Unfortunately, SCD may be the initial and only Family history is particularly important in patients with manifestation of disease, with the diagnosis of HCM made suspected HCM. A history of early-onset cardiac disease or, at postmortem examination after the death of a previously 76 B.S. Kelly et al.

Fig. 5 Normal sinus rhythm with LVH and deep narrow Q waves in the lateral leads V5 and V6. healthy individual [1,2,5,6]. Despite increased awareness, may be helpful for ruling out other conditions. Unlike many HCM remains one of the most common causes for sudden other forms of cardiomyopathy, HCM generally is associ- death in young athletes. ated with a normal heart size on chest x-ray because the Physical examination is more likely to be abnormal in hypertrophy is primarily limited to the septum. Because of patients with LVOT obstruction. Palpation of the carotid the complex, time-consuming, and cost-prohibitive nature of artery may reveal an initial brisk upstroke with subsequent genetic testing, its use is limited to investigational research collapse and secondary rise in late systole because of and is not available for routine clinical practice [2]. dynamic LVOT obstruction. Examination of the precordium Most patients with HCM will have an abnormal ECG may reveal a nondisplaced, hyperdynamic point of maximal [1,8,9]. It is important to note that the ECG may be abnormal impulse or a double apical impulse during systole as the even when echocardiography fails to demonstrate hypertro- hypertrophied noncompliant left ventricle contracts force- phy [3,8,10]. The ECG findings present in patients with fully. Less commonly, a triple apical impulse may occur HCM depend on the pattern and extent of myocardial secondary to the addition of a palpable atrial gallop. An S4 hypertrophy. Characteristic findings of LVH are common gallop may be heard during atrial systole. and include high-voltage R waves in the anterolateral leads The classic description of the cardiac murmur in HCM (V4,V5,V6, I, and aVL) [3,6,9,11]. Prominent R waves may referred to patients with subaortic obstruction only and may be present in leads V1 and V2 as well [12]. The most frequent be present in as few as 30% to 40% of patients [7]. In these abnormalities found are large-amplitude QRS complexes patients, auscultation may reveal a harsh crescendo- consistent with LVH and associated ST-segment and T-wave decrescendo systolic ejection murmur, which is best changes [1,13]. Deep, narrow Q waves are common in the appreciated at the apex and radiates to the sternal notch. inferior (II, III, aVF) and lateral (I, aVL, V5, and V6) leads in The intensity of the murmur is increased by maneuvers that patients with septal hypertrophy and may mimic myocardial decrease preload such as standing, Valsalva maneuver, or infarction [3,6,9,11]. The morphology of the Q waves—deep amyl nitrite administration. The murmur of HCM is unique and very narrow—is perhaps the most characteristic and from nearly all other cardiac murmurs in that it increases specific finding of HCM. Although often these Q waves are with Valsalva, making the performance of this maneuver mistaken for those of myocardial infarction, it should be very important during cardiac examination. In patients with recognized that infarction-related Q waves are wider, at least concurrent mitral valve abnormalities, a holosystolic 0.04 seconds (ie, at least 1 mm wide on standard ECG murmur that radiates to the axilla may be appreciated, paper). The Q waves of HCM are rarely this wide. In the reflecting mitral regurgitation. authors’ experience, deep narrow Q waves in the lateral leads 3.3. Diagnosis are much more common than in the inferior leads. These Q waves may appear in all of the lateral leads—I, aVL, V5, Once the diagnosis of HCM is considered, the workup and V6 (Fig. 1); they may predominate in just leads I and should generally include a chest x-ray, an ECG, and a aVL (Fig. 4); or they may predominate in just leads V5 and transthoracic echocardiogram with Doppler examination. Of V6 (Fig. 5). Although large-amplitude QRS complexes are these examinations, the ECG and echocardiogram have the the most common finding in HCM, deep narrow Q waves in highest yield, as the chest x-ray will likely be normal, but the lateral leads are the most specific finding and should Hypertrophic cardiomyopathy 77

Fig. 6 Apical hypertrophic cardiomyopathy. Normal sinus rhythm with LVH and deeply inverted T waves in the mid and lateral precordial leads. T wave abnormality is also present in the inferior leads in this patient. immediately prompt a high degree of suspicion when they echocardiogram examination after appropriate provocation are found. A less common variant of HCM involves isolated to evaluate for latent obstruction [3]. apical hypertrophy. Patients with apical HCM will frequently have deeply inverted T waves in the mid and lateral 3.4. Sudden cardiac death precordial leads (Fig. 6) [1,3,6,9]. Sudden cardiac death is the most feared complication of Shimizu et al examined the chronological ECG changes HCM. It is an evolving topic of intense investigation. Several in carriers with the deletion of lysine 183 (Lys183del) in the clinically useful risk factors have been identified. The cardiac troponin I (cTnI) gene compared with noncarriers predictive value of each factor is low; thus, a collective risk and found abnormal Q waves in leads I, II, III, aVF, V5, and profile, created by accurate assessment of individual risk V6 were frequently observed during the early teenage years. factors, is the best method to identify high-risk patients [5]. In contrast, carriers of the Lys183del in cTnI older than Individual risk factors for SCD include prior cardiac 20 years were more likely to demonstrate abnormal Q waves arrest, episodes of sustained ventricular tachycardia, recur- primarily in I, aVL, and other lateral leads. They concluded rent syncope, especially when associated with exertion, a that sensitivity, specificity, positive predictive value, and family history of 1 or more SCDs, extreme LVH (maximum negative predictive value of abnormal Q waves in the teens wall thickness N30-35 mm), the presence of LVOT gradient of their study population was 67%, 100%, 100%, and 78%, N30 mm Hg, an abnormal blood pressure response during respectively [13]. exercise, and the presence of nonsustained ventricular The transthoracic echocardiogram is one of the most tachycardia during ambulatory ECG monitoring [2,4,5,14]. useful tests in patients with HCM. It can document a Collectively, a risk profile with 2 or more individual risk number of findings including the degree of hypertrophy, factors suggests a high annual risk of SCD. systolic or diastolic dysfunction, the presence of LVOT or The risk profile can be used to guide therapy. In patients mid-ventricular obstruction, and mitral valve abnormalities with zero or one risk factor, the annual SCD rate is including the degree of SAM. Most patients with HCM have approximately 1%. This group represents most patients varying degrees of diastolic dysfunction. Systolic dysfunc- with HCM. In patients with 2 or more risk factors, the tion is rare in the early course of the disease, but as the annual incidence of SCD is significant and prophylactic disease progresses to generalized dilated cardiomyopathy therapy with automatic implantable cardioverter/defibrillator over the course of many years, systolic dysfunction (AICD) or amiodarone is warranted [1,4,5]. becomes common as well. The direction and degree of mitral regurgitation, if present, can also be documented. The 3.5. Management options addition of Doppler studies is important to correctly identify outflow obstruction and mitral regurgitation. Patients Hypertrophic cardiomyopathy remains the most common without obstruction at rest should undergo repeat Doppler cause of sudden death during athletic endeavors [2,4].Itis 78 B.S. Kelly et al. therefore extremely important to recommend that any Unlike surgical septal myomectomy, the long-term effects of patient suspected of having HCM avoid all intense, alcohol septal ablation are unknown. This procedure has competitive, or isometric activity. In general, light aerobic generally been reserved for patients who are older and have exercise such as walking or noncompetitive swimming, is a higher surgical risk [1]. considered safe [1]. Treatment strategies aim to improve or relieve the 3.6. Complications symptoms caused by LVOT obstruction or abnormal Arrhythmias are common in patients with HCM. ventricular function due to diastolic dysfunction. Tradition- Hypertrophic cardiomyopathy is associated with an in- ally, b-blocker medications have been the initial treatment creased incidence of Wolf-Parkinson-White syndrome. In of exertional dyspnea or chest pain [1,4]. The beneficial addition, as many as 10% to 40% of patients with HCM effects of b-blockers in patients with HCM are because of a experience chronic or paroxysmal atrial fibrillation [2,16]. decrease in heart rate, an improvement of ventricular filling Patients with HCM who develop atrial fibrillation are at during diastole, a decrease in myocardial oxygen demand, increased risk for systemic thromboembolism, heart failure, and a reduction of sympathetic tone [4]. Alternatively, and death [2,4,16]. Predictors of atrial fibrillation in patients verapamil has been used effectively in patients who either with HCM include left atrial enlargement and advanced age cannot tolerate a b-blocker (eg, severe asthmatic) or have no [16]. Although b-blockers or verapamil are usually effective initial response to beta-blockade. Verapamil should be at controlling the heart rate, patients with recurrent or prescribed with caution to patients with a large outflow chronic atrial fibrillation may benefit from amiodarone or gradient or pulmonary hypertension as its use in these atrioventricular node ablation [2]. Anticoagulant therapy patients may lead to severe hemodynamic compromise [4]. should be started on all patients without contraindications Disopyramide, a class IA antiarrhythmic drug, used most who display recurrence of even brief episodes of atrial often in conjunction with a b-blocker, has been used as an fibrillation [2,4]. alternative to verapamil and may relieve symptoms through Turbid blood flow in areas of obstruction as well as its negative inotropic properties [2,4]. If a patient develops mitral valve anomalies place patients with HCM at moderate signs of left ventricular systolic dysfunction (ie, congestive risk for infective endocarditis [17]. Appropriate antibiotic heart failure), afterload-reducing agents, b-blockers, digox- prophylaxis is warranted for patients undergoing dental or in, and judicious use of diuretics are indicated [1]. surgical procedures [1,17]. It is important to note that the administration of a b-blocker or verapamil to patients with HCM does not protect them from sudden death [2,4]. These agents should 4. Conclusion be used only to improve symptoms. Nonpharmacologic therapies are reserved for the minority Hypertrophic cardiomyopathy is one of the most common of patients with HCM who fail medical management. inherited primary cardiac disorders. Despite previous over- Surgical septal myomectomy has been performed to relieve estimations of the annual mortality attributed to HCM, it LVOT obstruction since its introduction in 1958. Currently, a remains the most common cause of SCD in young athletic modified Morrow procedure termed extended myomectomy individuals. The presenting symptoms associated with HCM is performed at specialized centers with the additional goal of are variable owing in part to the significant genetic correcting the SAM of the mitral valve that contributes to the heterogeneity. Obtaining a detailed family history is essential pathophysiology [15]. The benefits to surgical septal to aid in risk assessment. The ECG will be abnormal in most myomectomy are significant and include relief of symptoms patients with HCM. The most specific ECG abnormalities and improved quality of life; however, the procedure that emergency physicians should be aware of are concurrent requires a sternotomy, cardiopulmonary bypass, and carries LVH with deep narrow Q waves in the lateral leads. The a 1-month mortality of approximately 0%to 6% [15]. diagnosis is currently confirmed by Doppler echocardiogra- Alcohol septal ablation is a newer procedure that phy, although ECG changes may be present before the involves injection of a small amount of absolute alcohol echocardiogram is abnormal. into a septal perforator coronary artery after successful All patients suspected of having HCM should be referred cannulation using a percutaneous transluminal catheter to a cardiologist for further testing and risk assessment. technique. This procedure produces a modest infarction of Patients must be instructed to refrain from intense exercise the myocardial septum that results in subsequent akinesis, while awaiting confirmation of the diagnosis. Medical septal thinning, and fibrosis [1,15]. Follow-up studies therapy aims to improve the symptoms associated with indicate significant reduction in LVOT pressure gradients. LVOT obstruction and diastolic dysfunction, but does not Complications include a short-term mortality of 0% to 4%, protect against SCD with the exception of amiodarone. undesired backflow of alcohol into the left anterior Patients with medically refractive symptoms may be descending coronary artery with subsequent larger infarc- considered for a surgical septal myomectomy or alcohol tion, and a 9% to 38% incidence of complete heart block septal ablation. High-risk patients need prompt referral and necessitating placement of a permanent pacemaker [1,15]. evaluation for placement of a potentially life-saving AICD. Hypertrophic cardiomyopathy 79

References [10] Konno T, Shimizu M, Ino H, et al. Diagnostic value of abnormal Q waves for identification of preclinical carriers of hypertrophic cardiomyopathy based on molecular genetic diagnosis. Eur Heart J [1] Popjes ED, Sutton M. Hypertrophic cardiomyopathy. Pathophysiolo- 2004;25(3):246-51. gy, diagnosis and treatment. Geriatrics 2003;58(3):41-6. [11] Cardiomyopathies. In: Wagner GS, editor. Marriott’s: practical [2] Maron BJ. Hypertrophic cardiomyopathy. Lancet 1997;12(350): electrocardiography. 10th ed. Philadelphia7 Lippincott Williams & 127-33. Wilkins; 2001. p. 205-6. [3] Wigle ED. The diagnosis of hypertrophic cardiomyopathy. Heart [12] Mattu A, Brady WJ, Perron AD, et al. Prominent R wave in V1: 2001;86:709-14. electrocardiographic differential diagnosis. Am J Emerg Med 2001; [4] Spirito P, Seidman CE, McKenna WJ, et al. The management of 19:504-13. hypertrophic cardiomyopathy. N Engl J Med 1997;336:775-85. [13] Shimizu M, Ino H, Yamaguchi M, et al. Chronologic electrocardio- [5] Frenneaux MP. Assessing the risk of sudden cardiac death in a patient graphic changes in patients with hypertrophic cardiomyopathy with hypertrophic cardiomyopathy. Heart 2004;90:570-5. associated with cardiac troponin I mutation. Am Heart J 2002; [6] Khan IA, Ajatta FO, Ansari AW. Persistent ST segment elevation: 143(2):289-93. a new finding in hypertrophic cardiomyopathy. Am J Emerg Med [14] Spirito P, Bellone P, Harris K, et al. Magnitude of left ventricular 1999;17(3):296-9. hypertrophy and risk of sudden death in hypertrophic cardiomyopathy. [7] The Cardiomyopathy Association. Which Cardiomyopathy? Available N Engl J Med 2000;342:1778-85. at: http://www.cardiomyopathy.org/html/which_card_hcm.htm [15] Surgery for obstructive hypertrophic cardiomyopathy. Available at: [Accessed April 28, 2005]. http://www.hcmny.org/hcmprog/cardiomyopathy.htm [Accessed May [8] Dipchand AI, McCrindle BW, Gow RM, et al. Accuracy of surface 2, 2005]. electrocardiograms for differentiating children with hypertro- [16] Ozdemir O. P-wave durations as a predictor for atrial fibrillation phic cardiomyopathy from normal children. Am J Cardiol 1999;83: development in patients with hypertrophic cardiomyopathy. Int J 628-30. Cardiol 2004;94(2-3):163-6. [9] Surawicz B, Knilans TK. Hypertrophic cardiomyopathy. Chou’s: [17] Dajani AS, Taubert KA, Wilson W, et al. Prevention of bacterial electrocardiography in clinical practice, 5th ed. Philadelphia7 W.B. endocarditis: recommendations by the American Heart Association. Saunders Company; 2001. p. 258-61. JAMA 1997;277(22):1794-801. American Journal of Emergency Medicine (2007) 25,80–86

www.elsevier.com/locate/ajem

Clinical Notes Ultrasound-guided hip arthrocentesis in the ED

Kalev Freeman MD, PhD, Andreas Dewitz MD, William E. Baker MD*

Department of Emergency Medicine, Boston Medical Center—Boston University School of Medicine, Boston, MA 02118, USA

Received 30 March 2006; revised 27 July 2006; accepted 6 August 2006

Abstract In patients presenting with atraumatic joint pain and swelling, diagnosis is typically made by synovial fluid analysis. Management of an acute suspected hip joint arthritis can present a challenge to the emergency physician (EP). Hip joint effusions are somewhat more difficult to identify and aspirate than effusions in other joints that are commonly managed by EPs. Identification and aspiration of a hip joint effusion under ultrasound guidance is a well-established procedure in the fields of orthopedic surgery and interventional radiology. Here, we report 4 cases of ultrasound-guided hip arthrocentesis at the bedside by EPs; relevant technical details of the procedure are reviewed. These cases demonstrate the feasibility of ultrasound-guided hip arthrocentesis in the emergency department (ED) by EPs. With increasing availability of bedside ultrasound in the ED, suspected hip joint arthritis or infection may be evaluated and managed by the trained EP in a fashion similar to other joint arthritides. D 2007 Elsevier Inc. All rights reserved.

1. Introduction aspirations is most advantageous, early arthrocentesis is no longer debated: bPrompt evaluation of purulent joint fluid Acute hip pain may arise from a variety of inflammatory appears to be crucial both for preservation of articular processes involving the hip joint. Of particular concern for cartilage and for resolution of the infection [1].Q the emergency physician (EP) is the patient with a suspected Evaluation of suspected hip joint sepsis presents a septic hip, an orthopedic emergency because of the potential particular problem for the EP because it is inherently more for rapid and permanent joint destruction. Other possible difficult to perform arthrocentesis of the hip than other causes of hip pain in adults include trauma, intraarticular peripheral joints. Delayed treatment of a septic hip joint may fracture or hematoma, crystal deposition, degenerative lead to serious sequelae, including avascular necrosis of the arthritis, avascular necrosis, osteomyelitis, bone tumors, femoral head, proximal femoral and/or pelvic osetomyelitis, and connective tissue disorders. Acute septic arthritis of any and systemic sepsis. When the diagnosis of septic hip joint, including the hip joint, is typically managed with arthritis is considered, immediate orthopedic surgery con- adequate drainage of the infected joint space, administration sultation is recommended [2]. Typically, the hip joint is of antibiotics, and joint rest in a position of stability [1]. aspirated either at the bedside or in the operating room by While controversy exists over whether open surgical large-bore needle aspiration, arthroscopy, or open drainage. drainage, arthroscopic drainage, or treatment by multiple Ultrasound guidance has been used to assist in arthrocentesis of the hip joint for over 20 years, and its use has been described in the emergency medicine literature [3,4]. * Corresponding author. Department of Emergency Medicine, Boston Medical Center—Boston University School of Medicine, Boston, MA However, careful review of the literature review reveals 02118, USA. only one reported case documenting ultrasound-guided hip E-mail address: [email protected] (W.E. Baker). arthrocentesis performed by an EP [5].

Fig. 3 Line diagram of image obtained when scanning as described in Figure 1; femoral head, and femoral neck, with joint capsule in close opposition. Space between arrows indicates anterior synovial recess. array and sector transducers. The patient should be placed in a supine position, with the ultrasound transducer located just below the inguinal ligament, lateral to the femoral Fig. 1 Demonstration of ultrasound guided hip arthrocentesis vessels, with the transducer orientation marker angled with transducer orientation marker angled superomedially, drape 8 and probe cover omitted for purpose of demonstration. superomedially at an approximately 45 angle, aiming toward the umbilicus (Fig. 1). This scanning position This report reviews 4 cases of ultrasound-guided hip places the transducer such that the long axis of the arthrocentesis performed by EPs in our emergency department ultrasound beam is aligned with the long axis of the (ED) and includes a brief discussion of the technical aspects femoral neck (Fig. 2). The acetabular labrum, femoral head, of the procedure and its use in the ED evaluation and and neck (collum) provide strong echo reflections that are management of the patient with atraumatic hip pain. readily identified on the ultrasound monitor at about 3 to 5 cm below the skin surface. As per usual scanning convention, the curved echo of the bony femoral head will 2. Methods be located on the left side of the sonogram. The articular capsule of the hip joint is a moderately reflective fibrous Sonographic evaluation of the hip may be successfully structure extending from the acetabular labrum to its point accomplished with a variety of transducers, including linear of fixation on the lower femoral neck. Synovial fluid cannot typically be seen in a normal hip joint (Fig. 3). When a hip effusion is present, it will appear as an anechoic or hypoechoic collection that elevates the joint capsule. The most prominent area of fluid accumulation occurs in the area termed the anterior synovial recess,a potential space located just anterior to the femoral neck and immediately posterior to the articular capsule (Fig. 4). This is the area that should always be visualized in order to confirm or exclude a hip effusion. Small effusions are more easily identified with the knee in slight flexion and the hip in slight internal rotation; this is the recommended patient

Fig. 2 Anterior view of hip; parallel lines represent the Fig. 4 Space between arrows indicates joint effusion in anterior ultrasound plane, perpendicular to the long axis of the femoral synovial recess, with displacement of anterior joint capsule from neck with arrows indicating probe orientation. the femoral neck. 82 K. Freeman et al.

Fig. 5 See above for details. Key: a, acetabular labrum; fh, femoral head; fn, femoral neck; e, effusion; jc, joint capsule; n, needle tip. position for scanning usually suggested in the literature [6]. technique occurs in the patient with chronic rheumatic Extension of the knee and external rotation of the leg, in disease, where ultrasound measurement of the anterior contrast, will tighten the articular capsule and displace collum to capsule distance may not distinguish a thickened synovial fluid posteriorly [6]. One report suggested placing capsule from thickened synovium [10]. Once an effusion the hip in extension and abduction improved ultrasound- has been identified, it may be aspirated with ultrasound guided visualization of the joint, but this positioning was guidance or mapping. The transducer should first be placed compared to neutral position rather than the more frequent- just below the inguinal ligament in a transverse orientation, ly described flexion with internal rotation [7]. with the probe marker oriented to the left. The goal is to The inflamed hip joint should always be examined in clearly identify the location of the medially located conjunction with a sonogram of the contralateral hip. common femoral vein, artery, and nerve in relation to the Diagnosis of a hip effusion is based upon measurement joint space in order that these structures may be avoided of the widest distance of the anechoic or hypoechoic fluid during the aspiration. The hip should then be prepped and strip located between the anterior surface of the femoral draped for sterile needle aspiration. The ultrasound neck and the echogenic joint capsule [8]. Criteria for transducer may be placed in a sterile glove or covered defining an effusion include a convex bulging joint capsule with a commercially available sterile sheath; aspiration can with a fluid strip of greater than 5 mm or greater than 2 mm then be performed with an 18-gauge spinal needle and a of fluid strip asymmetry, compared to the contralateral hip 20-mL syringe under direct visualization. The procedure is [9]. An important caveat to this standard measurement generally performed with topical or local anesthesia. With Ultrasound-guided hip arthrocentesis in the ED 83 the transducer orientation marker now aimed superiome- crystals were seen. Gram stain was negative for organisms. dially along the long axis of the femoral neck, and while The patient reported decreased pain and improved range of directly visualizing the hip effusion, the spinal needle is motion after arthrocentesis. He was started on intravenous advanced within the long axis of the ultrasound beam from antibiotics, treated with analgesics and admitted to the its skin entry site at the inferior border of the transducer. hospital. Rheumatology recommended continued antibiot- The characteristic reverberation artifact of the advancing ics, and further workup included an erythrocyte sedimenta- spinal needle can be used to guide the needle tip directly tion rate of 14 mm/h, a C-reactive protein of less than into the largest region of fluid accumulation at the level of 0.5 mg/dL, and a negative Lyme titer. Final culture results the anterior synovial recess. The synovial fluid is then were negative, and the patient was comfortable and aspirated and sent for analysis. ambulatory on a regimen of nonsteroidal anti-inflammatory Four cases of patients who had arthrocentesis of a hip medication. He was presumed to have had reactive joint in the ED, performed by an EP under ultrasound monoarticular arthritis, likely secondary to a viral gastroen- guidance and recorded on SVHS videotape or digital teritis, and was discharged from the hospital. format, were identified. Cases where an effusion of the hip joint was identified by ultrasound but not aspirated by 3.2. Case 2 the EP were not included. Ultrasound examination of the hips was performed with either a Biosound 3000D A 42-year-old man was referred from a community equipped with a 5-MHz linear array probe (Biosound health center with a chief complaint of left hip pain. His pain Esaote, Indianapolis, Ind) or a Hitachi EUB 525 equipped began 3 days before arrival and was associated with with a 3.5 MHz curvilinear probe (Hitachi Medical Systems subjective fever. There was no associated trauma. His America, Twinsburg, Ohio). In each case, the patient was medical history was significant for type I diabetes, placed in a supine position, and both hips were assessed by hypertension, chronic renal insufficiency, and chronic ultrasound to identify the presence of an effusion. Where a anemia. He had been evaluated by his primary care provider fluid collection in the affected joint space was visualized, 3 months before for bilateral hand pain and swelling and had the appropriate hip was prepped and draped, local been prescribed acetaminophen with codeine. Temperature anesthesia was administered, and an 18-gauge spinal needle in the ED was 98.98F. Physical examination was significant was used for arthrocentesis under real-time ultrasound for tenderness over the hip joint as well as pain with range guidance, as previously described [6,11,12]. Fig. 5 dem- of motion. Plain radiography of the left hip was normal. His onstrates representative findings from each of the described complete blood count and serum glucose were normal; the cases. Cases without recorded ultrasound images of the hip BUN and creatinine were elevated at 56 and 8.4 mg/dL, joint were not included. respectively. The erythrocyte sedimentation rate was elevated at 116 mm/h. Bedside ultrasound demonstrated a hip effusion with a thickened joint capsule, and 8 mL of cloudy fluid was 3. Illustrative cases obtained on arthrocentesis. Synovial fluid analysis revealed a WBC of 78000/mm3.Nocrystalswereseen,andGramstain 3.1. Case 1 was negative for organisms. The patient was treated with A 28 year-old male presented to the ED complaining of cefazolin and ketorolac and admitted to the hospital for 2 days of worsening pain in the right thigh. The pain started further workup. Antinuclear antibody titer was mildly posi- upon awakening, extended throughout the proximal femur, tive at a ratio of 1:80, and rheumatoid factor was elevated at b and was associated with a btightQ feeling in the hip joint. 63 IU/mL (normal 20 IU/mL). Lyme serology was negative. There had been no recent trauma or repetitive activities. He He remained afebrile; his pain and ambulation improved with reported abdominal bloating, diarrhea, and fever on the day narcotics and nonsteroidal anti-inflammatory medication. before the onset of pain, but the associated symptoms had Final synovial fluid culture results were negative, and he resolved. He had no other medical history. He was afebrile, waspresumedtohavehadareactivemonoarthropathy.He and physical examination was significant only for marked was discharged to follow-up with the rheumatology service. pain with even slight range of motion of the right hip joint. 3.3. Case 3 Plain radiography of the right hip was normal, as were the complete blood count and serum glucose. A 42-year-old female was transferred by ambulance Sonography revealed a hip joint effusion and a thickened from a local community health center complaining of right joint capsule, and arthrocentesis yielded approximately lower quadrant abdominal pain that had begun upon 20-mL of cloudy, straw-colored fluid. Synovial fluid awakening that day. She pointed to her right inguinal analysis revealed a white blood cell (WBC) count of region, and the referring physician was concerned about a 42000/mm3, with 90% neutrophils, 1% lymphocytes and possible inguinal hernia or ovarian torsion. Her pain was 9% monocytes, a joint fluid glucose of 65 mg/dL, lactate sharp, radiated to the mid thigh, and was worse with dehydrogenase of 548 IU/L, and a protein of 5.2 g/dL. No ambulation. She was unable to sit or stand comfortably, but 84 K. Freeman et al. the pain was somewhat lessened by a stooped posture. She were tender to palpation. Left knee extension, hip flexion, denied fever, nausea, trauma, or prior episodes of similar and internal rotation were limited by pain. Sensation and discomfort. Her medical history was significant only for a reflexes were intact. Pulses were easily palpable and cesarean section. On physical examination, she had normal symmetrical. His complete blood count was notable for a vital signs but appeared tearful and uncomfortable. Her WBC of 12.1 Â 103/lL with a normal differential. Plain abdomen was soft and nontender. She had pain on radiographs of the left knee, hip, and pelvis showed palpation of the right inguinal region and significant sclerosis and deformity of the left femoral head without discomfort with right hip flexion or internal rotation. fracture or dislocation. A bedside ultrasound of the left hip Pulses were symmetric in both lower extremities, and joint revealed a large effusion, and bedside arthrocentesis reflexes, sensation, and motor function were intact. Pelvic yielded approximately 20 mL of dark yellow fluid. examination was normal. The patient had minimal relief Synovial fluid analysis revealed a WBC count of 48500/ with multiple does of morphine. A complete blood count mm3 with 95% neutrophils, 0% lymphocytes, and 5% was performed and demonstrated a peripheral white blood monocytes; a joint fluid glucose of 45 mg/dL; a lactate cell count of 11.8 Â 103/lL, hemoglobin level of 13.2 g/dL, dehydrogenase of 864 IU/L; and a protein of 5.5 g/dL. His hematocrit of 38.3%, and a platelet count of 390 Â 103/lL. erythrocyte sedimentation rate was 49 mm/h. The patient Serum glucose was 96 mg/dL. Plain films of the hip and received an initial dose of IV antibiotics and was admitted lumbar spine were normal. A computed tomographic scan to the hospital. Although plasma cells were not seen on the of the abdomen/pelvis was remarkable only for the peripheral blood smear, multiple myeloma was considered presence of a right hip effusion. as a possible diagnosis. The orthopedic surgery service Bedside ultrasound was then used to localize the was consulted and felt that the radiographic images and effusion and approximately 10-mL of turbid, yellow- clinical picture were most consistent with avascular colored fluid was obtained under ultrasound guidance. necrosis of the femoral head. The patient’s pain was Synovial fluid analysis revealed a WBC count of 43600/ controlled with nonsteroidal anti-inflammatory medica- mm3 with 85% neutrophils, 0% lymphocytes, and 15% tions, and his final synovial fluid cultures were negative. monocytes; a joint fluid glucose of 34 mg/dL; lactate His peripheral white blood cell count decreased to 4.7 Â dehydrogenase of 854 IU/L; and protein of 4.2 g/dL. No 103/lL, and he was soon ambulating comfortably with the crystals were seen, and Gram stain was negative for assistance of a cane. He was discharged with a diagnosis organisms. The patient received an initial dose of intrave- of avascular necrosis of the left femoral head and possible nous antibiotics and was admitted to the hospital. Antinu- multiple myeloma. He was started on metoprolol and clear antibody (ANA) screen and Lyme disease serology scheduled to follow-up with both the orthopedic surgery were negative. She remained afebrile, and final synovial and internal medicine services. fluid culture results were negative. She was believed to have had reactive monoarticular arthritis, and her antibiotics were discontinued. By hospital day 3, she was ambulatory 4. Discussion without pain, and she was discharged on ibuprofen to follow-up with rheumatology. Prior to this manuscript, there has been only 1 published 3.4. Case 4 case report of hip arthrocentesis performed by an EP [5]. The case involved a patient with acute hip pain and fever. A 51-year-old man with a history of hypertension, Orthopedic surgery was consulted, and after several asthma, and alcoholism presented to the ED complaining unsuccessful attempts at blind aspiration, hip arthrotomy of left knee pain. He reported sustaining an injury during a and irrigation in the operating room was planned. The collision with a bicyclist approximately 6 months before, operation was avoided by needle aspiration of the hip which had resulted in intermittent discomfort in his left effusion under direct ultrasound guidance by the EP, and knee. During the week before his presentation, his knee synovial fluid analysis subsequently revealed calcium pain had returned, with radiation of the pain to his left hip pyrophosphate crystals consistent with a diagnosis of and back. The patient reported that the pain had gradually pseudogout. Another published report documents the use progressed to the point where he could only get around his of bedside ultrasound by an EP to detect the presence of a apartment by crawling on the floor. His pain was constant, hip effusion; subsequent arthrocentesis on this patient was exacerbated by movement, and extended from the knee to performed under fluoroscopic guidance, however, and the his lower back. He denied fever, nausea, weakness, patient was taken to the operating room for open drainage paresthesias, or incontinence. He last used his inhaled [13]. Review of the medical literature yielded no other albuterol 1 day before arrival and had not been taking his published cases of hip ultrasound by EPs, although the antihypertensive medications. On physical examination, diagnosis of hip effusion by ultrasound and ultrasound- he was afebrile, with a pulse of 83 and blood pressure of guided arthrocentesis of the hip joint has previously been 173/102 mm Hg. He was in no acute distress and had discussed in the emergency medicine literature [4] as well as ethanol on his breath. His left knee, femur, and lateral hip in an ultrasound textbook written for EPs [3]. Ultrasound-guided hip arthrocentesis in the ED 85

