Volume 17 Issue 3 by Western Journal of Emergency Medicine

SEPTEMBER 20TH HYATT REGENCY ORANGE COUNTY, CA

Volume XVII, Number 3, May 2016

Open Access at www.westjem.com

ISSN 1936-900X

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE

ENDEMIC INFECTIONS 238 Identify-Isolate-Inform: A Tool for Initial Detection and Management of Zika Virus Patients in the Emergency Department

KL Koenig, A Almadhyan, MJ Burns

VOLUME XVII, NUMBER 3, May 2016

2016

Western Journal of Emergency Medicine

PRESENTS

West

SOCIETAL IMPACT ON EMERGENCY CARE 245 Emergency Medical Treatment and Labor Act (EMTALA) 2002-15: Review of Office of Inspector General Patient Dumping Settlements

N Zuabi, LD Weiss, MI Langdorf

TREATMENT PROTOCOL ASSESSMENT 252 Prospective Validation of Modified NEXUS Cervical Spine Injury Criteria in Low-risk Elderly Fall Patients

J Tran, D Jeanmonod, D Agresti, K Hamden, RK Jeanmonod

PRACTICE VARIABILITY 258 Quality Improvement Initiative to Decrease Variability of Emergency Physician Opioid Analgesic Prescribing

JH Burton, JA Hoppe, JM Echternach, JM Rodgers, M Donato

CRITICAL CARE 264 Anticoagulation Reversal and Treatment Strategies in Major Bleeding: Update 2016

S Christos, R Naples

271

Academic Emergency Medicine Physiciansâ&#x20AC;&#x2122; Knowledge of Mechanical Ventilation

SR Wilcox, TD Strout, JI Schneider, PM Mitchell, J Smith, L Lutfy-Clayton, EG Marcolini, A Aydin, TA Seigel, JB Richards

DIAGNOSTIC ACUMEN 280 Anti-N-Methyl-D-Aspartate Receptor Encephalitis, an Underappreciated Disease in the Emergency Department

DR Lasoff, J Corbett-Detig, R Sell, M Nolan, G Wardi

PAGES 238-383

Contents continued on page iii

CALIFORNIAACEP.ORG > INFO@CALIFORNIAACEP.ORG A Peer-Reviewed, International Professional Journal

TAB COPY GOES HERE

SAVE THE DATE!

November 1 - 5, 2016 Hilton Union Square, San Francisco, CA

ACOEP’s flagship conference brings you more than ever before. Highlights include: • Specialized Tracks in Critical Care, Pain Management, and Wellness

• Advanced Airway Course • Active Shooter Training • Advanced EKG Course • FOEM Legacy Gala: Dinner & Awards Ceremony

• Meeting of ACOEP’s Council for Women in Emergency Medicine • 1-on-1 Access to Expert Speakers

ON SALE NOW! Visit www.acoep.org/scientific for more information. Paramount Sponsor

Prime Sponsor

Supporting Sponsor

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE Mark I. Langdorf, MD, MHPE, Editor-in-Chief University of California, Irvine School of Medicine Niels K. Rathlev, MD, Associate Editor Tufts University Medical School and Baystate Medical Center

Shahram Lotfipour, MD, MPH, Managing Associate Editor University of California, Irvine School of Medicine Rick A. McPheeters, DO, Associate Editor Kern Medical Center

Joel M. Schofer, MD, MBA, Associate Editor Naval Medical Center Portsmouth

Edward Michelson, MD, Associate Editor Case Western University

Thomas Terndrup, MD, Associate Editor Ohio State University

Section Editors AAEM/RSA

ED Administration

Andrew W. Phillips, MD, MEd Stanford/Kaiser Emergency Medicine

James Langabeer II, MBA, EMT, PhD Rolando Valenzuela, MD University of Texas Medical School University of Southern California

Behavioral Emergencies

Emergency Cardiac Care

Legal Medicine

Leslie Zun, MD, MBA Chicago Medical School

Amal Mattu, MD University of Maryland

Methodology and Biostatistics

Emergency Medical Services

Elizabeth Burner, MD, MPH University of Southern California

Michael Gottlieb, MD Cook County Hospital

Michael P. Wilson, MD, PhD University of California, San Diego

Clinical Practice

Michael Abraham, MD University of Maryland Eric Snoey, MD Alameda County Medical Center David Thompson, MD University of California, San Francisco Kenneth S. Whitlow, DO Kaweah Delta Medical Center

Gary Johnson, MD Upstate Medical University

Michael Kurz, MD University of Alabama at Birmingham

Christopher Kahn, MD, MPH University of California, San Diego Derek Cooney, MD State University of New York Upstate Medical University, New York Joshua B. Gaither, MD University of Arizona, Tuscon

Shira A. Schlesinger, MD, MPH University of California, Irvine Joseph Shiber, MD University of Florida - College of Medicine Geriatrics Teresita M. Hogan, MD Todd Slesinger, MD University of Chicago Aventura Hospital and Medical Center Kathleen Walsh, DO, MS Christopher â&#x20AC;&#x153;Kitâ&#x20AC;? Tainter, MD University of Wisconsin University of California, San Diego

Critical Care

International Medicine

Chris Mills, MD, MPH Santa Clara Valley Medical Center

Resident/Student/Fellow Forum John Ashurst, DO Lehigh Valley Health Network Cecylia Kelley, DO Inspira Health Network

Technology in Emergency Medicine

Greg P. Moore, MD, JD Madigan Army Medical Center Craig Anderson, MPH, PhD University of California, Irvine

Sanjay Arora, MD University of Southern California Robert L. Rogers, MD University of Kentuky

Trauma

William Paolo, MD SUNY Upstate

Christian McClung, MD MPhil University of Southern California

David Peak, MD Massachusetts General Hospital/Havard Medical School

Musculoskeletal

Juan F. Acosta DO, MS Pacific Northwest University

Toxicology

Jeffrey R. Suchard, MD University of California, Irvine

Neurosciences

Edward P. Sloan, MD, MPH University of Illinois at Chicago William D. Whetstone, MD University of California, San Francisco

Pediatric Emergency Medicine

Judith Klein, MD University of California, San Francisco Paul Walsh, MD, MSc University of California, Davis

Brandon Wills, DO, MS Virginia Commonwealth University

Ultrasound

Gavin Budhram, MD Tufts University Laleh Gharahbaghian, MD Stanford University

Disaster Medicine

Infectious Disease

Gentry Wilkerson, MD University of Maryland

Sukhjit S. Takhar, MD Harvard Medical School

Education

Injury Prevention

Shane Summers, MD Muhammad Waseem, MD Brooke Army Medical Center Lincoln Medical & Mental Health Center J. Matthew Fields, MD Public Health Thomas Jefferson University Jeremy Hess, MD, MPH Emory University

Bharath Chakravarthy, MD, MPH University of California, Irvine

Chadd Kraus, DO, DrPH, MPH University of Missouri - Columbia

Jeffrey Druck, MD University of Colorado

Wirachin Hoonpongsimanont, MD University of California, Irvine

Trevor Mills, MD, MPH Northern California VA Health Care

Christopher Kang, MD Madigan Army Medical Center

Douglas Ander, MD Emory University

Robert Derlet, MD University of California, Davis

Michael Epter, DO Maricopa Medical Center

Official Journal of the California Chapter of the American College of Emergency Physicians, the America College of Osteopathic Emergency Physicians, and the California Chapter of the American Academy of Emergency Medicine

Available in MEDLINE, PubMed, PubMed Central, CINAHL, SCOPUS, Google Scholar, eScholarship, Melvyl, DOAJ, EBSCO, EMBASE, Medscape, HINARI, and MDLinx Emergency Med. Members of OASPA. Editorial and Publishing Office: WestJEM/Depatment of Emergency Medicine, UC Irvine Health, 333 City Blvd, West, Rt 128-01, Orange, CA 92868, USA Office: 1-714-456-6389; Email: Editor@westjem.org

Volume XVII, no. 3 : May 2016

Western Journal of Emergency Medicine

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE

Editorial Board

Erik D. Barton, MD, MBA University of California, Irvine

Edward Panacek, MD, MPH University of South Alabama

Peter A. Bell, DO, MBA Ohio University, Heritage College of Osteopathic Medicine

Niels K. Rathlev, MD Tufts University Medical School and Baystate Medical Center

Barry E. Brenner, MD, MPH Case Western Reserve University

Robert M. Rodriguez, MD University of California, San Francisco

David F.M. Brown, MD Massachusetts General Hospital/ Harvard Medical School Francis Counselman, MD Eastern Virginia Medical School

Advisory Board

Peter A. Bell, DO, MBA American College of Osteopathic Emergency Physicians Ohio University, Heritage College of Osteopathic Medicine John B. Christensen, MD California Chapter Division of AAEM

Mark I. Langdorf, MD, MHPE UC Irvine Health SOM

Peter Sokolove, MD University of California, San Francisco Samuel J. Stratton, MD, MPH Orange County, CA, EMS Agency Robert Suter, DO, MHA UT Southwestern Medical Center

Shahram Lotfipour, MD, MPH UC Irvine Health SOM

Steven Gabaeff, MD Clinical Forensic Medicine

Thomas Terndrup, MD Ohio State University

Brent King, MD, MMM University of Texas, Houston

Scott Zeller, MD Alameda County Medical Center

William Mallon, MD California ACEP American College of Emergency Physicians University of Southern California Keck School of Medicine

Edward Michelson, MD Texas Tech University

Leslie Zun, MD, MBA Chicago Medical School

Daniel J. Dire, MD University of Texas Health Sciences Center San Antonio

Trevor Mills, MD, MPH California Chapter Division of AAEM LSU Medical Center

Linda S. Murphy, MLIS University of California, Irvine School of Medicine Librarian

Aimee Moulin, MD California ACEP American College of Emergency Physicians University of California, Davis

Jonathan Olshaker, MD Boston University

Editorial Staff

International Editorial Board Arif A. Cevik, MD Eskişehir Osmangazi University Medical Center, Eskişehir, Turkey

Amin A. Kazzi, MD The American University of Beirut, Lebanon

Francesco D. Corte, MD Azienda Ospedaliera Universitaria “Maggiore della Carità”, Novara, Italy

Steven H. Lim, MD Changi General Hospital, Singapore

Vijay Gautam, MBBS University of London, United Kingdom Wirachin Hoonpongsimanont, MD University of California, Irvine

Jan Wachtler American College of Osteopathic Emergency Physicians

Amal Khalil, MBA UC Irvine Health SOM

Scott Rudkin, MD, MBA University of California, Irvine

Elena Lopez-Gusman California ACEP American College of Emergency Physicians

Robert W. Derlet, MD University of California, Davis

Robert Suter, DO, MHA American College of Osteopathic Emergency Physicians UT Southwestern Medical Center

Kobi Peleg, PhD, MPH Tel-Aviv University, Israel Rapeepron Rojsaengroeng, MD Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

Nadeem Albadawi, BS Associate Publishing Director

Vincent Lam, BS Publishing Director

Chelsey Bithell, BS Associate Editorial Director

Ryan Nguyen, BS Marketing Director

June Casey, BA Copy Editor

Patrick Penalosa, BS Editorial Director

Emily DeVillers CAL AAEM WestJEM Liaison

Diane Shin, BS Associate Publishing Director

Joyce Y. Kim, BS Editorial Director

Alex Trinh, BS Website Manager

Available in MEDLINE, PubMed, PubMed Central, CINAHL, SCOPUS, Google Scholar, eScholarship, Melvyl, DOAJ, EBSCO, EMBASE, Medscape, HINARI, and MDLinx Emergency Med. Members of OASPA. Editorial and Publishing Office: WestJEM/Depatment of Emergency Medicine, UC Irvine Health, 333 City Blvd, West, Rt 128-01, Orange, CA 92866, USA Office: 1-714-456-6389; Email: Editor@westjem.org

Western Journal of Emergency Medicine

Volume XVII, no. 3 : May 2016

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health JOURNAL FOCUS Emergency medicine is a specialty which closely reflects societal challenges and consequences of public policy decisions. The emergency department specifically deals with social injustice, health and economic disparities, violence, substance abuse, and disaster preparedness and response. This journal focuses on how emergency care affects the health of the community and population, and conversely, how these societal challenges affect the composition of the patient population who seek care in the emergency department. The development of better systems to provide emergency care, including technology solutions, is critical to enhancing population health.

Table of Contents continued 283

ACE-I Angioedema: Accurate Clinical Diagnosis May Prevent Epinephrine-Induced Harm

RM Curtis, S Felder, R Borici-Mazi, I Ball

BEHAVIORAL HEALTH 290 Case Series of Synthetic Cannabinoid Intoxication from One Toxicology Center

KD Katz, AL Leonetti, BC Bailey, RM Surmaitis, ER Eustice, S Kacinko, SM Wheatley

295

Gender Differences in Emergency Department Visits and Detox Referrals for Illicit Drug Use and Nonmedical Use of Opioids

H Ryoo, EK Choo

302 Variations in Substance Use Prevalence Estimates and Need for Interventions among Adult Emergency Department Patients Based on Different Screening Strategies Using the ASSIST

RC Merchant, T Liu, JR Baird

HEALTH OUTCOMES 315 Undertriage of Trauma-Related Deaths in U.S. Emergency Departments

JA Holst, SM Perman, R Capp, JS Haukoos, AA Ginde

324

Vital Signs Predict Rapid-Response Team Activation Within Twelve Hours of Emergency Department Admission

331

Trauma Triage and Trauma System Performance

333

Inadequate Sensitivity of Laboratory Risk Indicator to Rule Out Necrotizing Fasciitis in the Emergency Department

JM Walston, D Cabrera, SD Bellew, MN Olive, CM Lohse, MF Bellolio G Johnson

E Burner, SO Henderson, G Burke, J Nakashioya, JR Hoffman

EDUCATION 337 Ten Tips for Engaging the Millennial Learner and Moving an Emergency Medicine Residency Curriculum into the 21st Century

SL Toohey, A Wray, W Wiechmann, M Lin, M Boysen-Osborn

344

Survey of Individual and Institutional Risk Associated with the Use of Social Media

M Garg, DA Pearson, MC Bond, M Runyon, MT Pillow, L Hopson, RR Cooney, J Khadpe, JT Nomura, PC Inboriboon

Policies for peer review, author instructions, conflicts of interest and human and animal subjects protections can be found online at www.westjem.com. Volume XVII, no. 3 : May 2016

iii

Western Journal of Emergency Medicine

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE

Table of Contents continued 350

Impact of Doximity Residency Rankings on Emergency Medicine Applicant Rank Lists

355

Emergency Medicine Resident Perceptions of Medical Professionalism

362

Introducing a Fresh Cadaver Model for Ultrasound-guided Central Venous Access Training in Undergraduate Medical Education

WJ Peterson, LR Hopson, S Khandelwal, M White, FE Gallahue, J Burkhardt, AM Rolston, SA Santen J Jauregui, MO Gatewood, JS Ilgen, C Schaninger, J Strote

R Miller, H Ho, V Ng, M Tran, D Rappaport, WJA Rappaport, SJ Dandorf, J Dunleavy, R Viscusi, R Amini

POPULATION HEALTH RESEARCH DESIGN 367 Authorship Trends of Emergency Medicine Publications over the Last Two Decades

R Lammers, T Simunich, J Ashurst

EMERGENCY MEDICAL SERVICES 372 Out-of-Hospital Surgical Airway Management: Does Scope of Practice Equal Actual Practice?

M Furin, M Kohn, R Overberger, D Jaslow

TECHNOLOGY IN EMERGENCY MEDICINE 377 Pilot Study to Determine Accuracy of Posterior Approach Ultrasound for Shoulder Dislocation by Novice Sonographers

S Lahham, B Becker, A Chiem, LM Joseph, CL Anderson, SP Wilson, M Subeh, A Trinh, E Viquez, JC Fox

383

Point-of-Care Ultrasound Diagnosis of Left-Sided Endocarditis

CW Bugg, K Berona

Policies for peer review, author instructions, conflicts of interest and human and animal subjects protections can be found online at www.westjem.com. Western Journal of Emergency Medicine

Volume XVII, no. 3 : May 2016

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE

This open access publication would not be possible without the generous and continual financial support of our society sponsors, department and chapter subscribers. Professional Society Sponsors American College of Osteopathic Emergency Physicians California Chapter Division of American Academy of Emergency Medicine

California ACEP

Academic Department of Emergency Medicine Subscriber Advocate Christ Medical Center Oak Lawn, IL Allegheny Health Network Pittsburgh, PA American University of Beirut Beirut, Lebanon Baylor College of Medicine Houston, TX Baystate Medical Center/Tufts University Springfield, MA Boston Medical Center Boston, MA Brigham and Womenâ&#x20AC;&#x2122;s Hospital Department of Emergency Medicine Boston, MA Brown University Providence, RI California State University Fresno Fresno, CA Carl R. Darnall Army Medical Center Fort Hood, TX Conemaugh Memorial Medical Center Johnstown, PA Eastern Virginia Medical School Norfolk, VA Emory University Atlanta, GA Florida Hospital Medical Center Orlando, FL Georgia Regents University Emergency Medicine Residency and Fellowship Augusta, GA Good Samaritan Hospital Medical Center West Islip, NY Henry Ford Medical Center Detroit, MI Highland Hospital Oakland, CA INTEGRIS Health Oklahoma City, OK

Kaweah Delta Health Care District Visalia, CA Kennedy University Hospitals Turnersville, NJ Kern Medical Center Bakersfield, CA Lakeland HealthCare St. Joseph, MI Lehigh Valley Hospital and Health Network Allentown, PA Louisiana State University Health Sciences Center New Orleans, LA Madigan Army Medical Center Tacoma, WA Maimonides Medical Center Brooklyn, NY Maricopa Medical Center Phoenix, AZ Massachusetts General Hospital Boston, MA Mayo Clinic College of Medicine Rochester, MN Mt. Sinai Medical Center Miami Beach, FL New York Medical College at Metropolitan Hospital Center New York, NY New York Methodist Hospital Brooklyn, NY North Shore University Hospital Manhasset, NY Ohio State University Medical Center Columbus, OH Oklahoma University Norman, OK Penn State Milton S. Hershey Medical Center Hershey, PA Presence Resurrection Medical Center Chicago, IL

Regions Hospital/ Health Partners Institute for Education and Research St. Paul, MN Robert Wood Johnson Hospital New Brunswick, NJ Southern Illinois University Carbondale, IL Stanford University Palo Alto, CA SUNY Downstate Medical Center New York City, NY SUNY Upstate Medical Center Syracuse, NY Temple University Philadelphia, PA University Hospitals Case Medical Center Cleveland, OH University of Alabama, Birmingham Birmingham, AL University of Arizona Tucson, AZ University of California, Davis Medical Center Sacramento, CA University of California, San Francisco San Francisco, CA University of California, San Francisco, Fresno Fresno, CA University of California Irvine Orange, CA University of California, Los Angeles Los Angeles, CA University of California San Diego La Jolla, CA University of Colorado & Denver Health Denver, CO University of Florida Jacksonville, FL

State Chapter Subscriber

Arizona Chapter Division of the American Academy of Emergency Medicine California Chapter Division of the American Academy of Emergency Medicine Florida Chapter Division of the American Academy of Emergency Medicine Great Lakes Chapter Division of the American Academy of Emergency Medicine

University of Illinois at Chicago Chicago, IL University of Iowa Iowa City, IA University of Kansas Hospital Kansas City, KS University of Louisville Louisville, KY University of Maryland Baltimore, MD University of Michigan Ann Arbor, MI University of Missouri Columbia, MO University of Nebraska Medical Center Omaha, NE University of Nevada Las Vegas, NV University of Oklahoma Norman, OK University of Southern California/Keck School of Medicine Los Angeles, CA University of Tennessee, Memphis Memphis, TN University of Texas Houston, TX University of Washington Seattle, WA University of Wisconsin Hospitals and Clinics Madison, WI University of Utah Salt Lake City, UT Virginia Commonwealth University Medical Center Richmond, VA York Hospital York, ME

Tennessee Chapter Division of the American Academy of Emergency Medicine Uniformed Services Chapter Division of the American Academy of Emergency Medicine Virginia Chapter Division of the American Academy of Emergency Medicine

International Society Partners

Sociedad Argentina de Emergencias Thai Association for Emergency Medicine Sociedad Chileno Medicina Urgencia Emergency Medicine Association of Turkey To become a WestJEM departmental sponsor, waive article processing fee, receive print and copies for all faculty and electronic for faculty/residents, and free CME and faculty/fellow position advertisement space, please go to http://westjem.com/subscribe or contact: Emily DeVillers WestJEM Staff Liaison Phone: 1-800-884-2236 Email: sales@westjem.org

Volume XVII, no. 3 : May 2016

Western Journal of Emergency Medicine

Concept Paper

Identify-Isolate-Inform: A Tool for Initial Detection and Management of Zika Virus Patients in the Emergency Department Kristi L. Koenig, MD* Abdulmajeed Almadhyan, MD*‡ Michael J. Burns, MD†

*University of California Irvine Medical Center, Department of Emergency Medicine, Center for Disaster Medical Sciences, Orange, California † University of California Irvine Medical Center, Department of Emergency Medicine and Department of Medicine, Division of Infectious Diseases, Orange, California ‡, Qassim University, Department of Emergency Medicine, Saudi Arabia

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted February 24, 2016; Accepted March 21, 2016 Electronically published April 4, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.30188

First isolated in 1947 from a monkey in the Zika forest in Uganda, and from mosquitoes in the same forest the following year, Zika virus has gained international attention due to concerns for infection in pregnant women potentially causing fetal microcephaly. More than one million people have been infected since the appearance of the virus in Brazil in 2015. Approximately 80% of infected patients are asymptomatic. An association with microcephaly and other birth defects as well as Guillain-Barre Syndrome has led to a World Health Organization declaration of Zika virus as a Public Health Emergency of International Concern in February 2016. Zika virus is a vector-borne disease transmitted primarily by the Aedes aegypti mosquito. Male to female sexual transmission has been reported and there is potential for transmission via blood transfusions. After an incubation period of 2-7 days, symptomatic patients develop rapid onset fever, maculopapular rash, arthralgia, and conjunctivitis, often associated with headache and myalgias. Emergency department (ED) personnel must be prepared to address concerns from patients presenting with symptoms consistent with acute Zika virus infection, especially those who are pregnant or planning travel to Zika-endemic regions, as well as those women planning to become pregnant and their partners. The identify-isolate-inform (3I) tool, originally conceived for initial detection and management of Ebola virus disease patients in the ED, and later adjusted for measles and Middle East Respiratory Syndrome, can be adapted for real-time use for any emerging infectious disease. This paper reports a modification of the 3I tool for initial detection and management of patients under investigation for Zika virus. Following an assessment of epidemiologic risk, including travel to countries with mosquitoes that transmit Zika virus, patients are further investigated if clinically indicated. If after a rapid evaluation, Zika or other arthropod-borne diseases are the only concern, isolation (contact, droplet, airborne) is unnecessary. Zika is a reportable disease and thus appropriate health authorities must be notified. The modified 3I tool will facilitate rapid analysis and triggering of appropriate actions for patients presenting to the ED at risk for Zika. [West J Emerg Med. 2016;17(3):238–244.]