In the cases reported here, each of the 4 patients had arthrocentesis of the hip can contribute to this goal. As with historical or physical examination features concerning for other acutely inflamed, painful monoarthropathies, arthro- possible acute septic arthritis. Emergency physician–per- centesis of the hip is important, first, to rule out septic formed ultrasound revealed an effusion of the affected hip arthritis [17]. Second, if a septic arthritis is present, adequate joint in each case, and bedside arthrocentesis was success- drainage through arthrocentesis may be therapeutic, and if fully performed under direct sonographic visualization by clinical improvement occurs on antibiotics, further interven- 2 of the authors (WB, cases 1 and 2; AD, cases 3 and 4). tion may not be necessary [1]. Third, regardless of cause, There were no complications. In each case, synovial fluid effusions are painful because of pressure exerted on the analysis demonstrated evidence of inflammation, with synovial capsule; release of this pressure by arthrocentesis polymorphonuclear leukocytes seen on microscopy, but often provides immediate pain relief and improvement in without organisms on gram stain. All 4 patients were range of motion. This has been well documented in cases of admitted, and final culture results were uniformly negative. hip effusions [6,14]. Fourth, synovial fluid analysis may lead Three patients were discharged with the diagnosis of to alternative diagnoses, such as crystal arthropathies, reactive synovitis of unknown etiology, and the fourth rheumatoid arthritis, or reactive arthritis [5]. was felt to have had avascular necrosis of the femoral head Although children and the elderly are the patient groups and a secondary reactive hip effusion. All 4 patients who will most commonly present with atraumatic hip pain, showed clinical improvement within 2 to 3 days, without acute septic arthritis in the pediatric population carries a operative intervention. higher risk of severe complications. Because of the These 4 cases demonstrate the feasibility of ultrasound- developing femoral head circulation, the pediatric septic guided hip arthrocentesis in the ED. The safety and hip is at high risk for avascular necrosis, and epiphyseal tolerability of this procedure in the hands of radiologists involvement with potential separation and pathologic dislo- has previously been well-documented [6,14]. Visualization cation may occur. Although transient synovitis is the most of a hip joint effusion by ultrasound was first reported in common final diagnosis in pediatric hip pain, the early 1980 [15]. Although ultrasound-guided arthrocenteses of the presentation of this childhood diagnosis is nearly identical to hip were initially performed under general anesthesia or early septic arthritis: spontaneous, progressive pain in the conscious sedation, by 1990, a case series of ultrasound- hip, groin, or thigh; difficulty with weight-bearing or guided hip aspirations in children under local anesthesia was ambulation; fever; and irritability. Clinical prediction algo- published [14]. Since then, published series of as many as rithms for differentiation between the diagnosis of transient 800 aspirations have been described by radiologists, without synovitis and acute septic arthritis of the hip have not been a single complication, including iatrogenic infection [6]. supported when parameters such as the temperature, The potential risk of damage to the femoral neurovascular leukocyte count, erythrocyte sedimentation rate (ESR), complex is small, and with careful visualization of the and weight-bearing status are relied upon [17]. To defini- relationship between of the joint space and the femoral tively diagnose septic arthritis of the hip, analysis and artery, vein and nerve this risk is negligible. Although there culture of synovial fluid is necessary. Hip ultrasono- is a risk of pain from the arthrocentesis, multiple reports graphy and arthrocentesis are generally advocated as suggest that the procedure is well tolerated, with only adjunctive diagnostic modalities in the assessment of topical anesthesia or local anesthetic infiltration used before pediatric hip pain [16-18]. the aspiration [6,14,16]. Other risks are bleeding, vasovagal Protocols for evaluation of acute hip pain with ultrasound reaction, and failure to obtain fluid. assistance have been established in the pediatric population. Emergency physician–performed ultrasound of the pain- Ultrasound-guided evaluation of the irritable pediatric hip ful hip at the bedside can expedite early diagnosis and allows for early diagnosis of septic arthritis and other serious treatment. Ultrasound of the hip is the preferred modality for pathology, avoids unnecessary hospital admissions, and demonstration of fluid collections around the hip joint, and helps relieve pain by joint decompression [16].Ina early detection of an effusion in the evaluation of a patient prospective study of 50 pediatric patients presenting with with atraumatic hip pain may direct the differential diagnosis acute hip pain, a pelvic radiograph and initial blood work [4]. Making a timely diagnosis of possible septic arthritis is including a complete blood count with differential, as well as most important, however, because irreversible hip destruction an erthryocyte sedimentation rate, were obtained for each begins within 2 to 4 days [1,16]. In cases of acute hip pain patient [16]. One fracture was identified. Bilateral hip with an effusion, clinical prediction algorithms are inade- ultrasound was subsequently performed, and in 36 patients, quate to predict infection, and arthrocentesis is clinically an effusion was identified, and immediate ultrasound-guided indicated. Although on-call orthopedics or interventional aspiration was performed. Patients with significant systemic radiology consultants are capable of performing hip arthro- symptoms, with or without an effusion present, were centesis, their availability may be limited and may delay the admitted for further evaluation. In one case, Gram stain diagnosis, especially during boff-hoursQ. Timely diagnosis was positive, and the patient was admitted for parenteral and antibiotic administration are critical to outcome in septic antibiotics and monitoring. When the immediate Gram arthritis, and the ability of EPs to perform bedside stain was negative, and there were no systemic symptoms, 86 K. Freeman et al. the patient was discharged, provided that provision could References be made for urgent recall, should cultures later prove positive. Patients without an effusion on ultrasound imag- [1] Park AL, Dlabach JA. Infectious arthritis. Campbell’s operative ing or without significant systemic symptoms were also orthopaedics. St. Louis, MO7 Mosby; 2003. p. 687-700. discharged. In either case, follow-up was scheduled in 1 week [2] Burton JH. Acute disorders of the joints and bursae. In: Tintinalli J.E., Kelen GD, Stapczynski JS, editors. Emergency medicine. Tintinalli, for repeat ultrasound and for bone scintigraphy if pain New York7 McGraw-Hill; 2004. p. 1795-801. persisted. With the increasing availability and use of ED [3] Dewitz A, Frazee BW. Soft tissue applications. In: Ma OJ, Mateer JR, bedside ultrasound, a similar protocol might logically be editors. Emergency ultrasound. Tintinalli, New York7 McGraw-Hill; extended to the evaluation of adult patients in the ED. 2003. p. 361-90. Clinical validation of such protocols should be the subject of [4] Valley VT, Stahmer SA. Targeted musculoarticular sonography in the detection of joint effusions. Acad Emerg Med 2001;8(4):361-7. future investigation. [5] Smith SW. Emergency physician–performed ultrasonography-guided The 4 cases presented here demonstrate how bedside hip arthrocentesis. Acad Emerg Med 1999;6(1):84-6. ultrasound can be incorporated as part of a rational approach [6] Berman L, Fink AM, Wilson D, et al. Technical note: identifying and to the ED management of atraumatic hip pain and how aspirating hip effusions. Br J Radiol 1995;68(807):306-10. detection and aspiration of hip effusions may be included [7] Chan YL, Cheng JC, Metreweli C. Sonographic evaluation of hip effusion in children. Improved visualization with the hip in extension among the armamentarium of bedside procedures performed and abduction. Acta Radiol 1997;38(5):867-9. by the trained EP. [8] Koski JM, Anttila P, Hamalainen M, et al. Hip joint ultrasonography: After the preparation of this manuscript, an additional correlation with intra-articular effusion and synovitis. Br J Rheumatol case of atraumatic hip pain presented in our emergency 1990;29(3):189-92. department. The patient was a 42 year-old male with 2 days [9] Moss SG, Schweitzer ME, Jacobson JA, et al. Hip joint fluid: detection and distribution at MR imaging and US with cadaveric of progressive hip pain, worse with motion, without correlation. Radiology 1998;208(1):43-8. associated fever or nausea. Ultrasonography revealed a hip [10] Soini I, Kotaniemi A, Kautiainen H, et al. US assessment of hip joint joint effusion (not shown), and arthrocentesis was per- synovitis in rheumatic diseases. A comparison with MR imaging. Acta formed by the emergency physician. The synovial fluid was Radiol 2003;44(1):72-8. found to have greater than 95,000 white blood cells, [11] Mayekawa DS, Ralls PW, Kerr RM, et al. Sonographically guided arthrocentesis of the hip. J Ultrasound Med 1989;8(12):665-7. antibiotics were started, and orthopedic surgery was con- [12] Fessell DP, Jacobson JA, Craig J, et al. Using sonography to reveal sulted. The patient underwent operative debridement and and aspirate joint effusions. AJR Am J Roentgenol 2000;174(5): irrigation of the infected hip joint. Culture of the synovial 1353-62. fluid later revealed Streptococcus pneumoniae. The patient [13] Lee FM, Fox JC. Diagnosis of hip effusion with bedside ultrasound in recovered uneventfully. This case, along with the 4 the emergency department. Calif J Emerg Med 2003;4(1):51-4. [14] Hill SA, MacLarnon JC, Nag D. Ultrasound-guided aspiration described in the report, show the feasibility of using for transient synovitis of the hip. J Bone Joint Surg Br 1990;72(5): ultrasound-guided hip arthrocentesis in the ED to evaluate 852-3. possible septic arthritis of the hip. [15] Seltzer SE, Finberg HJ, Weissman BN. Arthrosonography–technique, sonographic anatomy, and pathology. Invest Radiol 1980; 15(1):19-28. 5. Summary [16] Fink AM, Berman L, Edwards D, et al. The irritable hip: immediate ultrasound guided aspiration and prevention of hospital admission. Ultrasound-guided arthrocentesis of the hip joint is a Arch Dis Child 1995;72(2):110-3. well-established procedure, and the cases reported here [17] Luhmann SJ, Jones A, Schootman M, et al. Differentiation between illustrate its feasibility in the hands of trained EPs. In a septic arthritis and transient synovitis of the hip in children with clinical prediction algorithms. J Bone Joint Surg Am 2004;86-A(5): fashion similar to the workup of other painful monoarthro- 956-62. pathies, this procedure allows for timely diagnosis and [18] Bickerstaff DR, Neal LM, Booth AJ, et al. Ultrasound examination of management of the patient with of atraumatic hip pain. the irritable hip. J Bone Joint Surg Br 1990;72(4):549-53. American Journal of Emergency Medicine (2007) 25, 87–123

www.elsevier.com/locate/ajem

Case Reports Unusual presentation of herpes simplex virus the CSF revealed no abnormality and protein and glucose encephalitis: bilateral thalamic involvement and normal levels were normal. imaging of early stage of the disease On the third day of hospitalization, the test on CSF revealed 40 white blood cells with 30% segmented Herpes simplex virus encephalitis (HSE) is one of the neutrophils and 70% lymphocytes, erythrocytes, protein most common forms of encephalitis characterized by level of 65 mg/dL, and glucose level of 45 mg/dL. The CSF progressive cerebral necrosis and edema [1,2].Herpes culture and pharyngeal cultures were sterile. The CSF was simplex virus (HSV) 1 causes more than 90% of all cases negative for adenovirus, measles virus, the coxsackie virus, of HSE. HSV-2 encephalitis occurs mostly during the and Epstein-Barr virus–specific antibodies. HSV-1 IgM was neonatal period. present in both serum and CSF, and the diagnosis was Patients with HSE frequently present with a rapid onset confirmed by detection of HSV-1 DNA in the CSF. of neurologic symptoms and with rapid progression from A repeat CSF examination demonstrated a positive result confusion to coma within a few days [3,4]. Antiviral therapy for antibody and PCR results were positive for HSV-1 DNA. must be started during the acute stage to achieve any clinical The next morning, cranial MRI was performed. Lesions in benefit [5]. Magnetic resonance imaging (MRI) is a more the bilateral thalamus and temporal and frontal area lesions sensitive noninvasive test for early diagnosis of HSE than were detected by T2-weighted, diffusion-weighted, and computed tomography (CT) [6]. However, using conven- FLAIR images. Antiviral therapy (acyclovir) and cerebral tional T1-weighted and T2-weighted images, some patients edema procedures were administered for suspected HSE on have been observed to be affected with HSE that caused the third night of the hospital stay. On the 20th day of temporal lesions that were not clearly detectable in the acute hospitalization, the patient’s level of consciousness was not stage after onset of seizures [7]. improved. She was discharged on the 45th day with major Recently, fluid-attenuated inversion recovery (FLAIR) neurologic deficits. and diffusion imaging has proven very useful in differen- A wide range of pathogenic organisms, the most common tiating between healthy and pathologic tissues in brain of which are viruses that include HSV-1 and HSV-2, herpes diseases in childhood. The current study suggests that zoster virus, arboviruses, and enteroviruses, cause acute FLAIR pulse sequence allows for earlier detection of encephalitis. In general, acute viral encephalitis causes temporal lesions caused by HSE. The advantage of early diffuse parenchymal infiltration of inflammatory cells, and diagnosis of temporal lesions is the possibility of starting this leads to chromatolysis and pyknosis of neurons and, at early antiviral therapy for HSE and preventing the devel- times, extensive necrosis. These pathologic findings are opment of necrotizing lesions. Sometimes MRI and reflected by areas of low attenuation on CT, low signal diffusion images may not show abnormal tissues in the intensity on T1-weighted MRI, and high signal intensity early stage of disease. However, we should perform on T2-weighted MRI, depending on the degree and severity polymerase chain reaction (PCR) on the cerebrospinal fluid of inflammation. (CSF) to determine HSV and must start antiviral therapy [8]. HSE is the most common type of acute viral infection, An 8-year-old girl was admitted with a 2-day reported with a prevalence of 1 in 250000. Early diagnosis is history of headache, fever of 398C, refractory focal clonic essential because the prognosis is dependent on early seizure, and impaired consciousness. The patient was treatment [9,10]. Clinical signs might be nonspecific, so unconscious and responded only to pain. Routine blood diagnosis is based on the proof of evidence of viral DNA by chemistry findings were normal. CSF analysis showed no PCR in the cerebrospinal liquor. In herpetic encephalitis, abnormality. Blood and urine cultures were negative for MRI usually shows local temporal or frontal changes, but pathogens. A slow-wave focus of high voltage was observed MRI of the brain may also show normal findings. on the right side of the parietal lobe and temporo-occipital Electroencephalography (EEG) usually shows abnormali- lobe on the electroencephalogram. On the day of admission, ties, showing focal slow activity or periodic complexes. MRI (T1, T2, FLAIR and diffusion imaging) was per- EEG results, similar to MRI of the brain, may also be formed, but no abnormal signal was detected in temporal normal. In our patient, MRI of the brain showed no and other areas (Fig. 1A and B). On the day of admission, morphologic changes consistent with encephalitis. Thus,

0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. 88 Case Reports for an early diagnosis, it is important to suspect HSV Although no signs are pathognomonic for HSE, this infection and to perform appropriate radiologic and cere- disease should be considered in a patient with a progres- brospinal liquor examinations [11]. sively deteriorating level of consciousness, fever, abnormal Typically, the disease is localized on the temporal lobe. CSF indices, and focal neurologic findings in the absence of MRI findings typically include focal edema in the medial other causes. Acyclovir should be administered intrave- cortical region of the temporal and orbital frontal lobes, nously at 30 mg/kg per day for 14 to 21 days for the insular cortex, and angular gyrus. The basal ganglia and treatment of HSE. Even with appropriate administration of lobar white matter tend to be relatively spared. HSE should antiviral therapy, substantial mortality and morbidity from not be ruled out if MRI or CT shows sparing of the temporal HSE remain, with 19% of patients dying and 62% of lobe, including involvement of extra temporal areas or survivors having residual neurologic sequelae [12]. normal findings. Bilateral thalamic involvement is extreme- The diagnosis of HSV infections has been revolutionized ly rare in childhood. by the application of PCR technology to clinical specimens, including CSF and blood. Many reports have now demonstrated that PCR of CSF is both highly sensitive and specific for the rapid diagnosis of HSE. The clinical diagnosis, however, is often unreliable because numerous neurologic syndromes may mimic HSE. For a favorable prognosis of the disease, it is important to achieve a rapid diagnosis because mortality rate can only be decreased when antiviral therapy is initiated soon after the onset of symptoms. Positive PCR signals could be achieved up to 12 days after the appearance of neurologic symptoms [13]. Encephalitis and meningitis describe inflammatory processes involving the brain and meninges. The cardinal features of meningitis are headache, fever, and neck stiffness, whereas encephalitis is characterized by headache, fever, and neurologic or psychiatric alteration. Although the most common cause in both cases is infectious, the differentiation between meningitis and encephalitis in the emergency department (ED) is not always clear. Although the 2 conditions may be differentiated with further testing including MRI, specialized CSF analysis, and EEG, these modalities are often unavailable in the time required to make decisions about empiric therapy in the ED. Our findings suggest that although empiric therapy for bacterial pathogens occurred regularly in the ED, empiric treatment of encephalitis with acyclovir did not [14]. Therefore, emergency physicians should consider empiric acyclovir for patient with findings suggestive of viral encephalitis. Our patient was admitted for evaluation of acute onset of seizures. She had fever and no CSF pleocytosis and imaging abnormality on admission to the hospital. In conclusion, HSE remains one of the most devastating infections of the central nervous system despite available antiviral therapy. Clinical diagnosis is suggested in the encephalopathic, febrile patient with focal neurologic signs. However, these clinical findings are not pathognomonic because other numerous infections in the central nervous system can mimic HSE. Support for the diagnosis from a neurodiagnostic perspective is aided by the demonstration of disease of the temporal lobe by MRI scan and spike and slow-wave activity on EEG [3,4,7]. We believe that Fig. 1 A, On admission, diffusion-weighted image shows no this patient illustrates several important points about abnormal enhancement. B, On the third day of hospitalization, pediatric HSE. diffusion-weighted image shows abnormal hyperintensity in both The PCR detection of DNA in the CSF proved to be a thalami, temporal, and frontal area. sensitive and specific tool for the diagnosis of HSE; Case Reports 89 however, the most important factor in the diagnosis of [11] Niaz FE, Abou-Khalil B, Tesauro T, Aragon E. Meningoencephalitis: herpes virus infection in our patient was the consideration of an atypical presentation of herpes simplex type 2 central nervous the diagnostic possibility. system infection. Tenn Med 1998;91:183-5. [12] Whitley RJ, Alford CA, Hirsch MS, et al. Vidarabine versus acyclovir This report also illustrated that HSE could be present therapy in herpes simplex encephalitis. N Engl J Med 1986;314:144-9. with bilaterally thalamic involvement. CSF and MRI results [13] Wasay M, Mekan SF, Khelaeni B, Saeed Z, Hassan A, Cheema Z, et al. may be normal at the onset of the disease. Extra temporal involvement in herpes simplex encephalitis. Eur J Neurol 2005;12:475-9. Hakan GqmqY MD [14] Benson PC, Swadron SP. Empiric acyclovir is infrequently initiated in Y the emergency department to patients ultimately diagnosed with Sefer Kumanda MD encephalitis. Ann Emerg Med 2006;47:100-5. Department of Pediatric Neurology Erciyes University Faculty of Medicine 38039 Kayseri, Turkey E-mail address: [email protected] Subcortical stroke presenting as micrographia Hqseyin Per MD Department of Nephrology Acute ischemic stroke often presents with various motor Erciyes University Faculty of Medicine and sensory symptoms including hemiparesis, facial droop, 38039 Kayseri, Turkey aphasia, or hemiparesthesia [1]. Upon presentation, patients Fulya Tahan MD often manifest more than one symptom or clinical sign, which Esad Kfklq MD often assists in diagnosis. Reported below is the unusual case Department of Pediatrics of an acute small vessel infarction involving the left basal Erciyes University Faculty of Medicine ganglia that manifested as isolated micrographia. 38039 Kayseri, Turkey A 60-year-old white female with untreated hypertension presented to the emergency department (ED) with a 2-day Musa Karakqkc¸q MD history of small handwriting. The patient noted that the Department of Pediatric Hematology writing in her checkbook had become visibly smaller Erciyes University Faculty of Medicine despite attempts to write normally; other documents over 38039 Kayseri, Turkey the ensuing 2 days also revealed evidence of smaller than usual handwriting when compared with her usual penman- doi:10.1016/j.ajem.2006.05.007 ship. The patient denied headache, aphasia, vertigo, dysphagia, limb weakness or paresthesia, or ataxia. Her References only medical problem was hypertension; however, she had not refilled her antihypertensive agent in the last year. The [1] Heiner L, Demaerel P. Diffusion-weighted MR imaging findings in a patient took no medications, had never smoked, and only patient with herpes simplex encephalitis. Eur J Radiol 2003;45:195-8. occasionally consumed alcohol. There was no family history [2] Bode MK, Tikkakoski T, Tuisku S. Serial MR imaging findings of of cerebrovascular disease. HSV-2 encephalitis in a 50-year-old male patient. Eur J Radiol Extra The patient’s vital signs were the following: pulse rate, 2003;45:1-4. 82; respirations, 18; temperature, 97.28F; and blood pres- [3] Damasio AR, Van Hoesen GW. The limbic system and the localisation of herpes simplex encephalitis. J Neurol Neurosurg Psychiatry 1985; sure, 156/93 mm Hg. General physical examination was 48:297-301. normal. Neurologic examination was normal with the ex- [4] Meyding-Lamade U, Lamade W. HSV and the CNS. Neuroscientist ception of minimal dysarthria and micrographia when the 1996;2:44-54. patient was asked to write, as compared with documents [5] Lester JW, Carter MP, Reynolds TL. Herpes encephalitis: MR from prior to the onset of symptoms. Computed tomography monitoring of response to acyclovir therapy. J Comput Assist Tomogr 1988;12:941-3. of the brain showed minimal small vessel ischemic changes. [6] Osborn AG. Infections of the brain and its linings. In: Osborn Serum chemistries, complete blood count, urinalysis, sedi- AG, editor. Diagnostic neuroradiology. St. Louis (Mo)7 Mosby; mentation rate, electrocardiogram, and chest radiograph 1994. p. 694-6. were normal. Serum low-density lipoprotein was 107 mg/dL. [7] Rose JW, Stroop WG, Matsuo F, Hnkel J. Atypical herpes simplex Carotid ultrasound and transthoracic echocardiogram were encephalitis: clinical, virologic, and neuropathologic evaluation. Neurology 1992;42:1809-12. normal. The patient was felt to have an acute subcortical [8] Kato T, Yamanouchi H, Sugai K, Takashima S. Improved detection of lacunar infarction and was admitted to the hospital. Mag- cortical and subcortical tubers in tuberous sclerosis by fluid-attenuated netic resonance imaging of the brain revealed an acute inversion recovery MR imaging. Neuroradiology 1997;39:378-80. infarction involving the left superior basal ganglia and [9] Jubelt B, Miller JR. Viral infections. In: Rowland LP, editor. Merritt’s corona radiata; magnetic resonance angiography was nor- textbook of neurology. 9th ed. Philadelphia7 Williams &Wilkins; 1995. p. 142-79. mal. The patient was treated with aspirin, simvastatin, and [10] Adams RD, Victor M, Ropper AH. Principles of neurology, 6th ed. lisinopril, and at 2-week follow-up, was asymptomatic and New York7 McGraw-Hill; 1997. p. 749-59. had normal handwriting. 90 Case Reports

Acute ischemic stroke is commonly encountered in the [6] Pick A. Ueber eine eigenthumliche Schreibstorung, Mikrographie, in ED and often presents with hemiparesis, hemiparesthesia, Folge cerebraler Erkrankung. Prag Med Wochenschr 1903;28:1-4. aphasia, or transient visual loss [1]. Most of the cases of [7] Nakamura M, Hamamoto M, Uchida S, et al. A case of micrographia after subcortical infarction: possible involvement of frontal lobe func- ischemic stroke occur in patients with hypertension, tion. Eur J Neurol 2003;10:593-6. diabetes mellitus, or hyperlipidemia, often in the setting of [8] Derkinderen P, Dupont S, Vidal JS, Chedru F, Vidailhet M. Micro- structural cardiac or large artery disease such as atrial graphia secondary to lenticular lesions. Mov Disord 2002;17:835-7. fibrillation, left ventricular dysfunction, or carotid athero- [9] Kim JS, Im JH, Kwon SU, Kang JH, Lee MC. Micrographia after sclerosis. The patient in this report presented with the thalamo-mesencephalic infarction: evidence of striatal dopaminergic hypofunction. Neurology 1998;51:625-7. dominant complaint of small handwriting (micrographia), which classically is associated with Parkinson disease, a neurodegenerative condition characterized by depletion of dopamine in the substantia nigra [2,3]. Acute paraplegia caused by an accidental ingestion of Micrographia has rarely been reported in patients with hydrogen peroxide focal cerebral lesions such as intraparenchymal hematoma [4], syphilis [5], central nervous system lymphoma [6], and The danger of hydrogen peroxide results from a cerebral infarction [7-9]. Interestingly, all of the previously widespread embolism of oxygen bubbles in the vasculature. reported cases of micrographia in the setting of subcortical There have been reports of the toxicity after wound infarction have involved the left basal-ganglia-thalamic irrigation with hydrogen peroxide during surgery and even region. The reason for this is unclear but may involve of fatal complications after ingestion of a concentrated ischemic disruption of the frontal-subcortical circuits, which solution [1-4]. We present what we believe to be the first include the putamen/basal ganglia and supplemental motor case reported in the literature of acute paraplegia as an initial areas [7,9]. In addition, failure in automatic execution of manifestation of the toxicity by ingesting the solution. motor programs by the extrapyramidal neurons occurs in A previously healthy 47-year-old man ingested acciden- Parkinsonism and may also be the etiology of micrographia tally, while drunk, about 100 mL of 50% hydrogen due to infarctions involving the basal ganglia [9]. Why left peroxide, which was stored as a bleaching agent for basal ganglia pathology seems to predominate is unclear, cuttlefish in unmarked containers in the family refrigerator. but it should be noted that micrographia only rarely Soon afterward, he was found with foam around the mouth. complicates strokes in this area, making it difficult to offer Both lower limbs exhibited weakness, and he walked with a a definitive explanation. staggering gait. Although he was taken to the nearest In summary, clinicians should consider the diagnosis of hospital for help, his symptoms continued to progress basal ganglia infarction (especially left-sided) in patients during the first 3 hours, regardless of immediate decontam- with sudden-onset micrographia. Magnetic resonance im- ination by performing gastric lavage. He exhibited shortness aging of the brain is useful in this setting to diagnose not of breath before the transfer, and on arrival, his vital signs only acute cerebral infarction but also other types of mass were the following: blood pressure 184/118 mmHg; pulse lesions as well. 149 beats per minute, respiratory rate 22 breaths per minute, Mark A. Marinella MD body temperature 38.28C, oxygen saturation (by oximeter) Wright State University School of Medicine 96%, and Glasgow Coma scale score of E3V5M6. On Dayton, OH 45409, USA neurologic examination, the muscle power of both lower E-mail address: [email protected] limbs was grade 0/5, and the sensory test showed numbness below the level of the xiphoid process. His abdomen was doi:10.1016/j.ajem.2006.03.033 distended, and abdominal films showed amorphous intramural gas over the left upper abdomen. The white cell count was 22800/mm3, and the other laboratory studies showed References results which were within normal limits. Working under the assumption of gas embolism, the [1] Brust JCM. Cerebral infarction. In: Rowland LP, editor. Merritt’s patient underwent emergent hyperbaric oxygen (HBO) Neurology, 10th ed. Philadelphia7 Lippincott, Williams, and Wilkins; therapy 5 hours after ingestion. Just before entering the 2000. p. 232-40. hyperbaric chamber, he had became drowsy, and both upper [2] Fahn S, Przedborski S. Parkinsonism. In: Rowland LP, editor. Merritt’s Neurology, 10th ed. Philadelphia7 Lippincott, Williams, and Wilkins; limbs had exhibited gradual weakening. He was intubated 2000. p. 679-93. and transferred to the intensive care unit because of a [3] Sapira JD. The neurologic examination. In: Orient JM, editor. The art deteriorated mental status and respiratory movement after and science of bedside diagnosis. Baltimore7 Williams and Wilkins; the HBO treatment. The muscle power of both upper limbs 1990. p. 451-527. was measured at grade 2/5. Subsequently, a head magnetic [4] Martinez-Vila E, Artieda J, Obeso JA. Micrographia secondary to lenticular hematoma. J Neurol Neurosurg Psychiatry 1988;51:1353-6. resonance imaging (MRI) revealed multiple embolic infarcts [5] Lewitt P. Micrographia as a focal sign of neurologic disease. J Neurol over the supratentorial and the infratentorial region (Fig. 1). Neurosurg Psychiatry 1983;46:1152-3. In addition, a cervical spine MRI also revealed multiple Case Reports 91

Fig. 1 Magnetic resonance images of the brain showing multiple high-signal changes in the diffusion-weighted images involving the supratentorial bilateral cerebral hemisphere (A) and the infratentorial left cerebellar hemisphere (B), indicating areas of embolic infarct. embolic infarcts over the second to seventh cervical and the first to second thoracic sections of the spinal cord (Fig. 2). The patient gradually gained consciousness and was extubated successfully on the sixth day in hospital. Muscle power returned to both upper limbs, although some neurologic sequelae remained after a regimen of HBO therapy and physical rehabilitation. He was discharged with paraparesis and a neurogenic bladder. It has been calculated that 1 mL of 50% hydrogen peroxide can produce approximately 165 mL of oxygen gas [5]. In this case, the patient had ingested around 100 mL of 50% hydrogen peroxide, and so, 16.5 L of oxygen would have been produced shortly after the ingestion. Its toxic effect may occur by arterial embolism through the previously proposed mechanisms suggested by Sherman et al [6]. Our case presented multiple infarctions resulting from numerous gas bubbles showing on subsequent MRI scans of the brain and the cervical spine. Ijichi et al [7] and Ashdown et al [8] also reported the same cerebral insult after ingestion of concentrated hydrogen peroxide. Howev- er, our report of an initial manifestation of acute paraplegia resulting from spinal infarctions has not been reported in other similar cases. The use of HBO in treating gas embolism from decompression sickness is well known. There are 2 reports of the successful use of HBO as a treatment option for Fig. 2 Magnetic resonance images of the cervical spine showing hydrogen peroxide caused embolism in the literature [9,10]. discrete segmental high-signal lesions in the T2-weighted images The oxygen bubbles in capillaries can block the blood involving the second, third, sixth, and seventh cervical and the first supply to tissues, and the severity of tissue damage depends to second thoracic sections of the spinal cord. 92 Case Reports on both the hypoxia tolerance of damaged tissues and the [8] Ashdown BC, Stricof DD, May ML, et al. Hydrogen peroxide absorptive rate of oxygen bubbles. Absorption of the bigger poisoning causing brain infarction: neuroimaging findings. AJR Am J bubbles is a difficult process, and the pathophysiology of Roentgenol 1998;170:1653-5. [9] Mullins ME, Beltran JT. Acute cerebral gas embolism from hydrogen applying HBO is to reduce the size of bubbles by the peroxide ingestion successfully treated with hyperbaric oxygen. principle of the Boyle-Mariotte law. In this case, we found J Toxicol Clin Toxicol 1998;36:253-6. that the later observed symptoms, such as decreased level of [10] Vander Heide SJ, Seamon JP. Resolution of delayed altered mental consciousness and upper limb weakness, resolved com- status associated with hydrogen peroxide ingestion following hyper- pletely after HBO therapy. However, nerve cells are a more baric oxygen therapy. Acad Emerg Med 2003;10:998-1000. [11] Dunbar EM, Fox R, Watson B, et al. Successful late treatment of hypoxia-sensitive tissue than others, and prolonged expo- venous air embolism with hyperbaric oxygen. Postgrad Med J 1990; sure to a hypoxic environment will result in apoptotic 66:469-70. neuronal cells. We hypothesize that multiple spinal infarc- [12] Bitterman H, Melamed Y. Delayed hyperbaric treatment of cerebral air tions had occurred before HBO therapy was performed and embolism. Isr J Med Sci 1993;29:22-6. had already resulted in irreversible neurologic sequelae. According to the above mechanisms, HBO can block the progression of hypoxia in tissues in such cases. There would arguably have been a better neurologic outcome if he had Takotsubo cardiomyopathy: an unusual syndrome received earlier interruption with HBO therapy. However, mimicking an ST-elevation myocardial infarction some authors have nevertheless concluded that delayed HBO therapy is still beneficial even 15 to 60 hours after onset of Coronary artery atherosclerosis and subsequent plaque symptoms of acute gas embolism, suggesting that it is not too rupture are thought to be responsible for most myocardial late to perform HBO treatment on such a critical patient, even infarctions (MIs). Recent cardiology literature, primarily if this requires transferring the patient to the nearest describing elderly female patients, has introduced an hyperbaric facility when none is available locally [11,12]. unusual subset of patients who have ST-segment elevation anterior MI with no evidence of angiographic disease on Te-Ming Liu MD cardiac catheterization [1-4]. These patients have profound Kuo-Chin Wu MD and peculiar reversible cardiac wall motion abnormalities Ko-Chi Niu MD, PhD and dysfunction. This cardiomyopathy, coined the Takot- Hung-Jung Lin MD subo cardiomyopathy or Takotsubo syndrome in Japan Department of Emergency Medicine because of the heart’s appearance on the end-systolic Department of Hyperbaric Oxygen Therapy ventriculogram, is beginning to shed some light on other Chi-Mei Medical Center etiologies as suggested, contributing factors in some acute Yung-kang City MIs, including stress, hormones, and microvascular spasm. Tainan 710, Taiwan, ROC A 77-year-old white woman presented to the emergency department (ED) via emergency medical services with doi:10.1016/j.ajem.2006.05.014 severe respiratory distress. According to her husband, she became acutely dyspneic at home after a heated argument, with a subsequent decrease in her level of consciousness. Her medical history was significant for chronic obstructive References pulmonary disease and coronary artery disease, with a stent apparently placed to an unknown vessel 2 years prior at a [1] Shaw A, Cooperman A, Fusco J. Gas embolism produced by different hospital. On arrival to the ED, she was found to be hydrogen peroxide. N Engl J Med 1967;277:238-41. in respiratory failure requiring immediate rapid sequence [2] Bassan MM, Dudai M, Shalev O. Near-fatal systemic oxygen intubation. Her initial electrocardiogram (ECG) tracing embolism due to wound irrigation with hydrogen peroxide. Postgrad Med J 1982;58:448-50. demonstrated sinus rhythm with more than 1 mm ST [3] Morikawa H, Mima H, Fujita H, et al. Oxygen embolism due to elevation in leads V2 to V5 (Fig. 1). Physical examination hydrogen peroxide irrigation during cervical spine surgery. Can J revealed an intubated and sedated patient with blood Anaesth 1995;42:231-3. pressure of 98/60 mm Hg, heart rate of 110, jugular venous [4] Christensen DW, Faught WE, Black RE, et al. Fatal oxygen pressure of 11 mm Hg, an S gallop, and bibasilar rales. A embolization after hydrogen peroxide ingestion. Crit Care Med 4 1992;20:543-4. portable chest radiograph after intubation revealed flattened [5] Neff SP, Zulueta L, Miller R. Hydrogen peroxide: an unusual cause of diaphragms consistent with emphysema and pulmonary arterial and venous gas embolism. Anaesthesia 1996;51:683-4. vascular fullness, suggestive of pulmonary edema. A [6] Sherman SJ, Boyer LV, Sibley WA. Cerebral infarction immedi- bedside ultrasound in the ED demonstrated anterior wall ately after ingestion of hydrogen peroxide solution. Stroke 1994;25: hypokinesis and an ejection fraction of 30%. 1065-7. [7] Ijichi T, Itoh T, Sakai R, et al. Multiple brain gas embolism after After her intubation, she was given aspirin per orogastric ingestion of concentrated hydrogen peroxide. Neurology 1997;48: tube and urgently rushed to the cardiac catheterization 277-9. laboratory, where her coronary arteries were found to Case Reports 93

Fig. 1 Electrocardiogram with ST elevations in leads V2 to V5. be clear of significant disease. Her left ventriculogram recently, this syndrome had only been described in Asia; demonstrated apical, mid-anteroseptal, mid-anterior, mid- however, there are increasing numbers of documented case inferoseptal, and lateral wall hypokinesis with basilar reports elsewhere in the world in non-Asian patients [1,8,9]. hyperkinesis, creating transient left ventricular apical The typical patient presentation is an elderly woman ballooning, otherwise known as takotsubo cardiomyopathy with mild to moderate chest pain, dyspnea, an ECG (Fig. 2). The patient’s initial creatine kinase–MB and demonstrating ST-segment elevations in leads V2 through troponin T were 5.7 and less than 0.01 ng/mL, with levels V5, and a modest rise in cardiac enzyme markers [5]. peaking at 24 and 0.63 ng/mL, respectively. She was treated Almost invariably, the patient presenting with Takotsubo with furosemide, carvedilol, ramipril, and aspirin. Warfarin syndrome experiences some type of emotional or physical was added to prevent left ventricular thrombus formation. stress trigger just before the disease onset [1-3].Someof She was extubated on hospital day 2, her symptoms these documented triggers have included a surgical or resolved, her pro-brain natriuretic peptide trended toward medical procedure, a death in the family, panic disorder, or normal, and subsequent echocardiography performed 4 days respiratory distress [1-3]. Echocardiography typically after admission revealed a normal ejection fraction of 60% to 65% with grossly normal wall motion and a relaxation abnormality. The patient was discharged home at her baseline functional status 6 days after admission. Takotsubo cardiomyopathy is a transient cardiomyopathy characterized by marked apical hypokinesis and ballooning in the absence of significant coronary artery disease [5]. Takotsubo (tako [octopus] and tsubo [jar]) is a Japanese term for a round-bottomed, narrow-necked jar used to trap octopi in Japan (Fig. 3). Similar to this octopus jar, the heart’s appearance on a left ventriculogram in Takotsubo cardiomyopathy appears wide at the apex during end systole, with a narrowing where there is basilar hyperkinesis (Fig. 1). The exact mechanism is unknown; however, hormonal predisposition [6], an inciting stressor [1,2], and microvascular dysfunction resulting in simultaneous multi- vessel spasm [3,4,6,7] have been suggested as possible etiologies. Most patients with this condition are female. An Fig. 2 Left ventriculogram at end systole. Note the basilar association with an unusually long course of the left anterior hyperkinesis (white arrows) and left ventricular apical ballooning descending coronary artery has also been noted [5] Until (black arrows). 94 Case Reports

frequently now that providers are aware of the phenomenon. One recent case series reported that approximately 2.2% of ST-elevation acute coronary syndromes presenting to a referral hospital during 2002 to 2003 were considered to be consistent with Takotsubo cardiomyopathy [12]. Com- plications associated with the cardiomyopathy include congestive heart failure, ventricular fibrillation, and death. Recurrence seems to be rare in these patients when followed for 1 to 4 years, and the prognosis seems to be favorable with supportive therapy, including diuresis, beta-blockade, angio- tensin-converting enzyme inhibition, and anticoagulation [13]. It is unknown at this time whether thrombolysis has any therapeutic effect on this syndrome. The data to date do not support plaque rupture and subsequent thrombosis as the etiologies of Takotsubo, which seems to resolve spontaneously with medical management alone without angioplasty or thrombolysis. Current evidence suggests thrombolytics may only be harmful in these patients, with the inherent bleeding risks of the therapy without documented benefits. In conclusion, transient left ventricular apical ballooning—the Takotsubo cardiomyopathy—and an acute MI due to plaque rupture clinically present quite similarly, with ST-segment elevation, cardiac biomarker release, chest pain, and myocardial dysfunction. Key factors associated with the Takotsubo syndrome include an anterior MI pattern, b Q Fig. 3 Takotsubo jar. established elderly-female association, and a significant stressor as a trigger. Thrombolysis does not seem to be a demonstrates reversible anterior-apical hypokinesis or aki- cure for the Takotsubo syndrome, and therefore may nesis [7]. As opposed to a typical ST-elevation MI, Ogura possibly only cause the patient harm. These 2 cardiac [10] and Inoue [11] have both documented an association of disorders can only be carefully differentiated with coronary Takotsubo syndrome lacking the typical reciprocal ST angiography. Emergency medicine physicians should be depressions on the ECG. Ogura studied patients with aware of this disease, especially in the setting of an elderly confirmed Takotsubo cardiomyopathy compared with con- female patient presenting with an anterior MI pattern, no trols with anterior MI. He found that factors unique to the reciprocal ST-segment depressions, and a distinct stress Takotsubo syndrome included (1) the absence of reciprocal trigger. Current cardiology literature suggest that cardiac changes, (2) absence of abnormal Q waves, ST elevation in catheterization be a definite first choice ahead of thrombo- V4-6/V1-3 [termed sigmaSTeV(4-6)/V (1-3)] of 1 or higher, lytics, even if a small delay in getting to the catheterization and (3) a prolonged QTc all showing a high sensitivity and laboratory would make these interventions otherwise equal. specificity for diagnosing Takotsubo cardiomyopathy. If cardiac catheterization is unavailable or further delayed, Ogura determined that the combination of the absence of thrombolysis must still be advocated in the acute setting reciprocal changes and a sigmaSTeV(4-6)/V (1-3) of 1 or because distal left anterior descending infarctions (distal to higher had a greater specificity (100%) and overall accuracy D1 and S1) have been shown to appear electrocardiograph- (91%) than either criteria [10]. Inoue performed a similar ically similar to Takotsubo syndrome [11] in the acute comparison, further breaking down the anterior MI control phase, and the former is still far more common and group into those with culprit lesions proximal to either S1 or definitely amenable to thrombolytic therapy. Aside from D1 (groups A and B) and those with culprit lesions distal to immediate cardiac catheterization to make the diagnosis, S1 and D1 (group C). In patients with Takotsubo syndrome, typical medications used to treat MI such as aspirin, reciprocal ST-segment depression in the inferior leads was morphine, nitroglycerine, b-blockers, and heparin should observed less frequently than in patients in groups A ( P b still be initiated when patients suspected to have Takotsubo .0001) and B ( P = .0002), and abnormal Q waves and syndrome are treated in the ED. ST-segment elevation in the inferior leads were observed more frequently than in group A or B ( P = .0007 and P = Tom E. Kolkebeck MD .0057, respectively). The ECG findings in Takotsubo Department of Emergency Medicine syndrome did not differ from those in group C [11]. Wilford Hall Medical Center Although this syndrome was considered a fairly rare Lackland Air Force Base, TX 78236, USA entity in the cardiology community, it is being noticed more E-mail address: [email protected] Case Reports 95

Casey L. Cotant MD reactions increased from 1.5% in 1989 to 1990 to 9.2% in Department of Internal Medicine 1998 to 2000 [1]. We report a unique case of anaphylaxis to Wilford Hall Medical Center topical bacitracin ointment after application to a fresh tattoo. Lackland Air Force Base, TX 78236, USA A 42-year-old white man with no significant medical history was brought to the ED by his spouse. On physical Richard A. Krasuski MD examination, the patient was disoriented and unable to Department of Cardiology follow commands. The Initial vital signs were the following: Wilford Hall Medical Center temperature, 96.5; heart rate, 46; respiratory rate, 14; blood Lackland Air Force Base, TX 78236, USA pressure, 60/40; and pulse oximetry, 85%, on room air. The patient’s skin appeared flushed, with breath sounds clear to doi:10.1016/j.ajem.2006.08.001 auscultation bilaterally. History obtained from the patient’s spouse revealed that the patient had a fresh tattoo placed on his anterior chest wall the night before, and approximately References 10 to 15 minutes before arrival in the ED, the patient had applied bacitracin to the tattoo. [1] Connelly KA. Stress, myocardial infarction, and the btako-tsuboQ The patient received normal saline solution, Solumedrol phenomenon. Heart 2004;90(9):52. [2] Ueyama T. Estrogen attenuates the emotional stress-induced cardiac (125 mg), Benadryl (25 mg), Pepcid (20 mg), and 0.3 mL of responses in the animal model of tako-tsubo cardiomyopathy. 1:10,000 epinephrine intravenously. Within 10 minutes, blood J Cardiovasc Pharmacol 2003;42(Suppl 1):S117-9. pressure increased to 123/67, heart rate increased to 71, and [3] Kurisu S. Tako-tsubo–like left ventricular dysfunction with pulse oximetry was 100% on 2 L of oxygen. After 4 hours of ST-segment elevation: a novel cardiac syndrome mimicking acute observation in the ED and with symptoms completely myocardial infarction. Am Heart J 2002;143(3):448-55. [4] Nyui N. dTako-tsuboT transient ventricular dysfunction: a case report. resolved, the patient was discharged home in stable condition. Jpn Circ J 2000;64(9):715-9. Bacitracin was awarded the title of Contact Allergen of [5] Iban˜ez B. Tako-tsubo syndrome associated with a long course of the Year 2003 by the American Contact Dermatitis Society the left anterior descending coronary artery along the apical dia- because of its sensitizing capacity. It is the ninth most phragmatic surface of the left ventricle. Rev Esp Cardiol 2004;57(3): common allergen as of 2000 [2]. Although this is common 209-16 [Spanish]. [6] Upadya S. Tako-tsubo cardiomyopathy (transient left ventricular knowledge in the specialty of dermatology, there is a paucity apical ballooning)—a myocardial perfusion echocardiogram study. of reports in the emergency medicine literature. Case reports Chest 2004;67:345. of allergic reactions to bacitracin have been reported after [7] Yasuga Y. Tako-tsubo–like transient left ventricular dysfunction and application to abrasions, chronic ulcers, clean surgical apical thrombus formation: a case report. J Cardiol 2004;43(2):75-80 procedures, intranasal packing, and even mediastinal irriga- [Japanese]. [8] Desmet WJ. Apical ballooning of the left ventricle: first series in white tion with bacitracin solution in the operating room [3-5]. patients. Heart 2003;89(9):974-6. However, we present a unique case of anaphylaxis after [9] Gaspar J. Tako-tsubo syndrome (transient antero-apical dyskinesia): application of bacitracin ointment to a fresh tattoo. first case reported in Latin America and review of literature. Arch The widespread use of bacitracin has led to an increased Cardiol Mex 2004;74(3):205-14 [Spanish]. recognition of allergic contact dermatitis, with a potential [10] Ogura R. Specific findings of the standard 12-lead ECG in patients with dtakotsuboT cardiomyopathy: comparison with the findings of for even life-threatening anaphylaxis. Emergency medicine acute anterior myocardial infarction. Circulation 2003;57(8):687-90. physicians need to be conscious of both immediate (IgE) [11] Inoue M. Differentiation between patients with takotsubo cardiomy- and delayed (cell-mediated) reactions to bacitracin. Physi- opathy and those with anterior acute myocardial infarction. Circula- cians and patients need to be aware of these potential serious tion 2005;69(1):89-94. side effects when using bacitracin on damaged skin or [12] Bybee KA. Clinical characteristics and thrombolysis in myocardial infarction frame counts in women with transient left ventricular apical mucosal surfaces. With continued use of bacitracin in these ballooning syndrome. Am J Cardiol 2004;94:343-6. circumstances, it has the potential to become the number [13] Akashi YJ. The clinical features of takotsubo cardiomyopathy. QJM one topical allergen in North America. 2003;96(8):563. Karen Greenberg DO Anaphylaxis to topical bacitracin ointment James Espinosa MD Victor Scali DO Bacitracin is used on a daily basis for treatment of local Emergency Medicine Department infections and is a popular topical antibiotic for wound University of Medicine and prophylaxis because of its low cost and presumed low toxicity Dentistry of New Jersey-School of Osteopathic Medicine [1]. Historically, bacitracin has been thought to rarely cause Kennedy Health System-Stratford Division allergic reactions. In fact, most physicians are not aware of any Stratford, NJ 08084, USA side effects of the drug. However, in 2003, bacitracin was E-mail address: [email protected] awarded bAllergen of the YearQ by the North American Contact Dermatology Group because its incidence of allergic doi:10.1016/j.ajem.2006.04.010 96 Case Reports

References

[1] Lio PA, Kaye ET. Topical antibacterial agents. Infect Dis Clin North Am 2004;18:717-33. [2] Aberer W. Bacitracin for lubrication: an allergen for more convenience? J Am Acad Dermatol 2005;52. [3] Saryan JA, Dammin TC, Bouras AE. Anaphylaxis to topical bacitracin zinc ointment. Am J Emerg Med 1998;16:512-3. [4] Gall R, Blakley B, Warrington R, Bell DD. Intraoperative anaphylactic shock from bacitracin nasal packing after septorhinoplasty. Anesthesi- ology 1999;91:1545. [5] Blas M, Briesacher KS, Lobato EB. Bacitracin irrigation: a cause of anaphylaxis in the operating room. Anesth Analg 2000;91:1027-8.