INTRODUCTION Following closely after the 2014 Ebola outbreak, due to concerns for rapid spread and a link to birth defects, the World Health Organization (WHO) declared Zika virus a Public Health Emergency of International Concern (PHEIC) on February 1, 2016, making it only the fourth time in history for such a declaration.1 A PHEIC is defined as an extraordinary Western Journal of Emergency Medicine

event which is determined to include the following: 1. Constitute a public health risk to other states through the international spread of disease, and 2. Potentially require a coordinated international response. It implies a situation that is – 1. Serious, unusual or unexpected, 2. Carries implications for public health beyond the

238

Volume XVII, no. 3 : May 2016

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management

affected stateâ&#x20AC;&#x2122;s national border, and 3. May require immediate international action. WHO declared prior PHEICs for the April 2009 H1N1 pandemic, in May 2014 with the resurgence of polio after its near-eradication, and in August 2014 in response to the Ebola outbreak in West Africa. The Zika declaration is unprecedented as it is the first time that a vector-borne disease, i.e. an infection that is not generally transmitted from person-to-person, has been deemed a PHEIC. Vector-borne diseases, including malaria and dengue, account for more than 17% of all infectious diseases and cause more than one million deaths annually.2 While diseases spread by vectors are among the most complex of all infectious diseases to prevent and control, identification and management in the emergency department (ED) setting differs dramatically from preparation for other infectious diseases such as Ebola, measles, and Middle East Respiratory Syndrome (MERS), each of which require immediate implementation of various types of isolation (contact, airborne, and droplet, respectively). Zika virus is a RNA flavivirus related to dengue, yellow fever, West Nile, and Japanese encephalitis viruses, but not to chikungunya, which is a togavirus. It was first isolated in 1947 from the serum of a febrile sentinel rhesus monkey kept in a cage on a wooden platform in the tree canopy of the Zika forest in Uganda, in which the investigators were studying yellow fever transmission.3 The following year it was isolated from Aedes africanus mosquitoes in the same forest. Sporadic human cases were reported from Africa and Southeast Asia over the next several decades. The first recognized outbreak occurred on Yap Island in the Federated States of Micronesia in 2007, in which an estimated 73% of the population greater than three years of age was infected and 80% of infections were subclinical. A larger outbreak then occurred in French Polynesia in 2013-2014 affecting 66% of the population, with spread to other South Pacific islands in 2014. In the French Polynesia outbreak, there was an associated large increase in the reported incidence of Guillain-Barre syndrome (GBS). A case-control study of this outbreak found a GBS attack rate of 0.24 per 1,000 Zika virus infections, with GBS developing rapidly after the onset of infection with a generally favorable outcome.4 Finally, the outbreak in the Western Hemisphere, which began in northeastern Brazil in 2015, with explosive spread to many countries and territories, has infected more than one million persons (Figure 1). After microcephaly was reported in the Zika virus epidemic in South America, a retrospective analysis of the French Polynesia outbreak found that the number of microcephaly cases associated with Zika virus infection was 95 (95% CI [34-191]) per 10,000 women infected in the first trimester, which is about 50 times higher than the baseline rate in the population.5 Zika virus RNA can be detected in amniotic fluid as well

as in brain tissue of fetuses with microcephaly at a time when the maternal serum and urine are negative for RNA, supporting the association between Zika virus infection and microcephaly.6 As with all emerging infectious diseases,7 healthcare workers must keep up to date with information about how to detect and manage Zika virus. This paper describes the adaptation of the identify-isolateinform (3I) tool (initially developed for Ebola virus disease8,9 and modified for measles10 and Middle East Respiratory Syndrome (MERS))11 for use in the detection and management of potential Zika virus patients presenting to the ED, including women who are pregnant or contemplating pregnancy, and their partners. Using this novel 3I tool, emergency physicians will be better prepared to detect and manage patients presenting to the ED with concerns about Zika virus.

Volume XVII, no. 3 : May 2016

CLINICAL PRESENTATION Zika virus is of particular concern to women who are pregnant or considering becoming pregnant (and their partners) as it has been detected in fetuses, amniotic fluid, and semen. These patients, especially if they have been in, or are considering travel to, areas with documented Zika transmission, may present to the ED requesting information and testing. Signs and Symptoms Zika infection is asymptomatic in 80% of cases. When symptomatic, it is typically described as a mild dengue-like illness, manifested by two or more of the following: sudden onset fever, conjunctivitis, arthralgias, and maculopapular rash (Figure 2). Myalgias and headache are also common. The WHO interim case definition describes suspected, probable and confirmed cases.12 Unlike in dengue, death from acute Zika virus infection is rare, but has been reported in a child with sickle cell disease in Colombia.13 The incubation period for Zika virus after a person is bitten by

Figure 1. Countries with confirmed Zika cases as of April 4, 2016.* *Source: http://www.cdc.gov/zika/geo/active-countries.html.

239

Western Journal of Emergency Medicine

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management

testing is not recommended for asymptomatic persons, with the exception of pregnant women 2 to 12 weeks after travel to areas with ongoing Zika virus transmission.17 In symptomatic patients, serum should be obtained for polymerase chain reaction (RT-PCR), IgM, and IgG if within seven days of symptom onset. If more than seven days has elapsed, only IgM and IgG should be assessed; RT-PCR is not indicated.18 Note that serological cross reactivity may occur between Zika and other flaviviruses (e.g., dengue, yellow fever, St. Louis encephalitis, Japanese encephalitis, West Nile). Public health departments may decline to test if the clinical scenario is not suggestive of Zika virus or associated complications. In particular, due to the high cross-reactivity with related flaviviruses, serologic test interpretation is complex and can be difficult, leading to false positive results. Conversely, a negative Zika IgM or RT-PCR test result does not rule out Zika virus infection. Emergency clinicians should consult with public health experts to assist with interpretation of test results.

Figure 2. Typical maculopapular rash of Zika virus.* *Photograph courtesy of Dr. David C. Pigott, University of Alabama at Birmingham.

an infected mosquito ranges from 2-7 days, and may be up to 14 days. This is the reason for screening for risk factors within 14 days of symptom onset. The incidence of GBS or microcephaly appears to be similar between symptomatic and asymptomatic cases. Transmission Zika virus is transmitted by the bite of the Aedes aegypti mosquito and likely also Aedes albopictus. Mosquito bites account for the vast majority of infections, and thus prevention from bites is the principle protective measure. In addition, Zika has been detected for prolonged periods in semen and documented cases of sexual transmission are increasing.14,15 Intrauterine/perinatal transmission also occurs. Furthermore, laboratory exposure and other blood-borne transmissions are possible; thus, blood banks have developed screening and protection criteria to mitigate spread.16 Theoretically, Zika virus could also be contracted from infected breast milk or organ/tissue transplants. Emergency Department Patient Evaluation Patients should be assessed for pregnancy as well as for neurological symptoms and signs indicating possible GBS. While the situation is rapidly evolving, as of March 2016, laboratory testing is only available through the Centers for Disease Control and Prevention (CDC) and several state and territory health departments. Generally, Western Journal of Emergency Medicine

Differential Diagnosis The differential for Zika virus infection is broad as many diseases present similarly. In patients who have traveled to, or lived in, an area with ongoing Zika transmission within 14 days, chikungunya and dengue viruses, as well as malaria, leptospirosis, and rickettsial infections, are among the other infections that should be considered. Co-infection with dengue and Zika has been reported. It is likely that some travelers to Zika endemic areas will be found to be co-infected with other pathogens acquired from arthropod bites, especially malaria. Distinguishing these infections from one another based on clinical symptoms and signs can be difficult, but malaria is not typically associated with rash or conjunctivitis. Treatment Treatment for Zika virus infection is entirely supportive as no specific antiviral therapy is yet available. Aspirin and nonsteroidal anti-inflammatory drugs should be avoided until dengue can be ruled out, to reduce the risk of hemorrhage. As noted, the vast majority of cases (approximately 80%) are asymptomatic; most of the remainder of patients have self-limited illnesses, with symptom resolution within seven days.19 Patients who develop GBS may require intensive care, including mechanical ventilation. Pregnant women who test positive for Zika virus should be offered ultrasound assessments every 3-4 weeks to assess for microcephaly with consideration for amniocentesis for Zika virus RNA testing on amniotic fluid. Fetal microcephaly is often not detectable in the first trimester of pregnancy but is most easily diagnosed in the late second and early third trimester. In some cases, it can be missed on prenatal ultrasound and not detected until after birth. Prevention While under development, as of early 2016, no vaccine

240

Volume XVII, no. 3 : May 2016

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management

Identify, Isolate, Inform

Emergency Department Evaluation and Management of Patients Under Investigation (PUIs) for Zika Virus Information Current as of March 22, 2016

IDENTIFY

EXPOSURE

• • •

SYMPTOMS

Exposure PLUS Symptoms

Within 14 Days (incubation period 2-7 days)

* 80% of infected persons are asymptomatic

Travel to endemic areas, including Brazil (see map) PLUS Inadequate mosquito bite protection

Suspected case:

Maculopapular Rash and/or Sudden Fever PLUS ≥ 1 sign or symptom

Transmitted by bites from infected Aedes aegypti or Aedes albopictus mosquitoes, however o male to female sexual transmission reported o blood borne transmission possible via transfusion o perinatal and in utero transmission possible

• • •

Headache and Myalgias also common * Consider Chikungunya and Dengue

ISOLATE?

INFORM

Use Standard Precautions

• •

•

Arthralgia Arthritis Non-purulent/hyperemic Conjunctivitis

Contact, Droplet, and Airborne Isolation precautions NOT necessary Avoid blood donation+, sexual contact^, mosquito bites during viremic phase

• •

Consult blood bank for up-to-date policies Virus may be present in semen for prolonged periods; review current guidelines for duration of abstinence + ^

• •

* Association with • Congenital microcephaly/other malformations

Hospital Infection Prevention Health Department (local or state)

Health Department coordinates testing

Test ONLY symptomatic patients EXCEPTION: Can be offered to pregnant women 2 to 12 weeks after travel to areas with ongoing Zika virus transmission * Obtain Zika serum PCR, IgM and IgG * No PCR if > 7 days since symptom onset

The Identify-Isolate-Inform tool was conceived by Dr. Kristi L Koenig, Director, Center for Disaster Medical Sciences, UC Irvine

Figure 3. Koenig’s identify-isolate-inform tool adapted for Zika virus.

Volume XVII, no. 3 : May 2016

241

Western Journal of Emergency Medicine

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management against Zika virus exists. Avoidance of travel to Zikaendemic areas or protection against mosquito bites is the best prevention. Mosquitoes that spread Zika bite mostly during the daytime. Preventative measures include the following: 1. Use approved insect repellents (DEET and permethrin are safe and effective for pregnant women when used in accordance with the product label) 2. Use window and door screens, or stay in airconditioned buildings 3. Wear long-sleeved shirts and long pants 4. Use a mosquito bed net. For patients with known or suspected Zika virus, prevent spread to others by avoiding mosquito bites (as this could infect the mosquito, which could then bite and spread the virus). Furthermore, infected men should take measures to avoid sexual transmission of virus to their partners by abstaining, using condoms, and refraining from anal intercourse and fellatio. Only male to female sexual transmission has been reported as of early 2016. Patient Disposition Admission criteria for patients who are at risk for Zika virus are similar to those for any other patient. Most patients can be managed conservatively as outpatients; however, those with complications such as GBS may require hospitalization. Identify-Isolate-Inform The identify-isolate-inform (3I) tool, initially developed for Ebola virus disease and subsequently adapted for measles and MERS, can be modified for the ED evaluation and management of patients under investigation for Zika (Figure 3). As most cases of Zika virus are mild and self-limited and would not be contagious in the ED setting, the tool does not need to be immediately applied on initial patient contact (as it would be for highly contagious infectious diseases). The first step in using the Zika 3I tool is to identify patients with a possible Zika exposure within 14 days of symptom onset. This involves taking a travel history and assessing whether the patient was bitten by mosquitoes or has had sex with a patient with Zika virus. If the patient lives in, or has traveled to, areas with ongoing Zika virus transmission within two weeks, assess whether two or more of the following symptoms are present: sudden fever, maculopapular rash, arthralgias, and conjunctivitis. If the epidemiologic and symptom screens are positive, or if a pregnant patient requests testing within 2 to 12 weeks after potential exposure, investigate for the presence of Zika virus by serologic (antibody) testing. While every patient presenting to the ED should be assessed for the potential to transmit disease (the “vital sign zero” concept)20, if only arboviruses, malaria or rickettsia are being considered in the differential diagnosis, isolation is not necessary. As with all patients, standard precautions should be maintained; however, contact, droplet and airborne Western Journal of Emergency Medicine

precautions (as would be needed for selected other infectious diseases) are not required. Isolation is NOT necessary for Zika (due to its transmission characteristics, e.g. which are different from MERS and Ebola). The second “I” in the 3I tool can be considered to be “investigate” rather than “isolate” if diseases that are contagious from person-to-person are not being considered. “Isolate,” however, remains a useful term to remind the healthcare providers who initially assesses the patient to consider whether isolation precautions are needed, as patients may present with similar symptoms associated with other infectious diseases that are contagious from person to person. In addition, to avert acquiring disease, people should “isolate” from mosquito vectors to prevent infection and, if infected, to prevent potential transmission to others. Patients presenting in high-risk areas of the U.S. should be advised to use precautions such as staying indoors in an air-conditioned environment and wearing long-sleeved shirts and long pants. On March 16, 2016, the U.S. National Center for Atmospheric Research published a public release on potential Zika virus risk estimated for 50 U.S. cities indicating that “summertime weather conditions are favorable for populations of the mosquito along the East Coast as far north as New York City and across the southern tier of the country as far west as Phoenix and Los Angeles.”21 Thus, there is a heightened concern for an increased incidence of Zika virus in the U.S. The final action of the tool is to “inform.” On January 21, 2016, U.S. government regulations authorized CDC to receive case notifications for arboviral illnesses and Zika virus, adding these entities to the list of nationally notifiable infectious diseases. In addition to notifying the hospital infection prevention and control team, emergency physicians should ensure that suspected Zika cases are reported to appropriate health authorities. Patients who do not meet medical criteria for admission should be informed about measures to take to limit the risk of virus transmission, e.g. to avoid sexual activity, mosquito bites, and blood donation. Relevance to Emergency Medicine Zika virus disease will likely have substantially less direct impact on ED operations than other emerging infectious diseases such as the 2009 H1N1 influenza epidemic and 2014 Ebola virus disease outbreak. Nevertheless, due to worldwide media attention and an evolving situation, it is important for emergency personnel to be familiar with the basic principles of assessment and management of Zika virus. In particular, women who are pregnant (or considering pregnancy) and have been in (or are considering travel to) areas with ongoing Zika virus transmission, as well as women who have had sex with a male who has travelled to a region of endemic transmission, may present to the ED seeking testing and other advice.

242

Volume XVII, no. 3 : May 2016

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management

CDC issued an unprecedented travel advisory for pregnant women22, and health authorities have recommended that women living in affected areas delay pregnancy.23 In addition, patients may seek information on the potential for development of microcephaly and other birth defects and the possibility of sexual transmission of Zika virus. It is possible that in the future Zika virus will be added to the list of “TORCH” infections, which include toxoplasmosis, other (syphilis, varicella-zoster, parvovirus B19), rubella, cytomegalovirus (CMV), and herpes infections, that are associated with congenital anomalies. Patients requiring emergency blood transfusions will need to be counseled about the risks and what measures have been taken to prevent them from contracting Zika virus. Pregnant women are likely to be particularly concerned about receiving blood. Finally, there is the potential for shortages to the blood supply due to a decrease in eligible donors, particularly as Zika virus becomes more widespread over time and during warmer weather.

Bodden, Linda Dickey, and Dr. Shruti Gohil of the UC Irvine Health, Epidemiology and Infection Prevention team for their critical review of the Zika 3I Tool. Address for Correspondence: Kristi L. Koenig, MD, University of California Irvine, Department of Emergency Medicine, Center for Disaster Medical Sciences, 333 The City Boulevard West, Suite 640, Rt 128-01, Orange, CA, 92868. Email: kkoenig@uci.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Koenig et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

Additional Considerations As with all contagious infectious diseases, the question of when to use the public health tools of quarantine and isolation is critical.24,25 While at least one U.S. politician has suggested that patients returning home from areas with ongoing Zika virus transmission such as Brazil should be quarantined, there is no scientific basis for this approach.26,27 As noted previously, Zika virus is not readily transmitted from person to person, either before or after symptom onset. Caution is needed to avoid negative psychosocial consequences like stigmatization such as was seen in the 2014 Ebola virus disease outbreak.28,29 Zika virus has a massive economic impact in many sectors, including healthcare. Thus, there may be a shift of resources away from other critical areas of healthcare operations and research. For example, while not approved, in the U.S. alone, President Obama requested 1.8 billion USD in emergency Zika funding.30

REFERENCES 1. WHO statement on the first meeting of the International Health Regulations (2005) (IHR 2005) Emergency Committee on Zika virus and observed increase in neurological disorders and neonatal malformations. Available at: http://www.who.int/mediacentre/news/ statements/2016/1st-emergency-committee-zika/en/. Accessed Feb 19, 2016. 2. WHO Vector-borne diseases. Available at: http://www.who.int/ mediacentre/factsheets/fs387/en/. Accessed Feb 19, 2016. 3. Fauci AS and Morens DM. Zika virus in the Americas — Yet another arbovirus threat. N Engl J Med. 2016;374:601-4. 4. Cao-Lormeau V-M, Blake A, Mons S, et al. Guillain-Barre syndrome outbreak associated with Zika virus infection in French Polynesia: a case-control study. Lancet. Published online Feb 29, 2016. 5. Cauchemez S, Besnard M, Bompard P, et al. Association between Zika virus and microcephaly in French Polynesia, 2013-15: a retrospective study. Lancet. Published online Mar 15, 2016.

CONCLUSION Zika is an emerging infectious disease that has gained worldwide attention in large part due to its association with fetal microcephaly and rapid global spread. As with any novel infection, it is important not only to identify and treat individual patients, but also to protect healthcare providers and the public health. The identify-isolate-inform (3I) tool is an instrument that can be used real-time on the front lines to rapidly detect and manage patients at risk for Zika virus disease presenting to the ED. Use of the 3I tool will assist emergency physicians in performing rapid and appropriate screening and management and counseling for patients concerned about Zika virus.

6. Centers for Disease Control and Prevention: Webinar—Update on interim Zika virus clinical guidance and recommendations. 2016. Available at: http://emergency.cdc.gov/coca/transcripts/2016/calltranscript-022516.asp. Accessed Mar 16, 2016. 7. Koenig KL and Schultz CH. The 2014 Ebola Virus Outbreak and Other Emerging Infectious Diseases 2014. Available at: http://www. acep.org/uploadedFiles/ACEP/practiceResources/issuesByCategory/ publichealth/The%202014%20Ebola%20Virus%20Outbreak.pdf. Accessed Feb 20, 2016. 8. Koenig KL. Identify, Isolate, Inform: A 3-pronged Approach to Management of Public Health Emergencies. Disaster Med Public Health Prep. Available at: http://journals.cambridge.org/ action/displayAbstract?fromPage=online&aid=9396817&fileId

ACKNOWLEDGMENTS The authors thank Dr. Susan Huang, Heather ReasinVolume XVII, no. 3 : May 2016

=S1935789314001256. Accessed Feb 20, 2016. 9. Koenig KL, Majestic C, Burns MJ. Ebola Virus Disease: Essential

243

Western Journal of Emergency Medicine

Koenig et al.

Identify-Isolate-Inform: Zika Virus Detection and Management Public Health Principles for Clinicians. West J Emerg Med. 2014.

journals.cambridge.org/action/displayAbstract?fromPage=online&aid

Available at: http://www.escholarship.org/uc/item/1bh1352j#page-1.

=9587810&fulltextType=AC&fileId=S1935789314001207&specialArti

Accessed Feb 20, 2016.

cle=Y. Accessed Feb 20, 2016.

10. Koenig KL, Burns MJ, Alassaf W. Identify-Isolate-Inform: A Tool

21. Potential Zika virus risk estimated for 50 U.S. Cities. National Center

for Initial Detection and Management of Measles Patients in the

for Atmospheric Research public release. 2016. Available at: http://

Emergency Department. West J Emerg Med. 2015. Available at:

www.eurekalert.org/pub_releases/2016-03/ncfa-pzv031516.php.

http://westjem.com/volume-16-issue-2/identify-isolate-inform-a-tool-

Accessed Mar 17, 2016.

for-initial-detection-and-management-of-measles-patients-in-the-

22. CDC issues interim travel guidance related to Zika virus for 14

emergency-department.html. Accessed Feb 20, 2016.

Countries and Territories in Central and South America and the

11. Koenig KL. Identify-Isolate-Inform: A modified tool for initial detection

Caribbean. Available at: http://www.cdc.gov/media/releases/2016/

and management of Middle East Respiratory Syndrome patients in

s0315-zika-virus-travel.html. Accessed Feb 20, 2016.

the emergency department. West J Emerg Med. 2015. Available at:

23. Darlington, S. CNN: Brazil warns against pregnancy due to spreading

http://escholarship.org/uc/item/3k27v8g1. Accessed Feb 20, 2016.

virus. Available at: http://www.cnn.com/2015/12/23/health/brazil-zika-

12. WHO Emergency Preparedness and Response. Zika virus disease.

pregnancy-warning/. Accessed Feb 20, 2016.

Interim case definition. 2015. Available at: http://www.who.int/csr/

24. Barbisch D, Shih F, Koenig KL. Is There a Case for Quarantine?

disease/zika/case-definition/en/. Accessed Feb 24, 2016.

Perspectives from SARS to Ebola. Disaster Medicine and Public Health

13. Arzuza-Ortega L, Polo A, Pérez-Tatis G, et al. Fatal Zika virus

Preparedness. 2015. Available at: http://journals.cambridge.org/action/di

infection in girl with sickle cell disease, Colombia [letter]. Emerg Infect

splayAbstract?fromPage=online&aid=9620407&fulltextType=RA&fileId=

Dis. 2016. Available at: http://dx.doi.org/10.3201/eid2205.151934.

S1935789315000385&specialArticle=Y. Accessed Feb 20, 2016.

Accessed Feb 24, 2016.

25. Koenig KL. Health Care Worker Quarantine for Ebola: To Eradicate

14. Musso D, Roche C, Robin E, et al. Potential sexual transmission of

the Virus or Alleviate Fear? Ann Emerg Med. 2014. Available at:

Zika virus. Emerg Infect Dis. 2015;21(2):359–361.

http://www.annemergmed.com/article/S0196-0644(14)01571-6/

15. Zika Virus (CDC Website). Available at: http://www.cdc.gov/zika/

fulltext. Accessed Feb 20, 2016.

transmission/. Accessed Feb 20, 2016.

26. Maron DF. Why we shouldn’t quarantine travelers because of Zika.

16. FDA issues recommendations to reduce the risk for Zika virus blood

Scientific American. 2016. Available at: http://www.scientificamerican.

transmission in the United States. 2016. Available at: http://www.fda.

com/article/why-we-shouldn-t-quarantine-travelers-because-of-zika/.

gov/NewsEvents/Newsroom/PressAnnouncements/ucm486359.htm.

Accessed Feb 20, 2016.

27. Koenig KL. Quarantine for Zika virus: Where’s the science? Disaster

17. Update: Interim guidelines for health care providers caring for

Med Public Health Prep. Early On Line, March 2016. Available at:

pregnant women and women of reproductive age with possible Zika

http://journals.cambridge.org/action/displayAbstract?fromPage=onl

virus exposure — United States, 2016. MMWR. 2016;65(05):122-7.

ine&aid=10264573&fulltextType=AC&fileId=S1935789316000562.

Available at: http://www.cdc.gov/mmwr/volumes/65/wr/mm6505e2er.

Accessed Apr 4, 2016.

htm. Accessed Feb 20, 2016.

28. IFRC Battling fear and stigma over Ebola in West Africa. Available at:

18. CDC, Division of Vector-Borne Diseases Memorandum on Revised

http://www.ifrc.org/en/news-and-media/news-stories/africa/guinea/

diagnostic testing for Zika, chikungunya, and dengue viruses in US

battling-fear-and-stigma-over-ebola-in-west-africa-65367/. Accessed

Public Health Laboratories. 2016. Available at: http://www.cdc.gov/

Feb 20, 2016.

zika/pdfs/denvchikvzikv-testing-algorithm.pdf. Accessed Feb 20,

29. WHO Zika Strategic Response Framework & Joint Operations

2016.

Plan. 2016. Available at: http://www.who.int/emergencies/zika-virus/

19. Lucey DR and Gostin LO. The Emerging Zika Pandemic: Enhancing

strategic-response-framework.pdf. Accessed Feb 20, 2016.

Preparedness. JAMA. 2016. Available at: http://jama.jamanetwork.

30. Korte G and Szabo L. Obama asks for $1.8 billion in emergency

com/article.aspx?articleid=2485361. Accessed Feb 20, 2016.

Zika funding. USA TODAY. 2016. Available at: http://www.usatoday.

20. Koenig KL. Ebola Triage Screening and Public Health: The New “Vital Sign Zero”. Disaster Med Public Health Prep. Available at: http://

Western Journal of Emergency Medicine

com/story/news/politics/2016/02/08/obama-asks-emergency-zikafunding/80002570/. Accessed Feb 20, 2016.

244

Volume XVII, no. 3 : May 2016

Original Research

Emergency Medical Treatment and Labor Act (EMTALA) 2002-15: Review of Office of Inspector General Patient Dumping Settlements Nadia Zuabi, BS* Larry D. Weiss, MD, JD† Mark I. Langdorf, MD, MHPE*

*University of California Irvine, Department of Emergency Medicine, Irvine, California † University of Maryland, Department of Emergency Medicine, Baltimore, Maryland

Section Editor: Gregory Moore, MD, JD Submission history: Submitted January 5, 2016; Revision received February 23, 3016; Accepted March 13, 2016 Electronically published May 4, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29705

Introduction: The Emergency Medical Treatment and Labor Act (EMTALA) of 1986 was enacted to prevent hospitals from “dumping” or refusing service to patients for financial reasons. The statute prohibits discrimination of emergency department (ED) patients for any reason. The Office of the Inspector General (OIG) of the Department of Health and Human Services enforces the statute. The objective of this study is to determine the scope, cost, frequency and most common allegations leading to monetary settlement against hospitals and physicians for patient dumping. Methods: Review of OIG investigation archives in May 2015, including cases settled from 20022015 (https://oig.hhs.gov/fraud/enforcement/cmp/patient_dumping.asp). Results: There were 192 settlements (14 per year average for 4000+ hospitals in the USA). Fines against hospitals and physicians totaled $6,357,000 (averages $33,435 and $25,625 respectively); 184/192 (95.8%, $6,152,000) settlements were against hospitals and eight against physicians ($205,000). Most common settlements were for failing to screen 144/192 (75%) and stabilize 82/192 (42.7%) for emergency medical conditions (EMC). There were 22 (11.5%) cases of inappropriate transfer and 22 (11.5%) more where the hospital failed to transfer. Hospitals failed to accept an appropriate transfer in 25 (13.0%) cases. Patients were turned away from hospitals for insurance/ financial status in 30 (15.6%) cases. There were 13 (6.8%) violations for patients in active labor. In 12 (6.3%) cases, the on-call physician refused to see the patient, and in 28 (14.6%) cases the patient was inappropriately discharged. Although loss of Medicare/Medicaid funding is an additional possible penalty, there were no disclosures of exclusion of hospitals from federal funding. There were 6,035 CMS investigations during this time period, with 2,436 found to have merit as EMTALA violations (40.4%). However, only 192/6,035 (3.2%) actually resulted in OIG settlements. The proportion of CMScertified EMTALA violations that resulted in OIG settlements was 7.9% (192/2,436). Conclusion: Of 192 hospital and physician settlements with the OIG from 2002-15, most were for failing to provide screening (75%) and stabilization (42%) to patients with EMCs. The reason for patient “dumping” was due to insurance or financial status in 15.6% of settlements. The vast majority of penalties were to hospitals (95% of cases and 97% of payments). Forty percent of investigations found EMTALA violations, but only 3% of investigations triggered fines. [West J Emerg Med. 2016;17(3):245–251.]

INTRODUCTION

The Emergency Medical Treatment and Active Labor Act

Volume XVII, no. 3 : May 2016

(EMTALA) of 1986 was enacted to prevent discrimination of patients in hospital emergency departments (ED). During 245

Western Journal of Emergency Medicine

Zuabi et al.

Review of Office of Inspector General Patient Dumping Settlements its debates and public hearings, Congress expressed its intent to ban financial discrimination and resultant “dumping” of uninsured patients on public hospitals. However, in its final legislated form, EMTALA bans discrimination of ED patients for any reason. It was enacted as part of the Consolidated Omnibus Budget Reconciliation Act (initially designated COBRA, 1985), 1 and modified as the Omnibus Budget Reconciliation Act (OBRA) of 1989. Increasing instances of refusal of care to patients with emergency medical conditions (EMC) prompted Congress to pass this unfunded mandate that required hospitals to: 1. Provide appropriate medical screening examinations (MSE) to the point of identifying or excluding an EMC 2. Stabilize EMCs according to the hospital’s capabilities 3. Provide timely consultation, treatment and hospitalization for the EMC within the “capacity” of the treating hospital and medical staff 4. Appropriately transfer unstable patients to a higher level of care (HLOC) if benefits outweigh the risks of transfer 5. Report known violations by hospitals and physicians receiving such transfers In addition, the statute provided civil penalties for violation upon both hospitals and physicians. It is investigated by the Office of the Inspector General (OIG) of the Department of Health and Human Services’ (HHS) Centers for Medicaid and Medicare Services (CMS).2 EMTALA defines an EMC as either (1) a medical condition manifesting itself by acute symptoms of sufficient severity (including severe pain, psychiatric disturbances and/ or symptoms of substance abuse) such that the absence of immediate medical attention could reasonably be expected to result in: placing the individual’s health (or, with respect to a pregnant woman, the health of the woman or her unborn child) in serious jeopardy; or serious impairment to bodily functions; or serious dysfunction of any bodily organ or part; or (2) with respect to a pregnant woman who is having contractions, that there is inadequate time to effect a safe transfer to another hospital before delivery, or that the transfer may pose a threat to the health or safety of the woman or the unborn child.3 Potential penalties for both hospitals and physicians are severe, despite the fact that neither is compensated for the cost of providing care to uninsured or underinsured patients. Physicians and hospitals are fined up to $50,000 per incident for failing to comply with EMTALA and are also at risk of exclusion from federally funded Medicare and Medicaid programs for repeated or flagrant violations.1 Furthermore, physicians and hospitals are liable regardless of intent, as determined in Roberts v. Galen, an EMTALA case that reached the U.S. Supreme Court in 1999.4 The regional offices of HHS and CMS are responsible Western Journal of Emergency Medicine

for investigating complaints of alleged EMTALA violations and forwarding confirmed violations to the OIG for possible imposition of civil monetary fines.2 The regional CMS office usually delegates the initial onsite investigation to a state department of health. Compliance with the EMTALA statute is difficult and costly, as every patient who has been deemed to have “come to the ED” must have an evaluation to exclude an EMC. Even the U.S. General Accounting Office (GAO) outlined some of these difficulties in a 2001 report. They include overcrowding from increasing ED patients with non-urgent complaints and reduced ED on-call specialist panels due to the inadequate compensation to care for a large number of uninsured patients. Furthermore, this report highlighted ambiguous regulations, including which satellite hospital sites fall under the EMTALA mandate.2 Despite 300 federal court decisions expanding interpretation of the statute with case law, the HHS maintains no ongoing transparent and public reporting system for potential violations. Even the U.S. GAO has criticized the OIG enforcement as inconsistent and weak.5 While the OIG publicly discloses settlements on its website, these have not been compiled or analyzed since 2006.6 This paper provides an update on the activity of EMTALA settlements reported by the OIG, as well as the scope, cost, frequency, regional location, and most common allegations leading to settlements by hospitals and physicians for EMTALA violations. The descriptions of these settlements uniformly include a statement that settlements were made without admission of guilt on the part of hospitals or physicians. METHODS We reviewed the OIG EMTALA archives in May 2015, which included EMTALA cases settled from 2002-2015.7 (https://oig.hhs.gov/fraud/enforcement/cmp/patient_dumping. asp accessed May 2015 with most recent report 3/17/15.) Each settlement includes a one-paragraph synopsis of the case, with varying detail. The website uniformly names the hospital, state, physician name (if applicable), and amount of fine. None mentioned loss of federal funding as a consequence. The description often included the general clinical diagnosis of the patient, mechanism of injury and age, and most importantly, category of alleged violation. Age was not uniformly reported, and at times the description included reference to multiple patients per report without ages. Narrative descriptions were codified by one of the authors (NZ), and entered into an Excel spreadsheet (version 7.0, Microsoft Corporation, Redmond, WA, 2007). We recorded the following 12 categories of alleged violations:

246

1. 2. 3. 4. 5.

Failure to screen for an EMC Failure to stabilize a patient with an EMC Inappropriate transfer of a patient with an EMC Failure to transfer a patient with an EMC Patient turned away for insurance or financial status Volume XVII, no. 3 : May 2016

Zuabi et al.

Review of Office of Inspector General Patient Dumping Settlements

6. 7. 8. 9.

Patient in active labor On-call physician refused to see patient with EMC Patient with EMC inappropriately discharged Hospital did not accept referral for transfer in of patient with EMC 10. No specialist physician available upon patient with EMC arrival 11. ED on ambulance diversion 12. Hospital where patient presented had capacity to care for EMC but refused Each of the 192 entries in the OIG database was allocated to as many categories (1-12) as appropriate. Therefore, the number of allegations described here exceeds the number of OIG reports by design. We obtained information regarding total number of EMTALA allegations (denominator) in the U.S. from three sources. The first was a website maintained by the Association of Healthcare Journalists (AHCJ).8 This database covers January 6, 2011, to May 13, 2015, (4.5 years). We used the search term 489.24, which is the CMS code designation for an allegation of EMTALA violation. We designate this as “AHCJ Website.” Second, we also derived information on the total number of CMS EMTALA investigations from the AHCJ database from this same organization, which requires membership, and which we designate “AHCJ Database.” Finally, we sought personal communication from the federal administrator designated “EMTALA Technical Lead - Hospital Program Analyst Survey & Certification Group – Division of Acute Care Services Centers for Medicare & Medicaid Services.” Since this was an analysis of publicly available data, no human subjects’ approval was required. RESULTS There were 192 settlement agreements (14 per year average for more than 4,000 hospitals in the U.S.). Fines against both hospitals and physicians totaled $6,357,000 (hospital and physician average $33,435 and $25,625 respectively). There were 184 (95.8%, $6,152,000) settlements involving hospitals and eight (4.2%) against physicians ($205,000). Therefore, 97% of monetary penalties were levied against hospitals. There were 392 alleged categories (1-12) of EMTALA violations identified in these 192 brief reports from the OIG website (average 2.04 per settlement). The categories of settlements are detailed in Figure 1. The most common settlements were for failing to screen (144/192, 75%) and stabilize (82/192, 42.7%) for EMCs. Other factors are listed in Figure 1. Although loss of Medicare/Medicaid federal funding is an additional possible penalty for EMTALA violation, there were no reported cases where hospitals were excluded from federal reimbursement. There was no information on EMTALA investigations that were not subject to settlement. Volume XVII, no. 3 : May 2016

The number of settlements per year varied during the study period (Figure 2), from a high of 30 in 2003 to lows of seven per year in 2009 and 2010 (2015 is a partial year). We sought to determine the proportion of EMTALA investigations that resulted in fines, requiring a denominator of EMTALA investigations from 2002-15. This was derived from three sources, which yielded different results. The AHCJ Website (found at www.Hospitalinspections. org) returned 359 records of hospitals investigated from 201115. We scrutinized each to assure there was indeed at least one EMTALA-based allegation recorded, including the categories above, and eliminated duplicates. This yielded 338 unique EMTALA allegations in 4.5 years. Earlier data were not compiled from this source. The AHCJ Database downloaded directly from the subscription to the organization listed 527 instances of “investigation for violation of EMTALA” between January 2011 and January 2015 (4 years), among 14,516 total CMS hospital offenses that led to investigations for all causes (3.6%). Of the 14,516 offenses listed in the AHCJ Database that were investigated by CMS, 10 specific EMTALA violation categories were listed. These were not the same as the 12 categories of violations identified by the authors of this paper, which were contained in the OIG narrative descriptions. For example, the authors of this paper identified four additional categories of violations not in the AHCJ list: 1) active labor, 2) turned away for financial status, 3) no specialty physician available, and 4) ED on ambulance diversion. Conversely, the AHCJ Database had two categories that the authors did not find in the OIG website: 1) ED log (a required list of all patients who present to any ED in chronological order) and 2) posting of signage regarding EMTALA obligations. The sum of all allegations in the AHCJ Database was 1,386 (multiple allegations in each of 527 investigations), and we report the breakdown of these alleged violations by category in Figure 3. Therefore, depending on the denominator taken from the AHCJ Database (n=527) or the AHCJ Website (n=338), there were between approximately 75-130 EMTALA investigations per year from 2011-2015. Extrapolating these 4 to 4.5 years of data over the 12-year period of OIG listed EMTALA settlements would amount to approximately 100 per year, or 1,200 investigations. Given 192 monetary settlements, this indicates that, at most, approximately 16% of EMTALA investigations result in fines. The third source of information (and perhaps most reliable) on the scope of EMTALA allegations and investigations is CMS’ own database, operational since 2004. Between then and 2015, there were 6,316 complaints received (approximately 574 per year), 6,035 investigations done by CMS (549/year), of which 2,436 found EMTALA violations (221/year).9 This source documents more than four times as many investigations per year as the AHCJ website or database contain. Using these data to extrapolate over the 12-year period during which OIG settlements are publicly reported,

247

Western Journal of Emergency Medicine

Review of Office of Inspector General Patient Dumping Settlements 160

would drop the proportion of investigations that result in OIG fines to 3.2% (192/6,035). Furthermore, the proportion of even CMS findings of liability (n=2,436) that resulted in OIG settlements occurred in only 7.9% (192/2,436) of cases. There was significant variation in the number of settlements by CMS region (Figure 4), from a high of 68 in Region 4 (Southeast) to a low of 0 in Region 10 (Pacific Northwest), shown in Figure 5.

144

Alleged Violations

140 120 100 80

60 40 20

Figure 1. Factors associated with monetary settlement with U. S. Office of Inspector General for allegations of violation of Emergency Medical Treatment and Labor Act (EMTALA), 2002-15 (n= 392 for all categories of violations).

Figure 2. Number of monetary settlements with the U.S. Office of Inspector General for allegations of violation of Emergency Medical Treatment and Labor Act (EMTALA), 2002-15 (n=192).

Figure 3. Emergency Medical Treatment and Labor Act (EMTALA) violations (2011-15) by category, from the Association of Healthcare Journalists Database. These 1,386 categories of alleged violations were contained in 527 separate CMS (Centers for Medicare and Medicaid Services) investigations. ED, emergency department

Western Journal of Emergency Medicine

Zuabi et al.

DISCUSSION This paper adds unique data to the emergency medicine literature, as it provides the first review of recent CMS EMTALA investigations and civil monetary penalties. As the investigation process is complaint-driven, these data reveal that most complaints relate to allegations of improper screening examinations, followed by improper stabilization, and then improper transfers, but only a minority of allegations result in fines. The EMTALA mandate began in 1986 in response to high-profile cases where patients were denied care, or were transferred from the ED without receiving necessary stabilizing care. Prior to EMTALA, in many communities the under- and uninsured received care in public hospitals (usually academic medical centers), while privately insured patients often received care in privately-owned community hospitals. When patients without insurance presented with medical emergencies or active labor to private hospitals, they were often turned away, some with bad outcomes in transit, or after arrival at public hospitals. Rather than dealing directly with inequitable funding of providers and hospitals for healthcare, EMTALA placed an unfunded mandate on EDs and on-call specialists to provide screening and stabilizing care to everyone, without regard for insurance or ability to pay. Consequently, regional surveys of community hospital EDs in California documented the erosion of on-call specialty panels from 2000 to 2006. The EMTALA mandate was an important driver, as specialists refused to take ED call to avoid being subjected to EMTALA mandates to care for the unfunded.10 This further shifted underfunded and unfunded care to university and public hospitals. This trend accelerated after November 2003 when CMS significantly amended the EMTALA regulations. The new amendments significantly changed the duty of hospitals to provide panels of on-call physicians to their EDs. Prior to that time, the on-call panel had to include all specialties represented in the organized medical staff of the hospital. After November 2003, hospitals only had to provide an on-call panel that reasonably met the needs of its community.11 Very quickly, surgical subspecialists began disappearing from hospital call schedules. The â&#x20AC;&#x153;needs of the communityâ&#x20AC;? were then served by transferring patients to academic medical centers and other tertiary care facilities. Therefore, for a variety of reasons, the post-EMTALA provision of emergency care no longer involves

248

Volume XVII, no. 3 : May 2016

Zuabi et al.