Laceration of the popliteal artery and compartment syndrome resulting from stingray envenomation Fig. 2 One of 2 fasciotomy sites 21 days after the injury (photo courtesy of Sarasota Memorial Hospital). A 46-year-old man was releasing a stingray from a fishing line when its tail barb punctured the patient’s wader boots and entered his left medial thigh. When the para- (Fig. 2). The severed popliteal arterial segment and a swab medics arrived, the patient was hypotensive and tachycardic. from the wound were sent for culture and sensitivity but He complained of severe pain surrounding the wound and grew no organisms. The patient was discharged home from numbness in his left foot. No pedal pulse could be palpated the hospital on postoperative day number 8 and was in the patient’s lower extremity. continued on intravenous and oral antibiotics. A wound In the emergency department, the patient required several vacuum was used on the envenomation site, and the patient units of packed red blood cells to maintain his blood regained full function of his lower extremity. pressure. Wound exploration in the operating room revealed There are approximately 2000 stingray injuries each year a lacerated left popliteal artery. No barb or other foreign in the United States [1]. However, stingrays are extremely body was found within the wound. The lacerated segment of docile and nonaggressive (Fig. 3). It is not until the the left popliteal artery was removed and repaired with a left stingray’s wings are disturbed that its tail reflexively whips saphenous vein bypass. The surgical incision was closed, upward [1,2]. At the base of the tail is a caudal spine that but the area of envenomation was left to heal by secondary contains a venom organ and an enveloping integumentary intention (Fig. 1). During the procedure, the patient required sheath [1] (Fig. 4). The spine contains sharp retroserrated a 4-compartment fasciotomy for compartment syndrome, teeth that can produce a puncture wound or laceration which developed distal to the site of the envenomation containing a film of venom, mucus, and pieces of the sheath or spine [1]. The venom has vasoconstrictive properties that may cause cyanosis, necrosis [3], poor wound healing, and

Fig. 1 The puncture wound caused by the stingray barb and, Fig. 3 The cownose stingray (Rhinoptera bonasus) is commonly beneath it, the closed operative incision site 21 days after the injury caught by fisherman off the West Coast of Florida (photo courtesy (photo courtesy of Sarasota Memorial Hospital, Sarasota, Fla). of Dr Carl A. Luer and Mote Marine Laboratory, Sarasota, FL). Case Reports 97

[3] Barss P. Wound necrosis caused by the venom of stingrays. Pathological findings and surgical management. Med J Aust 1984; 141:854-5. [4] Perkins RA, Morgan SS. Poisoning, envenomation, and trauma from marine creatures. Am Fam Physician 2004;69(4):885-90. [5] Russell FE, Panos TC, Kang LW, et al. Studies on the mechanism of death from stingray venom: a report of two fatal cases. Am J Med Sci 1958;235:566-83. [6] Jain AL, Robertson GJ, Rudis MI. Surgical issues in the poisoned patient. Emerg Med Clin North Am 2003;21:1117-44.

An unusual primary blast injuryB Traumatic brain injury due to primary blast injury

Fig. 4 This spine is from a southern stingray (Dasyatis The reports on primary blast injuries focus on damage in americana) and measures 22 cm (photo courtesy of Dr Carl A. gas-containing organ systems; however, the likelihood of Luer and Mote Marine Laboratory). blast-induced neurotrauma is often overlooked or related to secondary or tertiary injury cause [1]. We report a case of infection [4]. Life-threatening and fatal injuries are rare but traumatic brain injury due to primary blast injury caused by have occurred after abdominal or thoracic penetration, work-related accident. injury to the femoral artery, and tetanus [2,5]. A 36-year-old male worker with blast injury due to steam Patients with significant blood loss should be treated boil explosion was admitted to the ED. On arrival, the with aggressive fluid resuscitation and transfusion if patient’s vital signs were normal. The Glasgow Coma Scale necessary. Wounds require irrigation, exploration, and (GCS) score was recorded as 15/15. There were no blunt or debridement. Caution should be used when removing an penetrating wound sign on head and neck examinations embedded spine because this can cause further release of other than left side tympanic membrane rupture. Second- venom [6]. The wound should be allowed to heal through degree burns (about 12%) were found on the left upper arm. delayed primary closure [1]. Patients should be placed on a Mental status was depressed when chest x-ray was prophylactic antibiotic that covers Staphylococcus, Strepto- performed. The patient was intubated when his GCS coccus, Vibrio vulnificus, and Mycobacterium marinum[4]. declined to 8/15 within several minutes and his pupils A tetanus vaccine or tetanus toxoid should be given, if became anisocoric. Computed tomography of the head indicated [2]. Patients with wounds that may have penetrat- displayed mild cerebral edema, minimally depressed linear ed the thoracic or abdominal cavities should have further petrose bone fracture, left temporoparietal epidural hemor- assessment to rule out other associated life-threatening rhage, and diffuse subarachnoid hemorrhage (Fig. 1). The injury [2]. It is recommended that patients be observed for patient had surgery as quickly as possible. On the following at least 3 hours for systemic adverse effects [1]. The patient days, he was observed in the intensive care unit and died should be advised to return in 48 hours for wound on postoperative 25th day. reevaluation and to return sooner for signs and symptoms Explosions can result in unique injury patterns to the of infection. central nervous system (CNS) [2]. Neurological impairment due to blast injury was initially attributed to air emboli in the Charlotte Derr MD cerebral circulation. In addition, studies on animals suggest Barbara J. O’Connor RN, MBA that the overpressure wave is transferred to the central Sandra L. MacLeod MD, MPH nervous system, causing diffuse axonal injury. Higher levels Sarasota County Fire Department of blast overpressure can cause skull fractures or coup- Emergency Services, Sarasota counter-coup injuries [3-5]. FL 34236, USA There have been several theories related to the mechanism of brain damage after primary blast exposure. First, doi:10.1016/j.ajem.2006.04.016 shock waves are transmitted to the brain via viscera, muscles, bones, blood vessels, and cerebrospinal fluid [6,7]. Second, the blast may create a higher pressure in References

[1] Auerbach PS. Envenomation by aquatic invertebrates. In: Auerbach PS, B editor. Wilderness medicine. 4th ed. St. Louis (Mo)7 Mosby Inc; 2001. Authors contributions: Serkan YVV lmaz did patient’s care and p. 1488-92. management and conceived the paper. Manuscript was written by Serkan [2] Fenner PJ, Williamson JA, Skinner RA. Fatal and non-fatal stingray YVV lmaz and Murat Pekdemir. All authors drafted the manuscript and envenomation. Med J Aust 1989;151:621-5. contributed substantially to its revision. 98 Case Reports

Fig. 1 Computed tomography image showed mild cerebral edema, minimally depressed linear petrose bone fracture, left temporoparietal epidural hemorrhage, and diffuse subarachnoid hemorrhage. the body except in the cranium; therefore, the blood from References the largest vessels may be squeezed into this low-pressure area [7]. When these reach the CNS, it is likely that they [1] DePalma RG, Burris DG, Champion HR, et al. Blast injuries. N Engl J generate high-pressure differentials, resulting in rupture of Med 2005;352:1335-45. the blood vessels in the meninges and brain tissue. [2] Lavonas E, Pernardt A. Blast injuries. [Emedicine Web site]. January 17, 2006. Available at: http://www.emedicine.com/emerg/topic63.htm This case appears to have been caused by primary blast [Accessed March 6, 2006]. injury. The patient’s head and neck examinations were [3] Singer P, Cohen J, Stein M. Conventional terrorism and critical care. unimpressive, and he had no neurological deficit on arrival, Crit Care Med 2005;33:61-5. with a GCS recorded as 15/15. With the rising risk of [4] Mayorga M. The pathology of blast overpressure injury. Toxicology terrorist bombings in many parts of the world, emergency 1997;121:17-28. [5] Sutphen SK. Blast injuries: a review. [Medscape Web site]. November physicians need to consider initially occult traumatic brain 9, 2005. Available at: http://www.medscape.com/viewprogram/4714 injury due to blast wave. [Accessed March 10, 2006]. [6] Cernak I, Wang Z, Jiang J, et al. Ultrastructural and functional characteristics of blast injury-induced neurotrauma. J Trauma 2001;50: Serkan YVlmaz MD 695-706. Murat Pekdemir MD [7] Siri KK, Egil O. Blast induced neurotrauma in whales. Neurosci Res Department of Emergency Medicine 2003;46:377-86 School of Medicine, Kocaeli University 41380 Kocaeli, Turkey E-mail addresses: [email protected] Gastric outlet obstruction in an infant: lactobezoar [email protected] A nearly 4-month-old infant, former product of an doi:10.1016/j.ajem.2006.04.014 uncomplicated term, twin gestation, presented to the ED Case Reports 99 with 1 day of nonbloody, nonbilious emesis, frequent loose stools, and irritability. Vital signs showed a respiratory rate of 50 to 60 breaths/min, 100% oxygen saturation on room air, sinus tachycardia of 200 to 210 beats/min, and blood pressure of 105/60. On examination, the infant appeared in significant respiratory distress with clear breath sounds, clinical evidence of dehydration, poor peripheral perfusion, and a distended abdomen. Radiography (Fig. 1) revealed a large air-filled stomach with absent distal bowel gas. Ultrasound findings were not consistent with a pyloric stenosis, intussusception, or obvious volvulus. Upper gastrointestinal series (Fig. 2) demonstrated an intragastric filling defect consistent with a blactobezoar.Q The infant improved with nasogastric decompression and intravenous fluids. Further history uncovered that the powdered infant formula was incorrectly prepared with half the necessary water. Lactobezoars represent the most common type of bezoar of infancy and toddlerhood but are rarely considered in a differential diagnosis [1]. Case reviews suggested an association with prematurity, low birth weight, and hyper- caloric or predominantly casein-based formulas [2]. The occurrence in term infants or those receiving breastmilk or Fig. 2 Upper gastrointestinal contrast study revealing an intra- soy-based formulas, however, supports a multifactorial gastric filling defect without evidence of intestinal obstruction etiology including possible delayed gastric emptying and or malrotation. altered gastric secretions [3]. Given its protean manifes- conservative therapy, practitioners should have an index of tations, potential morbidity, and typical resolution with suspicion and pursue radiographic support before more invasive measures. Robert J. Graham MD Division of Critical Care Department of Anesthesia Children’s Hospital Boston Boston, MA 02115, USA

Phil Stein MD Department of Pediatrics The Floating Hospital for Children Tufts-New England Medical Center Boston, MA 02111, USA

doi:10.1016/j.ajem.2006.03.034

References

[1] Gittelman MA, Racadio J, Gonzalez del Rey J. Radiological case of the month. Lactobezoar. Arch Pediatr Adolesc Med 1999;153(5):541-2. [2] Grosfeld JL, Schreiner RL, Franken EA, Lemons JA, Ballantine TV, Weber TR, et al. The changing pattern of gastrointestinal bezoars in infants and children. Surgery 1980;88(3):425-32. [3] DuBose V TM, Southgate WM, Hill JG. Lactobezoars: a patient series and literature review. Clin Pediatr (Phila) 2001;40(11):603-6.

An unusual case of pediatric abdominal distension

A 19-month-old boy presented to the ED with a Fig. 1 Portable bbabygramQ radiograph demonstrating massive 7-month history of a progressively enlarging abdominal gastric distention with no distal bowel gas. girth. He had nighttime tactile fevers for the past month 100 Case Reports and decreased appetite with some weight loss. There was no abdominal pain, vomiting, diarrhea, or jaundice. The patient weighed 14.7 kg, was afebrile, had a heart rate of 150 bpm, and a respiratory rate of 30 breaths/min. His lungs were clear, and there were no adventitious cardiac sounds. He had good bowel sounds. His abdomen was nontender but was protuberant with a girth of 66 cm. Fluid was ballotable throughout the entire extent of the abdomen, but there was no defined palpable mass and no hepatosplenomegaly. Furthermore, the patient had no peripheral edema. The patient’s white blood cell count was 4.7K/lL, with 19 neutrophils and 1 eosinophil. His hemoglobin level was 14 g/dL, platelet count was 234000/lL, aspartate aminotransferase level was 48 U/L, alanine aminotransferase level was 46 U/L, and total bilirubin level was 0.2 mg/dL. Plain abdominal films showed marked abdominal distension with diffuse opacification suggestive of ascites. A computed Fig. 2 Computed tomographic scan image of the abdomen again tomographic scan of the abdomen (Figs. 1 and 2) revealed a showing medial displacement of the bowel contents. massive amount of fluid that displaced the diaphragm superiorly and the small bowel and colon medially. This computed tomographic scan finding was initially felt to be in the head and neck, whereas intra-abdominal locations are consistent with massive ascites. However, the medial rare [1]. Abdominal lymphangiomas may develop in the displacement of the bowel prompted suspicion for a mass. mesentery of the intestine, omentum, or retroperitoneum Surgical consultants performed an exploratory laparotomy. [2,3]. Cysts may be unilocular or multilocular and may A large omental cyst containing 2 L of fluid was found contain serous, serosanguinous, or chylous fluid [4]. (Fig. 3). An omentectomy was performed, and pathology Presentation of such cysts depends on size and location. confirmed the diagnosis. The patient did well postoperative- Omental cysts may present as an abdominal mass with vague ly, and the abdominal distension resolved. pain. Retroperitoneal cysts may go undetected until there is Omental cysts are cystic lymphatic malformations of the compression on the bladder and kidneys [2,5]. Mesenteric abdomen. These benign congenital tumors arise from cysts tend to present more acutely with abdominal pain from sequestered lymphatic tissue failing to communicate with bleeding, compression, torsion, or infection [6]. the lymphatic system. Most cystic lymphangiomas are found Plain films may show displacement of bowel loops or diffuse opacification. Ultrasound is able to define the cystic mass, its septations, and wall thickness. Computed tomography is more specific in defining anatomic location, adjacent organ involvement, and size [4].Classically,computed

Fig. 1 Computed tomographic scan image of the abdomen showing a large amount of homogenous fluid with medial displacement of the bowel contents. The liver and spleen are also well visualized. Fig. 3 Intraoperative picture of the omental cyst. Case Reports 101 tomographic scan findings will depict a multiseptated cystic Massive tissue emphysema after mass with homogenous contents. Medial displacement of the cardiopulmonary resuscitation bowel and colon with visible septations is indicative of an abdominal lymphangioma and not ascites [7]. A 77-year-old male patient was referred to our intensive Treatment of abdominal cystic lymphangiomas is with care unit after cardiopulmonary resuscitation (CPR) in a complete surgical excision [8]. Recurrence is rare for peripheral rehabilitation hospital. He was initially admitted to omental cysts, but this is not often the case with mesenteric the rehabilitation hospital for weaning from long-term and retroperitoneal cysts. respirator treatment because of chronic respiratory failure The differential diagnosis of abdominal distension is vast. secondary to a severe pneumonia. Although rare, cystic lymphangiomas should be considered The tracheotomized patient had a history of severe when children have associated vague abdominal complaints. chronic obstructive pulmonary disease, lung emphysema, Computed tomographic scan findings of septations and renal failure, and critical illness polyneuropathy. medial displacement of bowel contents are classic for cystic He suddenly became unconscious and pulseless. An lymphangiomas and should not be confused with ascites. electrocardiogram showed ventricular fibrillation. Defibril- lation and CPR were immediately started, resulting in a sufficient cardiac output after approximately 15 minutes of Christopher T.Y. Hsu MD resuscitation. He was then transported to our hospital for Department of Pediatrics emergency cardiac catheterization assuming cardiac ischemia as the cause for ventricular fibrillation. When the patient Maria Carmen G. Diaz MD was admitted to our intensive care unit, he was ventilated Department of Pediatrics with 100% oxygen, sedated, and showed an increasing Division of Emergency Medicine demand for intravenous catecholamine treatment. In addi- E-mail address: [email protected] tion, the patient displayed a palpable skin emphysema around his chest and upper abdomen. Suspecting a David Rappaport MD pneumothorax, we immediately performed a chest x-ray Department of Pediatrics that confirmed the skin emphysema and demonstrated a Alfred I. duPont Hospital for Children minor apical left-sided pneumothorax (Fig. 1). Meanwhile, Wilmington, DE 19899, USA the subcutaneous emphysema rapidly spread to the head and the lower abdomen, including the scrotum, which became doi:10.1016/j.ajem.2006.05.035 markedly distended (up to a diameter of approximately 25 cm). We also performed a computer tomography (CT)

Acknowledgment

The authors would like to thank Chris Giaquinto, PA for the intraoperative photo.

References

[1] Konen O, Rathus V, Dlugy E, et al. Childhood abdominal cystic lymphangiomas. Pediatr Radiol 2002;32:88-94. [2] Lin JI, Fisher J, Caty MG. Newborn intraabdominal cystic lymphatic malformation. Semin Pediatr Surg 2000;9:141-5. [3] Kosir MA, Sonnino RE, Gauderer MW. Pediatric abdominal lymphangiomas: a plea for early recognition. J Pediatr Surg 1991;26: 1309-13. [4] Losanoff JE, Richman BW, El-Sherif A, et al. Mesenteric cystic lymphangioma. J Am Coll Surg 2003;196:598-603. [5] Luo CC, Huang CS, Chao HC, et al. Intra-abdominal cystic lymphangiomas in infancy and childhood. Chang Gung Med J 2004; 27:509-13. [6] Chung MA, Brandt ML, St-Vil D, Yazbeck S. Mesenteric cysts in children. J Pediatr Surg 1991;26:1306-8. [7] Lugo-Olivieri CH, Taylor GA. CT differentiation of large abdominal lymphangioma from ascites. Pediatr Radiol 1993;23:129-30. Fig. 1 A chest x-ray performed immediately after arrival in the [8] Hancock BJ, St-Vil D, Luks FI, et al. Complications of lymphangiomas intensive care unit demonstrating subcutaneous emphysema and a in children. J Pediatr Surg 1992;27:220-6. minor apical pneumothorax on the left side. 102 Case Reports

unstable and died in refractory cardiogenic shock before cardiac catheterization could be performed. In summary, we describe a patient with massive tissue emphysema after CPR. Most likely, either fractured ribs due to chest compressions or a ruptured bulla had caused the pneumothorax in this patient with known pulmonary emphysema. Of note, in the initial chest x-ray, the pneumothorax was underestimated because the lung was still partly adherent to the lateral wall, and the tissue emphysema mimicked lung vessels in the chest. Only by CT scan the extensive bilateral pneumothoraces and pleural adhesions were visible and facilitated the immediate decompression by chest tube insertion. Auscultation of the lung had not helped finding the diagnosis in this emergency situation, because no breathing sound was audible on both sides (probably because of the massive skin emphysema). We assume that the pneumothorax in combination with high-pressure ventilation had pressed air not only to the chest wall but also through the diaphragm into the abdomen. The present case report demonstrates the diagnostic pitfalls and the fatal consequences of a pneumothorax developing after CPR. Prompt treatment with a chest tube should be considered in a patient with massive tissue emphysema even in the absence of clear evidence for an extensive pneumothorax on the chest x-ray.

Raffi Bekeredjian MD F. Joachim Meyer MD Department of Cardiology E-mail address: [email protected]

Fig. 2 High-resolution CT scan demonstrates (A) bilateral, Boris Radeleff MD mostly ventral, pneumothoraces with adhesion of visceral and Department of Radiology parietal pleura on both lateral chest walls; (B) tissue emphysema in head (left side) and pelvis (right side); (C) free air in the abdomen Norbert Frey MD (transversal plane left and sagittal plane right). Department of Cardiology University of Heidelberg scan of the thorax and abdomen to exclude tracheal or Heidelberg 69120, Germany esophageal rupture. On the contrary, the CT scan demonstrated bilateral ventral pneumothorax predominantly on the doi:10.1016/j.ajem.2006.05.034 left side with adhesions of the visceral and parietal pleura (Fig. 2A). In addition, subcutaneous emphysema could be detected in the head and pelvis (Fig. 2B), and large amounts Succinylcholine-induced masseter muscle rigidity of free air in the peritoneum (Fig. 2C). As a consequence of associated with rapid sequence intubation prolonged chest compressions during the CPR, fractured ribs were observed bilaterally. Succinylcholine is the only depolarizing neuromuscular After immediate insertion of a chest tube on the left side, blocker available in the United States, and it has long been ventilation pressures required for sufficient tidal volumes the drug of choice for rapid sequence intubation for and scrotum diameter instantaneously decreased. Mean- emergency physicians. Succinylcholine has rapid time-to- while, laboratory results showed elevated c-troponin T onset of action and short duration of action that the (0.54Ag/L) and normal creatinine kinase (75U/L). The nondepolarizing neuromuscular blockers lack. electrocardiogram demonstrated a sinus rhythm with a right Unfortunately, there are a number of potential side effects bundle-branch block and no ST elevation. Therefore, we associated with succinylcholine. These include hyperkale- assume that a non–ST-elevation myocardial infarction had mia, malignant hyperthermia, prolonged blockade, fascicu- caused the malignant arrhythmia. Despite continuous lation, histamine release, dysrhythmias, and increased pressure support, the patient remained hemodynamically intracranial, intraocular, and intragastric pressures. Case Reports 103

Of note, succinylcholine has been associated with the Succinylcholine-induced MMR is a potentially life- rare side effect of masseter muscle rigidity (MMR). The threatening complication. MMR is related to several incidence of MMR with pediatric patients is estimated to be myopathic disorders including malignant hyperthermia, less than 1% [1]. Although MMR is widely discussed in Becker dystrophy, and generalized myotonia. MMR as well specialties such as pediatrics and anesthesiology, there is as malignant hyperthermia are thought to be caused by little data available to the emergency medicine physician membrane hyperexcitability secondary to either sodium or concerning the incidence of MMR in adults. chloride channel dysfunction [2]. A 36-year-old black male patient was brought in by EMS Masseter spasm has been linked as an early indicator to the with altered mental status due to an overdose of a bhandfulQ hypermetabolic state of malignant hyperthermia. Early of clonidine 0.1 mg tablets and an unknown amount of clinical findings in malignant hyperthermia include muscle alcohol approximately 3 hours before evaluation. On arrival, rigidity, sinus tachycardia, increased carbon dioxide produc- his blood pressure was 118/82 mm Hg, heart rate 56 beats/ tion, and skin cyanosis with mottling [3]. Marked hyperther- min, respiratory rate 8, oxygen saturation 100% on 10 L by mia (up to 458C[1138F]) occurs minutes to hours later; core face mask, rectal temperature 96.58F, Glascow Coma Scale body temperature tends to rise 18C every 5 to 60 minutes. (GCS) 7, and a finger stick glucose 122. Emergency medical Hypotension, complex dysrhythmias, rhabdomyolysis, elec- system (EMS) reported that the patient was found in his trolyte abnormalities, disseminated intravascular coagula- living room couch, unresponsive to verbal or painful stimuli. tion, and mixed acidosis commonly accompany the elevated The EMS inserted two 18-gauge intravenous catheters in the temperature. Rarely, biochemically proven malignant hyper- bilateral antecubital fossa, placed the patient on non- thermia may present solely with rhabdomyolysis in the rebreather face mask, and administered 4 mg of Narcan. absence of hyperthermia [4,5]. In these cases, dantrolene They reported no response to their therapy. sodium should be administered at 1 mg/kg intravenously until In the ED, the patient was given additional 4 mg of Narcan, symptoms resolve or a maximum dose of 10 mg/kg is given. and a decision was made to orally intubate after a repeat GSC Masseter muscle rigidity is typically managed with of 3. The patient had normal external anatomy, and on initial cessation of the paralytic agent and rescheduling the evaluation had a Mallampati type II airway. The patient was procedure after an evaluation for malignant hyperthermia preoxygenated with a bag valve mask and was given etomidate has been completed. This is not a feasible option in the ED, 30 mg intravenously followed by 1.5 mg/kg of succinylcho- where most of the rapid sequence intubation are emergent. line. Cricoid pressure was applied, and after 60 seconds, an Rather, early recognition of MMR is the best defense attempt was made to intubate the patient with an 8.0-mm against poor outcomes. Anticipation of such side effects can endotracheal tube using a Macintosh 4.0 blade. This was prompt rapid secondary rescue airway attempts. Immediate complicated by an inability to fully open the mouth. The first administration of nondepolarizing agent while ventilation is attempt at endotracheal intubation was complicated by continued can result in a successful intubation. intubation of the esophagus. After withdrawing the endotra- Emergency medicine physicians should be aware of this cheal tube, the patient was ventilated to oxygen saturation of uncommon complication with succinylcholine and be 100%. The second attempt was met by the same difficulty in prepared to manage it. opening the mouth and an inability to visualize the vocal cords. Based on a presumed diagnosis of MMR, 10 mg of Christopher S. Roman DO vecuronium was administered intravenously with resolution Alex Rosin MD of the masseter rigidity. The patient was successfully intubated Department of Emergency Medicine on a third attempt with a Miller 4.0 blade and 8.0 ETT. Repeat Carl R. Darnall Army Medical Center blood pressure ranged 140-173/103-120 mm Hg, pulse 65 to Fort Hood TX 76544, USA 109, and rectal temperature 97.28F. Physical examination was significant for an intubated patient but was otherwise doi:10.1016/j.ajem.2006.05.032 unremarkable. Postintubation chest radiograph confirmed good position of the endotracheal (ET) and nasogastric (NG) tube as well as an arterial blood gas with a pH 7.36, Pco2 References À 40 mm Hg, Po2 131 mm Hg, and HCO3 22 mm Mg. The rest of the laboratory data include a complete blood count, [1] Carroll JB. Increased incidence of masseter spasm in children with strabismus anesthetized with halothane and succinylcholine. Anesthe- chemistry, liver function tests, Tylenol and Aspirin levels, siology 1987;67:559-61. and cardiac enzymes were within normal limit. Urine drug [2] Vital GM, et al. Masseter muscle rigidity associated with glycine sup screen was positive for cocaine. Electrocardiogram showed a 1306-to-alanine mutation in the adult muscle sodium channel alpha normal sinus rhythm with early depolarization. The patient subunit gene. Anesthesiology 1995;82(5):1097-103. was treated with 70 g of charcoal via the NG, whole bowel [3] Litman RS, Rosenberg H. Malignant hyperthermia: update on susceptibility testing. JAMA 2005;293:2918. irrigation with Go-litely, and was admitted to the intensive care [4] Fierobe L, Nivoche Y, Mantz J, et al. Perioperative severe rhabdo- unit. Before being transferred, there were no signs of myolysis revealing susceptibility to malignant hyperthermia. Anesthe- hypermetabolism, and the patient’s ED was unremarkable. siology 1998;88:263. 104 Case Reports

[5] Harwood TN, Nelson TE. Massive postoperative rhabdomyolysis after uneventful surgery: a case report of subclinical malignant hyperthermia. Anesthesiology 1998;88:265.