Review of Office of Inspector General Patient Dumping Settlements

Figure 4. Number of monetary settlements with U.S. Office of Inspector General by Centers for Medicare and Medicaid (CMS) region, for Emergency Medical Treatment and Labor Act (EMTALA) violations, 2002-15.

Figure 5. Center for Medicare and Medicaid (CMS) regions relevant to reporting of U.S. Office of Inspector General Emergency Medical Treatment and Labor Act (EMTALA) enforcement.

many specialists in some communities. However, during this time, the number of settlements appears to have decreased. After 30 years of implementation, a whole generation of physicians has been exposed to safer evaluation of potential medical emergencies prior to consideration for transfer, and to the prohibition of inquiry regarding method of payment prior to MSE. To further understand this, it is instructive to examine the CMS’s “EMTALA Physician Review Worksheet,” which gives good insight into the factors involved in an investigation (Appendix). First, there are instructions defining the medical screening examination, which can range from a “brief history and physical examination,” to a “complex process” involving ancillary studies, diagnostic procedures (such as lumbar puncture), advanced imaging, consultation and procedures performed by consultants. The standard asked of the CMS physician reviewer Volume XVII, no. 3 : May 2016

is one of “reasonable clinical confidence,” sufficient to determine whether or not an EMC…existed.” The form asks the reviewer to identify “inappropriately long delay” prior to the MSE. “In labor” is defined as “having contractions,” without further specifying interval or severity. The reviewer is asked to determine, “with reasonable medical certainty” whether there would be time to transfer the pregnant patient safely. Patient outcome after transfer is noted to not be a determinant of appropriate stabilization, but this could be a “red flag.” However, regarding transfer of a woman in labor, delivery prior to arrival at the receiving hospital is considered a “marker of instability” for transfer. The obligation for care includes that which is “within the full capabilities” of a hospital’s staff and facilities and includes access to on-call specialists. The form asks the investigator to state the reason the EMC was not stabilized prior to transfer, and notes that the patient can refuse stabilizing treatment. During the transfer, the form asks whether this was done with the best equipment and personnel, whether the benefits outweighed the risks of transfer, whether the physician documented the benefits and risks in the medical record, and if the patient was sent with all medical records. Designation as a specialty center is not limited to, for example, trauma/burn/neonatal intensive care unit (ICU), but rather asks the investigator to opine on other hospital capabilities. On the receiving end, hospital’s ability to accept a transfer can be limited by “lack of capacity,” but this is not further defined. Disputes arise whether a receiving hospital can refuse a transfer because of lack of inpatient, ICU or operating rooms, in addition to ED beds. This element is vague in all EMTALA guidance from CMS and the statute itself. The investigator worksheet similarly leaves this undefined. The worksheet instructs the investigator specifically to not render an opinion regarding EMTALA violation. Finally, it requires the listing of the specialty of any potential violator physician, implying this is not limited to the actions of the emergency physician, but includes on-call specialists as well. There is a small chance that any individual case will lead to an EMTALA investigation and fine. However, this could have devastating consequences for a physician or hospital. A $50,000 fine may have a minimal impact for a hospital, but a significant impact on a physician, as malpractice insurance policies do not cover the cost of civil penalties. Also, when a regional CMS office authorizes an on-site investigation, they can also investigate other previous cases involving the hospital and the physician. If CMS finds repeated or flagrant violations, each violation may result in an additional fine and may result in permanent exclusion of the providers from the Medicare and Medicaid programs. Even though CMS may fine both hospitals and physicians for EMTALA violations, a patient may only legally sue the hospital for alleged injuries due to an EMTALA violation. Even though physicians do not face tort liability for alleged EMTALA violations, the duties enumerated by EMTALA quickly became 249

Western Journal of Emergency Medicine

Zuabi et al.

Review of Office of Inspector General Patient Dumping Settlements national standards of practice. Therefore, in medical malpractice litigation, patients may allege negligence due to failure of a physician to adhere to national standards of practice, including screening, stabilization, and transfer of ED patients. A discussion of alleged EMTALA violations must include the three major obligations of hospitals and physicians, the duties to appropriately screen, stabilize, and transfer patients. Congress failed to provide a definition of appropriate MSE, and this led to more litigation than any other aspect of the statute. Gradually, the federal courts of appeals developed the “comparability test.” A hospital provides an appropriate MSE when it provides an examination comparable to a similar patient.12 This reflects the fact that EMTALA is an antidiscrimination statute. If the MSE does not reveal an EMC, then EMTALA obligations are fulfilled. However, the diagnosis of an EMC gives rise to the duty to stabilize.12,13 A patient is stable if it is reasonably likely the patient will not deteriorate en route during a transfer, including patients “transferred” to home (discharged from the ED).14 When providers stabilize the patient’s EMCs the obligations under EMTALA are fulfilled.14,15 The EMTALA transfer obligations only apply to unstable patients. Therefore, if an ED cannot stabilize a patient, then it may transfer the patient if “appropriate.”16 The referring hospital must stabilize the patient to its maximum potential, must secure acceptance from a provider at a receiving hospital, and send appropriate records, personnel and equipment. The referring physician must sign a certification, in actuality an oath, documenting that the medical benefits of transfer outweigh the risks. In addition to these “appropriate” transfer requirements, patients may request or demand transfer after proper disclosure of the risks. A receiving hospital has a duty to accept all “appropriate” transfers if it has capacity (an available appropriate bed) and the capability (appropriate staff), and if the receiving hospital serves as a regional referral center or has unique capabilities not available at the referring hospital.16 Hospitals do not have a duty to accept lateral transfers.17 We were not able to find any recent hospital or physician loss of federal funding in our investigation. The last information from 2007 from CMS reported 13 hospitals had been terminated from Medicare for EMTALA violations.6 We had substantial difficulty discovering the true number of EMTALA allegations that may result in penalties. We believe the federal government’s CMS data to be most accurate, suggesting a very low (<8%) incidence of OIG monetary penalties even after CMS determines there has been an EMTALA violation.9

status in 15.6% of settlements, the original intent of the statute. The vast majority of penalties were to hospitals (95% of cases and 97% of fines). Forty percent of investigations found EMTALA violations, but only 3% of investigations triggered fines.

Address for Correspondence: Mark I. Langdorf, MD, MHPE, University of California Irvine, 333 City Blvd W, Suite 640, Orange, CA 92868. Email: milangdo@uci.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Zuabi et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

CONCLUSION Of 192 hospitals and physicians settling with the OIG from 2002-15, most were for failing to provide screening (75%) and stabilization (42%) to patients with EMCs. The reason for patient “dumping” was due to insurance or financial Western Journal of Emergency Medicine

250

1. EMTALA. American College of Emergency Physicians Web site. Available at: http://www.acep.org/News-Media-top-banner/EMTALA/. Accessed Dec 30, 2015. 2. Emergency Care. EMTALA Administration and Enforcement Issues. United States General Accounting Office. Available at: http://www. gao.gov/new.items/d01747.pdf Published June 2001. Accessed Dec 30, 2015. 3. 42 U.S.C. §1395dd(E). (See 42 CFR 489.24(b)) Reference: PT31 EMT 5D Worksheet, CMS/OIG investigative report form. Accessed Jun 26, 2015. 4. Roberts v. Galen of Virginia- The Supreme Court Misses the Boat on EMTALA. Emtala.com Website. Available at: http://www.emtala.com/ roberts.htm. Accessed Dec 30, 2015. 5. Rosenbaum S, Cartwright-Smith L, Hirsh J, et al. Case studies at Denver Health: ‘patient dumping’ in the emergency department despite EMTALA, the law that banned it. Health Aff (Millwood). 2012;31:1749-56. 6. www.ehcca.com/presentations/uninscong1/yeh_1.ppt PowerPoint presentation from Charlotte Yeh MD, Regional Administrator, Centers for Medicare and Medicaid Services, given to the National Congress on the Un and Under Insured, December 10, 2007, Washington, D.C. 7. Patient Dumping. Office of Inspector General U.S. Department of Health and Human Services Website. Available at: https://oig.hhs. gov/fraud/enforcement/cmp/patient_dumping.asp. Accessed May 2015. 8. HospitalInspections.org. Association of Health Care Journalists Website. Available at: http://www.hospitalinspections.org/. Accessed Oct 22, 2015. 9. Personal communication from Mary Ellen Palowitch MHA RN, EMTALA Technical Lead - Hospital Program Analyst Survey &

Volume XVII, no. 3 : May 2016

Zuabi et al.

Review of Office of Inspector General Patient Dumping Settlements

Certification Group – Division of Acute Care Services Centers for

13. 42 U.S.C. §13955dd(E)

Medicare & Medicaid Services, 12/16/15.

14. See, e.g.: Green v. Touro Infirmary, 992 F.2d 537, (5th Cir. 1993)

10. Rudkin SE, Langdorf MI, Oman JA, et al. The worsening of ED

15. 42 U.S.C. §1395dd(C)

on-call coverage in California: 6-year trend. Am J Emerg Med.

16. 42 U.S.C.§1395dd(F)

2009;27:785-91.

17. Available at: https://www.cms.gov/Medicare/Provider-Enrollment-

11. 42 C.F.R. §489.24(j)(2)(iii).

and-Certification/SurveyCertificationGenInfo/downloads/

12. See, e.g.: Vickers v Nash, 78 F.3d. 139, (4th Cir. 1994)

SCLetter08-15.pdf, page 42.

Volume XVII, no. 3 : May 2016

251

Western Journal of Emergency Medicine

Brief Research Report

Prospective Validation of Modified NEXUS Cervical Spine Injury Criteria in Low-risk Elderly Fall Patients John Tran, MD* Donald Jeanmonod, MD* Darin Agresti, DO* Khalief Hamden, MD† Rebecca K. Jeanmonod, MD*

*St. Luke’s University Hospital, Department of Emergency Medicine, Bethlehem, Pennsylvania † Carilion Clinic, Roanoke, Virginia

Section Editor: Kathleen E. Walsh, DO, MS Submission history: Submitted January 5, 2016; Revision received February 8, 2016; Accepted March 9, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29702

Introduction: The National Emergency X-radiography Utilization Study (NEXUS) criteria are used extensively in emergency departments to rule out C-spine injuries (CSI) in the general population. Although the NEXUS validation set included 2,943 elderly patients, multiple case reports and the Canadian C-Spine Rules question the validity of applying NEXUS to geriatric populations. The objective of this study was to validate a modified NEXUS criteria in a low-risk elderly fall population with two changes: a modified definition for distracting injury and the definition of normal mentation. Methods: This is a prospective, observational cohort study of geriatric fall patients who presented to a Level I trauma center and were not triaged to the trauma bay. Providers enrolled non-intoxicated patients at baseline mental status with no lateralizing neurologic deficits. They recorded midline neck tenderness, signs of trauma, and presence of other distracting injury. Results: We enrolled 800 patients. One patient fall event was excluded due to duplicate enrollment, and four were lost to follow up, leaving 795 for analysis. Average age was 83.6 (range 65-101). The numbers in parenthesis after the negative predictive value represent confidence interval. There were 11 (1.4%) cervical spine injuries. One hundred seventeen patients had midline tenderness and seven of these had CSI; 366 patients had signs of trauma to the face/neck, and 10 of these patients had CSI. Using signs of trauma to the head/neck as the only distracting injury and baseline mental status as normal alertness, the modified NEXUS criteria was 100% sensitive (CI [67.9-100]) with a negative predictive value of 100 (98.7-100). Conclusion: Our study suggests that a modified NEXUS criteria can be safely applied to low-risk elderly falls. [West J Emerg Med. 2016;17(3):252–257.]

INTRODUCTION As the population ages, elder patients presenting to U.S. medical centers are becoming increasingly frequent. Older individuals are more likely to be hospitalized after sustaining a traumatic injury and now account for up to 25% of trauma admissions.1 According to a retrospective review of Medicare data between 2007 and 2011, the rate of elders presenting with cervical fractures has increased from 4.6 per 10,000 to 5.3 Western Journal of Emergency Medicine

per 10,000, while rates of hip fractures have decreased during the same time period.2 Associated with these cervical injuries is significant morbidity and mortality, with 30-day mortality rates of 13% in those without spinal cord injury and 28.4% in those with spinal cord injury.2 One-year mortality rates are respectively 24.5% and 41.7%.2 The National Emergency X-Radiography Utilization Study (NEXUS) criteria is a valuable clinical decision-

252

Volume XVII, no. 3 : May 2016

Tran et al. making tool to rule out cervical spine injuries (CSI) in the general population without radiographic imaging.3 In the elderly population, however, there has been some reluctance to accept the reliability of NEXUS criteria4-7 despite it having a demonstrated sensitivity of 99.6%3 overall and a sensitivity of 100% (95% CI [97.1-100]) in the elderly cohort of the validation population.8 One of the criticisms of the original NEXUS criteria is the subjective nature of some of the criteria used to distinguish an interpretable patient. A paper comparing resident and attending interpretation of the NEXUS criteria found poor to fair agreement when interpreting altered mental status, focal neurologic abnormality, and distracting injury, but was limited by its relatively small sample size.9 Other groups have included patients with Glasgow Coma Scale (GCS) of ≥13, despite the original criteria specifying “normal alertness.”10 Because a large proportion of elderly patients have a GCS of 14 despite having a normal level of alertness,11,12 it is important to determine if the NEXUS criteria can be applied to this cohort. Evans et al.13 demonstrated the validity of the NEXUS criteria in high-risk geriatric falls while maintaining a sensitivity of 100%. A modified NEXUS criteria was used, with more strict definitions of normal alertness and distracting injury. Normal alertness was substituted with the patient’s baseline mental status, and physical exam findings of trauma to the head or face were considered the only “distracting injuries.” With a specificity of only 12.9%, the NEXUS criteria has room for improvement to reduce unnecessary imaging.3 There are other studies that report a higher specificity when applying the NEXUS criteria, which range from 13%-46%.14 Evans et al. demonstrated an increased specificity with their modified NEXUS criteria compared to the original NEXUS criteria.13 However, this study was a retrospective review of higher risk elderly falls that were triaged to the trauma bay and probably represent a more injured group than those presenting to the average emergency department (ED). Our study aims to validate this modified NEXUS criteria in a prospective study of low-risk elderly fall patients who are not triaged to the trauma bay. MATERIALS AND METHODS Study Design This is a prospective observational cohort study of elderly fall patients at a single facility. We enrolled a convenience sample of patients and subsequently reviewed their charts. Patients or their family members or chronic care facility personnel provided verbal consent at the time of enrollment to participate in telephone contact follow up. The research protocol was reviewed and approved by the institutional review board at the study facility. Study Setting and Population The study site is a Level I community trauma center Volume XVII, no. 3 : May 2016

Prospective Validation of Modified NEXUS Criteria that hosts an emergency medicine residency with 40 total residents. The annual ED census is about 75,000. There are about 2,100 trauma alerts annually, and 130-150 of these are for geriatric fall patients. Resident and attending physicians were educated regarding the study with monthly announcements made during weekly mandatory education time. Educational posters regarding the study were hung in physician documentation areas in the ED as well as in the nurses’ stations, and email reminders were sent to all ED medical providers at least bi-monthly. Patients were eligible for enrollment in the study if they were 65 years of age or older and presented to the ED with a complaint related to a fall. Additionally, patients were required to be at baseline neurologic status as per their family member or chronic care facility staff. Patients were excluded if they met major trauma criteria and were triaged to the trauma bay or if they were determined to have an acute change in baseline neurologic functioning as per the physician caring for the patient, including clinical intoxication. Patients were not excluded due to dementia, aphasia, or any cognitive or neurologic deficit that was determined by the physician caring for the patient to be the patient’s baseline. Study Protocol and Measurements Patients eligible for this study were identified by attending and resident physicians working in the ED. When an eligible patient presented for care, the physician caring for the patient would assess whether the patient was at baseline neurologic function. Then he or she would ask for verbal consent from the patient, caretaker, or chronic care facility personnel for research associates to contact the patient, caretaker, or chronic care facility personnel by phone in follow up. The physician caring for the patient then completed a data collection form regarding presence or absence of NEXUS criteria, and signs and location of head trauma. The data collection form contained a closed list of possibilities for each question, and the provider caring for the patient was instructed to circle his or her responses. Research associated entered data into a standardized Microsoft Excel 2007 spreadsheet (Microsoft Corporation, Redmond, WA). Study patients were evaluated and dispositioned at the sole discretion of the treating physician team. Research associates retrospectively reviewed each patient’s medical record after his or her ED visit to determine the results of any diagnostic testing, specifically radiographic imaging, the disposition decision and service, and any neurosurgical interventions during the hospitalization. Other significant traumatic injuries were also recorded. Significant traumatic injuries included visceral injuries or bony injuries. Soft tissue injuries such as abrasions, contusions, skin tears, and lacerations were not recorded. At 4-6 weeks after the initial ED visit, a research associate called study patients or their caregivers in follow 253

Western Journal of Emergency Medicine

Tran et al.

Prospective Validation of Modified NEXUS Criteria up. This was done to assure that any patients who were neither admitted and observed nor imaged were in fact uninjured by the fall. Patients who were called were queried as to how they were feeling globally as well as specifically queried as to neck pain, numbness, tingling, weakness, and the presence of other neurologic symptoms. Patients were queried about interval ED visits and their outcome. Patients with new or ongoing symptoms were encouraged to return to the ED for further evaluation. Date of follow up and patient responses were recorded. A patient was determined to have no significant acute neck injury if the following criteria were met: 1) He/she had a negative neck computed tomography (CT) or magnetic resonance imaging (MRI) performed; 2) the patient was admitted to the hospital and had no sequelae at discharge; 3) review of his/her medical record revealed repeat hospital visits unrelated to falls with no sequelae or complaints related to the index visit; or 4) the patient had no complaints at 30 days post-injury in telephone follow up. Data Analysis We analyzed data using descriptive statistics and chi square. Data were analyzed using MedCalc (ÂŠ1993-2013, Ostend, Belgium), VassarStats: Website for Statistical Computation (vassarstats.net, author Richard Lowry, PhD, Professor of Psychology Emeritus, Vassar College, Poughkeepsie, NY, ÂŠ 1998-2013), and Microsoft Excel 2007 (Microsoft Corporation, Redmond, Washington). With the expectation of an injury rate similar to the original NEXUS study (4.6%), we anticipated that our confidence intervals for our sensitivity calculation would be about 88-100%, with a negative predictive value of 99%. RESULTS Demographic data We enrolled 800 patients with fall events over a 16-month period in 2011-2012. One patient fall event was excluded because the patient was enrolled in the study twice during a single visit by two different providers, leaving 799 for analysis. Four patients were lost to follow up. These four patients were included in analyses of demographic and mechanistic data, but were excluded for all NEXUS calculations and outcome data. The majority of falls were unwitnessed (62.3%) and occurred at home. The reported mechanism and position at the time of fall as well as patient demographics are shown in Table 1. Since most falls were unwitnessed, mechanism and position were largely reported by the patients themselves. NEXUS criteria: normal alertness, no intoxication, and no focal neurologic deficits All enrolled patients were at their own baseline mental status, and none had new focal neurologic deficits as per protocol. Clinically intoxicated patients were excluded from Western Journal of Emergency Medicine

the study. Breath alcohol and blood alcohol testing was not routinely performed in this patient cohort, and was not recorded if it was performed. NEXUS criteria: absence of neck tenderness Of 678 patients for whom follow up was obtained, 85.3% had an absence of neck tenderness; 95 patients had neck tenderness (11.9%). In 22 patients (2.8%), the exam was equivocal or the patient was unable to verbalize tenderness. These patients were conservatively estimated to be NEXUS positive, and were considered to have neck tenderness for the purposes of this study. NEXUS criteria: absence of distracting injury Distracting injury, defined as signs of trauma to the head or neck only, was present in 366 patients (46%). An additional 114 patients with no signs of trauma to the head or neck had orthopedic injuries (most commonly hip fracture n=31,upper extremity fracture n=27, and rib fracture n=11). These orthopedic injuries, as per protocol, were not considered distracting. Therefore, 429 geriatric fall patients (54.0%) were categorized as having no distracting injury. Cervical spine injuries Four patients were lost to follow up, 329 patients (46%) underwent cervical spine computed tomography, and the remainder were either admitted and observed, called by telephone, or seen in follow up (Figure). Three patients died, one of whom had a negative CT, two of whom did not have cervical spine imaging and were excluded from further analyses. Eleven cervical injuries were found: six patients had isolated injuries to C1 or C2, one had an injury to C1 and C7, and the remainder had lower CSI (Table 2). None of the patients required operative intervention. Seven of the injured patients had cervical spine tenderness and 10 had signs of trauma to the head, most commonly to the face or frontal area (Table 3). Using the patientâ&#x20AC;&#x2122;s personal baseline mental status rather than GCS and using signs of trauma to the head as the only distracting injury, NEXUS performed well in this population, with a sensitivity of 100% (67.9-100%) and a negative predictive value of 100% (98.7100%). The specificity of NEXUS in this population was 47.7 (44.2-51.3). DISCUSSION Approximately 2.5 million elderly falls are treated in the ED every year, but only about 10% of these falls result in significant injury.15,16 Although it is important to not miss injuries when present, the majority of evaluated elderly patients will have negative evaluations. The NEXUS criteria aid in determining which patients need imaging, but case reports4,5 of missed injuries in elderly patients to whom the criteria have been applied have led to a number of providers who reflexively apply the Canadian C-Spine Rule of imaging

254

Volume XVII, no. 3 : May 2016

Tran et al.