Pontine hemorrhage and isolated abducens nerve palsy

A 29-year-old man presented to the ED with a chief complaint of diplopia when looking toward the left. His symptoms were present for 5 days. He denied experiencing headache, neck pain, or vomiting. He had no dysphagia or difficulty with speaking. He denied weakness or numbness in the face or extremities. He had no past medical history or any allergy to medications. He admitted to using cocaine occasionally, with his last use being 10 days before the onset of his diplopia. The patient’s physical examination revealed a young healthy man in no acute distress. His vital signs were as follows: blood pressure, 134/90 mm Hg; pulse rate, 82 beats/ min; respiratory rate, 18 breaths/min; and body temperature, 37.08C (98.48F). His pupils were equal and reactive. His visual acuity was 20/20 bilaterally, with no visual field deficit. His funduscopic examination findings were normal. Fig. 2 Computed tomographic scan image of the head showing Evaluation of his extraocular movements revealed left pontine hemorrhage (arrow). lateral gaze palsy (Fig. 1). No other abnormality was appreciated on the remainder of his examination. A computed tomographic scan of the head revealed a left nerve within the pons is rare because of its close anatomical pontine hemorrhage (Fig. 2). Magnetic resonance imaging/ relationship to other structures [3]. angiography confirmed the CT findings found no evidence The workup of a patient with abducens nerve palsy of tumor, dissection, aneurysm, or arteriovenous malforma- depends on the likely underlying cause and on the presence tion. The patient was admitted to the hospital after of other neurologic findings. An abducens nerve palsy is consultation with the neurosurgery service. considered isolated if it is present in the absence of orbital Palsy of the sixth cranial (abducens) nerve is more disease, severe headache, and other neurologic deficits. The common than that of the other nerves that control most common cause of isolated sixth nerve palsy in extraocular movements, accounting for almost 50% of cases patients older than 55 years who have a history of [1,2]. The causes of abducens nerve palsy are numerous and hypertension or diabetes is microvascular disease. In most range from being benign to becoming life-threatening of these patients, the prognosis is good and neuroimaging (Table 1). in the ED is not recommended [4,5]. Close follow-up The abducens nerve originates from the pons in the floor should be arranged because some patients will be diag- of the fourth ventricle, where its nucleus is near the medial longitudinal fasciculus and seventh cranial nerve. From Table 1 Etiology of abducens nerve palsy there, it descends through the pontine tegmentum and emerges at the caudal pons. Isolated injury to the abducens Cause Percentage Microvascular disease (diabetes mellitus/hypertension) 35 Unknown 26 Trauma 12 Multiple sclerosisa 7 Tumor 5 Aneurysm 2 Otherb 15 a Abducens nerve palsy is the presenting feature of multiple sclerosis in 0.5% of cases. b Other causes include increased intracranial pressure (pseudotumor cerebri), herpes zoster, syphilis, Wernicke’s encephalopathy, meningitis, sarcoidosis, post–lumbar puncture complications, operative complications, migraine, petrous apicitis, and sinusitis. Fig. 1 Left abducens nerve palsy. Case Reports 105 nosed with an intracranial neoplasm (2%), neurosarcoidosis ED treatment of brainstem anesthesia after (1%), or multiple sclerosis (0.5%). retrobulbar block In contrast, patients with a nonisolated abducens nerve palsy should receive neuroimaging in the ED. In these Brainstem anesthesia is a serious, but rarely reported, patients, cerebrovascular accident (21%), intracranial neo- complication of retrobulbar blocks performed for ophthal- plasm (18%), and aneurysm (11%) are more common. mologic procedures. Retrobulbar blocks are frequently used Patients with isolated abducens nerve palsy who are younger to provide adequate anesthesia, akinesia, and control of than 55 years and do not have vasculopathic risk factors (ie, intraocular pressure during various ophthalmologic proce- diabetes mellitus or hypertension) are also more likely to have dures, as well as postoperative analgesia. This is the first case a significant pathology that requires neuroimaging. A central of emergency department (ED) treatment after this procedure nervous system mass lesion is present in one third of young caused sustained brainstem anesthesia with seizures. patients with nonisolated abducens nerve palsy [6]. A 75-year-old man was transported to the ED from an Cocaine use is a risk factor for intracerebral hemorrhage outpatient surgical center where he was scheduled for a and may cause hemorrhage within the pons [7-11]. Our trabeculectomy on his right eye. The patient had received patient warranted neuroimaging within the ED because his 1 mg of midazolam and then received 50 mg of propofol with young age, lack of vasculopathic risk factors, and cocaine 20 mg of lidocaine while an ophthalmologist performed a use made it more likely that a pathologic lesion would be retrobulbar block using 4 mL of 0.75% bupivacaine mixed present. However, the finding of a pontine hemorrhage in a with an unknown concentration of lidocaine. Immediately, patient with isolated abducens nerve palsy has not been the patient became unresponsive and apneic. An anesthesi- reported on previously. ologist from an adjacent operating suite administered a small amount of esmolol for hypertension and tachycardia, Scott C. Sherman MD flumazenil (0.4 mg) without effect, and succinylcholine Babak Saadatmand MD (20 mg) and intubated the patient. Because of seizure activity, Department of Emergency Medicine 400 mg of sodium thiopental was administered in the Cook County Hospital (Stroger) ambulance; a small amount of phenylephrine was given Chicago, IL 60612, USA for hypotension. E-mail address: [email protected] Upon arrival in the ED, the patient had a blood pressure of 185/101, sinus rate of 68/min, rectal temperature of 35.48C, doi:10.1016/j.ajem.2006.05.031 and was still unresponsive with mild generalized tonic-clonic activity. The remainder of his examination was significant for an unreactive 5-mm left pupil (presumably due to prior References surgery) and an unreactive 2-mm right pupil (presumably due to the block). Between his episodes of seizure-like activity, he [1] Campis IP, Frenkiel S, Glikstein R, Mohr G. Unilateral sixth cranial nerve palsy caused by skull base mass lesions: case series. withdrew his limbs to pain and received a Glasgow Coma J Otolaryngol 2001;30(3):184-6. Score of 6 T. He had clear lungs, a regular cardiac rhythm, dry [2] Richards BW, Jones Jr FR, Younge BR. Causes and prognosis in 4,278 cases of paralysis of the oculomotor, trochlear, and abducens cranial nerves. Am J Ophthalmol 1992;113(5):489-96. Table 1 Known complications of retrobulbar anesthetic [3] Atilla H, Isikay CT, Kansu T. Isolated sixth nerve palsy from pontine block infarct. Acta Neurol Belg 2000;100(4):246-7. [4] Patel SV, Mutyala S, Leske DA, Hodge DO, Holmes JM. Incidence, Allergic reaction associations, and evaluation of sixth nerve palsy using a population- Amaurosis based method. Ophthalmology 2004;111(2):369-75. Brainstem anesthesia [5] Miller RW, Lee AG, Schiffman JS, Prager TC, Garza R, Jenkins PF, Bruising et al. A practice pathway for the initial diagnostic evaluation of isolated Cardiac arrest sixth cranial nerve palsies. Med Decis Making 1999;19(1):42-8. Cardiovascular depression [6] Peters III GB, Bakri SJ, Krohel GB. Cause and prognosis of Central retinal artery/vein occlusion nontraumatic sixth nerve palsies in young adults. Ophthalmology Chemosis 2002;109(10):1925-8. Corneal injury [7] Wojak JC, Flamm ES. Intracranial hemorrhage and cocaine use. Direct nerve trauma Stroke 1987;18(4):712-5. Ocular myotoxicity [8] McEvoy AW, Kitchen ND, Thomas DG. Intracerebral haemorrhage and Ocular perforation drug abuse in young adults. Br J Neurosurg 2000;14(5):449-54. Oculocardiac reflex (due to hemorrhage) [9] Nolte KB, Brass LM, Fletterick CF. Intracranial hemorrhage Optic atrophy associated with cocaine abuse: a prospective autopsy study. Neurol- Optic nerve sheath hematoma ogy 1996;46(5):1291-6. Pain [10] Egido-Herrero JA, Gonzalez JL. Pontine hemorrhage after abuse of Postoperative ptosis cocaine. LetterRev Neurol 1997;25(137):137-8. Retrobulbar/peribulbar hemorrhage [11] Ramadan NM, Levine SR, Welch KM. Pontine hemorrhage following Toxic regional nerve effect bcrackQ cocaine use. Neurology 1991;41(6):946-7. 106 Case Reports

adequate respiratory and hemodynamic support usually Table 2 Manifestations of brainstem anesthesia results in a full recovery within 48 hours. Aphasia Brainstem anesthesia probably occurs by direct injection Apnea into the optic nerve sheath during a deep retrobulbar Bradycardia injection but may occur via intraarterial injection of Cardiac arrest Confusion anesthetic with retrograde flow through the ophthalmic Diaphoresis artery into the cerebral circulation and midbrain area. Dilatation of the contralateral pupil As ophthalmologic procedures are increasingly per- Drowsiness formed in outpatient surgical centers where full anesthesio- Dysphagia logical support may not be available, emergency physicians Facial paralysis Gaze palsy should be prepared to rapidly diagnose and stabilize Hypertension affected patients. Loss of consciousness Nausea and vomiting James M. Dahle MD Seizures Kenneth V. Iserson MD Shivering Tachycardia Department of Emergency Medicine University of Arizona Tucson, AZ 85724, USA skin, and no evidence of emesis. Remarkable laboratory tests doi:10.1016/j.ajem.2006.05.030 included: glucose 221, lactate 3.7, pH 7.35, Pco2 25, and Po2 417. A head computed tomography was unremarkable. In the ED, the patient was given standard doses of References lorazepam, phenytoin, and phenobarbital to control seizures. His labile blood pressure was controlled with [1] Donlon J, Doyle D, Feldman M. Anesthesia for eye, ear, nose, and alternating nitroprusside drips and boluses of 0.9% saline. throat surgery. In: Miller R, editor. Miller’s anesthesia, 6th ed. Intensivists later used a propofol drip for sedation. A Philadelphia (PA)7 Elsevier Churchill Livingstone; 2005. p. 2527-31. magnetic resonance image showed chronic ischemic [2] Mayer AS, O’Connor RE. Respiratory arrest after local anesthesia for outpatient cataract surgery: a dramatic but transient complication. Ann changes and old lacunar infarcts. The sedation was weaned Emerg Med 1993;22:1357-9. 48 hours later, and the patient was successfully extubated. [3] Nicoll JM, Acharya PA, Ahlen K, Baguneid S, Edge KR. Central He remained confused for another 24 hours, subsequently nervous system complications after 6000 retrobulbar blocks. Anesth becoming alert and oriented. On day 4, he was discharged Analg 1987;66:1298-302. from the hospital in his usual state of health. [4] Wittpenn JR, Rapoza P, Sternberg P, Kuwashima L, Saklad J, Patz A. Respiratory arrest following retrobulbar anesthesia. Ophthalmology Retrobulbar blocks anesthetize the ciliary ganglion; ciliary 1986;93:867-70. nerves (sensory innervation of the cornea, iris, and ciliary [5] Ahmad S, Ahmad A. Complications of ophthalmologic nerve blocks: a body); and cranial nerves II, III, and VI. The (parasympa- review. J Clin Anesth 2003;15:564-9. thetic) ciliary ganglion sits between the lateral surface of the [6] Javitt JC, Addiego A, Friedberg HL, Libonati MM, Leahy JJ. optic nerve and the ophthalmic artery, approximately 1 cm Brainstem anesthesia after retrobulbar block. Ophthalmology 1987; 97:718-24. from the posterior boundary of the orbit. [7] Rodman DJ, Notaro S, Peer GL. Respiratory depression following A retrobulbar block is typically performed with the patient retrobulbar bupivacaine: three case reports and literature review. supine and looking forward. A 25-gauge, 3-cm needle is Ophthalmic Surg 1987;18:768-71. inserted at the inferolateral border of the bony orbit and directed toward the apex of the orbit. After entering the orbital muscle cone, the physician injects 2 to 4 mL of anesthetic [1]. Fatal afebrile streptococcal meningitis in a chronic The most commonly used local anesthetic agents are a 1:1 alcoholic patient mixture of 2% lidocaine with 0.5 or 0.75% bupivacaine, occasionally with added epinephrine or hyaluronidase. A 53-year-old man with a history of alcoholism and Complications of retrobulbar anesthetic blocks (Table 1) hypertension presented to the emergency department (ED) are rare but can be severe. One case of transient brainstem complaining of a severe thunderclap-like headache, coupled anesthesia has been reported in the emergency medicine with nausea and a single episode of vomiting. He had been literature [2]. encouraged to seek medical attention by his supervisor at The frequency of brainstem anesthesia after retrobulbar work, who noted the patient to blook different.Q On blocks varies from 0.13% to 0.79% [3,4]. Although usually admission, he had a temperature of 35.68C, pulse of occurring within 15 minutes, it can be delayed up to 114 beats per minute, blood pressure of 158/83 mmHg, 40 minutes after the block. Although different organ and respiratory rate of 18 breaths per minute. The physical systems can be affected [1,3,5-7],(Table 2), prompt, examination was remarkable for pale, diaphoretic skin, but Case Reports 107

drain placement, the cerebrospinal fluid (CSF) was noted to be turbid, whereas the brain appeared pale. The intracranial pressure was 10 mm Hg intraoperatively, and the drain was kept on continuous drainage to maintain an intracranial pressure of 0 mm Hg. Subsequent analysis of the CSF demonstrated the presence of Streptococcus pneumoniae. The CSF also contained 1053 white blood cells per milliliter, 115 red blood cells per milliliter, 33 mg/dL of glucose, and 161 mg/dL of protein. A postoperative magnetic resonance image demonstrated a small fourth ventricle consistent with aqueductal stenosis. The pneumococcus infection was treated with ceftriaxone and vancomycin. He was also given dexamethasone, mannitol, divalproex sodium, and phenytoin. The patient was maintained on mechanical ventilation postoperatively because of an inability to protect his airway and remained in a deep coma. Neurologically, the patient had a GCS score of 5 T after surgery, was unresponsive, and assumed a decerberate posture. His pupils were sluggish to react to light and he had diminished brain stem reflexes for the remainder of his hospitalization. In addition, on the first postoperative day, the patient’s peripheral blood white blood cell count rose to 18600 cells per milliliter, which peaked the following day at 23400 cells per milliliter, steadily decreasing thereafter. The cerebral spinal fluid white blood cell count returned to Fig. 1 Preoperative axial computed tomographic scan demon- normal within 72 hours after initiating antibiotic therapy. strating marked dilation of the lateral ventricles. Seven days after his initial presentation to the ED, with no change in the patient’s condition, the family elected to terminate mechanical ventilation, and unable to maintain otherwise normal. During the initial neurologic examina- respirations on his own, the patient immediately expired. tion, the patient was alert and oriented to person, place, and Although bacterial meningitis occurs fairly frequently, it time; lacked motor or sensory deficits; and had a score of has rarely occurs with such complete absence of signs and 15 on the Glasgow Coma Scale (GCS). He also had an symptoms upon immediate presentation to the ED. In fact, absence of both Kernig’s and Brudzinsky’s signs and denied this is the first case report to describe a patient with headache any neck pain or stiffness. The initial white blood cell count and emesis but an absence of any other overt signs of demonstrated 8600 cells per milliliter. He did not report any meningitis. Likely confounding the patient’s case and recent illness or trauma. possibly obscuring several of the signs of meningitis is the An initial computed tomography of the head demon- patient’s history of chronic alcoholism. Alcoholism is known strated the absence of any acute intracranial hemorrhage, to place individuals at a greater risk for developing a infarction, or mass effect (Fig. 1). However, the presence of multitude of infections as well as alters the normal immune prominent dilation of the lateral ventricles was noted. A response to infection [1-3]. In addition, many of the classic subsequent, magnetic resonance imaging and magnetic signs of meningitis depend on the presence of an inflamma- resonance angiography of the brain showed dilated lateral tory response. Fever is driven by a rise in inflammatory ventricles, with normal vascular anatomy and no evidence cytokines [4], whereas nuchal rigidity, Kernig’s sign, and of vasculitis. Four hours after initially arriving at the ED, the Brudzinski’s sign are indications of meningeal inflammation patient had become increasingly confused with a GCS score [5,6]. However, chronic alcoholics initially experience a of 13. Shortly thereafter, he had a tonic-clonic seizure hypoinflammatory response to infection, which would lasting 30 seconds and was given loreazepam and phenyt- preclude the development of fever by limiting the production oin. The GCS score after the seizure was 10. of cytokines [2]. Alcohol consumption also decreases the Given that the ventricle size could not adequately production of chemokines and depresses leukocyte migration explain the progressive course of the patient’s deterioration [7,8], which can limit the inflammatory mediated meningeal and a concern that lumber puncture could have been irritation that results in nuchal rigidity and a positive Kernig’s deleterious in the face of ventricular enlargement, the and Brudzinsky’s signs. These untoward effects of alcohol- patient was taken to the operating room for an emergency ism are likely to have contributed to the dirth of signs of right frontal external ventricular drain placement. During infection that this patient initially presented with. 108 Case Reports

Ramsis F. Ghaly MD [5] Verghese A, Gallenmore G. Kernig’s and Brudzinski’s signs revisited. Ghaly Neurosurgical Associates Rev Infect Dis 1987;1187-92. Aurora, IL 60504, USA [6] Reid H, Fallon RJ. Bacterial infections. In: Adams JH, Duchen LW, editors. Greenfield’s neuropathology. 5th ed. New York7 Oxford John Stroger Hospital of Cook County University Press; 1992. p. 311-2. Chicago, IL, USA [7] Szabo G, Chavan S, Mandrekar P, Catalano D. Acute alcohol E-mail address: [email protected] consumption attenuates interleukin-8 (IL-8) and monocyte chemo- attractant peptide (MCP-1) induction in response to ex vivo stimula- William P. Gibbons MD tion. J Clin Immunol 1999;19:67-76. [8] Saeed RW, Varma S, Peng T, Tracey KJ, Sherry B, Metz CN. Ethanol Rush Copley Medical Center Department of Pathology blocks leukocyte recruitment and endothelial cell activation in vivo and Aurora, IL, USA in vitro. J Immunol 2004;173:6376-83.

Timothy P. Plackett DO Chicago College of Osteopathic Medicine Atrial flutter with cardiac tamponade as initial Downers Grove, IL 60612, USA presentation of tuberculosis pericarditis doi:10.1016/j.ajem.2006.05.026 Cardiac tamponade is a true medical emergency requiring early recognition and access. Clinical clues of cardiac tamponade include tachypnea, dyspnea on exertion, water- References bag feature of heart on chest radiograph, low-amplitude [1] Friedman H, Newton C, Klein TW. Microbial infections, immunomo- QRS complex on electrocardiogram (ECG), pulsus para- dulation, and drugs of abuse. Clin Microbiol Rev 2003;16:209-19. doxus and unexplained pulseless electrical activity. Atrial [2] Frank J, Witte K, Schrodl W, Schutt C. Chronic alcoholism causes flutter is a rare presenting rhythm in cardiac tamponade. deleterious conditioning of innate immunity. Alcohol 2004;39:386-92. A healthy 61-year-old man presented with a 10-day [3] Messingham KA, Faunce DE, Kovacs EJ. Alcohol, injury, and cellular immunity. Alcohol 2002;28:137-49. history of left-side chest tightness, poor appetite, general [4] Koedel U, Scheld WM, Pfister H-W. Pathogenesis and pathophysiology malaise, and dyspnea on exertion. He had no previous of pneumococcal meningitis. Lancet Infect Dis 2002;2:721-36. cardiovascular illness history. Because of fatigue, he went

Fig. 1 Initial ECG presents as atrial flutter. Case Reports 109

Fig. 2 ECG presents as electrical alternans and atrial flutter. to the emergency department (ED) for help. There were no globulin M, thyroid stimulation hormone, free T4, and HIV fever, no tarry stool, no limb edema, no abdominal pain, antibody showed all negative findings. Cytology of pericar- and no cough during the last 10 days. Only mild runny dial effusion was negative for malignancy. Pericardial nose and sore throat were noticed before that. A chest effusion culture finally confirmed the pathogen Mycobacte- radiograph was ordered as an initial survey, which rium tuberculosis. Follow-up echocardiography demonstrat- demonstrated cardiomegaly. Electrocardiogram showed ed few residual effusions with thickened pericardium. The atrial flutter (Fig. 1). The blood investigations at ED patient was discharged with oral drugs for M.tuberculosis revealed white blood cell count, 10900/mm3; platelet count, 459000/mm3; hemoglobin, 11.4 g/dL; serum urea nitrogen, 19 mg/dL; serum creatinine, 1.1 mg/dL; sugar, 173 mg/dL; creatine kinase–MB, 1.70 ng/mL; Troponin I, b0.02 ng/mL; serum amylase, 54 U/L; and lipase, 101 U/L. Heart failure was impressed at first so admission was advised. Unfortunately, the condition of patient deteriorated rapidly in ED observation area, with manifestation as cold sweating, respiratory distress, and shock. Repeated ECG still revealed atrial flutter (Fig. 2). Emergent 2-dimensional cardiac echocardiography with Doppler demonstrated large amount of pericardial effusion with diastolic right ventricle compression (Fig. 3). The clinical cardiac echocardiography features were consistent with cardiac tamponade. The patient received catheter pericardiocentesis soon and was admitted to intensive care unit. Improvement in patient’s condition resulted in a stable hemodynamic status and smooth respiratory pattern. Laboratory results of antinuclear antibodies, C3, C4, cyto- Fig. 3 Cardiac echo reveals massive pericardial effusion with megalovirus immunoglobulin G, cytomegalovirus immuno- right ventricle diastolic compression. 110 Case Reports after a 41-day of hospitalization. Electrocardiogram after References discharge revealed normal sinus rhythm. Cardiac tamponade is caused by fluid accumulation [1] Palacios IF. Pericardial effusion and tamponade. Curr Treat Options around the cardiac chambers that influence cardiac output Oncol 1999;1(1):79-89. significantly [1]. It is also an important differential diagnosis [2] Naclerio EA. Penetrating wounds of the heart. Experience with in patients of pulseless electrical activity. There are many 249 patients. Dis Chest 1964;46:1-22. causes of acute cardiac tamponade, including trauma, post– [3] Merce J, Sagrista Sauleda J, Permanyer Miralda G, Carballo J, Olona J, Soler Soler J. Pericardial effusion in the elderly: a different disease? myocardial infarct rupture, and aortic dissection. The Rev Esp Cardiol 2000;53(11):1432-6. incidence of acute pericardial tamponade is approximately [4] Pousset F, Le Heuzey JY, Pialoux G, et al. Cardiac lymphoma 2% of penetrating trauma [2]. Malignancy, tuberculosis, and presenting as atrial flutter in an AIDS patient. Eur Heart J 1994; idiopathic pericarditis are common etiologies of subacute 15(6):862-4. tamponade [3]. [5] Ogimoto A, Hamada M, Shigematsu Y, et al. Cardiac tamponade with paroxysmal atrial flutter controlled by antituberculous therapy. Nippon Atrial flutter is a rare manifestation of cardiac tamponade Ronen Igakkai Zasshi 2004;41(1):112-6. [4,5]. It is usually associated with organic heart disease, [6] Bruch C, Schmermund A, Dagres N, et al. Changes in QRS voltage in such as ischemic heart disease, acute myocardial infarction, cardiac tamponade and pericardial effusion: reversibility after peri- caridomyopathy, myocarditis, and, less common, pulmonary cardiocentesis and after anti-inflammatory drug treatment. J Am Coll embolism. The most common ECG sign of large pericardial Cardiol 2001;38(1):219-26. [7] Maisch B, Seferovic PM, Ristic AD, et al. Guidelines on the diagnosis effusion is low voltage of the QRS complexes [6]. Sinus and management of pericardial diseases executive summary; the task tachycardia with combination of low voltage may imply force on the diagnosis and management of pericardial diseases of the pericardial effusion with tamponade. Chronic obstructive European society of cardiology. Eur Heart J 2004;25(7):587-610. pulmonary disease, cardiomyopathy, pleural effusion, and [8] Cheitlin MD, Armstrong WF, Aurigemma GP, et al. ACC/AHA/ASE condition after open heart surgery, however, could also have 2003 guideline update for the clinical application of echocardiography: summary article: a report of the American College of similar patterns. Electrical alternans with sinus tachycardia Cardiology/American Heart Association Task Force on Practice is a more specific sign in cardiac tamponade, but the Guidelines (ACC/AHA/ASE Committee to Update the 1997 Guide- absence of it does not exclude cardiac tamponade. lines for the Clinical Application of Echocardiography). Circulation Echocardiography now is a standard diagnostic tool for 2003;108(9):1146-62. pericardial effusion and cardiac tamponade [7-9]. Diastolic [9] Cheitlin MD, Alpert JS, Armstrong WF, et al. ACC/AHA Guidelines for the Clinical Application of Echocardiography. A report of the compression of right atrium and right ventricle on 2-dimen- American College of Cardiology/American Heart Association Task sional and Doppler echocardiography suggests cardiac Force on Practice Guidelines (Committee on Clinical Application of tamponade. Elective treatment may be done earlier if there Echocardiography). Developed in collaboration with the American is no sign of shock but echocardiographic findings of Society of Echocardiography. Circulation 1997;95(6):1686-744. tamponade present. Fluid resuscitation is still the first step [10] Salem K, Mulji A, Lonn E. Echocardiographically guided pericardiocentesis—the gold standard for the management of peri- of treatment in hemodynamic unstable patients with cardiac cardial effusion and cardiac tamponade. Can J Cardiol 1999;15(11): tamponade, and next is removal of pericardial effusion, 1251-5. whether by catheter pericardiocentesis or immediate surgery [7,10]. Catheter pericardiocentesis now is a favorable therapy in most clinical physicians because of its accessi- bility and low complication rate under cardiac echocardi- Importance of bedside echocardiography for detection ography guidance. of unsuspected isolated right ventricular infarction as a Our case emphasis that, although very rare, atrial flutter cause of cardiovascular collapse can be the ECG sign of cardiac tamponade. Thus, urgent further survey, especially cardiac echocardiography, is Isolated right ventricular infarction is a rare and important for patients with new onset of atrial flutter. Atrial commonly overlooked condition, because of scarce electro- flutter should be considered as a possible ECG finding in cardiographic changes. We present 2 cases of isolated right patients with cardiac tamponade. ventricular infarction that went unrecognized in the ED. Both patients presented with signs of circulatory collapse, I-Chuan Chen MD third-degree atrioventricular block, and without prominent Te-Fa Chiu MD ischemic changes in the electrocardiogram. Diagnosis was Jih-Chang Chen MD further obfuscated by the absence of chest pain and atypical Department of Emergency Medicine clinical presentation with left-hand paresis in one patient Chang Gung Memorial Hospital and hypothermia in the second. Echocardiography was Chang Gung University College of Medicine performed as part of the evaluation of unexplained Taoyuan, Taiwan, ROC hypotension. In both cases, it revealed right ventricular E-mail address: [email protected] dysfunction, which suggested possible acute right ventricular infarction. Diagnosis was confirmed by the finding of doi:10.1016/j.ajem.2006.05.023 subtotal ostial thrombotic occlusions of the right coronary Case Reports 111

Fig. 1 Standard electrocardiogram recorded at the time of admission to the emergency department showing complete atrioventricular block and very subtle ST-segment elevations in inferior leads. artery in both cases. Echocardiography played a decisive unexplained neurologic deficit, he was admitted to the role in the diagnostic procedure, enabling prompt therapeu- intensive care unit. Cardiac necrosis markers on admission tic intervention and subsequent resolution of shock. were negative, and chest x-ray was normal. Arterial blood Isolated right ventricular infarction is a rare event gases were normal. We could not confirm any paresis in the present in less than 5% of autopsies [1]. History and left upper extremity. To clarify the patient’s hemodynamic clinical presentation are unremarkable in one third of cases condition, we performed echocardiography and found [2]. Electrocardiogram (ECG) usually shows ischemic normal left chambers without any regional wall motion changes in inferior wall and typical changes in V4R lead. defects, but dilated and hypokinetic right ventricle (Fig. 2). However, ECG may also lack evident signs of myocardial Acute isolated right ventricular infarction was suspected as a ischemia, which often obscures the diagnosis [3,4]. Echocardiography represents an easily accessible and noninvasive diagnostic option which often resolves clinical dilemma. We present 2 cases of acute isolated right ventricular infarction associated with circulatory collapse that were unrecognized until echocardiographically demonstrated severe right ventricular failure directed suspicion toward the correct underlying cause. A 78-year-old man without any chronic disease presented to the emergency department because of weakness. He reported paleness and sweating, while denying chest pain, palpitations, or dyspnea. In addition, his main complaint was paraesthesia and decreased strength in his left hand. On admission, his blood pressure was 70/35 mm Hg and pulse was weak with a rate of 35 per minute. Lung fields were clear on auscultation and heart sounds were normal. The emergency physician documented a slightly decreased strength in the patient’s left hand and flexion of the elbow. Fig. 2 Echocardiogram recording from the first patient. Record- Electrocardiogram demonstrated complete atrioventricular ing during systole shows dilated right ventricle with absent block and rather subtle ST-segment elevations in inferior contraction of the lateral wall. The right ventricle is larger than leads and minimal ST-segment depression in lateral leads the left ventricle. RV indicates right ventricle; LV, left ventricle; (Fig. 1). Because of suspected Adams-Stokes syndrome and RA, right atrium. 112 Case Reports

Fig. 3 Standard electrocardiogram from the second patient. It demonstrates junctional rhythm and unspecific widespread changes in ST segment and T waves in inferior and anterior leads. Note fine artefacts due to patient’s tremor associated with hypothermia. possible cause and confirmed with a right-sided ECG leads unit. Echocardiography was performed as part of the recording. Patient received isotonic saline and was referred evaluation of circulatory shock of uncertain etiology. It to the catheterization laboratory. We found single-vessel demonstrated decreased left ventricular systolic function disease with an ostial subtotal thrombotic lesion of the right with hypokinesia of the left ventricular posterior wall and, coronary artery (RCA). After direct stenting, TIMI 3 flow most notably, dilated and hypokinetic right ventricle (Fig. 4). resorted and right-sided ECG changes resolved. The Subsequent ECG recording of right-sided leads showed ST- remaining course was uneventful and the patient was segment elevations. Urgent coronarography demonstrated discharged without any sequelae. ostial occlusion of the RCA and chronic occlusion of the left The second patient was a 74-year-old man who was circumflex artery with autocollaterals (Fig. 5). After intra- admitted because of hypothermia, hypotension, and som- aortic balloon pump insertion, target lesion was successfully nolence. On the day of admission, he went to work in the forest and was found lying in the stream several hours later. On admission, he was somnolent, hypothermic at 318C, and his blood pressure was 70/50 mm Hg; his pulse was weak with a rate of 32 per minute. Neck veins were engorged, lung fields were clear on auscultation, and heart sounds were normal without any murmurs. Besides obtundation, there were no neurologic deficits. During the laboratory examination, we registered increased urea and creatinine; troponin I levels were normal at the time of admission. Arterial blood gas analysis demonstrated metabolic lactic acidosis (pH 7, 21; pCO2, 4.0 kPa; HCO3, 12.1 mmol/L; pO2, 10.6 kPa) and increased lactate to 9.7 mmol/L. Electrocardiogram showed junctional rhythm, with a ventricular rate of 25 per minute, slight ST-segment elevation, and inverted T waves in D III and aVF, as well as concave type ST-segment elevation in anterior leads (Fig. 3). Electrocardiographic changes were judged to be nonspecific, especially in the context of hypothermia and hypoten- Fig. 4 Echocardiogram recording from the second patient sion, and the patient was transferred to the intensive care demonstrating similar characteristics as in the first patient. Case Reports 113

Fig. 5 Coronary arteriogram of the RCA from the second patient. On the left panel, ostial occlusion of the RCA is evident. Right panel shows large RCA after successful balloon angioplasty and stent implantation.

resolved by percutaneous balloon angioplasty and stent at least partially, account for bradycardia and decreased implantation. After the procedure, the patient’s status slowly systolic function of the left ventricle. Nevertheless, exceed- improved; he was weaned from intra-aortic balloon counter- ingly compromised left and right ventricular function with pulsation on day 3. Subsequent course was uneventful. respect to the severity of hypothermia indicated an Determining the etiology of cardiovascular collapse in additional cause of compromise in myocardial contractility. critically ill patients is crucial to a rapid and appropriate In this patient, more widespread ischemic ECG changes and therapeutic response. Besides other possible mechanisms, additional posterior wall hypokinesia were a consequence of such as left ventricular dysfunction, or inadequate preload, double-vessel coronary occlusion. However, acute deterio- severe right ventricular dysfunction due to isolated myo- ration with shock was caused by acute right ventricular cardial infarction of the right ventricle is a rare but infarction due to ostial RCA occlusion, as left circumflex potentially reversible cause of hypotension. Usually, isolat- artery occlusion was old and was autocollateralized. ed right ventricular infarction does not present a diagnostic Echocardiography has been previously shown to provide dilemma because of ischemic changes in standard 12-lead valuable information on a hypotensive patient with regard to ECG. In the first patient, ECG did show subtle ischemic volume status, contractility, and valvular function [5-7]. changes in inferior leads, suggesting inferior wall ischemia. Evaluation of right ventricular function and dimensions In the second patient, ECG was even less conclusive owing provides information with regard to preload status, as well to more widespread ST-segment and T-wave changes. as a potential cause of hypotension, such as pulmonary However, in both patients, the complexity of clinical embolism and right ventricular dysfunction. As such, presentation and confusing findings at admission blurred echocardiography was also endorsed by emergency medi- the underlying pathologic event. Echocardiography was cine physicians who use it to perform and interpret goal- used as part of the diagnostic workup of a hemodynamically oriented examinations and reduce time to diagnosis [8]. compromised patient with no evident underlying cause of Right ventricular myocardial infarction is a known deter- his status. Although we cannot exclude the possibility of minant of short- and long-term prognosis which is amenable coincident involvement of the left ventricular inferior wall, to treatment [3,9,10]. In one study, clinical signs of right absence of any segmental left ventricular contractility ventricular dysfunction were present in only one tenth of defects by echocardiography indicates that, even if such patients admitted with diagnosis of acute inferior wall did exist, it was not clinically significant. Finding of myocardial infarction. However, subsequent echocardiogra- severely compromised right ventricular function with phy and ECG demonstrated associated right ventricular preserved left ventricular contractility in hypotensive infarction in 37% of these patients [9]. The acute in-hospital patients generally suggests pulmonary embolism or right mortality rate of patients with right ventricular infarction is ventricular infarction as possible causes. In our case, right 3 times higher than in those with inferior infarction [3,9]. ventricular dysfunction in combination with bradycardia and Rapid determination of correct diagnosis with timely absence of hypoxemia made diagnosis of pulmonary treatment is therefore crucial. Echocardiographic detection embolism much less likely and pointed toward a correct of isolated or predominant right ventricular dysfunction in underlying cause. In the second patient, hypothermia could, hypotensive patients is an important marker of possible 114 Case Reports infarction of the right ventricle and should lead to further disease [2,3]. Recent publications also confirm that myo- attempts to exclude or confirm this diagnosis. cardial infarctions (MIs) can be seen on MDCT scan, but These cases illustrate the diversity of clinical presentation enrolled patients in these studies were in the subacute or in acute isolated right ventricular infarction. We emphasize chronic stage of MI [4-6]. The clinical utility of MDCT to the decisive role of echocardiography in reducing the time demonstrate acute MI has been reported in some cases with to diagnosis and expedite patient care. It should represent a non–ST-elevation MI [1,7-9]. Herein, we present a case of standard investigation in patients with unexplained cardio- hyperacute stage of ST-elevation MI incidentally diagnosed vascular presentation. early by this imaging modality before the electrocardiogram (ECG) or cardiac biomarkers became positive. Mitja Lainscak MD, MSc A 52-year-old man presented to our emergency depart- General Hospital Murska Sobota ment (ED) with 20-minute symptoms of persistent anterior Murska Sobota, Slovenia chest pain radiating to back associated with diaphoresis. He had a history of malignant melanoma of the right big toe Andrej Pernat MD, PhD postoperation. Physical examination was unremarkable. The Department of Cardiology initial ECG showed normal sinus rhythm with no evidence of University Medical Centre Ljubljana ischemic changes, and cardiac enzyme analysis was within SI-1000 Ljubljana, Slovenia normal limits, with a creatine kinase (CK) of 34 U/L, CK-MB E-mail address: [email protected] of 1 U/L, and troponin I of 0.04 ng/mL. Initial treatment included oxygen, aspirin, nitroglycerin, and morphine. A doi:10.1016/j.ajem.2006.05.022 cardiologist was consulted because of unresolved symptoms, and bedside echocardiogram was interpreted as normal left ventricular function with an ejection fraction of 67%. A References 40-slice, non–ECG-gated MDCT angiography (Philips, Cleveland, OH; 40 Â 0.625-mm collimation, 500-millisec- [1] Andersen HR, Falk E, Nielsen D. Right ventricular infarction: ond rotation, 120 kV) was ordered to evaluate for potential frequency, size, and topography in coronary heart disease. J Am Coll Cardiol 1987;10:1223-32. aortic dissection, which demonstrated normal aorta but an [2] Kinch JW, Ryan TJ. Right ventricular infarction. N Engl J Med 1994; area of transmural hypodensity with normal wall thickening 330:1211-7. along the anterior wall of the left ventricular and poor [3] Zehender M, Kasper W, Kauder E, et al. Right ventricular infarction as contrast-perfused left anterior descending coronary artery an independent predictor of prognosis after acute inferior myocardial (LAD) suggesting an occlusion (Fig. 1). The repeated ECG infarction. N Engl J Med 1993;328:981-8. [4] Antonelli D, Schiller D, Kaufman N, Barzilay J. Isolated right 10 minutes after MDCT scan showed ST-segment elevation ventricular infarction: a diagnostic challenge. Cardiology 1984;71: in leads V2 through V5. Primary coronary angiography was 273-6. performed immediately and revealed a total occlusion of the [5] Sanfilippo A, Weyman A. The role of echocardiography in managing middle LAD (Fig. 2). The patient underwent successful critically ill patients. J Crit Illn 1988;3:27-44. percutaneous transluminal coronary angioplasty with stent [6] Oh JK, Seward JB, Khandheria BK, et al. Transesophageal echocardiography in critically ill patients. Am J Cardiol 1990;66:1492-5. placement. Serial cardiac enzyme analysis confirmed a MI [7] Heidenreich PA, Stainback RF, Redberg RF, et al. Transesophageal with an early peaking CK value of 2224 U/L (7% MB form) at echocardiography predicts mortality in critically ill patients with 6 hours after onset of symptoms. The rest of the hospitali- unexplained hypotension. J Am Coll Cardiol 1995;26:152-8. zation course was uneventful. [8] Durham B. Emergency medicine physicians saving time with Evaluating patients who present with symptoms sugges- ultrasound. Am J Emerg Med 1996;14:309-13. [9] Chockalingam A, Gnanavelu G, Subramaniam T, et al. Right tive of acute coronary syndrome (ACS) remains a clinical ventricular myocardial infarction: presentation and acute outcomes. challenge in the ED. Acute coronary syndrome constitutes a Angiology 2005;56:371-6. variety of clinical presentations, and other differential [10] Skali H, Zornoff LA, Pfeffer MA, et al. Prognostic use of echocardi- diagnosis such as abnormalities of the great vessels, ography 1 year after a myocardial infarction. Am Heart J 2005;150: pericardial disease, or gastrointestinal disorders may mimic 743-9. it. Early diagnosis of ST-elevation MI is essential goal because it indicates the potential for a substantial irreversible infarction and is the primary indication for emergent Acute myocardial infarction diagnosed early by reperfusion therapy to salvage the myocardium. multidetector computed tomography The 12-lead ECG and cardiac enzyme levels are used to screen a suspected patient for ACS. It is known that a The development of multidector computed tomography normal initial ECG does not preclude the diagnosis of acute (MDCT) has allowed assessment of not only coronary artery MI [10], and cardiac enzymes may require 4 to 8 hours to stenoses and occlusion, but also coronary artery plaques and have elapsed after the onset of coronary artery occlusion. myocardial perfusion [1]. Previous literature has empha- The subsidiary noninvasive echocardiography is highly sized its usefulness for detecting obstructive coronary artery dependent on the operator and reader, and may be less Case Reports 115

Fig. 1 A 40-slice non–ECG-gated contrast-enhanced MDCT delineated the myocardium, LAD, and left circumflex coronary artery. A, A transmural hypodense area along the anterior wall of the left ventricle (arrowhead). B, Poor contrast-perfused LAD (arrowhead). C, Poor contrast-perfused LAD (arrowhead) compared with the left circumflex coronary artery (arrow). useful in patients who have preexisting wall motion Considering the fast rotation time, multislice acquisition, abnormalities. Therefore, serial follow-up of ECG and and the high spatial resolution, MDCT is able to provide a cardiac enzymes every 2 to 4 hours is a validated practice comprehensive assessment about cardiac morphology and in managing patients with ACS. coronary angiography. In our case, an MDCT was ordered to

Fig. 2 The right anterior oblique view of conventional coronary angiography delineated the LAD in the pre- and postangioplasty stages. A, A total occlusion of the middle LAD (arrowhead). B, LAD in postangioplasty stage showing normal contrast perfusion (arrowhead). 116 Case Reports exclude aortic dissection and incidentally demonstrated [2] Nieman K, Rensing BJ, van Geuns RJM, et al. Usefulness of early the hyperacute stage of ST-elevation MI before the multislice computed tomography for detecting obstructive coronary ECG or cardiac biomarkers became evidence. Patel et al [8] artery disease. Am J Cardiol 2002;89:913-8. [3] Nieman K, Cademartiri F, Lemos PA, et al. Reliable non-invasive have reported incidentally discovered MI in 6 patients coronary angiography with fast submillimeter multislice spiral receiving MDCT scans for non–cardiac-related indications. computed tomography. Circulation 2002;106:2051-4. Myocardial infarction can be visible on contrast-enhanced [4] Gosalia A, Haramati LB, Sheth MP, et al. CT detection of acute MDCT scan even when cardiac gating is not used [8]. The myocardial infarction. AJR Am J Roentgenol 2004;182:1563-6. infarcted region is visualized as a segmental area of [5] Nikolaou K, Knez A, Sagmeister S, et al. Assessment of myocardial infarctions using multidetector-row computed tomography. J Comput hypoattenuation in comparison with the normal enhanced Assist Tomogr 2004;28:286-92. myocardium during the arterial phase [4–6]. The normal [6] Francone M, Carbone I, Danti M, et al. ECG-gated multi-detector row wall thickness of the infarcted segment is consistently spiral CT in the assessment of myocardial infarction: correlation with associated with a recent event and might be a useful non-invasive angiographic findings. Eur Radiol 2006;16:15-24. criterion to differentiate recent (b1 month) from nonrecent [7] Paul JF, Dmabrin G, Caussin C, et al. Sixteen-slice computed N tomography after acute myocardial infarction: from perfusion defect ( 1 month) MI [6]. In the report of Fancone et al [6], using to the culprit lesion. Circulation 2003;108:373-4. MDCT in the detection of MI had a diagnostic accuracy, [8] Patel R, Lewis D, Dubinsky TJ. Myocardial infarction non-ECG sensitivity, and specificity of 91%, 83%, and 91%, synchronized contrast-enhanced multi-detector computed tomography. respectively. Sato et al [1] have indicated the feasibility of Emerg Radiol 2005;11:301-5. MDCT in the ED for detecting patients with ACS who did [9] Zeina AR, Orlov I, Blinder J, et al. Atypical presentation of acute myocardial infarction in a young man diagnosed by multidetector not manifest ECG and enzymatic evidence. computed tomography. Isr Med Assoc J 2006;8:69-70. In conclusion, further research should be aimed at [10] Menown IB, Mackenzie G, Adgey AA. Optimizing the initial 12-lead determining the consistency and accuracy of MDCT for electrocardiographic diagnosis of acute myocardial infarction. Eur ACS. When the heart is included in a CT study, one should Heart J 2000;21:275-83. be alert to potential findings that indicate acute or chronic myocardial disease. Patients with symptoms suggestive of ACS, negative initial ECG, and missing cardiac enzymes may benefit from this efficient diagnostic modality. Life-threatening epistaxis from a rare treatable cause

Epistaxis after head injury is a common occurrence; Ju-Sing Fan MD however, epistaxis caused by a traumatic pseudoaneurysm Chorng-Kuang How MD of the internal carotid artery (ICA) is rare. This condition Chii-Hwa Chern MD can be potentially fatal if unrecognized or diagnosed Lee-Min Wang MD late, and its management differs from that of epistaxis of Chun-I Huang MD other causes. Emergency Department Taipei Veterans General Hospital Taipei 112, Taiwan, ROC National Yang-Ming University School of Medicine Taipei 112, Taiwan, ROC E-mail address: [email protected]

Jen-Dar Chen MD Radiology Department Taipei Veterans General Hospital Taipei 112, Taiwan, ROC National Yang-Ming University School of Medicine Taipei 112, Taiwan, ROC doi:10.1016/j.ajem.2006.05.021

References

[1] Sato Y, Matsumoto N, Ichikawa M, et al. Efficacy of multislice computed tomography for the detection of acute coronary syndrome in Fig. 1 Digital subtraction angiogram showing an aneurysm in the emergency department. Circ J 2005;69:1047-51. the cavernous part of the ICA. Case Reports 117

of the left ICA projecting into the sphenoid sinus with opacification of the sphenoid sinus around a pseudoaneurysm, which was consistent with a blood clot. An urgent coil embolization of the left ICA was performed on the same day with rapid occlusion of the left ICA. No residual filling of the pseudoaneurysm occurred, and there was good cross-flow into the left hemisphere. The patient was closely monitored, had no complication, and made an uneventful recovery from the procedure. Craniofacial trauma is the second most common cause of internal carotid aneurysms, after arteriosclerosis. Although most cases occur after severe head trauma associated with skull base fractures, postoperative and spontaneous nontraumatic aneurysms have been reported. Most traumatic aneurysms of the ICA are based in the cavernous portion of the ICA and rupture in the sphenoid sinus. In most documented cases, the aneurysm involved the intracavernous ICA with erosion and rupture into the sphenoid sinus [1], typically presenting as massive delayed Fig. 2 Digital subtraction angiogram showing an embolic coil in epistaxis. Recurrent epistaxes occur as a result of a small situ and occlusion of the aneurysm. rupture that tamponades from a hematoma in the closed space of the sphenoid sinus [2]. The third, fourth, and sixth cranial nerves as well as the first and second divisions of the We present the case of an 18-year-old male patient who V cranial nerve, passing through the cavernous sinus, can be sustained a skull base fracture and blindness in one eye after involved [3]. experiencing a road traffic accident. He presented with Aneurysm of the ICA should be suspected in all patients repeated bouts of epistaxis refractory to treatment until with head injury who present with massive epistaxis. The computed tomography (CT) angiography was conducted, hallmark of this condition is delayed, repetitive, and profuse revealing a traumatic pseudoaneurysm in the cavernous epistaxis occurring within 3 weeks of the initial injury; portion of the ICA (Fig. 1). This was successfully treated however, massive hemorrhage may be delayed by months or with balloon embolization of the ICA (Fig. 2). years [4]. The early recognition of the condition, consideration In a review of 100 patients presenting with epistaxis of the etiology of delayed and repetitive epistaxis particu- resulting from a pseudoaneurysm, Chambers et al [4] found larly in relation to head injury, and prompt management that 54% had initial epistaxis after 1 month and that there are emphasized. was a 4-month delay between the occurrence of the epistaxis An 18-year-old male patient was brought to the AED and diagnosis in 75% of the cases. with a Glasgow Coma Scale score of 3/15 and a left Clinically, the finding of the classic triad of epistaxis, pupillary defect. He was intubated and ventilated; a CT scan head injury, and monocular blindness, constituting the triad revealed a skull base fracture and a small left frontal set forth by Maurer et al [5], is pathognomonic of a extradural hematoma. traumatic pseudoaneurysm of the ICA. Immediate loss of The patient was managed conservatively and recovered, vision may be caused by permanent damage to the optic except for left blindness, owing to optic nerve injury. nerve or the ophthalmic artery [2]. One week later, the patient had an episode of epistaxis, Although carotid angiography is the gold standard for which was managed with anterior nasal packing, and suspected vascular trauma, CT is the investigation most subsequently underwent nasal endoscopy, which did not often done for head injuries. Computed tomography scans reveal any obvious source of bleeding. He was discharged with fractures involving the sphenoid sinus, sella turcica, with ophthalmologic follow-up. and orbital apex or unilateral opacification of the sphenoid Over the next 2 months, the patient had multiple bouts of sinus should arouse suspicion for an ICA aneurysm, and a epistaxis managed with anterior and posterior nasal packing; carotid angiography should follow the CT scan. Interest- eventually, however, he was readmitted with torrential ingly, 6% of initial angiograms may be normal [2] and an epistaxis and a hemoglobin level of 5.0 g/L. He underwent aneurysm is demonstrated only at a subsequent study an emergency anterior ethmoidal artery ligation, which performed at the time of epistaxis. temporarily settled his epistaxis, and was then taken up for Early treatment of the ICA aneurysm cannot be septoplasty and left sphenopalatine artery ligation 2 weeks emphasized enough. The mortality rate of posttraumatic later. He consequently underwent repeat CT scan and CT aneurysms with rupture into the sphenoid sinus ranges from angiography, which revealed a traumatic pseudoaneurysm 30% to 50% [2]. 118 Case Reports

Emergency control of epistaxis will require anterior and Score (GCS): E1V1M5, right hemiparesis, and blood posterior nasal packing, which will buy time for definitive pressure of 199/134 mm Hg. Fundus examination showed neuroradiologic investigations and treatment. papilledema with flame-shaped hemorrhage. Emergent brain Open surgical techniques, such as ligation of the computed tomography showed intracranial hemorrhage common carotid artery, the external carotid artery, or the (ICH) (about 15 Â 20 Â 30 mm) with mild perifocal edema ICA in the neck, are historical and have no place in near the left side foramen of Monro. Rupture of the hema- the management of ICA traumatic aneurysms. With the toma into the ventricle and hydrocephalus with mild midline advent of endovascular techniques, open clipping of the shift to right were also noted (Fig. 1). Emergent neurosurgery aneurysm has taken a backseat. with intraventricular intracranial pressure (ICP) monitor Endovascular occlusion of the aneurysm with detachable was performed. coils or balloons is the order of the day. A meta-analysis On neurologic examination, her GCS score was suggested that balloon occlusion and coil embolization are E1VEM4. Right hemiplegia with right positive Babinski reasonably safe and result in occlusion of the aneurysm in sign was found. Laboratory investigations were all normal. most patients [6]. Her chest x- ray showed mild cardiomegaly, and the electrocardiogram (ECG) showed evidence of left Mohammed Iqbal Syed MD ventricular hypertrophy (Fig. 2). Abdominal ultrasound Andy Chin MD showed a small left kidney. Doppler study showed no Louise Clark MD evidence of renal artery stenosis. Abdominal computed Southern General Hospital tomography angiography showed a small left kidney (right South Glasgow University Hospitals 10.4 cm vs left 4.9 cm) with irregular size and compensatory National Health Service Trust enlargement of right kidney, suggesting chronic renal Glasgow G51 3PZ, UK parenchymal disease. E-mail address: [email protected] Testing for antinuclear antibody was negative. Serum complement was normal. The other findings showed doi:10.1016/j.ajem.2006.05.020 the following: plasma renin, 49.4 pg/mL; aldosterone, 303 pg/mL; urinary vanillylmandelic acid, 13.2 mg/d. Urine References dipstick testing revealed 2+ protein and 2+ occult blood.