Prospective Validation of Modified NEXUS Criteria

Table 1. Baseline characteristics of geriatric patients presenting to a Level I trauma center after a fall. Characteristics Total (n=799) Median age (IQR)

Table 3. Head trauma location and cervical spine injury. Location N=795 C spine injuries No signs head trauma

85 (79-90)

429(54.0)

244(30.7)

Head trauma location

Gender (%)

Face/frontal

Male

265 (33.2)

Parietal/occipital

52(6.5)

Female

534 (66.8)

Occipital

65(8.2)

Unknown

5(0.6)

Living environment (%) Home

450 (56.3)

Assisted living and nursing home

327 (40.9)

Other

20 (2.5)

799 Patients

Mechanism of fall (%) Trip

249 (31.2)

Loss of balance

219 (27.4)

Weakness

39 (4.9)

Dizziness

30 (3.8)

Syncope

40 (5.0)

Unsure

222 (27.8)

Position prior to fall (%) Standing

536 (67.1)

Seated

116 (14.5)

Neck CT

No Neck CT

n=329

n=470

Admitted

Discharged

n=189

n=281

Lying

49 (6.1)

Follow-up Phone Interview

Climbing stairs

10 (1.3)

n=273

Unknown

88 (11.0)

Repeat Hospital Visit n=4

Table 2. Cervical spine injuries in elderly fall patients. Patient

Injury

Dens fracture

C1 fracture

C1 fracture, occipital condyle fracture

C1 fracture, C7 burst

C4 fracture

C5 fracture

C7 fracture

to anyone greater than age 64. When this is applied to an exclusively elderly population, the specificity of the test approximates the disease prevalence and because of the low incidence of disease in a low-risk geriatric population, application of liberal imaging criteria where everyone is imaged results in an unacceptably low specificity. Volume XVII, no. 3 : May 2016

Lost to Follow-Up n=4 Figure. Patient imaging and follow up. CT, computed tomography

Similar to the NEXUS cohort, our study population had a low rate of CSI with only 1.4% of the subjects with abnormalities seen on imaging. In the cohort of the NEXUS criteria that included only geriatric patients, the reported incidence of 4.6% was approximately twice that of the larger all-age cohort. Our reported incidence is likely lower because we specifically looked at a low-risk population that was not triaged to the trauma bay. Evans et al. reported an injury rate of 7.8%, but their sample was from a high-risk elderly fall population that was triaged to the trauma bay.13 We think that the incidence of injury in our cohort accurately reflects the experience of most U.S. EDs. Using a modified NEXUS criteria, no CSI was missed resulting in a 100% sensitivity for our low-risk elderly fall 255

Western Journal of Emergency Medicine

Tran et al.

Prospective Validation of Modified NEXUS Criteria population. In addition, our study demonstrated an improved specificity of 47.7%, compared to the 12.9% of the original NEXUS study,3 the 14% in the elderly cohort of NEXUS8, and the 7.4% reported in a higher risk cohort of elder patients.13 Also significant was the demonstration that when the modified NEXUS criteria were applied to a low-risk elderly patient population, the specificity was significantly higher than the 14% when the modified NEXUS criteria that was applied to elders in a trauma bay.13 This reported improved specificity is likely the result of the modified criteria being applied to an already narrowed cohort that excluded patients with altered mental status from their baseline and patients with focal neurologic deficits. Forty-six percent of the subjects in this cohort had physical evidence of trauma to the face or head, and only 14% of the subjects had neck tenderness. Neck tenderness was the more specific criteria seen in patients with CSI, in which 6.0% of patients with neck tenderness were found to have CSI on imaging, while only 2.7% of subjects with trauma to the head or face had CSI on imaging. Because the cohort in this study only included patients who were at their normal level of alertness and without focal neurologic deficits and did not include any patients who were deemed to be clinically intoxicated, there is no information about these remaining NEXUS criteria. Because elderly patients are more likely to injure the higher cervical spine8 there is concern that these patients may be missed because of absence of midline tenderness since the cervical spine is less superficial in this area. Of the seven patients in our dataset who had an injury to C1 or C2 (64% of all fractures), only one did not have midline neck tenderness on exam. Interestingly, only one of the 11 patients with a fracture had no signs of head trauma and the majority of visible trauma identified was found to be face or frontal area (9 out of 10 patients with evidence of trauma). Perhaps the finding of trauma to the face and head can be used as a surrogate marker to identify those with the potential for significant hyperextension that causes these injuries. The idea that all elder patients with any trauma need to have their cervical spine imaged just based on their age defies common sense. The NEXUS criteria have been criticized because of the subjective nature of some of their diagnostic criteria. Hopefully, this study adds to the body of knowledge on the evaluation of cervical spine injury in elderly patients by suggesting that we can safely apply NEXUS and can more specifically apply it to patients who are at their normal mental status (but not necessarily GCS 15) and narrow the definition of distracting injury to include only evidence of trauma to the face and head. Importantly for emergency physicians, we were able to demonstrate that application of the NEXUS criteria and these modified NEXUS criteria to elderly patients results in a negative predictive value of 100%.

elderly population enrolled in this study included only injury due to falls, but did not include other mechanisms such as motor vehicle collision or assault. Therefore, the data may not be extrapolated to all-cause trauma. This population was also a low-risk patient cohort who did not meet trauma alert criteria (Appendix A), which may differ at other facilities. As well, this study was done at a single tertiary care trauma center and the geriatric population may not reflect the experience of other facilities. With our data gathering, we attempted to be as specific as possible with closed-end choices, but with any data collection, there is always the possibility of misidentification or misclassification of variables such as tenderness or the presence or location of trauma. As stated in the discussion, our reported test specificity of a modified NEXUS criteria likely overestimates the actual test performance, because the cohort excluded those with altered mental status and focal neurologic findings (who were presumptively imaged) but should not have resulted in decreased sensitivity. Finally, and perhaps most importantly, because of the relative scarcity of CSI in this cohort the 95% confidence intervals are fairly wide, and ultimately this study deserves to be repeated prospectively across multiple institutions, including significantly more patients. CONCLUSION Our study validates the use of a modified NEXUS criteria in a low-risk elderly fall population. Using “variation from baseline mental status” and “evidence of injury to the face or head” as a substitute of distracting injury resulted in a sensitivity and negative predictive value of 100%.

Address for Correspondence: Rebecca K. Jeanmonod, MD, St. Luke’s University Hospital, Department of Emergency Medicine, St. Luke’s University Health Network, 801 Ostrum St, Bethlehem PA 18015. Email: rebeccajeanmonod@yahoo.com. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Tran et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

LIMITATIONS Several limitations were identified in this study. The Western Journal of Emergency Medicine

1. Haas B, Gomez D, Xiong W, et al. External benchmarking of trauma center performance: have we forgotten our elders? Ann Surg.

256

Volume XVII, no. 3 : May 2016

Tran et al.

Prospective Validation of Modified NEXUS Criteria

2011;253(1):144-50.

2015;48(4):445-9.

2. Cooper Z, Mitchell SL, Lipsitz S, et al. Mortality and readmission

10. Konstantinidis A, Plurad D, Barmparas G, et al. The presence of

after cervical fracture from a fall in older adults: Comparison

nonthoracic distracting injuries does not affect the initial clinical

with hip fracture using national Medicare data. J Am Geriatr

examination of the cervical spine in evaluable blunt trauma patients: a prospective observational study. J Trauma. 2011;71(3):528-32.

Soc. 2015;63(10):2036-42. 3. Hoffman JR, Mower WR, Wolfson AB, et al. Validity of a set of clinical

11. Hustey FM and Meldon SW. The prevalence and documentation of

criteria to rule out injury to the cervical spine in patients with blunt

impaired mental status in elderly emergency department patients.

trauma. N Engl J Med. 2000;343(2):94-9.

Ann Emerg Med. 2002;39(3):248-53.

4. Barry TB and McNamara RM. Clinical decision rules and cervical

12. Hustey FM, Meldon SW, Smith D, et al. The effect of mental status

spine injury in an elderly patient: a word of caution. J Emerg Med.

screening on the care of elderly emergency department patients. Ann

2005;29(4):433-6.

Emerg Med. 2003;41(5):678-84.

5. Collins NC and McKenzie JV. The NEXUS criteria: do they withstand

13. Evans D, Vera L, Jeanmonod D, et al. Application of National

the test of time? Eur J Emerg Med. 2013;20(1):58-60.

Emergency X-Ray Utilizations Study low-risk c-spine criteria in high-

6. Morrison J and Jeanmonod R. Imaging in the NEXUS-negative

risk geriatric falls. Am J Emerg Med. 2015;33(9):1184-7.

patient: when we break the rule. Am J Emerg Med. 2014;32(1):67-70.

14. Michaleff ZA, Maher CB, Verhagen AP, et al. Accuracy of the

7. Goode T, Young A, Wilson SP, et al. Evaluation of cervical spine

Canadian C-spine Rule and NEXUS to screen for clinically important

fracture in the elderly: can we trust our physical examination? Am

cervical spine injury in patients following blunt trauma: A systematic

Surg. 2014;80(2):182-4. 8. Touger M, Gennis P, Nathanson N, et al. Validity of a decision rule

review. CMAJ. 2012;184(16):E867-76. 15. Centers for Disease Control and Prevention, National Center for

to reduce cervical spine radiography in elderly patients with blunt

Injury Prevention and Control. Web–based Injury Statistics Query

trauma. Ann Emerg Med. 2002;40(3):287-93.

and Reporting System (WISQARS). Available at: http://www.cdc.gov/

9. Matteucci MJ, Moszyk D, Migliore SA. Agreement between resident and faculty emergency physicians in the application of

injury/wisqars/index.html. Accessed Nov 11, 2015. 16. Tinetti ME and Williams CS. Falls, injuries due to falls, and the risk of

NEXUS criteria for suspected cervical spine injuries. J Emerg Med.

Volume XVII, no. 3 : May 2016

admission to a nursing home. N Engl J Med. 1997;337(18):1279-84.

257

Western Journal of Emergency Medicine

Original Research

Quality Improvement Initiative to Decrease Variability of Emergency Physician Opioid Analgesic Prescribing John H. Burton, MD* Jason A. Hoppe, DO†‡ Jeff M. Echternach, BBA* Justin M. Rodgers, MS* Michael Donato, DO*

*Carilion Clinic, Department of Emergency Medicine, Roanoke, Virginia † University of Colorado Denver School of Medicine, Department of Emergency Medicine, Aurora, Colorado ‡ Rocky Mountain Poison and Drug Center, Denver, Colorado

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted January 4, 2016; Revision received March 2, 2016; Accepted March 8, 2016 Electronically published May 2, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29692

Introduction: Addressing pain is a crucial aspect of emergency medicine. Prescription opioids are commonly prescribed for moderate to severe pain in the emergency department (ED); unfortunately, prescribing practices are variable. High variability of opioid prescribing decisions suggests a lack of consensus and an opportunity to improve care. This quality improvement (QI) initiative aimed to reduce variability in ED opioid analgesic prescribing. Methods: We evaluated the impact of a three-part QI initiative on ED opioid prescribing by physicians at seven sites. Stage 1: Retrospective baseline period (nine months). Stage 2: Physicians were informed that opioid prescribing information would be prospectively collected and feedback on their prescribing and that of the group would be shared at the end of the stage (three months). Stage 3: After physicians received their individual opioid prescribing data with blinded comparison to the group means (from Stage 2) they were informed that individual prescribing data would be unblinded and shared with the group after three months. The primary outcome was variability of the standard error of the mean and standard deviation of the opioid prescribing rate (defined as number of patients discharged with an opioid divided by total number of discharges for each provider). Secondary observations included mean quantity of pills per opioid prescription, and overall frequency of opioid prescribing. Results: The study group included 47 physicians with 149,884 ED patient encounters. The variability in prescribing decreased through each stage of the initiative as represented by the distributions for the opioid prescribing rate: Stage 1 mean 20%; Stage 2 mean 13% (46% reduction, p<0.01), and Stage 3 mean 8% (60% reduction, p<0.01). The mean quantity of pills prescribed per prescription was 16 pills in Stage 1, 14 pills in Stage 2 (18% reduction, p<0.01), and 13 pills in Stage 3 (18% reduction, p<0.01). The group mean prescribing rate also decreased through each stage: 20% in Stage 1, 13% in Stage 2 (46% reduction, p<0.01), and 8% in Stage 3 (60% reduction, p<0.01). Conclusion: ED physician opioid prescribing variability can be decreased through the systematic application of sharing of peer prescribing rates and prescriber specific normative feedback. [West J Emerg Med. 2016;17(3):258–263.]

INTRODUCTION Pain is a major reason for patients to seek care in the emergency department (ED), accounting for approximately Western Journal of Emergency Medicine

42% of ED visits.1-3 Emergency physicians have been scrutinized for inadequately addressing analgesia.3-6 National pain treatment initiatives have contributed to a substantial 258

Volume XVII, no. 3 : May 2016

Burton et al.

Initiative to Decrease Opiod Analgesic Prescribing Variability

increase in the prescribing of opioid analgesics in the United States. From 1999 to 2008, retail sales of opioid analgesics nearly doubled; unfortunately, prescription opioid abuse and overdose rates have paralleled this rise in opioid prescribing.7 Deaths in the U.S. attributed to prescription opioid overdose have surpassed deaths due to motor vehicle crashes as the leading cause of accidental death.8 Much of the dialogue regarding possible solutions to the epidemic of opioid abuse and overdose includes addressing the physicianâ&#x20AC;&#x2122;s role in prescribing opioids. The decision as to whether opioids are necessary, appropriate, and safe is complex. In their contemplation of how to address pain, healthcare providers attempt to balance the patientâ&#x20AC;&#x2122;s needs and goals with the benefits and potential side effects of available treatments. Despite the heightened awareness of harm from opioids and recent interventions, physician opioid prescribing practices remain highly variable.7,9,10 Tamayo-Sarver et al. found the decision to prescribe opioids to be very inconsistent, even when physicians are provided identical patient scenarios. Moreover, respondents in this investigation reported that the same clinical information (e.g. a patient requesting a strong analgesic) changed the likelihood of prescribing opioids in opposite directions for different physicians.11 It is not known whether providing objective feedback to physicians regarding their own opioid analgesic prescribing and that of their peers can decrease the variability of opioid prescribing within an ED practice environment. To our knowledge there have been no studies assessing the effect of quality improvement (QI) initiatives on physician opioid prescribing practices. To address this question, we used electronic health record (EHR) prescription information to generate aggregate and prescriber-specific data for our ED physician group. During the implementation of a QI project, we measured the impact of using sequential feedback of this prescribing data on the variability of opioid prescribing within the group. We hypothesized that sharing of peer opioidprescribing rates with normative feedback to physicians would decrease prescribing variability among an emergency medicine physician group. METHODS Study Design and Setting This QI initiative was implemented in a large, hospitalemployed physician group staffing seven EDs with approximately 265,000 combined annual visits. The setting included a variety of practice locations including both academic and private as well as urban and rural locations. All sites share a common EHR (Epic 2010 Verona, WI). All prescriptions are ordered electronically via the EHR. The local institutional review board (IRB) determined this project was QI and waived further IRB oversight. Subjects The physician group includes both EM boarded and Volume XVII, no. 3 : May 2016

non-EM boarded physicians working fulltime in the seven EDs. Pediatric EM physicians, mid-level providers and residents were excluded. We included for analysis all opioid prescriptions written at ED discharge over the 15-month data collection period from February 2012 to April 2013. Prior to study initiation, an institution-specific controlled medication prescription policy and an Internet-based statewide prescription drug monitoring program (PDMP) existed. The institution specific policy was instituted in 2009 and was identical across all practice sites with no changes during the data collection period for this project. The PDMP in our state includes mandated pharmacist entry for all controlled substance prescriptions at the time the prescription is filled. The PDMP was accessible to all physicians in the ED practice group and it remained unchanged with regards to entry of patients and prescriber access during the data collection period. There was no formal policy in the physician practice group or at a state level that structured or directed use of this program. Therefore, physicians used this database at their own discretion with influences on their prescribing patterns unique to each physician. Methods and Measurements We abstracted clinical data for all ED discharges from the EHR via a computer algorithm. These included visit date opioid analgesic prescriptions (medication and quantity of pills), and ED provider identity. Tramadol and liquid cough preparations were excluded from the analysis a priori as the focus was on the most common prescription opioids. No patient identifiers (medical record number or patient identity) were recorded in the database or reported to the prescribing physicians. This QI project incorporated three sequential stages. Study stage 1 was a nine-month baseline period during which physicians were unaware of the evaluation or planned report for this stage or future stages. The project lead investigator was the only physician aware of the data gathering, the intention to perform a QI intervention, and the initial results. In stage 2, physicians were informed via physician practice meetings and protected institutional email of the following: (1) a three-month prospective opioid prescribing report would be collected for all ED discharges; (2) prescribing data would be tracked; and (3) the intent of the initiative was to inform physicians of their practices and practice group mean opioidprescribing rates. Physicians were informed that they would privately receive their own data at the conclusion of this period and that there would be no punitive use of the data. They were unaware of the baseline data set. At the conclusion of stage 2 each ED physician received their own data via email demonstrating their individual rate of opioid analgesic prescribing. Each physician also received the calculated mean rates of opioid prescribing for discharged ED patients for their peer groups, both at their specific practice site and all other practice sites in the physician group. Therefore, all physicians were aware of their own frequency

259

Western Journal of Emergency Medicine

Initiative to Decrease Opiod Analgesic Prescribing Variability

Burton et al.

of prescribing opioids at discharge, but remained blinded to all other individuals in the group except for the means for each practice site and the group mean. Lastly, physicians were informed via the same methods (meetings and email) that a second three-month interval (stage 3) would begin after their receipt of the stage 2 data report. In stage 3, the same data would be collected for each physician using the same methods. Physicians were informed that at the completion of this stage, they would again receive their personal opioid prescribing rates and the group means both overall and at each site. In addition, each physician would also receive the unblinded data results for each physician in the group.

after import from the EHR data report. Descriptive statistics for continuous variables were presented using means, quartiles and standard deviations (SD). We presented categorical variables as the percentage of the occurrence frequency with p values generated using chi-square tests and t-tests, as appropriate. A repeated measures analysis of variance model was used to test for differences in mean rates across three stages. The mixed model included a fixed effect of stage (baseline, stage 2, and stage 3 modeled categorically). We used a general unstructured variance-covariance matrix to account for the correlation of repeated assessments on each participating prescribing physician, and to allow for potentially unequal variances across stages. A Tukey-Kramer multiple comparison procedure was used to compare all pairs of means. We also used statistical inference to determine if the variance parameters were different across the three stages.

Outcomes The primary outcome goal of the QI project was to decrease variability of the physician groupâ&#x20AC;&#x2122;s opioid prescribing for discharged ED patients. This process endpoint was chosen given that QI projects that improve system defects have the potential to maximize the effect on the care of all patients.12 Group prescribing variability was measured with sequential interventions, both when physicians became aware of their own frequency of prescribing relative to their site and group practice mean (stage 2) and when they were aware that unblinded peer physician-prescribing data would be provided (stage 3). We hypothesized that transparency of prescribing behaviors with provider-specific feedback would decrease variability in physician frequency of opioid prescribing, resulting in the physician group normalizing around the group mean during sequential assessments. Secondary outcomes for each stage were changes in opioid prescribing frequency and changes in the quantity of pills written per opioid analgesic prescription. The calculation of our opioid analgesic prescription reporting metrics was straightforward. Two metrics were calculated based on the data collected: (1) opioid prescribing rate, defined as the number of patients discharged from the ED with opioid prescriptions divided by total number of patient discharge encounters for each individual prescriber, and (2) opioid pill count, defined as the mean quantity of opioid analgesic pills written per prescription. Using these two metrics, we were able to generate both physician specific and aggregate data for each site and the complete physician group. Analysis We used Microsoft Excel 2002 (Microsoft Corporation, Seattle, WA) for data entry and simple descriptive statistics

RESULTS We included for analysis all 47 physicians who remained in the group during the measured interval. Twenty percent of the groupâ&#x20AC;&#x2122;s physicians were non-emergency medicine residency trained. Overall, there were 149,884 eligible patient discharges combined for all practice sites (Stage 1: 82,24; Stage 2: 35,525; Stage 3: 32,118). The variability in physician prescribing decreased through each stage of the initiative, as represented by the distributions of the standard error and standard deviations for each mean rate (Table 1). The overall results of the mixed-effects model for physician opioid prescribing rates suggest there is at least one difference between mean prescribing physician rates among the three separate stages (Figure 1). For a more granular evaluation of the differences in rates means, we applied a Tukey-Kramer multiple comparison procedure to all pairwise comparisons of mean rates. Differences between each pair of mean stage rates were demonstrated (Table 2). The mean opioid pill count decreased in each stage: 16 pills in Stage 1, 14 pills in Stage 2 (18% reduction, p<0.01), and 13 pills in Stage 3 (18% reduction, p<0.01). The group mean prescribing rate also decreased through each stage. Physician opioid prescribing rate: 20% in Stage 1 (baseline), 13% in Stage 2 (46% reduction, p<0.01), and 8% in Stage 3 (60% reduction, p<0.01) (Table 1). LIMITATIONS External validity for this QI initiative is limited by the use a single physician group in one state. Webster et al. reported

Table 1. Individual stage physician opioid prescribing rate mean with corresponding standard error and standard deviation. Project stage

Opioid prescribing rate (mean)

Standard error

0.200

0.012

0.084

0.130

0.010

0.071

0.083

0.007

0.048

Western Journal of Emergency Medicine

260

Standard deviation

Volume XVII, no. 3 : May 2016

Burton et al.

Initiative to Decrease Opiod Analgesic Prescribing Variability

Table 2. Tukey-Kramer pairwise comparison of physician opioid prescribing rate means between each stage to evaluate for differences between stages. Estimate

Adjusted p-value

Baseline to 2

Stage comparison

0.069

<0.0001

Baseline to 3

0.117

<0.0001

2 to 3

0.047

<0.0001

Figure 1. Mean stage physician prescribing rates and corresponding standard errors by stage of intervention.

that geographic variation in opioid prescribing for acute back pain was associated with state-level contextual factors framed by social conditions such as income, healthcare access and workers compensation.13 Our patient population may differ from those at other centers; however, the fundamental concept of reducing prescribing variability in the practice of emergency medicine remains valid. We chose not to compare practice sites given the overlap of physician practice between each site. In this large emergency medicine practice group, physicians work at different practice locations during the course of each year. Physicians typically have a “base” site with a second practice location occupying up to 50% of their clinical duties. It is uncommon, however, for physicians to practice at more than two sites. Not all physicians in the ED practice group are EM boarded. However, the percentage of emergency medicine board certification is similar to the most recent national emergency physician workforce estimates, and the intervention was directed at all ED physicians.14 The before-and-after project design is a limitation in that the validity of the results can be affected by secular trends, i.e. events in addition to the intervention may have caused a change in the outcome being studied. As noted in the methods, there were no changes in the facility or state prescribing guidelines during the data collection period and the PDMP was available Volume XVII, no. 3 : May 2016

and unchanged throughout the QI project. The nature of QI interventions and the necessity of implementing this change systemwide precluded the use of a control group; therefore, we were unable to account for the effects of other variables. The Hawthorne effect, in which individuals who know they are being observed modify their behavior while such monitoring is in effect, is a recognized limitation of interpreting QI initiatives. We attempted to minimize this during the initial stages by blinding the provider to the objective of the QI (to reduce variability). This phenomenon can actually be harnessed in QI projects to sustain positive changes in practice by committing to continuous and ongoing provider feedback.15 The reliability of the reception of individual prescribing information and the individual interpretation of this data was not measured. In the context of the larger issue of pain management and the role of opioid analgesics, the main limitation of this QI intervention was the chosen outcome of the process (opioid prescribing) rather than patient level outcomes. Measuring patient outcomes (the efficacy of opioid analgesics prescriptions, return visits for pain, patient satisfaction, abuse, misuse, and the long-term impact of this intervention) was outside the scope of this project. We did not link the individual prescriptions to diagnostic information and have no ability to measure the clinical appropriateness of each prescription. Ideally, any assessment of the appropriateness of an opioid analgesic prescription would weigh the risks against the benefit of their intended use. Our QI effort and analysis did not allow for detailed insight into the choices made by specific physicians that changed their prescribing practices during the study intervals. Lastly, we did not address the sustainability of the impact of this intervention beyond the study period. DISCUSSION Emergency physicians have perceptions of “normal” rates of opioid prescribing for ED patients. They also have a perception of how their prescribing practices compare to their peers. It is expected that these estimations influence their clinical decisions to prescribe opioid analgesics. We found that the use of provider-specific feedback to inform these perceptions can influence prescribing decisions and decrease opioid analgesic prescribing variability within an EM peer group. Social norms are beliefs and perceptions held by individuals regarding their understanding of normal behavior. Normative feedback involves informing individuals of their own behavioral patterns as compared to the behavior of their peers.16 This approach is distinct from the use of audit and feedback. Audit and feedback aims to improve practice by prompting providers (feedback) to modify behavior by comparing their past performance (audit) to professional standards or targets.17 Professional standards and targets for ED opioid analgesic prescribing are not clearly defined for the treatment of acute pain. Previous studies have shown that the 261

Western Journal of Emergency Medicine

Burton et al.

Initiative to Decrease Opiod Analgesic Prescribing Variability use of normative feedback to correct beliefs regarding social norms can influence an individual’s behavior. For example, Moreira and Latkin, respectively, demonstrated that the frequency of alcohol use in college students and the risk of human immunodeficiency virus transmission in intravenous drug abusers can be reduced through correcting social norms by informing study participants of how their behavior compares to peer behavior.16,18 Social norms may adversely affect opioid analgesic prescribing decisions if physician estimates regarding their own prescribing behaviors and the prescribing practices of their peers are not accurate. If, for example, a prescriber believes that the “normal” behavior of his or her ED peer group is to prescribe opioids to all patients with low acuity back pain, this will affect the provider’s decision to prescribe. If this assumption is inaccurate, then the decision is inappropriately influenced and the result is an action not consistent with the group. In this investigation, we used a QI initiative to collect prescribing data with normative feedback to accurately inform emergency physicians about their own prescribing rates compared to those of their ED peer group, thus correcting social norm misperceptions. This QI project is novel in its approach to informing the decision-making process. This intervention starts to address one source of poor quality regarding opioid analgesics: variability in physician opioid prescribing. A previous ED performance improvement program involving a departmental approach to the treatment of dental pain was associated with a reduction in the rate of opioid prescriptions and ED visits for dental pain through the use of a site-specific practice guideline.18 In that project, the decreased use of opioid analgesics was accomplished by specifically restricting the use of opioid analgesics for dental diagnoses. Our project did not restrict opioid use or dictate prescribing practices. Opioid analgesic prescribing to ED patients deemed appropriate for discharge has been relatively understudied. Most efforts to date have been directed at the efficacy, or complications, of specific opioid analgesics or the demographics of patients that may be either over- or under-prescribed analgesics for pain-related ED visits. Deciding whether or not opioids are the safe and appropriate choice for a given patient is wrought with physician preferences and perceptions. We have attempted to address physician perceptions and variability by providing objective prescribing rates. We observed that sharing of prescribing data among physicians was associated with a decrease in the overall opioid prescribing rate for the physician group. It is possible that these changes in prescribing habits may be due to other factors beyond this initiative or from the fact that they were being tracked (Hawthorne effect). This observation should not necessarily be interpreted as a call or attempt to decrease the number of opioid prescriptions provided to discharged ED patients, as it is not clear that decreased prescribing is the correct target. Indeed, efforts were made from the start Western Journal of Emergency Medicine

of the feedback stages to assure physicians that these data were intended to inform their decision-making and to serve as an opportunity to benchmark their individual practice by site and among their ED peers. It is possible that a reduction in prescribing of opioid analgesics may have been in conflict with the appropriate management of pain in some cases and may not be associated with better long-term outcomes. Our ultimate goal was to decrease prescribing variability via informed clinician decision-making through the use of open sharing of objective opioid prescribing data that address the possible influence of social norms. Future investigations should evaluate the effect of awareness, and open reporting among peers of physician prescribing rates and opioid pill counts on patient outcomes. This information may serve as valuable internal and external benchmark assessments for emergency medicine prescribing quality. CONCLUSION This QI initiative resulted in reduced variability among ED physicians with respect to frequency of opioid analgesic prescriptions. We also found a decrease in the frequency of opioid prescribing and number of pills per prescription. This low cost intervention that informs opioid prescribing decisions has the potential for wide-reaching impact. Emergency physician opioid prescribing practices and social norm-related prescribing behavior can be influences through the systematic application of prescriber-specific normative feedback and shared peer-prescribing rates.

Address for Correspondence: John H. Burton, MD, Carilion Clinic, Department of Emergency Medicine, 1906 Belleview Avenue, Roanoke, VA 24014. Email: JHBurton@carilionclinic.org. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Burton et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

262

1. Phillips DM. Joint Commission on Accreditation of Healthcare Organizations. JCAHO pain management standards are unveiled. JAMA. 2000;284:428-429. 2. Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioid prescribing by race/ethnicity for patients seeking care in US

Volume XVII, no. 3 : May 2016

Burton et al.

Initiative to Decrease Opiod Analgesic Prescribing Variability

emergency departments. JAMA. 2008;299:70-78.

emergency physician decision making about prescribing opioid

3. Weiner SG, Perrone J, Nelson LS. Centering the pendulum: the

analgesics. Ann Emerg Med. 2004;43(4):483-93.

evolution of emergency medicine opioid prescribing guidelines. Ann

12. Kelly JJ, Thallner E, Brioda RI, et al. Emergency Medicine Quality

Emerg Med. 2013;62(3):241-3.

Improvement and Patient Safety Curriculum. Academic Emerg Med.

4. Cantrill SV, Brown MD, Carlisle RJ, et al. American College of

2010;17:e110-129.

Emergency Physicians Opioid Guideline Writing Panel. Clinical

13. Webster BS. Geographic variation in opioid prescribing for acute, work-

policy: critical issues in the prescribing of opioids for adult patients in

related, low back pain and associated factors: A multilevel analysis. Am

the emergency department. Ann Emerg Med. 2012;60(4):499-525.

J Ind Med. 2009;52(2):162-171.

5. American College of Emergency Physicians. Policy statement.

14. Ginde AA, Sullivan AF, Camargo CA. National Study of

Optimizing the treatment of pain in patients with acute presentations.

the Emergency Physician Workforce. Annals Emerg Med.

Ann Emerg Med. 2010;56:77-79.

2009;54(3):349-359.

6. American College of Emergency Physicians. Policy statement.

15. Cumbler E, Zaemisch R, Graves A, et al. Improving Stroke Alert

Electronic prescription monitoring. Ann Emerg Med. 2012;59:241-

Response Time: Applying Quality Improvement Methodology to

242.

the Inpatient Neurologic Emergency. J Hospital Med. 2012;7:137-

7. Paulozzi LJ, Jones CM, Mack KA, et al. Centers for Disease Control

141.

and Prevention. Vital signs: overdoses of prescription opioid pain

16. Moreira MT, Smith LA, Foxcroft D. Social norms interventions to

relieversâ&#x20AC;&#x201D;United States, 1999-2008. Morb Mortal Wkly Rep.

reduce alcohol misuse in university or college students. Cochrane

2011;60:1487-1492.

Database of Syst Rev. 2009,Issue 3.

8. Warner M, Chen LH, Makuc DM, et al. Drug Poisoning Deaths in the

17. Ivers N, Jamtvedt G, Flottorp S, et al. Audit and feedback: effects on

United States, 1980â&#x20AC;&#x201C;2008 NCHS Data Brief. Number 81, December

professional practice and healthcare outcomes. Cochrane Database

2011. Available at: http://www.cdc.gov/nchs/data/databriefs/db81.htm. Accessed Jun 16, 2015.

Syst Rev. 2012;13;6:CD000259. 18. Latkin C, Donnell D, Liu T, Davey-Rothwell M, et al. The dynamic

9. Institute of Medicine (IOM). Relieving Pain in America: A Blueprint

relationship between social norms and behaviors: the results of

for Transforming Prevention, Care, Education, and Research. The

an HIV prevention network intervention for injection drug users.

National Academies Press, Washington, DC, June 29, 2011. 10. Manchikanti L, Helm S 2nd, Fellows B, et al. Opioid epidemic in the

Addiction. 2013:934-943. 19. Fox TR, Li J, Stevens S, et al. A performance improvement prescribing

United States. Pain Physician. 2012;15(3 Suppl):ES9-38.

guideline reduces opioid prescriptions for emergency department

11. Tamayo-Sarver JH, Dawson NV, Cydulka RK, et al. Variability in

Volume XVII, no. 3 : May 2016

dental pain patients. Ann Emerg Med. 2013;62(3):237-40.

263

Western Journal of Emergency Medicine

Review Article

Anticoagulation Reversal and Treatment Strategies in Major Bleeding: Update 2016 Steve Christos, DO, MS* Robin Naples, MD†

*Presence Resurrection Medical Center, Department of Emergency Medicine, Chicago, Illinois † Lewis Katz School of Medicine at Temple University, Department of Emergency Medicine, Philadelphia, Pennsylvania

Section Editor: Todd L. Slesinger, MD Submission history: Submitted November 17, 2015; Revision received March 1, 2016; Accepted March 18, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29294

[West J Emerg Med. 2016;17(3):264–270.]

INTRODUCTION Anticoagulation is the mainstay of medical treatment, prevention and reduction of recurrent venous thromboembolism, stroke prevention in patients with non-valvular atrial fibrillation, and it reduces the incidence of recurrent ischemic events and death in patients with acute coronary syndrome. Options for anticoagulation have been steadily increasing. Physicians need to be aware of the clinical profile of anticoagulation agents, reversal agents and treatment strategies in the face of major bleeding.

DOACs act directly upon Factors IIa (DTIs) or Xa (“Xabans”) without using antithrombin as a mediator.1,2 Note that each direct factor Xa inhibitor has the letters “Xa” in its spelling. See Figure 1 for an overview of the coagulation cascade and site of action of the anticoagulants.

ANTICOAGULATION TREATMENT OPTIONS Physicians have numerous anticoagulation treatment options in their arsenal including the traditional injectable indirect factor Xa and IIa (thrombin) inhibitors, which are mediated through an antithrombin dependent mechanism. Examples of injectable indirect factor Xa and IIa inhibitors include unfractionated heparin (UFH), which inhibits factors Xa and to a lesser extent IIa. UFH also inhibits factors XIIa, XIa and IXa. Enoxaparin (Lovenox), a low-molecular-weightheparin (LMWH), inhibits factors Xa and to a lesser extent IIa. Finally, Fondaparinux (Arixtra) is an exclusive indirect factor Xa inhibitor. Once an injectable indirect factor Xa and IIa inhibitor has been selected, an oral vitamin K antagonist (VKA), traditionally warfarin (Coumadin), is started on the same day. VKAs reduce the synthesis of factors II, VII, IX and X thereby reducing ability to generate active thrombin, rather than inhibition of activated enzymes. Target-specific oral anticoagulants (TSOACs) or direct oral anticoagulants (DOACs) were developed to provide more stable pharmacokinetic and pharmacodynamic options for oral anticoagulation. Examples of DOACs include the following: direct thrombin inhibitors (DTIs) dabigatran (Pradaxa) and direct factor Xa inhibitors (“Xabans”) rivaroxaban (Xarelto), apixaban (Eliquis) and edoxaban (Lixiana, Savaysa). The Western Journal of Emergency Medicine

CLINICALLY SIGNIFICANT ACUTE HEMORRHAGE The risk of major bleeding with DOACs is low; however, major life-threatening bleeding can occur. Examples of clinically significant major life-threatening bleeding include intracranial, intraspinal, intraocular, retroperitoneal, intraarticular, pericardial, intramuscular with compartment syndrome or a fall in hemoglobin >2g/dL leading to transfusion. Physicians need to have aggressive and comprehensive anticoagulation reversal and treatment strategies in the face of major bleeding. MANAGEMENT OF BLEEDING 1. Discontinue anticoagulant (know half-life) - DOACs have short half-lives (range from 5 to 17 hours), which suggests reversal drugs may not be needed in non-urgent situations; however, in emergency situations such as life-threatening major bleeding or non-elective major surgery anticoagulation reversal strategies should be established. 2. Control active bleeding. 3. Maintain adequate fluid, oxygen and hemodynamic support. 4. Transfuse packed red blood or initiate massive transfusion protocols, if necessary. 5. Consider platelets in patients who are thrombocytopenic or on anti-platelet therapy (aspirin (ASA), clopidogrel (Plavix), prasugrel (Effient), dipyridamole (Persantine) or ticlopidine (Ticlid)

264

Volume XVII, no. 3 : May 2016

Christos et al.

Anticoagulation Reversal and Treatment Strategies

Figure 1. Coagulation cascade and site of action of anticoagulants.

6. Order routine lab tests: complete blood count, basic metabolic panel, liver function tests and disseminated intravascular coagulation panel. 7. Order coagulation testing. (Utility of testing based on anticoagulant is discussed below.) COAGULATION ASSAYS Activated Partial Thromboplastin Time (aPTT) Activated partial thromboplastin time (aPTT) is a measure of the intrinsic pathway. Traditionally it has been useful to determine the anticoagulation status of patients receiving UFH. In patients who are taking DOACs, effects on the aPTT are variable. Prolongation of aPTT occurs in a curvilinear fashion in patients taking both IIa and Xa inhibitors; however, the degree of prolongation is dependent upon the reagent used.3,4,5 At therapeutic levels of dabigatran, the clinician should expect the aPTT to be prolonged. Recognize that there may still be clinical anticoagulation effects of dabigatran with a normal aPTT; however, the patient’s serum levels would fall below the therapeutic range (<80μg/L).4,5 At therapeutic levels of the direct oral “Xabans,” a PTT will not reliably be prolonged. The test is only useful in patients on Xa inhibitors at supra-therapeutic levels. Prothrombin Time (PT) and International Normalized Ratio (INR) Prothrombin time (PT) / international normalized ration Volume XVII, no. 3 : May 2016

(INR) is a measure of the extrinsic and common pathway; it is useful to determine the anticoagulation status of patients receiving VKAs. Similar to aPTT, PT/INR is variably affected by the DOACs. The degree of elevation is specific to the reagent as well as the calibration of the laboratory equipment.3-5 At therapeutic levels of dabigatran, an abnormality of the PT/INR is not expected. In a patient taking dabigatran, an elevated INR is an indication of serum levels three to four times the upper limit of normal therapeutic concentrations.4 Of the direct oral Xa inhibitors, rivaroxaban has the strongest effect on the PT/INR. At therapeutic serum concentrations, one would expect both rivaroxaban and edoxaban to cause elevation of the PT/INR.3,5,6 Apixaban weakly affects the PT/INR levels.5 Similar to stated above, a normal PT/INR does not exclude some degree of anticoagulant effect, but merely indicates a level below that expected at therapeutic dosing. Thrombin Time (TT); Also Known as Thrombin Clotting Time Thrombin time (TT) directly assesses the activity of thrombin. This test is useful in patients receiving dabigatran (Pradaxa). A normal TT excludes dabigatran activity; the degree of elevation of the TT is not a direct correlate of serum levels.4 Chromogenic Anti-Factor Xa Chromogenic anti-factor Xa testing measures the

265

Western Journal of Emergency Medicine

Anticoagulation Reversal and Treatment Strategies

Christos et al.

concentration of anticoagulants that inhibit factor Xa. This test is useful in patients receiving LMWHs, fondaparinux (Arixtra) and direct oral factor “Xabans”. ANTICOAGULATION REVERSAL AND TREATMENT OPTIONS IN MAJOR BLEEDING Anticoagulation reversal and treatment options in major bleeding include protamine, phytonadione (Vitamin K), hemodialysis, oral-activated charcoal, antifibrinolytic agents including tranexamic acid, desmopressin, blood

products including packed red blood cells (PRBCs) and platelets, prothrombin complex concentrates (PCCs), and specific reversal agents. See Figure 2 for reversal strategies of conventional anticoagulants in patients with significant bleeding. Given most clinicians’ apprehension and relative inexperience with reversal of the DOACs, below we will discuss in detail the risks and benefits of different treatment options. (See summary - reversal of direct oral anticoagulants in patients with significant bleeding, Figure 3,)7-16

Anticoagulant Heparin

Reversal agent Protamine sulfate Protamine sulfate is strongly basic and combines with acidic heparin forming a stable inactive complex 1mg per 100 units of heparin, not to exceed 50 mg Max infusion rate - 5mg/min Check aPTT 5-15min after initial dose and then at 2-8 hours

Enoxaparin

Protamine sulfate 1mg per mg of enoxaparin if last injection <8hrs 0.5mg per mg of enoxaparin if last injection >8hrs 0.5mg per mg of enoxaparin if bleeding persists after 4 hours of first dose Single dose not to exceed 50mg Max infusion rate – 5mg/min

Warfarin

4F-PCC (KCentra, Octaplex)^ If INR 2 to <4 25 units/kg; not to exceed 2500 units If INR 4 to 6 35 units/kg; not to exceed 3500 units If INR >6 50 units/kg; not to exceed 5000 units Single dose only OR If 4F-PCC is not available Fresh Frozen Plasma: 10-20mL/kg* PLUS Vitamin K: 5-10mg IVPB (20-60 minutes) with either PCC or FFP Repeat INR in 30-60 minutes after administration

Fondaparinux

4F-PCC (KCentra, Octoplex)^ # 50 units/kg; not to exceed 5000 units

Single dose only Figure 2. Reversal strategies for conventional anticoagulants in patients with significant bleeding. PCC, prothrombin complex concentrates; INR, international normalized ratio; IVPB, intravenous piggyback; FFP, fresh frozen plasma ^4F-PCC contains heparin and is contraindicated in patients with a history of heparin induced thrombocytopenia. *4F-PCC has factor concentrations 25 x greater than FFP. Reversal of INR with 4F-PCC is within 30-60 minutes versus 6-24 hours in FFP. Risk of volume overload and transfusion-related acute lung injury (TRALI) with FFP. Thrombotic event rate 8.7% for 4F-PCC vs. 5.5% with FFP. # Off-label use.