[1] Goleas J, Paige ML, Mikhael MA, Wolff AP. Intracavernous artery aneurysms presenting as recurrent epistaxis. Ann Otol Rhino Laryngol 1991;100:577-9. [2] Lee JP, Wang ADJ. Epistaxis due to traumatic intracavernous aneurysm. J Trauma 1990;30:619-22. [3] Kaplan HA, Browder J, Krieger AJ. Intracavernous connections of the cavernous sinus. J Neurosurg 1976;45:166-8. [4] Chambers EF, Rosenblum AE, Norman D, et al. Traumatic aneurysm of cavernous internal carotid artery with secondary epistaxis. AJNR Am J Neuroradiol 1981;2:405-9. [5] Maurer MY, Mills M, German WJ. Triad of unilateral blindness, orbital fracture and massive epistaxis in head injury. J Neurosurg 1961;18: 837-46. [6] Van Der Schaaf IC, Brilstra EH, Buskens E, Rinkel GJ. Endovascular treatment of aneurysms in the cavernous sinus: a systematic review on balloon occlusion of the parent vessel and embolisation with coils. Stroke 2002;33:313-8.

Spontaneous intracranial hemorrhage caused by sustained hypertension in a child

A previously healthy 10-year-old girl presented with history of sudden onset of headache with vomiting followed by unconsciousness within minutes. She was then immediately brought to our pediatric emergency department (ED) on Fig. 1 Brain computed tomography showed ICH (about 15 Â January 2, 2006. There was no preceding history of trauma, 20 Â 30 mm) with mild perifocal edema near the left side foramen seizures, diaphoresis, palpitations, or blurring of vision. At of Monro. Rupture of the hematoma into the ventricle and the ED, the patient was unconscious with Glasgow Coma hydrocephalus with mild midline shift to right were also noted. Case Reports 119

Fig. 2 ECG showed evidence of left ventricular hypertrophy.

Cerebral angiogram, obtained 1 week later, revealed no treatment and hypertension complicated with intracranial vascular a nomaly (F ig. 3A, B, C). Hype rtension was poorl y hemorrhage are rare in the literature. controlled with medication. After 1 month, improved blood Jainn-Jim Lin MD pressure and improved consciousness, as well as right side Chang-Teng Wu MD weakness were also noted. Shao-Hsuan Hsia MD Systemic hypertension complicated with spontaneous Division of Pediatric Critical and Emergency Medicine intracranial hemorrhage is a rare occurrence in children. The Chang Gung Children’s Hospital common cause of spontaneous intracranial hemorrhage in Chang Gung University Medical Collage children included vascular malformation (33.8%), hemato- Taoyuan 333, Taiwan, ROC logical (17.6%) or coagulation (14.7%) disorders, and E-mail address: [email protected] hemorrhagic brain tumors (13.2%) [1]. The 1999 Taiwanese Kuang-Lin Lin MD survey by Lin et al [2] revealed that the leading cause of Huei-Shyong Wang MD bleeding was arteriovenous malformation. Although sys- Division of Pediatric Neurology temic arterial hypertension is a common cause of brain Chang Gung Children’s Hospital hemorrhage in adults, it is rarely found in children with Chang Gung University Medical Collage brain hemorrhage. Taoyuan 333, Taiwan, ROC Sustained severe hypertension in children can almost always be related to a definite cause. In young children, doi:10.1016/j.ajem.2006.05.019 hypertension is often the result of renal parenchymal disease such as chronic glomerulonephritis, reflux nephropathy, and obstructive uropathy. Renal artery stenosis and cardiovas- References cular disorders, like coarctation of the aorta, which are less frequent causes of hypertension in this age group, are [1] Jarallah A, Rifai MT, Riela AR, Roach ES. Nontraumatic brain usually detected within the first decade of life [3,4]. hemorrhage in children: etiology and presentation. J Child Neurol Most patients with mild to moderate hypertension are 2000;15:284-9. asymptomatic. The manifestations of hypertension could be [2] Lin CL, Loh JK, Kwan AL, Howng SL. Spontaneous intracerebral related to complications of severe hypertension, such as hemorrhage in children. Kaohsiung J Med Sci 1999;15:146-51. congestive heart failure and hypertensive encephalopathy, or [3] Hari P, Bagga A, Srivastava RN. Sustained hypertension in children. Indian Pediatr 2000;37:268-74. may be nonspecific in a small proportion of children. In our [4] Gill DG, Mendis da Costa B, Cameron JD, Joseph MC, Ogg CS, case, the initial presentations are nonspecific. However, the Chantler C. Analyses of 100 children with severe and persistent continuous presence of sustained hypertension despite hypertension. Arch Dis Child 1976;51:951-6. 120 Case Reports

A rare diagnosis in ED: cerebral pyogenic ventriculitis due to infective endocarditis

Pyogenic ventriculitis is a rarely seen infection of the central nervous system, and it is characterized by the presence of suppurative fluid in the ventricles. Pyogenic ventriculitis is a well-known complication of shunting procedures, intraventricular surgery, and intrathecal chemo- therapy but is not secondary to infective endocarditis [1]. A 69-year-old woman presented with a complaint of acute confusion, nausea, and vomiting to our ED. She suffered ischemic stroke 5 weeks ago. She was discharged 3 weeks before with left sixth cranial nerve paralysis as a sequel. Her medical history was remarkable for hypertension, heart failure, mitral valve prolapse, and rheumatoid arthritis. She was on flucortolone, silazapril, acetyl salicylic acid, and rofecoxib therapy. On physical examination, she was lethargic with a temperature of 36.48C, pulse of 65/min, and blood pressure of 140/80 mm Hg. All extremity muscle strength was found 2-3/5 to the formal rating scale. Cranial nerve examination was intact except for left abducens palsy. She had no nuchal rigidity. Deep tendon reflexes were hypo- active, and there was no plantar response. On cardiac auscultation, she had 3/6 systolic murmur at left lower sternal border and apex. Initial complete blood count, urea, creatine, electrolytes, and liver function tests were normal except for mild leukocytosis (15.300 l/L). Computed tomography (CT) revealed asymmetric dilatation of left lateral ventricle and hypodensities in the right occipital horn (Fig. 1). Magnetic resonance (MR) showed hyperintense periventricular signal around both ventricles, and debris in posterior horn of right lateral ventricle with communicant-type hydrocephalus on T2-weighted images (Fig. 2). Contrast enhancement of the choroid plexus and complete ependymal enhancement were found at the inner side of the ventricular system on T1-weighted images (Fig. 3). The cerebrospinal fluid (CSF) obtained by lumbar puncture had the appearance of frank pus and contained many neutrophils (300 leukocyte/lL), elevated protein (256 mg/dL), reduced glucose (27 mg/dL), and Gram-positive cocci. Blood and CSF cultures were obtained, and intravenous ceftriaxone and metronidazole were admin- istrated for the diagnosis of pyogenic ventriculitis. Because the patient had a history of mitral valve prolapse, intermittent fever in the ED, and systolic murmur on examination, transesophageal echocardiography was performed. Severe mitral regurgitation and prolapse into left atrium due to rupture of the posterior leaflet, large vegetation on the posterior leaflet with apical punctative calcification were revealed. Infective endocarditis was considered as the focus of pyogenic ventriculitis. Vancomycin was added in the Fig. 3 Cerebral angiogram revealed no definite angiographic therapy, and she was admitted to the intensive care unit. An evidence of intracranial vascular malformation or aneurysm. (A) external ventricular drainage catheter was placed immediate- left interbal carotid artery AP view (B) left internal carotid artery ly, and patient status has improved. Her blood culture was lateral view (C) left verteral artery PA view. positive for coagulase-negative staphylococcus on the 11th Case Reports 121

Fig. 1 Hydrocephalus and irregular debris within the right ventricle on CT. Intermediate hypodensities, most likely represent- Fig. 3 Complete ependymal enhancement was found at the inner ing inflammatory debris, are layering in and occupying much of side of right ventricular system on T1-weighted images (1). the dependent portions of the ventricular cavity, especially in the occipital horns of the lateral ventricles (1). The lateral ventricles are paranasal sinusitis, hepatic abscess, and intraabdominal enlarged out of proportion to the size of the cerebral sulci, abscess associated with bowel perforation [1,2]. There is consistent with an obstructive hydrocephalus (2). no report about pyogenic ventriculitis associated with infective endocarditis. day. In the course of her hospital stay, she had suffered acute Microorganisms mostly identified to be associated with reinfection of the posterior parietal area. The patient’s pyogenic ventriculitis are Gram-negative bacteria, including condition has deteriorated progressively, and she died on Bacteroides, Enterobacter, Escherichia coli,andKleb- the 47th day of admission. siealla, followed by Gram-positive microorganism, com- Pyogenic ventriculitis is a rare type of intracranial monly Staphylococcus species [3,4]. In the literature, features infection. Without appropriate treatment, it is mostly fatal; of lumbar puncture and CSF analysis in patients with therefore, early diagnosis and treatment are essential. It is pyogenic ventriculitis had not been extensively discussed. generally associated with complicated brain surgery, rup- Decreased mental status, nuchal rigidity, cranial nerve tured parenchymal brain cerebritis, abscesses, extension of paralysis, focal neurological deficits, and extensor plantar meningitis into the ventricles, and infected brain malforma- response may be found in physical examination [1,2]. tions [2]. Pyogenic ventriculitis may originate from septic Symptoms, physical examination findings, and CSF analysis focus such as multiple dental caries, otitis media or of pyogenic ventriculitis are nonspecific. Therefore, differential diagnosis between ventriculitis and other central nervous system infection should only be made with imaging studies. The CT findings of pyogenic ventriculitis are hydrocephalus, pial enhancement, and debris in subarachnoid space and in ventricles. Subependymal enhancement on contrast- enhanced CT is a key diagnostic finding for ventriculitis [3]. The MR is superior to demonstrate enhancement and increased signal of pus into the ventricular system. The most common finding in MR is the presence of debris, especially seen in the occipital horn of the lateral ventricles. All of these diagnostic findings were detected in our patient. Half of the cases with infective endocarditis have at least one complication, among which, cardiac and neurological events predominate. Neurological complications develop in 20% to 40% of cases, most of which are stroke or transient ischemic attacks [5]. Central nervous system infections, Fig. 2 Hyperintense periventricular signal around both ventricle such as brain abscess and aseptic meningitis, are well- and debris in posterior horn of right lateral ventricle with communi- known complications of infective endocarditis but not cant-type hydrocephalus on T2-weighted fluid attenuated inversion pyogenic ventriculitis. recovery (FLAIR) images: 1 indicates ventricular debris; 2, hyperin- Emergency physicians should be aware of pyogenic tense periventricular signal; 3, asymmetrical left ventricle dilatation. ventriculitis in patients presenting with headache, altered 122 Case Reports mental status, and CT findings such as a hydrocephalus, tachycardia and a high DScv-mvo2 level prompted a debris in ventricles, and pial enhancement. Irregular technetium-99m–ethyl cysteinate dimer–labeled white blood ventricular debris and pial enhancement on MR should be cell scan, which revealed increased tracer uptake in the considered as diagnostic findings of pyogenic ventriculitis. myocardium that was suggestive of active myocardial In patients with a diagnosis of ventriculitis, infective inflammation. The patient’s condition improved consistently o endocarditis should be kept in mind as a septic focus; with mechanical ventilation and dobutamine. The DScv-mv 2 therefore, echocardiography should be performed especially decreased from 38 to 7 by day 6. The patient was weaned off in the absence of apparent source. ventilation and dobutamine therapy by day 12 and was discharged on day 21. She had no evidence of cardiomyop- Selahattin Kiyan MD athy 1 year after this episode of fulminant myocarditis. Ersin Aksay MD Case patient 2 was a 30-year-old man with a history Murat Ersel MD suggestive of unprotected sex who had presentation, early Department of Emergency Medicine management, and nuclear scan profiles similar to those of Ege University Hospital case patient 1. He had a pulmonary arterial wedge pressure of o o 35100, Bornova, Izmir, Turkey 22 mm Hg, an Smv 2 level of 45%, an Scv 2 level of 74%, o E-mail address: [email protected] DScv-mv 2 level of 29, cytomegalovirus ELISA–positive findings, HIV ELISA–negative findings, an absolute Sedat Yanturali MD CD4 count of 70 cells/lL, and indeterminate anti-HIV Department of Emergency Medicine antibody paper chromatography findings. The patient’s o Dokuz Eylul University Hospital DScv-mv 2 level continued to remain between 20 and Izmir, Turkey 25 for 5 days and decreased rapidly to 5 on day 6 with accompanying intractable hypotension. He died of his illness doi:10.1016/j.ajem.2006.05.018 on day 9. Fulminant myocarditis was probably secondary to acute cytomegalovirus infection and suspected acute References HIV infection. Venous oximetry has not been elucidated in fulminant [1] Barloon TJ, Yuh WT, Knepper LE, et al. Cerebral ventriculitis: MR myocarditis. A brief review of venous oximetry in shock findings. J Comput Assist Tomogr 1990;14:272-5. follows. In septic shock, the presence of a higher coronary [2] Bakshi R, Kinkel P, Mechtler L, et al. Cerebral ventricular empyema sinus oxygen saturation (Scso2) level in human beings [1], associated with severe adult pyogenic meningitis: computed tomogra- coexistence of organ dysfunction with increased blood flow phy findings. Clin Neurol Neurosurg 1997;99:252-5. indicating that global ischemia is not a principal mechanism [3] Fukui M, Williams R, Mudigonda S. CT and MR imaging features of of organ dysfunction [2], and an Smvo level rarely lower pyogenic ventriculitis. AJNR Am J Neuroradiol 2001;22:1510-6. 2 [4] Durand M, Calderwood S, Weber D, et al. Acute bacterial meningitis in than 65% [3] suggest that septic shock is not characterized adults: a review of 493 episodes. N Engl J Med 1993;328:21-8. by a high DScv-mvo2 level. In heart failure, causal [5] Hill EE, Herijgers P, Herregods MC, et al. Evolving trends in infective association of low Smvo2 with low Scvo2[4] and DScv- endocarditis. Clin Microbiol Infect 2006;12(1):5-12. mvo2 being lower than 7 [5] suggest that low Smvo2 in cardiogenic shock is secondary to low Scvo2, with very Should early venous oximetry be indicated in suspected o B minimal increase in DScv-mv 2. cases of fulminant myocarditis? Early diagnosis of fulminant myocarditis is a daunting task and no criterion reliably predicts mortality [6].Magnetic In this study, we explored the importance of an unreported resonance imaging and scintigraphy are emerging diagnostic o abnormality in venous oximetry, a very high DScv-mv 2 modalities in myocarditis [7,8], and high rates of apoptosis o level (central venous oxygen saturation, Scv 2; mixed with rapid loss of functional cardiomyocytes appear to venous oxygen saturation, Smvo2; difference between the influence the outcome, suggesting that apoptosis markers help o 2, DScv-mv 2), in the early diagnosis and management of with the prognosis [9]. Recently, an Smvo2 level higher than fulminant myocarditis. 60% has been proposed as the standard for ceasing percuta- Upon intensive care unit admission after a weeklong viral neous cardiopulmonary support in fulminant myocarditis [10]. prodrome, a 30-year-old woman had hypotension, breath- In both of our cases, acute respiratory distress syndrome o lessness, a Pa 2 level of 56 mm Hg, a serum lactate level with septic shock was the initial diagnosis and scintigraphy of 5.6 mmol/L, an ejection fraction of 45%, a pulmonary confirmatory of fulminant myocarditis was prompted by a o arterial wedge pressure of 25 mm Hg, an Smv 2 level of high DScv-mvo2 level as an early physiologic indicator of 42%, an Scvo2 level of 80%, DScv-mvo2 level of 38, normal myocardial inflammation. In case patient 2, acute hemody- cardiac enzymes, total leukocyte count finding of 4900/lL, namic worsening on day 6 was probably on the account of and a platelet count of 30000/lL. On day 2, persistent rapid myocardial apoptosis, which was reflected in a sharp decline in the DScv-mvo2. This report underscores the need B No financial or other potential conflicts of interest exist among for better understanding of venous oximetry in the diagnosis the authors. and management of fulminant myocarditis. Case Reports 123

Arvind K. Baronia MD References Afzal Azim MD Gagan Narula MD [1] Cunnion RE, Schaer GL, Parker MM, et al. The coronary circulation Mohan Gurjar MD in human septic shock. Circulation 1986;73:637-44. Banani Poddar MD [2] Giantomasso DD, May CN, Bellomo R. Vital organ blood flow during Department of Critical Care Medicine hyperdynamic sepsis. Chest 2003;124:1053-9. Sanjay Gandhi Postgraduate Institute of Medical Science [3] Krafft P, Steltzer H, Hiesmayr M, et al. Mixed venous oxygen Lucknow, Uttar Pradesh 226014, India saturation in critically ill septic shock patients: the role of defined E-mail addresses: [email protected] events. Chest 1993;103:900-6. [4] Ander DS, Jaggi M, Rivers E, et al. Undetected cardiogenic shock in [email protected], [email protected] patients with congestive heart failure presenting to the emergency [email protected], [email protected] department. Am J Cardiol 1998;82:888-91. [5] Chawla LS, Zia H, Gutierrez G, et al. Lack of equivalence bet- Sudeep Kumar DM ween central and mixed venous oxygen saturation. Chest 2004;126: Department of Cardiology 1891-6. Sanjay Gandhi Postgraduate Institute of Medical Science [6] McCarthy RE, Boehmer JP, Hruban RH, et al. Long-term outcome of fulminant myocarditis as compared with acute (nonfulminant) Lucknow, Uttar Pradesh 226014, India myocarditis. N Engl J Med 2000;342:690-5. E-mail address: [email protected] Sanjay [7] Cohen K, Claeys ES, Steg PG. MRI of acute myocarditis: a comprehensive approach based on various imaging sequences. Sanjay Gambhir MD Chest 2002;122:1638-48. Sukanta Barai MD [8] Sarda L, Colin P, Boccara F, et al. Myocarditis in patients with clinical Department of Nuclear Medicine presentation of myocardial infarction and normal coronary angio- Sanjay Gandhi Postgraduate Institute of Medical Science grams. J Am Coll Cardiol 2001;37(3):786-92. [9] Kytf V, Saukko P, Henn V, et al. Apoptotic cardiomyocyte death in Lucknow, Uttar Pradesh 226014, India fatal myocarditis. Am J Cardiol 2004;94:746-50. E-mail address: [email protected] [10] Aoyama N, Izumi T, Hiramori K, et al. National survey of fulminant [email protected] myocarditis in Japan—therapeutic guidelines and long-term prognosis of using percutaneous cardiopulmonary support for fulminant doi:10.1016/j.ajem.2006.05.017 myocarditis. Circ J 2002;66:133-44. American Journal of Emergency Medicine (2007) 25, 124–131

www.elsevier.com/locate/ajem

Correspondence Thiamine for the treatment of acute decompensated The primary outcome measure was admission to the heart failureB hospital. The decision to admit or discharge the patient was made by the treating attending physician 4 hours after To the Editor, enrollment. Our hospital does not have an observation unit. The secondary outcome measures were total in-hospital Thiamine deficiency is very common in patients with length of stay and change in dyspnea between enrollment heart failure [1]. Previous studies have suggested that and 4 hours postenrollment. Dyspnea was assessed by the thiamine supplementation may be an effective treatment of research staff on a 6-point ordinal scale at time of arrival, chronic heart failure [2,3]. It is not clear if it may also be enrollment, and 4 hours after enrollment. effective in the acute setting. Two other conditions Enrollment in this pilot study was limited to a total associated with thiamine deficiency, Wernicke’s encepha- of 50 patients because of financial constraints. This sample lopathy and beriberi, have been known to improve within size had a power of 80% to detect an absolute reduction in hours of thiamine supplementation [4,5]. hospitalization of 35%. Data were analyzed using STATA We conducted a study that was approved by the software version 8 and Cytel Studio software (StatXact & investigational review board at Baystate Medical Center LogXact) version 6 [6,7]. (Springfield, Mass). Consecutive emergency department A total of 65 met entry criteria of which 50 gave (ED) patients were enrolled during times when research consent. Complete data sets were available for 49 pa- staff were available. Adult patients with a primary diagnosis tients. Baseline characteristics were similar for both of acute decompensated heart failure were eligible if they groups with the possible exception for dyspnea score had a history of heart failure and were currently on a loop (Table 1). Primary and secondary outcome measures are diuretic. Acute decompensated heart failure was determined listed in Table 2. None was statistically significant (Fisher by the treating clinician based on worsening dyspnea within exact test and Wilcoxon-Mann-Whitney 2-sample rank the past 24 hours, rales, and chest x-ray findings of vascular sum test). redistribution. Patients were excluded if (1) they were On multivariable analysis using exact polytomous currently on thiamine supplementation, (2) had a diagnosis regression (adjacent category model), no statistical differ- of renal failure being treated with dialysis, (3) had a history ence was found for 4-hour change in dyspnea score of pulmonary disease (ie, asthma, chronic obstructive controlling for dyspnea score on arrival and at enrollment pulmonary disease, or emphysema), or (4) had a concurrent ED diagnosis other than heart failure. Patients were randomized to either the placebo or the thiamine group. Group allocation was stored in sequentially Table 1 Baseline variables numbered opaque envelopes. The patients, health care Thiamine group Control group providers, and research personnel were blinded to the group n2524 allocation. Age (y) 64 (54-73) 68 (62-75) Male sex (%) 64 71 Patients in the thiamine group received 100 mg thiamine Ethnicity/race (%) intravenously within 30 minutes of arrival to the ED. Caucasian 64 67 Patients randomized to the placebo group received an equal African American 24 29 volume of normal saline (1 mL). All other care was at the Hispanic 12 4 discretion of the treating clinician. In addition, at the time of Systolic blood pressure 140 (100-163) 138 (110-152) Serum enrollment, the treating attending physician was asked by Sodium 137 (131-146) 140 (134-149) the research staff if the patient would be admitted to the Creatinine 1.6 (1.4-2.0) 1.7 (1.5-2.1) hospital if the dyspnea had sufficiently resolved within Hemoglobin 11 (9-12) 10 (9-11) 4 hours regardless of the results of any diagnostic testing. Dyspnea score On arrival 4 (2-5) 3 (1-5) On enrollment 3 (2-5) 3 (1-4)

B Data are presented as percentages or medians (interquartile range). This study was presented at the SAEM 1999 (Boston, Mass).

Table 2 Outcome measures References Thiamine Control P group group [1] Hanninen SA, Darling PB, Sole MJ, Barr A, Keith ME. The prevalence of thiamin deficiency in hospitalized patients with n2524congestive heart failure. J Am Coll Cardiol 2006;47(2):354-61. Hospitalized (%) 92 96 N.99 [2] Seligmann H, Halkin H, Rauchfleisch S, et al. Thiamine deficiency in In-hospital length of stay (d) 4 (2-5) 3 (1.5-4) .11 patients with congestive heart failure receiving long-term furosemide Dyspnea score 4 h after enrollment 2 (1-3) 2 (1-3) .85 therapy: a pilot study. Am J Med 1991;91:151-5. 4-h Change in dyspnea score 1 (1-2) 1 (0-1) .09 [3] Shimon I, Almog S, Vered Z, et al. Improved left ventricular function Data are presented as percentages or medians (interquartile range). after thiamine supplementation in patients with congestive heart failure receiving long-term furosemide therapy. Am J Med 1995; 98:485-90. ( P = .052). In addition, no statistical difference was found [4] Brust J. Nutritional disorders of the nervous system. In: Goldman L, Bennett J, editors. Cecil textbook of medicine. 21st ed. W. B. for duration of hospitalization controlling for dyspnea score Saunders; 2000. on arrival ( P = .721) using exact regression. All models [5] Diltoer MW, Troubleyn J, Lauwers R, et al. Ketosis and cardiac controlled for dyspnea score at time of enrollment. failure: common signs of a single condition. Eur J Emerg Med 2004; At time of enrollment, when asked if the patient would be 11(3):172-5. admitted even if the dyspnea had sufficiently resolved [6] Stata statistical software [computer program]. Version 8. College 7 within 4 hours regardless of diagnostic testing results, the Station (TX) Stata Corporation; 2003. b Q [7] Cytel studio statistical software [computer program]. Version 6. treating clinician answered yes for 88% of the patients. Cambridge (MA)7 CYTEL Corporation; 2004. Thiamine deficiency has been shown to occur in about [8] Zenuk C, Healey J, Donnelly J, Vaillancourt R, Almalki Y, Smith S. 40% of patients with heart failure [1-3,8-15]. Inhibition of Thiamine deficiency in congestive heart failure patients receiving long the pyruvate dehydrogenase complex is the presumed term furosemide therapy. Can J Clin Pharmacol 2003;10(4):184-8. mechanism by which thiamine deficiency exacerbates heart [9] Brady J, Rock C, Horneffer M. Thiamin status, diuretic medications, and the management of congestive heart failure. J Am Diet Assoc failure [16]. Inhibition of pyruvate dehydrogenase complex 1995;95:541-4. has been postulated to decrease the efficiency of adenosine [10] Pfitzenmeyer P, Guilland JC, d’Athis P, Petit-Marnier C, Gaudet M. triphosphate production, increase cellular acidosis, and Thiamine status of elderly patients with cardiac failure including the increase free fatty acid levels [16-21]. effects of supplementation. Int J Vitam Nutr Res 1994;64(2):113-8. Animal models have demonstrated that thiamine defi- [11] Kwok T, Falconer-Smith J, Potter J, Ives D. Thiamine status of elderly patients with cardiac failure. Age Ageing 1992;21:67-71. ciency causes cardiac hypertrophy, depressed cardiac [12] Hardig L, Daae C, Dellborg M, Kontny F, Bohmer T. Reduced contractility, and dysrhythmias in the absence of beriberi thiamine phosphate, but not thiamine diphosphate, in erythrocytes in [16,22-32]. Three prospective human studies with chronic elderly patients with congestive heart failure treated with furosemide. congestive heart failure treated with thiamine yielded J Intern Med 2000;247:597-600. conflicting results, but all had significant methodological [13] Pepersack T, Garbusinski J, Robberecht J, Beyer I, Willems D, Fuss M. Clinical relevance of thiamine status amongst hospitalized elderly flaws [2,3,10]. patients. Gerontology 1999;45:96-101. Our pilot study had several limitations: (1) the primary [14] O’Keeffe ST, Tormey WP, Glasgow R, Lavan JN. Thiamine outcome measure is prone to significant confounding deficiency in hospitalized elderly patients. Gerontology 1994;40(1): issues; (2) the sample size was small; (3) we did not 18-24. measure thiamine levels; (4) and the groups were not [15] Levy W, Soine L, Huth M, Fishbein D. Thiamine deficiency in congestive heart failure. Am J Med 1992;93:705-6. balanced on dyspnea score on arrival to the ED, suggesting [16] Sutherland DJ, Jaussi AW, Gubler CJ. The effects of thiamine that the thiamine group had more severe heart failure. deprivation, and oxythiamine- and pyrithiamine-treatment on cardiac We did not detect a significant change in rate of function and metabolism in the rat. J Nutr Sci Vitaminol (Tokyo) hospitalization, 4-hour change in dyspnea, or length of 1974;20(1):35-54. hospitalization as a result of thiamine supplementation. [17] Singleton C, Martin P. Molecular mechanisms of thiamine utilization. Curr Mol Med 2001;1:197-207. However, thiamine supplementation appeared to cause a [18] Haas R. Thiamin and the brain. Annu Rev Nur 1988;8:483-515. trend for improvement in 4-hour change in dyspnea when [19] Rathanaswami P, Pourany A, Sundaresan R. Effects of thiamine baseline dyspnea score was controlled for. For most patients deficiency on the secretion of insulin and the metabolism of glucose in with acute decompensated heart failure, the decision to admit isolated rat pancreatic islets. Biochem Int 1991;25(3):577-83. was not based on the severity of dyspnea at time of admission. [20] Rathanaswami P, Sundaresan R. Effects of thiamine deficiency on the biosynthesis of insulin in rats. Biochem Int 1991;24(6):1057-62. [21] Iwata H. Catecholamine metabolism in thiamine-deficient rats. J Nutr Howard A. Smithline MD Sci Vitaminol (Tokyo) 1976;22(Suppl):25-7. Department of Emergency Medicine [22] Boullin DJ. Pharmacological responses of thiamine-deficient rat Baystate Medical Center tissues. Br J Pharmacol Chemother 1963;20:190-203. Tufts University School of Medicine [23] Gubler CJ. Enzyme studies in thiamine deficiency. Int Z Vitamin- forsch 1968;38(3):287-303. Springfield, MA, USA [24] Yamashita S. Relationship between thiamine deficient cardiac lesions and myocardial and blood pyruvic acid contents. Jpn Heart J 1971; doi: 10.1016/j.ajem.2006.05.008 12(5):474-85. 126 Correspondence

[25] Yamashita S. Effect of alcohol on thiamine deficient cardiac lesions. Jpn Heart J 1971;12(4):354-67. [26] Yoshitoshi Y, Shibata N, Yamashita S. Experimental studies on the beriberi heart. I. Cardiac lesions in thiamine deficient rats. Jpn Heart J 1961;2:42. [27] Aldinger EE. Effect of thiamine deficiency on potential myocardial contractility. Circ Res 1965;16:238-43. [28] Hirota Y, Kane R, Abelmann W. Cardiovascular effects of exercise in hamsters with experimental thiamine deficiency. Jpn Circ J 1979; 43:99. [29] Cohen EM, Abelmann WH, Messer JV, Bing HL. Mechanical properties of rat cardiac muscle during experimental thiamine deficiency. Am J Physiol 1976;231(5 Pt 1):1390-4. [30] Cappelli V, Bottinelli R, Polla B, Reggiani C. Altered contractile properties of rat cardiac muscle during experimental thiamine deficiency and food deprivation. J Mol Cell Cardiol 1990;22(10): 1095-106. [31] Rindi G, Rapuzzi G. The action potential of cardiac ventricular fibre cells during thiamine deficiency. Q J Exp Physiol Cogn Med Sci 1966;51(4):249-55. Fig. 2 Abdominal CT scan without contrast enhancement [32] Davies MJ, Jennings RB. The ultrastructure of the myocardium in the demonstrates an air-fluid level in the uterine cavity (arrow). thiamine-deficient rat. J Pathol 1970;102(2):87-95. Calcified spots were noted within the myometrium. Notice the distended rectosigmoid colon with stool impaction.