Western Journal of Emergency Medicine

266

Volume XVII, no. 3 : May 2016

Christos et al.

Anticoagulation Reversal and Treatment Strategies

Anticoagulant Dabigatran

Reversal/treatment options Idarucizumab 5.0g IV total dose (given as two separate doses of 2.5g 15 minutes apart) Alternatives if idarucizumab is unavailable Hemodialysis Activated charcoal 100g po/NG if ingestion time <2 hours 4F-aPCC (FEIBA)# 50 units/kg IV; not to exceed 5000 units (single dose only) Tranexamic acid 25mg/kg IV Desmopressin 0.3mcg/kg SQ or IV; limit to 2 IV doses given Increased risk of tachyphylaxis FFP: Not recommended rFVIIa: Not recommended

Apixaban

Activated charcoal 100g po/NG if ingestion time <6 hours 4F-PCC (KCentra / Octaplex)#^ 50 units/kg IV; not to exceed 5000 units (single dose only) Tranexamic acid 25mg/kg IV Desmopressin 0.3mcg/kg SQ or IV; limit to 2 IV doses given Increased risk of tachyphylaxis Andexanet alpha¥ 400mg IV bolus at 30mg/min followed by continous infusion at 4mg/min for 120 minutes FFP: Not recommended rFVIIa: Not recommended

Rivaroxaban

Activated charcoal 100g po/NG; if ingestion time <8 hours 4F-PCC (KCentra / Octaplex)# ^ 50 units/kg IV; not to exceed 5000 units (single dose only) Tranexamic acid 25mg/kg IV Desmopressin 0.3mcg/kg SQ or IV; limit to 2 IV doses given Increased risk of tachyphylaxis Andexanet alpha¥ 800mg IV bolus at 30mg/min followed by continuous infusion at 8mg/min for 120 minutes

FFP: Not recommended rFVIIa: Not recommended Figure 3. Reversal of direct oral anticoagulants (DOACs) in patients with significant bleeding. IV, intravenous; FFP, fresh frozen plasma; rFVIIa, Recombinant human Factor VIIa; PCC, prothrombin complex concentrates; FEIBA, Factor Eight Inhibitor Bypassing Activity; NG, nasogastric #Off label use. ^4F-PCC contains heparin and is contraindicated in patients with a history of heparin induced thrombocytopenia. ¥Not currently available on market. FDA trials ongoing. Dosing based on published Phase 3 trial.

Volume XVII, no. 3 : May 2016

267

Western Journal of Emergency Medicine

Anticoagulation Reversal and Treatment Strategies SPECIFIC REVERSAL AGENTS Idarucizumab (Praxbind) At the current time, idarucizumab is the only FDAapproved agent for reversal agent of the direct oral anticoagulants, and it only works on dabigatran. It is a monoclonal antibody fragment that binds with high affinity dabigatran (Figure 4).17 It has been shown in both in vitro studies in health volunteers to rapidly reverse the coagulation effects of dabigatran.7,8 In the RE-VERSE AD clinical trial in patients with active bleeding, idarucizumab was shown to reverse the anticoagulant effect in laboratory testing, an effect that was evident within minutes.8 Hemostasis was reported as being improved without significant thrombosis risk; however the improvement in patient outcomes (20% mortality in the group) is unclear.[8] If available, idarucizumab is recommended for reversal of anticoagulant effects for patients taking dabigatran. Andexanet Alpha Andexanet alpha acts as a decoy to target and sequester

Figure 4. Fab fragments (aDabi-Fab) reversal effects on dabigatran.

Figure 5. Andexanet alfa mechanism of action video.

Western Journal of Emergency Medicine

Christos et al. with high specificity both oral and injectable factor Xa inhibitors (Figure 5).18 It has been developed as an antidote to reverse the anticoagulant activity of oral direct (apixaban, edoxaban, and rivaroxaban) and injectable indirect (enoxaparin and fondaparinux) factor “Xabans.” Currently its manufacturer is pursuing FDA approval and it is in phase 4. ANNEXA-A and ANNEXA-R were trials published from phase 3 data that showed rapid reversal of anticoagulation effects in healthy volunteers given therapeutic doses of apixaban and rivaroxaban.9 This drug is not currently available. Aripazine Aripazine is a small, synthetic molecule with broad reversal activity administered as a single intravenous bolus dose. It binds to oral factor Xa inhibitors and DTIs, as well as UFH and LMWH via non-covalent bonding and charge-charge interactions to neutralize anticoagulation and bleeding.11 This drug is not currently available. Prothrombin Complex Concentrates (PCCs) Prothrombin complex concentrates have been developed to contain highly concentrated coagulation factors along with antithrombotic agents. Depending on the agent used, they may contain three factors (II, IX, X) or four factors (II, VII, IX, X). Four-factor (4F-) PCC may be inactive (4F-PCC; K-centra, Octaplex) or active (4F-aPCC; FEIBA); this is determined by the Factor VII component. PCCs have been shown to be successful in the reversal of VKAs and are considered first line treatment in patients with major bleeding on VKAs.11 Clotting factors are 25 times more concentrated than fresh frozen plasma (FFP) and INR reverses within minutes (15-20 minutes) ,while it may take FFP 6 to 24 hours. PCCs are available by IV immediately versus delay needed with FFP, and the risk of volume overload and transfusion-related acute lung injury (TRALI) with FFP is not seen with PCCs. Disadvantages of PCCs vs FFP include a small but real pro-thrombotic risk, availability at some institutions and cost. The cost for a course of treatment with FFP in a patient with warfarin-associated intracranial hemorrhage (ICH) and an INR of 3.0 has been estimated to be between $200 and $400 (U.S. dollars) vs $1,000 to $2,000 (U.S. dollars) for a course of treatment with PCC.19 In vitro human studies of healthy volunteers who have received oral direct IIa and Xa inhibitors have shown normalization of coagulation testing after receiving both 4F-PCC and 4F-aPCC.11,12 A variety of animal studies have been performed with dabigatran, rivaroxaban and apixaban.12-16 These studies have yielded variable results in improvement of bleeding times and overall blood loss. There have been no studies evaluating the effectiveness of four-factor PCCs in patients with active bleeding on DOACs. Given the available in vitro and animal literature, 4F-PCC (K-Centra, Octaplex) is probably beneficial in patients 268

Volume XVII, no. 3 : May 2016

Christos et al.

Anticoagulation Reversal and Treatment Strategies

taking oral direct “Xabans” with less benefit in patients on an oral direct IIa inhibitor (dabigatran). The contrary is true of 4F-aPCC (FEIBA); 4F-aPCC is probably beneficial in patients on dabigatran, with less benefit in those on the “Xabans.” With use of both four-factor PCCs, the risk of thrombosis needs to be considered against the benefit of hemorrhage control. Studies of 3F-PCC have not been performed in either in vitro or in animal models and thus, no recommendation can be made regarding its use in bleeding patient on DOACs.

It affects thrombosis by stimulating the release of von Willebrand factor (vWF) and increasing production of factor VIII. Similar to TXA, the overall cost and risk of administration of desmopressin is low. Therefore, it should be considered for use in the treatment of significant bleeding in patients on DOACs.

Recombinant Human Factor VIIa (rFVIIa, NovoSeven) Recombinant human Factor VIIa (rFVIIa) activates the coagulation cascade via the extrinsic pathway and has been used off label in the reversal of VKAs. Similar to PCCs, in vitro human and bleeding animal model studies have been performed to evaluated the benefit of this therapy in patients on DOACs.10,16 While the above-mentioned studies have shown improvement in laboratory testing, bleeding times and blood loss, the dose required to achieve benefit far exceeds the normal dosing of rFVIIa. Because of the high doses required and the concern for the possibility of thrombotic sequelae, rFVIIa is not recommended for the treatment of bleeding in patients on TSOAC therapy. Fresh Frozen Plasma (FFP) Fresh frozen plasma (FFP) is derived from whole blood and thus, contains all inactive components of the coagulation cascade in physiologic concentrations. FFP has traditionally been considered an option in patients on oral anticoagulants (e.g. VKAs) Unlike VKAs, the DOACs do not inhibit overall production of inactive coagulation components, but only bind to specific active factors. Therefore, treatment of significant bleeding in a patient taking DOACs is not aimed at repleting diminished factors. The relative concentration of coagulation factors in FFP is diminutive compared to the PCC products such that one would need over two liters of FFP to approach similar concentrations. The time to thaw and transfuse this volume of FFP also is a disadvantage. For these reasons, FFP is not recommended for reversal of DOACs unless no other agent is available.10,16 It should continue to be used as dictated by your institution’s massive transfusion protocol. Tranexamic Acid (TXA) Tranexamic acid inhibits fibrinolysis by inhibiting the binding of plasma to fibrin. While no studies have looked at the efficacy of TXA, the cost and overall risk of administering therapy (adverse reaction, thrombotic sequelae) is low. Therefore, the recommendation is that TXA should be considered for reversal of bleeding in patients taking DOACs. Desmopressin (DDAVP) Desmopressin is a synthetic analogue of vasopressin. Volume XVII, no. 3 : May 2016

Hemodialysis Dabigatran excretion is 80-85% renal which makes hemodialysis an option. Direct factor Xa oral inhibitors are mainly protein bound with only 25 to 35% renal excretion, therefore hemodialysis is not an option for the direct factor “Xabans”.10 Oral Activated Charcoal Oral activated charcoal, 100gm PO/NG x1, is an option to reduce absorption for all DOACs in the appropriate patient. Charcoal can be considered if the dose was taken within eight hours for rivaroxaban, six hours apixaban and within two hours of ingestion for edoxaban and dabigatran. CONCLUSION Physicians need to effectively manage major bleeding in patients on anticoagulants. Understanding the clinical profiles of anticoagulants and developing treatment and reversal strategies will help physicians effectively manage this lifethreatening emergency. While physicians are comfortable with the use of vitamin K, fresh frozen plasma and protamine, the use of prothrombin complex concentrates and future antidotes hold promise in improving the outcomes of patients who require emergent reversal of their anticoagulant therapy. ACKNOWLEDGEMENTS Sylvia Chen, PharmD, BCPS Regional Clinical Pharmacy Manager Clinical Specialist-Infectious Diseases/Critical Care Adrienne Perotti, PharmD, BCPS Clinical Specialist-Emergency Medicine/Toxicology Lindy Pryor, PharmD, BCCCP Clinical Specialist-Emergency Medicine

Address for Correspondence: Steve C. Christos, DO, MS, Presence Resurrection Medical Center, Department of Emergency Medicine, 7435 West Talcott, Chicago, IL 60631. Email: stevesfmc@gmail.com. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none.

269

Western Journal of Emergency Medicine

Anticoagulation Reversal and Treatment Strategies

Christos et al.

Copyright: ÂŠ 2016 Christos et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

10. Konstantinos NA and Hylek EM. Who, when and how to reverse non-vitamin K oral anticoagulants. J Thromb Thrombolysis. 2016;41(2):253-72. 11. Ferreira J and DeLosSantos M. The clinical use of prothrombin complex concentrate. J Emerg Med. 2013;44(6):1201-10. 12. Eerengberg ES, Kamphuisen PW, Sijpkens MK, et al. Reversal of rivaroxaban and dabigatran by prothrombin complex concentrate: a randomized, placebo-controlled, crossover study in healthy subjects.

REFERENCES

Circulation. 2011;124(14):1573-9.

1. Kearon C, Akl EA, Ornelas, J et al. Antithrombotic therapy for

13. Marlu R, Hodaj E, Paris A, et al. Effect of non-specific reversal

VTE disease: Chest guideline and expert panel report. Chest.

agents on anticoagulant activity of dabigatran and rivaroxaban: a

2016;149(2):315-52.

randomized ex vivo study in healthy volunteers. Thromb Haemost.

2. Harter K, Levine M, Henderson SO. Anticoagulation drug therapy: A

2012;108(2):217-24.

Review. West J Emerg Med. 2015;16(1):11-17. 3. Samama MM, Martinoli JL, LeFlem L, et al. Assessment of laboratory

14. Lee F MH, Chan A KC, Lau KK, et al. Reversal of new factor-specific

assays to measure rivaroxaban-an oral direct factor Xa inhibitor.

oral anticoagulants by rFVIIa, prothrombin complex concentrate and

Thromb Haemost. 2010;103(4):815-25.

activated prothrombin complex concentrate: a review of animal and human studies. Thromb Res. 2014;133:705-13.

4. Lindahl TL, Baghaei F, Blixter IF, et al. Effects of the oral, direct thrombin inhibitor dabigatran on five common coagulation assays.

15. Perzborn E, Heitmeier S, Laux V, et al. Reversal of rivaroxabaninduced anticoagulation with prothrombin complex concentrate,

Thromb Haemost. 2011;105(2):371-8.

activated prothrombin complex concentrate and recombinant

5. Turkoglu EI. NOACs and routine coagulation assays. How to

activated factor VII in vitro. Thromb Res. 2014;133:671-81.

interpret? Internation Journal Cardiovas Acad. 2015;1:41-2. 6. Adcock DM and Gosselin R. Direct Oral Anticoagulants (DOACs) in

16. Lazo-Langer A, Lang ES, Douketis J. Clinical management of new oral anticoagulants: a structured review with emphasis on the

the laboratory: 2015 Review. Thromb Res. 2015;136(1):7-12.

reversal of bleeding complications. Crit Care. 2013;17(3):230-42.

7. Glund S, Stangier J, Schmohl M, et al. Safety, tolerability, and efficacy of idarucizumab for the reversal of the anticoagulant effect

17. An antidote for dabigatran? Thinning out your blood thinner. Available at: http://doitwithscience.tumblr.com/. Accessed Mar 1, 2016.

of dabigatran in healthy male volunteers: a randomized, placebocontrolled, double-blind phase 1 trial. Lancet. 2015;386(9994):680-90.

18. Andexanet Alpha Mechanism of Action (MOA) Video. Available at: https://www.youtube.com/watch?v=LXA7aIAA7yQ. Accessed Mar 1,

8. Pollack CV, Reilly PA, Eikelboom J, et al. Idarucizumab for

2016.

Dabigatran reversal. N Engl J Med. 373(6); 511-20. 9. Siegal DM, Curnutte JT, Connolly SJ, et al. Andexanet alpha

19. Steiner T, Rosand J, Diringer M. Intracerebral hemorrhage associated

for the reversal of Factor Xa inhibitor activity. N Engl J Med.

with oral anticoagulant therapy: current practices and unresolved

2015;373(25):2413-24.

questions. Stroke. 2006;37(1):256.

Western Journal of Emergency Medicine

270

Volume XVII, no. 3 : May 2016

Original Research

Academic Emergency Medicine Physicians’ Knowledge of Mechanical Ventilation Susan R. Wilcox, MD* Tania D. Strout, PhD, RN† Jeffrey I. Schneider, MD‡ Patricia M. Mitchell, RN‡ Jessica Smith, MD§ Lucienne Lutfy-Clayton, MD¶ Evie G. Marcolini, MD|| Ani Aydin, MD# Todd A. Seigel, MD** Jeremy B. Richards, MD, MA††

*Medical University of South Carolina, Divisions of Emergency Medicine and Pulmonary, Critical Care and Sleep Medicine, Charleston, South Carolina † Maine Medical Center, Department of Emergency Medicine, Portland, Maine ‡ Boston Medical Center, Department of Emergency Medicine, Boston, Massachusetts § Alpert Medical School of Brown University, Department of Emergency Medicine, Providence, Rhode Island ¶ Baystate Medical Center, Department of Emergency Medicine, Springfield, Massachusetts || Yale University School of Medicine, Departments of Emergency Medicine and Neurology, Divisions of Neurocritical Care and Emergency Neurology and Surgical Critical Care, New Haven, Connecticut # Yale University School of Medicine, Department of Emergency Medicine, New Haven, Connecticut **Department of Emergency Medicine and Critical Care, Kaiser Permanente East Bay, Oakland and Richmond Medical Centers, California †† Medical University of South Carolina, Division of Pulmonary, Critical Care and Sleep Medicine, Charleston, South Carolina

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted December 15, 2015; Revision received January 19, 2016; Accepted February 5, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29517

Introduction: Although emergency physicians frequently intubate patients, management of mechanical ventilation has not been emphasized in emergency medicine (EM) education or clinical practice. The objective of this study was to quantify EM attendings’ education, experience, and knowledge regarding mechanical ventilation in the emergency department. Methods: We developed a survey of academic EM attendings’ educational experiences with ventilators and a knowledge assessment tool with nine clinical questions. EM attendings at key teaching hospitals for seven EM residency training programs in the northeastern United States were invited to participate in this survey study. We performed correlation and regression analyses to evaluate the relationship between attendings’ scores on the assessment instrument and their training, education, and comfort with ventilation. Results: Of 394 EM attendings surveyed, 211 responded (53.6%). Of respondents, 74.5% reported receiving three or fewer hours of ventilation-related education from EM sources over the past year and 98 (46%) reported receiving between 0-1 hour of education. The overall correct response rate for the assessment tool was 73.4%, with a standard deviation of 19.9. The factors associated with a higher score were completion of an EM residency, prior emphasis on mechanical ventilation during one’s own residency, working in a setting where an emergency physician bears primary responsibility for ventilator management, and level of comfort with managing ventilated patients. Physicians’ comfort was associated with the frequency of ventilator changes and EM management of ventilation, as well as hours of education. Conclusion: EM attendings report caring for mechanically ventilated patients frequently, but most receive fewer than three educational hours a year on mechanical ventilation, and nearly half receive 0-1 hour. Physicians’ performance on an assessment tool for mechanical ventilation is most strongly correlated with their self-reported comfort with mechanical ventilation. [West J Emerg Med. 2016;17(3):271–279.] Volume XVII, no. 3 : May 2016

271

Western Journal of Emergency Medicine

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation INTRODUCTION Although emergency physicians frequently intubate critically ill patients in the emergency department (ED), management of mechanical ventilation has traditionally not been emphasized in emergency medicine (EM) practice.1-4 In a previous study of EM residents, we found that while EM residents report caring for mechanically ventilated patients frequently in the ED, they received few hours of education on mechanical ventilation.5 We measured residents’ knowledge of mechanical ventilation and found that their performance on our assessment tool was only correlated with their self-reported comfort with caring for mechanically ventilated patients.5 Management of positive‐pressure ventilation can influence outcomes of critically ill patients for several conditions commonly encountered in EM practice.6-11 Patients with asthma are at high risk of complications and deterioration once intubated.8 Low-tidal volume ventilation improves mortality in patients with acute respiratory distress syndrome (ARDS).11 Careful management of oxygenation and ventilation by emergency care providers improves outcomes in intubated patients with traumatic brain injury.12,13 Furthermore, with increasing ED length of stays, emergency physicians may be responsible for management of mechanically ventilated patients for prolonged periods.14-16 Even for patients who are in the ED only briefly, ventilator-induced lung injury can occur in as little as 20 minutes.17 We designed this study to quantify academic EM attendings’ experience and knowledge regarding mechanical ventilation. We surveyed EM attendings to assess how frequently they receive education on mechanical ventilation, the frequency with which they care for mechanically ventilated patients in the ED, and their subjective comfort with managing mechanically ventilated patients. In addition, we used a knowledge assessment tool to characterize attendings’ knowledge regarding mechanical ventilation involving common emergency scenarios. We hypothesized that attendings with the most experience in managing mechanical ventilators in the ED would achieve higher scores on the assessment tool. METHODS Survey instrument development Previously, to quantify EM residents’ training experiences, we developed a five-point Likert scale survey tool to assess residents’ hours of education on mechanical ventilation, frequency with which they care for mechanically ventilated patients, and their comfort with managing ventilators. Details of survey development have been previously described.5 We modified the EM residents’ survey to better reflect attendings’ practice. Survey responses were dichotomized as affirmative or negative: responses “often” and “frequently” were defined as affirmative, while “never,” “rarely,” or “don’t know” were defined as negative. Any responses left blank were scored as “don’t know.” Western Journal of Emergency Medicine