Early diagnosis of gas-forming pyometra in an aged also complained of lower abdominal discomfort and nausea. patient can prevent mortality She had been admitted to a local hospital, and urinary tract infection was impressed. However, her symptoms To the Editor, persisted after 5-day antibiotic therapy, and her consciousness became drowsy gradually. She was transferred to Pyometra is an uncommon disease, referring to the our hospital for further management. Her history only accumulation of pus in the uterine cavity [1]. It often occurs included diabetes mellitus under oral hypoglycemic agents. in postmenopausal women and is associated with gyneco- On arrival, her consciousness was clear. Her vital logical malignancy [2]. Preoperative accurate diagnosis of signs were as follow: body temperature, 38.68C; heart rate, pyometra is difficult, and mostly, pyometra is found until 96 beats per minute; respiratory rate, 20 breath per minute; peritonitis develops [3]. We report a case with fever and and blood pressure, 116/78 mm Hg. The abdomen was soft, lower abdominal pain, diagnosed as pyometra by ultraso- but mild tenderness without rebounding pain was noted nography (US) preoperatively and treated properly without over lower abdomen. Others were unremarkable. Plain morbidity and mortality. abdominal radiograph showed nonspecific dilatation of A 79-year-old woman visited the emergency depart- small bowels and an intrauterine device (IUD) in ment (ED) because of intermittent fever for 2 weeks. She pelvic cavity. Laboratory results disclosed an elevated white blood cell count (37920/mm3) with left shift and an elevated C-reactive protein level of 28.43 mg/dL. Urinalysis disclosed pyuria (white blood cell N100/high- power field). Artery blood gas analysis showed metabolic acidosis. (pH 7.39, Pco2, 24.9 mm Hg; HCO3, 14.6 mEq/L; base excess, À9 mEq/L). Cefmetazole was prescribed. She was referred for abdominal US for her lower abdominal pain. The US revealed intrauterine turbid fluid accumulation with gas component (Fig. 1). Gas-forming pyometra was diagnosed. Computed tomography (CT) confirmed a distended uterus with air-fluid level inside (Fig. 2), and IUD was displaced. A gynecologist was consulted, and profuse yellowish discharge was found through the cervix during pelvic examination. Cervical dilatation and curettage were performed, and IUD was removed. Blood, urine, and pus cultures all yielded Pseudomonas aeruginosa. Antibiotics with ceftazidime Fig. 1 Ultrasonography showed intrauterine fluid accumulation was prescribed, and she was discharged later and followed with gas component (arrow). up at an outpatient clinic uneventfully. Correspondence 127

Pyometra is uncommon and defined as the accumulation References of pus in the uterine cavity [1]. It is thought as a result of interference with the natural drainage of the uterus. It is [1] Muram D, Drouin P, Thompson FE, et al. Pyometra. Can Med Assoc J commonly caused by malignant or benign gynecologic 1981;125:589-92. [2] Ikematsu Y, Kitajima T, Kamohara Y, Inoue K, Maeda J, Amano M, tumors, radiation cervicitis, atrophic cervicitis, puerperal et al. Spontaneous perforated pyometra presenting as pneumoperito- infection, congenital anomalies, and intrauterine device. neum. Gynecol Obstet Invest 1996;42(4):274-6. [4,5] It usually occurs in postmenopausal women with the [3] Nakao A, Mimura H, Fujisawa K, Ezawa K, Okamoto T, Iwagaki H, mean age of 69 years [2]. et al. Generalized peritonitis due to spontaneously perforated Preoperative accurate diagnosis of pyometra is usually pyometra presenting as pneumoperitoneum: report of a case. Surg Today 2000;30(5):454-7. difficult because the classical symptoms (vaginal discharge, [4] Kriplani A, Buckshee K, Relan S, Kapila K. dForgottenT intrauterine lower abdominal pain, and uterine enlargement) are all device leading to actinomycotic pyometra—13 years after menopause. nonspecific. More than 50% of all patients of nonruptured Eur J Obstet Gynecol Reprod Biol 1994;53(3):215-6. pyometra are asymptomatic [6]. Most diagnosis is made [5] Kaneko Y, Doi M, Kaibara M. Spontaneous perforation of pyometra: until the evidences of peritonitis occur or after laparotomy. a case report. Asia Oceania J Obstet Gynaecol 1994;20(3):263-7. [6] Hansen PT, Lindholt J. Spontaneously perforated pyometra. However, delayed diagnosis may lead to increased morbid- A differential diagnosis in acute abdomen. Ann Chir Gynaecol ity and mortality [7,8]. 1985;74(6):294-5. Ultrasound is a useful tool that provides emergency [7] Henriksen E. Pyometra associated with malignant lesions of the cervix physicians for rapid evaluation for patients with abdominal and the uterus. Am J Obstet Gynecol 1956;72(4):884-95. pain [9]. It is time-saving without radiation hazards and has [8] Chan LY, Yu VS, Ho LC, Lok YH, Hui SK. Spontaneous uterine perforation of pyometra. A report of three cases. J Reprod Med 2000; high sensitivity in assessing pyometra in previous reports 45(10):857-60. [8]. Ultrasonography showed loss of white line sign [9] Nasri MN, Coast GJ. Correlation of ultrasound findings and (endometrial stripe) [10] and hypoechoic fluid collection endometrial histopathology in postmenopausal women. Br J Obstet in the uterine cavity. [9] In our case, CT revealed a low- Gynaecol 1989;96(11):1333-8. density homogeneous mass in the peritoneal cavity, in [10] Davis PC, O’Neill MJ, Yoder IC, Lee SI, Mueller PR. Sonohystero- which an air-fluid level was noted. Gas component may be graphic findings of endometrial and subendometrial conditions. Radiographics 2002;22(4):803-16. found in imaging studies if it was caused by gas-forming [11] Akazawa K, Takamori H, Yasuda H. Clinico-statistical study on bacteria, including Escherichia coli, Klebsiella, Pseudomo- pyometra in high-aged outpatients. Nippon Sanka Fujinka Gakkai nas, Clostridium, streptococci, staphylococci, Bacteriodes, Zasshi 1991;43(11):1539-45. and others [3,11]. In summary, we report a case in which gas-forming pyometra was diagnosed preoperatively. Early diagnosis of Image pitfall of computed tomography in diagnosis of pyometra before rupture can avoid surgical exploration and aortic dissection decrease the morbidity and mortality. ED physicians should take it into consideration if an elderly woman present with lower abdominal pain and fever, especially in a high-risk To the Editor, group such as that with malignant or benign gynecologic tumors, radiation cervicitis, atrophic cervicitis, puerperal Accurate identification of aortic dissection is a challenge infection, and congenital anomalies. to an emergency physician. A myriad of symptoms, insufficient sensitivity of physical finding, and low specificity of chest radiography make prompt diagnosis difficult Chia-Chun Tai MD [1,2]. Furthermore, even if advanced diagnostic tests, such Wan-Ching Lien MD as computed tomography (CT), have been performed under Hsiu-Po Wang MD highly clinical suspicion, a variety of pitfall lurks in image Department of Emergency Medicine interpretation, sometimes leading to a so-called bVictim of National Taiwan University Medical Imaging TechnologyQ (VOMIT) case [3],as National Taiwan University Hospital described below. Taipei 100, Taiwan A 69-year-old woman, with history of hypertension and degenerative spondylosis, presented to the ED for an intense Kao-Lang Liu MD chest pain for several hours. The pain was described as Department of Medical Imaging sharp type radiating to the back. Vital signs included blood National Taiwan University pressure of 163/78 mm Hg, pulse of 64 beats/min, and National Taiwan University Hospital respiratory rate of 23 breaths/min. Physical examination and Taipei 100, Taiwan laboratory tests, including the chest radiography, electrocar- E-mail address: [email protected] diogram (ECG), and cardiac makers, were unremarkable. With the impression of aortic dissection, chest CT exami- doi: 10.1016/j.ajem.2006.04.012 nation was arranged. 128 Correspondence

A contrast-enhanced CT revealed a crescentic enhance- However, there were many pitfalls in CT images that ment in the aortic root among 5 cuts of films, considered as mimicked aortic dissection. These pitfalls were attributable type I aortic dissection with intimal flaps (Fig. 1). She then to technical factors, streak artifacts, periaortic structures, aortic received antihypertensive treatment and underwent explor- wall motion, and normal aortic sinuses [8]. For example, the atory cardiotomy immediately. To our surprise, at operation, dye column in the ascending aorta averaged over 2 points in the aorta was found to be normal. The postoperative course time of the ejection phase rendered a bfalse flap.Q was uncomplicated, and her initial symptom was thought as Artifacts were more associated with the conventional CT thoracic spondylosis during hospitalization. She was then machine (single-detector row CT). In our VOMIT case, the dismissed on hospital day 10. false-positive streak artifact caused by cardiac motion was Acute aortic dissection is a life-threatening condition. just generated by the kind of CT. Besides, the absence of The incidence of the disease ranges from 5 to 30 cases per hematoma in the mural and mediastinum, presenting at an million people per year, depending on the prevalence of risk atypical site for intimal dissection clinches the image factors in the study population [4-6]. Although the disease is diagnosis with aortic artifacts. Aortic motion artifacts mainly uncommon, its outcome is frequently fatal. Some physical affect the proximal section of the ascending aorta [9,10]. findings, including pulse deficits or focal neurological They appear as a localized duplication or pseudothickening deficits, increase the likelihood of an acute thoracic aortic of the aortic wall, sometimes generating the image of a dissection, and a completely normal chest radiograph result floating, intimal pseudomembrane. They are mostly located or lack of sudden onset chest pain lowers the likelihood. in the left anterolateral and right posterolateral positions in Overall, however, the clinical examination is insufficiently the transverse axial plane, as demonstrated in this case, and sensitive to exclude the diagnosis of aortic dissection [1]. are generally limited to only few contiguous sections. On the Computed tomography scanning was the most common other hand, true intimal tears occur in the regions of the aorta initial diagnostic test performed in patients enrolled in the that are subjected to the greatest dp/dt (momentary pressure International Registry of Acute Aortic Dissection because change) and pressure fluctuations [11], which should be most of its less invasive and rapid diagnosis in emergency situations marked in the right anterolateral ascending aorta and in the [4]. Sensitivities of 83% to 94% and specificities of 87% to first portion of the descending thoracic aorta. 100% have been reported with the use of CT scanning for the To dispel the diagnostic doubts at the ED, the following diagnosis of aortic dissection, except in cases of the dissecting strategies were suggested in facing with a dilemma of CT ascending aorta, in which its sensitivity is below 80% [7]. interpretation. Rescan the patient in a different table speed

Fig. 1 A crescentic enhancement in the aortic root among contiguous 5 cuts of CT scanning, considered as type I aortic dissection with intimal flaps. Correspondence 129 will generate artifact in different location, whereas a true [11] Wheat MW. Acute dissection of the aorta. Cardiovasc Clin 1987; dissection will be consistent. Reconstruction by a multi- 17:241-62. [12] Willoteaux S, Lions C, Gaxotte V, et al. Imaging of aortic dissection by detector row CT (helical CT) in a sagittal view will also give helical computed tomography (CT). Eur Radiol 2004;14:1999-2008. a more accurate assessment of aortic artifact [12,13]. [13] Kapustin AJ, Litt HI. Diagnostic imaging for aortic dissection. Semin Cardiac gating has been shown to reduce the artifact at Thorac Cardiovasc Surg 2005;17:214-23. the aortic root [14]. However, in emergency condition, its [14] Morgan-Hughes GJ, Owens PE, Marshall AJ, et al. Thoracic aorta at application has not yet been validated [12]. Transesophageal multi-detector row CT: motion artifact with various reconstruction windows. Radiology 2003;228:583-8. echocardiogram (TEE) could also assist in proper diagnosis. [15] Keren A, Kim CB, Hu BS, et al. Accuracy of biplane and multiplane The sensitivity and specificity of TEE have been reported to transesophageal echocardiography in diagnosis of typical acute aortic be as high as 98% and 63% to 96%, respectively [15]. The dissection and intramural hematoma. J Am Coll Cardiol 1996; main limitation of TEE is its strong dependence on the 28:627-36. experience of the investigator. Missed diagnosis of aortic dissection leads to a potentially lethal outcome, whereas misjudgment of artificial Other causes of unilateral pulmonary edema images results in unnecessary thoracotomy. Familiarity with these diagnostic pitfalls of CT facilitates correct recognition of aortic dissection. To the Editor,

Wen-Chu Chiang MD, MPH We read with interest the article bUnilateral pulmonary Pei-Chieh Kao MD edema related to massive mitral insufficiencyQ wherein the Chan-Ping Su MD authors describe the occurrence of unilateral pulmonary Department of Emergency Medicine edema after mitral insufficiency [1]. We describe a case of National Taiwan University Hospital, Yun-Lin Branch unilateral pulmonary edema related to end-stage renal Taipei 100, Taiwan disease that improved rapidly with hemodialysis and review E-mai address: [email protected] the literature for the causes of unilateral pulmonary edema. A 34-year-old man, with a known case of end-stage renal Juan Hsu MD disease (secondary to chronic glomerulonephritis) requiring Department of Cardiovascular Surgery maintenance hemodialysis, presented in the emergency National Taiwan University Hospital, Yun-Lin Branch department with severe dyspnea. Over the previous 12 hours, Taipei 100, Taiwan he experienced progressive orthopnea and cough with mucoid expectoration. The last session of hemodialysis doi: 10.1016/j.ajem.2006.04.011 was administered 24 hours ago, and the patient denied any history of undue salt or water intake, hemoptysis, or fever. References On examination, the patient was pale, in severe respiratory distress, and had to sit in an upright position as a result of [1] Klompas M. Does this patient have an acute thoracic aortic dissection? orthopnea. Blood pressure was 160/90 mm Hg, pulse rate JAMA 2002;287:2262-72. 110/min, respiratory rate 34/min, and oral temperature [2] Khan IA, Nair CK. Clinical, diagnostic, and management perspectives 37.28C. Auscultation revealed diminished breath sounds of aortic dissection. Chest 2002;122:311-28. over the lower two thirds of the left lung and crackles in [3] Trauma radiology misread/over-read images and reviews. Available the lower one third of the right lung. Auscultation of the at: http://www.trauma.org/radiology/index.html [accessed on Mar 30, 2006]. heart was normal. The progressive orthopnea led to the [4] Hagan PG, Nienaber CA, Isselbacher EM, et al. The International initial diagnosis of pulmonary edema, and the patient was Registry of Acute Aortic Dissection (IRAD): new insights into an old shifted to the respiratory intensive care unit where he was disease. JAMA 2000;283:897-903. treated with oxygen, intravenous diuretics, and morphine. [5] Spittell PC, Spittell Jr JA, Joyce JW, et al. Clinical features and differential diagnosis of aortic dissection: experience with 236 cases Preliminary investigations revealed a white blood cell count (1980 through 1990). Mayo Clin Proc 1993;68:642-51. of 12000/lL with 70% neutrophils, hemoglobin level of [6] Bickerstaff LK, Pairolero PC, Hollier LH, et al. Thoracic aortic 9.3 g/dL, creatinine level of 8.2 mg/dL, sodium level of aneurysms: a population-based study. Surgery 1982;92:1103-8. 141 mEq/L, potassium level of 5.8 mEq/L, and creatine [7] Nienaber CA, von-Kodolitsch Y, Nicolas V, et al. The diagnosis of phosphokinase (MB isozyme) of 20 IU/L. Arterial blood gas thoracic dissection by noninvasive imaging procedures. N Engl J Med o co 1993;328:1-9. analysis revealed a pH of 7.2, Pa 2 of 62 mm Hg, Pa 2 of [8] Batra P, Bigoni B, Manning J, et al. Radiographics 2000;20:309-20. 25 mm Hg, and HCO3 of 10 mEq/L, at Fio2 of 0.28. The [9] Duvernoy O, Coulden R, Ytterberg C. Aortic motion: a potential electrocardiogram demonstrated sinus tachycardia. Chest pitfall in CT imaging of dissection in the ascending aorta. J Comput x-ray revealed a unilateral homogenous pulmonary opacity Assist Tomogr 1995;19:569-72. [10] Qanadli SD, Hajjam M, Mesurolle B, et al. Motion artifacts of the occupying the right lower lobe and left pleural effusion aorta simulating aortic dissection on spiral CT. J Comput Assist (Fig. 1). Bedside echocardiography was normal with a left Tomogr 1999;23:1-6. ventricular ejection fraction of 56%. Although the patient’s 130 Correspondence temperature was normal, in the presence of a unilateral pulmonary infiltrate, as pneumonia could not be ruled out, treatment with intravenous cefpirome and azithromycin was initiated. In view of metabolic acidosis and hyperkalemia the patient underwent hemodialysis. After the dialysis, the patient’s disease course was characterized by rapid recovery: within a few hours, breathing improved dramatically and repeated arterial blood gas analysis revealed a pH of 7.32, Pao2 of 72 mm Hg, Paco2 of 35 mm Hg, and HCO3 of 18 mEq/L, at FiO2 0.21. Repeat chest x-ray obtained 12 hours after admission showed resolution of the pulmonary opacity with residual left pleural effusion (Fig. 2). These findings, together with the lack of fever, suggested that pneumonia was not a likely diagnosis. Moreover, sputum for culture and Gram stain was negative and we terminated the antibiotic treatment. The patient was investigated for the left pleural effusion. Pleural biopsy showed caseous granulomas with acid-fast bacilli. The patient was started on antituberculous therapy and discharged uneventfully after 7 days of hospitalization with advice for regular maintenance hemodialysis. Unilateral pulmonary edema is an uncommon, if not rare, Fig. 2 Chest radiograph shows complete clearing of the entity that can be mistaken for other causes of unilateral opacities on the right side. alveolar and interstitial infiltrates, especially pneumonia. It has been described after congestive heart failure [2], prolonged rest on one side in a patient with cardiac unilateral pulmonary disease such as MacLeod syndrome decompensation or receiving large amounts of fluids [3], [5] and unilateral pulmonary agenesis [6], after talc in patients with mitral valve insufficiency [1], in cases of pleurodesis [7], trauma [8], epilepsy [9], upper airway rapid expansion of the lung after pleural effusions and obstruction [10], pulmonary artery compression from aortic pneumothorax [4], in the normal lung in patients with dissection [11,12], pulmonary venous obstruction from mediastinal fibrosis [13] or postlobectomy [14], unilateral main stem intubation [15], neurogenic pulmonary edema [16], nitrogen mustard [17], amiodarone-related [18] and heroin-related [19] pulmonary edema, pregnancy [20], and in cases of fluid overload [21-23]. Unilateral pulmonary edema can result from myriad of causes. Thus, even if the pulmonary opacities are unilateral, if the clinical manifestation is compatible with pulmonary edema and not with pneumonia, early and aggressive treatment should be initiated for pulmonary edema.

Ritesh Agarwal MD, DM Ashutosh N. Aggarwal MD, DM Dheeraj Gupta Department of Pulmonary Medicine Postgraduate Institute of Medical Education and Research Chandigarh-160012, India E-mail addresses: [email protected] [email protected]

doi:10.1016/j.ajem.2006.05.025

References Fig. 1 Chest radiograph shows prominent right hila and dense homogenous opacity in the right lower lobe; also present are left [1] Legriel S, Tremey B, Mentec H. Unilateral pulmonary edema related pleural effusion and a central venous catheter for dialysis. to massive mitral insufficiency. Am J Emerg Med 2006;24(3):372. Correspondence 131

[2] Nitzan O, Saliba WR, Goldstein LH, Elias MS. Unilateral pulmonary [14] Gyves-Ray KM, Spizarny DL, Gross BH. Unilateral pulmonary edema: a rare presentation of congestive heart failure. Am J Med Sci edema due to postlobectomy pulmonary vein thrombosis. AJR Am J 2004;327(6):362-4. Roentgenol 1987;148(6):1079-80. [3] Leeming BW. Gravitational edema of the lungs observed during [15] Kramer MR, Melzer E, Sprung CL. Unilateral pulmonary edema after assisted respiration. Chest 1973;64(6):719-22. intubation of the right mainstem bronchus. Crit Care Med 1989; [4] Murat A, Arslan A, Balci AE. Re-expansion pulmonary edema. Acta 17(5):472-4. Radiol 2004;45(4):431-3. [16] Perrin C, Jullien V, Venissac N, Lonjon M, Blaive B. Unilateral [5] Saleh M, Miles AI, Lasser RP. Unilateral pulmonary edema in Swyer- neurogenic pulmonary edema. A case report. Rev Pneumol Clin 2004; James syndrome. Chest 1974;66(5):594-7. 60(1):43-5. [6] Nana-Sinkam P, Bost TW, Sippel JM. Unilateral pulmonary edema in [17] Goodman LR, Shanser JD. Unilateral pulmonary edema. An unusual a 29-year-old man visiting high altitude. Chest 2002;122(6):2230-3. complication of nitrogen mustard therapy. Radiology 1976; [7] Scalzetti EM. Unilateral pulmonary edema after talc pleurodesis. 120(1):166. J Thorac Imaging 2001;16(2):99-102. [18] Herndon JC, Cook AO, Ramsay MA, Swygert TH, Capehart J. [8] Dempster AG. Unilateral pulmonary edema complicating chest Postoperative unilateral pulmonary edema: possible amiodarone compression and cardiac avulsion. Am J Forensic Med Pathol 1986; pulmonary toxicity. Anesthesiology 1992;76(2):308-12. 7(4):350-3. [19] Sporer KA, Dorn E. Heroin-related noncardiogenic pulmonary edema: [9] Koppel BS, Pearl M, Perla E. Epileptic seizures as a cause of a case series. Chest 2001;120(5):1628-32. unilateral pulmonary edema. Epilepsia 1987;28(1):41-4. [20] Choi HS, Choi H, Han S, et al. Pulmonary edema during pregnancy: [10] Morisaki H, Ochiai R, Takeda J, Nagano M. Unilateral pulmonary unilateral presentation is not rare. Circ J 2002;66(7):623-6. edema following acute subglottic edema. J Clin Anesth 1990;2(1):42-4. [21] Balogun SA, Balogun RA. Acute unilateral pulmonary edema from [11] McTigue C, Scurry JP, Silberstein M. Unilateral pulmonary edema dietary salt and water load: a case report and review of the literature. associated with pulmonary arterial compression. Australas Radiol Conn Med 2001;65(11):653-6. 1988;32(3):390-3. [22] Di Benedetto C, Brunner W, Kuhn M. Unilateral pulmonary [12] Takahashi M, Ikeda U, Shimada K, Takeda H. Unilateral pulmonary edema in a dialysis patient with massive fluid overload and mitral edema related to pulmonary artery compression resulting from acute valve insufficiency. Schweiz Rundsch Med Prax 2003;92(29-30): dissecting aortic aneurysm. Am Heart J 1993;126(5):1225-7. 1265-8. [13] Routsi C, Charitos C, Rontogianni D, Daniil Z, Zakynthinos E. [23] Wong KS, Lin GJ, Lai CH, Lien R. Unilateral pulmonary edema: an Unilateral pulmonary edema due to pulmonary venous obstruction uncommon presentation of poststreptococcal glomerulonephritis. from fibrosing mediastinitis. Int J Cardiol 2006;108(3):418-21. Pediatr Emerg Care 2003;19(5):337-9.