Assessment instrument development A project team with backgrounds in EM, critical care, and educational survey development18,19 generated an assessment instrument with questions specific to EM. We created a series of questions involving key principles consistent with outlined objectives for resident education in mechanical ventilation,20 and the content was modified to be relevant to management of mechanically ventilated patients in the ED. Clinical scenarios emphasized emergency management of ventilated patients with asthma, ARDS, and traumatic brain injury, as evidence supports the importance of conscientious ventilator management for these conditions. 6,8,11,12,21-28 Questions were iteratively reviewed and edited by subject experts to optimize content, length, and relevance to the assessment tool’s goals, as previously described.5 Study Protocol Finalized versions of the survey and assessment tool were administered anonymously using REDCap (Nashville, TN) electronic data capture tools.29 REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing an interface for validated data entry. The survey and assessment tool were distributed by email to all EM attendings affiliated with the key teaching hospitals for seven EM residency training programs in the northeastern United States. The survey was sent via email invitation to attendings by a local site investigator once a week for three weeks in March and April 2015. The study protocol was approved by the institutional review boards of all participating institutions. Consent was obtained from participants at the time of participation, as the survey introduction stated that partaking of the survey indicated consent. Data Analysis Study data were exported into Microsoft Excel (Microsoft Corp., Redmond, WA) and then transferred into SPSS (v. 11.0, SPSS, Inc., Chicago, IL) for analysis. For all variables, we excluded missing data on a case-by-case basis. For the purposes of this study, we assumed the correct response rate for the assessment tool (test score) to be a surrogate for knowledge of mechanical ventilation. We examined the continuous outcome variable test score for normality in two ways. First, the outcome was examined visually using histograms and normal quantile-quantile plots. Then, Pearson’s second skewness coefficient was computed, revealing mild skew to the left, Sk2=-0.68. Survey data regarding study participants and characteristics of their training programs, mechanical ventilation educational experiences, and ventilator management experience were summarized using descriptive statistics. We used one-way analysis of variance to assess for differences in total test score across participating institutions. Tukey’s honest significant difference (HSD) was employed to assess for differences

272

Volume XVII, no. 3 : May 2016

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation

between institutional pairs. As our hypothesis was that attendings with the most exposure to managing mechanical ventilators in the ED would perform better on the knowledge assessment tool, we examined the relationship between these variables in several ways. Ordinary least squares regression analyses were performed with total test score serving as the outcome variable. The frequency with which attendings managed mechanically ventilated patients was the predictor variable. To examine the relationship between these variables after controlling for other variables significantly correlated to test score in simple correlation analysis (Spearman’s ρ), we employed hierarchical multiple regression models using the additional predictors program affiliation, completion of an EM residency training program, residency program emphasis on mechanical ventilator management, working in a setting where emergency physicians bear primary responsibility for ventilator management, and subjective comfort with managing mechanically ventilated ED patients. Exploratory regression analyses were then conducted to determine which variables, alone and in combination, were the strongest predictors of total test score. In addition to assessing normality, we evaluated additional linear regression assumptions using residual analyses and assessment of influence diagnostics. Multicollinearity was evaluated using variance inflation factors, which were all well below recommended cut points. We performed multivariate logistic regression analyses to evaluate the extent to which completion of an EM residency, program affiliation, residency program emphasis on mechanical ventilation management, working in a setting where an emergency physician bears primary responsibility for ventilator management, and level of comfort with managing ventilated patients influenced attendings’ self-reported comfort with managing mechanically ventilated patients. Coefficient estimates, adjusted odds ratios (aORs) and 95% confidence intervals (CIs) are reported for each variable. We accepted an alpha of less than 0.05 as statistically significant. RESULTS Characteristics of the Study Subjects Study surveys were distributed to 394 academic EM attendings, with 211 responding (response rate=53.6%). One physician completed the survey questions without answering any questions from the knowledge assessment tool, and seven other physicians did not fully complete the knowledge assessment tool. The response rate from the institutions ranged from 23.4 to 91.3%. The number of years as an attending emergency physician was well-distributed among the respondents, from 0-2 up to >15 years (Table 1). Educational Opportunities and Experience Managing Ventilated Patients Overall, study participants reported few educational Volume XVII, no. 3 : May 2016

opportunities regarding mechanical ventilation, as 158 attendings (74.5%) reported receiving three or fewer hours of ventilation-related education from EM sources over the past year and 98 (46%) reported receiving between 0-1 hour of education. Responses regarding educational experiences varied significantly among respondents from individual institutions (χ2=36.761, df =24, p=0.046). Similarly, only 29 (15%) respondents of those who completed an EM residency recalled mechanical ventilation being often or frequently emphasized in their training. Conversely, attendings reported frequently caring for mechanically ventilated patients in the ED. Sixty-four percent (n=136) reported that they care for four or more ventilated patients per month, and 18.5% (n=39) reported caring for 10 or more. Furthermore, 56% of respondents stated that mechanically ventilated patients rarely have changes made to their ventilator while they are in the ED. Sixty percent (n=126) of participants described feeling comfortable caring for mechanically ventilated ED patients “often” or “frequently”; whereas 38.2% (n=81) described “never” or “rarely” feeling comfortable managing these patients. Only 27.9% (n=59) described management of the ventilator as the responsibility of an emergency physician (resident or attending) at their institution, while 69% (n=145) identified the respiratory therapist (RT) as being primarily in charge of ventilator management (Table 1). Ventilator Management Knowledge The overall correct response rate for the nine-question assessment tool was 73.4%, standard deviation (SD)±19.9%. Of the 210 attendings who completed at least part of the assessment tool, 124 (59%) achieved a score of at least 70%. Significant differences in total test scores were noted between institutions (F=4.592, p<0.001). Post-hoc analysis revealed statistically significant differences in total test score between participants from the institution with the lowest mean score and those from three other institutions (p<0.001, p=0.015, and p=0.039). The relationship between participants’ years as an attending physician and scores on the knowledge assessment was not significant. Correlation analysis revealed statistically significant relationships between total test score and completion of an EM residency, program affiliation, residency program emphasis on mechanical ventilation management, working in a setting where an emergency physician bears primary responsibility for ventilator management, and level of comfort with managing ventilated patients. Relationships between total test score and having completed training in a non-EM residency, having completed a fellowship, hours of mechanical ventilation education, the frequency of managing ventilated patients, the frequency of ED-based ventilator changes were not significant (Table 2). Multivariate Results After adjusting for the effects of completion of an EM

273

Western Journal of Emergency Medicine

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation

Table 1. Emergency medicine (EM) attendings’ self-reported education and experience regarding mechanical ventilation. Survey question Respondents (%) How long have you been an EM attending? 0-2 years

36 (17.1)

3-5 years

33 (15.6)

5-10 years

56 (26.5)

10-15 years

37 (17.5)

>15 years

49 (23.2)

Have you completed an EM residency? Yes

189 (89.6)

22 (10.4)

Have you trained in another residency besides EM? No

171 (81.4)

Internal medicine

16 (7.6)

Surgery

8 (3.8)

Other

15 (7.1)

Did you complete a fellowship after EM residency?* No

111 (52.6)

Ultrasound

20 (9.5)

Toxicology

3 (1.4)

Pediatrics

10 (4.7)

Emergency medical services

9 (4.3)

Critical care

10 (4.7)

Wilderness medicine

2 (1.0)

Research

15 (7.1)

Other

32 (15.2)

Was mechanical ventilation an emphasized topic during your EM residency training? Never emphasized

8 (3.8)

Rarely emphasized

71 (33.6)

Sometimes emphasized

85 (40.3)

Often emphasized

21 (10.0)

Frequently emphasized

8 (3.8)

Not applicable - I did not do an EM residency

18 (8.5)

How many hours of instruction have you received on mechanical ventilation from other EM sources (EM articles, discussion in EM journal clubs, EM lectures/conferences, etc) in the last year? 0-1

98 (46.4)

2-3

60 (28.4)

4-5

13 (6.2)

More than 5

33 (15.6)

Don’t know

7 (3.3)

How often do you care for mechanically ventilated patients in the emergency department? Never

0 (0.0)

Rarely (1-3 patients/month)

72 (34.1)

Often (4-9 patients/month)

97 (46.0)

Frequently (>10 patients/month)

39 (18.5)

Don’t know *Some respondents completed more than 1 fellowship.

Western Journal of Emergency Medicine

3 (1.4)

274

Volume XVII, no. 3 : May 2016

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation

Table 1. Continued. Survey question

Respondents (%)

How often do mechanically ventilated patients in the emergency department (ED) have adjustments made to the ventilator while still in the ED? Never

0 (0)

Rarely (1-3 patients/month)

118 (55.9)

Often (4-9 patients/month)

68 (32.2)

Frequently (>10 patients/month)

14 (6.6)

Don’t know

11 (5.2)

How often do you feel comfortable managing mechanical ventilation and troubleshooting issues with ventilated patients in the ED? Never

3 (1.4)

Rarely

78 (37.0)

Often

96 (45.5)

Frequently

30 (14.2)

Don’t know

4 (1.9)

Who primarily directs changes to the mechanical ventilator for intubated patients in your ED? Respiratory therapist

145 (68.7)

Nurse

0 (0)

EM resident

15 (7.1)

EM attending

44 (20.9)

Physician not affiliated with the ED (ICU, pulmonologist, etc)

4 (1.9)

Don’t know ICU, intensive care unit

3 (1.4)

Table 2. Correlations between survey responses and total score on assessment tool. Characteristics of training program and experience Correlation with total test score (ρ)

P-value

Years as emergency medicine (EM) attending

-0.110

0.114

Training in another residency

-0.111

0.109

Fellowship training

0.023

0.743

Emphasis on mechanical ventilation in EM residency

0.260

<0.001*

Hours of mechanical ventilation education

0.093

0.182

Frequency of managing ventilated patients

0.105

0.129

Frequency of ventilator changes in the emergency department

0.055

0.431

Level of comfort with managing mechanically ventilated patients

0.356

<0.001*

EM management of mechanical ventilator

0.133

0.053

Completion EM residency

0.212

0.002*

0.152

0.027*

Program affiliation *Statistically significant correlation (2-tailed).

residency, program affiliation, residency program emphasis on mechanical ventilation management, working in a setting where an emergency physician bears primary responsibility for ventilator management, and level of comfort with managing ventilated patients, multivariable regression modelling determined that self-reported frequency of managing mechanically ventilated patients was not a significant predictor of total test score (t=0.163, p=0.871). Exploratory regression Volume XVII, no. 3 : May 2016

analyses revealed that the strongest and only significant predictor of total test score was attendings’ self-reported confidence in caring for mechanically ventilated patients (F=22.266, p<0.001). On average, test scores increased by approximately eight points (95% CI [4.5-10.9] points, p=0.001) when attendings reported feeling comfortable managing ventilated patients “often” or “frequently.” The addition of any other predictor variables, alone or in combination, did not 275

Western Journal of Emergency Medicine

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation

Table 3. Correlations between candidate predictor variables and attending physician comfort with managing mechanically ventilated patients. Characteristics of training program and experience Correlation with comfort (ρ) P-value Years as emergency medicine (EM) attending

-0.070

0.312

Training in another residency

0.092

0.185

Fellowship training

-0.066

0.342

Emphasis on mechanical ventilation in EM residency

0.261

<0.001*

Hours of mechanical ventilation education

0.250

<0.001*

Frequency of managing ventilated patients

0.134

0.052

Frequency of ventilator changes in the emergency department

0.147

0.033*

EM management of mechanical ventilator

0.258

<0.001*

Completion of EM residency

0.088

0.202

0.001

0.985

Program affiliation *Statistically significant correlation (2-tailed).

produce a more parsimonious model. Exploratory logistic regression modeling revealed that four variables were statistically significantly associated with attending physician comfort managing mechanically ventilated patients. Having completed an additional residency training program was most strongly associated with confidence, with an adjusted odds ratio (aOR) of 3.671 (p=0.013) for those completing an additional program as compared to those who did not. Next, working in a facility where an emergency physician bears primary responsibility for managing ventilator settings was associated with comfort, as attending physicians reporting having this role were more likely to report comfort than those who did not (aOR, 3.271, p=0.002). Working in a setting where the residency program “often” or “frequently” emphasizes ventilator management, as compared to settings that have less focus on the topic, was associated with an increased likelihood of reporting comfort (aOR, 1.732, p=0.002). Finally, attending physicians reporting four or more hours of curriculum-based ventilator education were more likely to report comfort than those reporting two or fewer hours (aOR 1.468, p=0.013). Other variables noted to be significantly correlated with physician comfort in simple correlation analysis (Table 3) did not produce significant improvements in the predictive power of the final model. DISCUSSION Emergency physicians increasingly care for critically ill, mechanically ventilated patients in the ED,30 and due to crowding the ED length of stay is increasing.15 Ventilator management decisions can directly affect patient outcomes, especially in asthma, ARDS, and traumatic brain injury,14-16 conditions commonly encountered in the ED. Although mechanical ventilation has been considered integral to EM practice by key EM organizations, including the American Board of Emergency Medicine, the hours required for training and required level of proficiency are not specified.31 Western Journal of Emergency Medicine

Our group has previously demonstrated that EM residents receive few hours of education on mechanical ventilation, yet report frequently caring for ventilated patients.5 Their self-reported comfort with caring for ventilated patients correlated with their score on the knowledge assessment instrument, and their post-graduate year, hours of residency education on mechanical ventilation, and frequency of caring for mechanically ventilated patients were associated with their comfort. To further evaluate EM physicians’ knowledge of mechanical ventilation, we subsequently assessed the attendings associated with the same residency programs previously queried. The knowledge assessment tool used in this study was designed to reflect educational objectives for management of mechanically ventilated patients, and tested knowledge in clinical scenarios commonly encountered by physicians in the ED.20,32 The instrument was rigorously designed, pre-tested, and pilot tested to optimize psychometric and performance characteristics, and is similar to a previously validated test.32 In this study, academic EM attendings report that curricular time dedicated to mechanical ventilation varied among their own EM residency training programs, but very few reported frequent emphasis on the topic. Additionally, current educational opportunities on ventilation are relatively minimal, as 75% of attendings responded that they had received three or fewer hours of education on mechanical ventilation in their in the past year. Although attendings reported few hours of education, 64% responded that they often or frequently care for intubated patients in the ED. Interestingly, while ventilated patients are common, 56% of respondents stated that patients rarely have any changes made to ventilator settings while in the ED. This may be factual, or may be a perception, as most attendings (69%) identified RTs as being primarily responsible for managing ventilators in the ED, and a minority of the attendings stated that management of the ventilator was the responsibility of an emergency physician. 276

Volume XVII, no. 3 : May 2016

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation

Overall, the attendings performed moderately well on the knowledge assessment, with a mean score of 73.4%±19. Notably, this is identical to the prior residents’ score of 73.3%±22.5 In the univariate analysis, emphasis on ventilation in the physicians’ EM residency correlated with test score. Additionally, level of comfort with managing mechanically ventilated patients, completion of an EM residency, and program affiliation also correlated with the overall score. EM management of the ventilator, as opposed to RTs, was not statistically significant (p=0.053). Multivariate analysis assessing for factors correlating with comfort in management of ventilators found that emphasis on mechanical ventilation in the attendings’ EM training, recent hours of mechanical ventilation education, frequency of ventilator changes in the ED, and EM management of mechanical ventilator all correlated with confidence in management of mechanically ventilated patients. However, frequency of managing ventilated patients did not reach statistical significance. These findings demonstrate that comfort with caring for ventilated patients may be an active process. Simply caring for ventilated patients in the ED without being a participant in decisions regarding ventilator management may not increase comfort, but EM management of the ventilator and increasing the frequency of ventilator changes in the ED does increase comfort. Prior work regarding mechanical ventilation in the ED is limited.33-35 Despite numerous studies showing improved outcomes with low tidal volume ventilation, prior surveys of ventilation in the ED have found that only 27.1%33 and 55.7%34 of patients received low tidal volumes, with one study finding a median tidal volume of 8.8 mL/kg of ideal body weight.34 Similar to our findings, a national survey of EDs from 2014 found that 73% of respondents reported that patients routinely received mechanical ventilation for several hours in the ED.35 Emergency physicians noted a lack of literature to guide mechanical ventilation specifically in the ED, but 100% of respondents were willing to adopt an intervention that could decrease the incidence of ARDS,35 such as low tidal volume ventilation. These prior findings demonstrate opportunities for improvement in mechanical ventilation in the ED, as well as a willingness to adopt such interventions on the part of emergency physicians. Throughout this current study, several parallels emerged between attendings and the previous study of EM residents. In addition to having the same score on the assessment instrument, a similar proportion of residents reported few hours of education with a high frequency of caring for ventilated patients. Also, 78% of residents felt that ventilator management was the responsibility of the RT. For the residents, level of comfort correlated with test score, much like the attendings. As with their attendings, the residents’ comfort increased with available hours of education on ventilation. These findings support the importance of education Volume XVII, no. 3 : May 2016

in residency, as increased hours of education improve the assessment score and increase comfort. Perhaps more importantly, education in residency appears to have lasting effects, as emphasis on mechanical ventilation in residency correlated with an improved attending performance on the assessment tool. Additionally, in our prior study, we found that residents had sufficient knowledge, but were not actively applying that knowledge by ceding opportunities to manage the ventilator to RTs. We hypothesized that there is an opportunity to improve emergency physicians’ familiarity and comfort with ventilators by encouraging more active involvement with ventilator management decisions in the ED. While emergency physicians performed well on the assessment instrument, there is potential benefit of active ventilator management and bedside education for both EM residents and attendings in improving knowledge and comfort with ventilated patients. This concept is supported by this study, as active management of the ventilator is associated with confidence in caring for these patients. LIMITATIONS Although our results are similar to a prior study of EM residents, our response rate in the current study was lower at 54%, with 211 respondents. Therefore, our results may have been influenced by non-responder bias and may be underpowered to detect significant differences. Physicians’ interest in the topic of mechanical ventilation may have influenced participation in this study, with interested physicians being more likely to complete the knowledge assessment tool and less-interested attendings being less likely to participate and complete the tool. A study of patient outcomes related to management of mechanical ventilation was beyond the scope of this limited study, and therefore, the impact on patients is not known. Finally, our multicenter study involved academic emergency physicians in the northeastern U.S., and the generalizability of our results to other settings, regions or countries is not known. This current study did not test the effects of an educational intervention on knowledge of mechanical ventilation. Assessing the value of education on ventilation for emergency physicians is an important future direction for study. CONCLUSION In this sample, we noted that attending physicians report caring for mechanically ventilated patients in the ED quite frequently; however, they also reporting having few educational opportunities regarding mechanical ventilation, with 75% stating that they received three hours or fewer over the last year. The majority of respondents identified a respiratory therapist as being primarily responsible for ventilator management, with few describing this role as belonging to an emergency physician. Performance on our mechanical ventilation knowledge

277

Western Journal of Emergency Medicine

Wilcox et al.

Academic Physicians’ Knowledge of Mechanical Ventilation assessment was moderate, with an average score of 73.4%. A higher score on the assessment portion correlated with prior emphasis on mechanical ventilation in the physician’s own residency, completion of an EM residency, and self-reported comfort in caring for ventilated patients. Physicians’ comfort was associated with the frequency of ventilator changes and EM management of ventilation, as well as hours of recent education. These findings suggest that education in residency may have lasting effects on future performance, and active participation in decisions regarding mechanical ventilation management, as well as educational opportunities, can increase confidence in caring for these critically ill ED patients.

ventilated patients. Intensive Care Med. 2005;31(7):922-6. 8. Hodder R, Lougheed MD, FitzGerald JM, et al. Management of acute asthma in adults in the emergency department: assisted ventilation. CMAJ. 2010;182(3):265-72. 9. Archambault PM and St-Onge M. Invasive and noninvasive ventilation in the emergency department. Emerg Med Clin North Am. 2012;30(2):421,49, ix. 10. Kilickaya O and Gajic O. Initial ventilator settings for critically ill patients. Crit Care. 2013;12;17(2):123. 11. Orebaugh SL. Initiation of mechanical ventilation in the emergency department. Am J Emerg Med. 1996;14(1):59-69. 12. Warner KJ, Cuschieri J, Copass MK, et al. Emergency department ventilation effects outcome in severe traumatic brain injury. J Trauma. 2008;64(2):341-7. 13. Shafi S and Gentilello L. Pre-hospital endotracheal intubation and

Address for Correspondence: Susan R. Wilcox, MD, Divisions of Emergency Medicine and Pulmonary, Critical Care and Sleep Medicine, Medical University of South Carolina, 96 Jonathan Lucas Street, Suite 812-CSB, MSC 630, Charleston, SC 29403. Email: wilcoxsu@musc.edu.

positive pressure ventilation is associated with hypotension and decreased survival in hypovolemic trauma patients: an analysis of the National Trauma Data Bank. J Trauma. 2005;59(5):1140,5; discussion 1145-7. 14. Herring A, Wilper A, Himmelstein DU, et al. Increasing length of stay

among adult visits to U.S. Emergency departments, 2001-2005. Acad Emerg Med. 2009;16(7):609-16. 15. Rose L, Gray S, Burns K, et al. Emergency department length of stay for patients requiring mechanical ventilation: a prospective observational study. Scand J Trauma Resusc Emerg Med.

Copyright: © 2016 Wilcox et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

2012;20:30,7241-20-30. 16. Lambe S, Washington DL, Fink A, et al. Trends in the use and capacity of California’s emergency departments, 1990-1999. Ann Emerg Med. 2002;39(4):389-96. 17. Hoegl S, Boost KA, Flondor M, et al. Short-term exposure to highpressure ventilation leads to pulmonary biotrauma and systemic inflammation in the rat. Int J Mol Med. 2008;21(4):513-9.

REFERENCES

18. Báez AA, Hanudel P, Perez MT, et al. Prehospital Sepsis Project

1. Couvillon L, Brandhurst R, DeBlieux P. Emegency Medicine Resident

(PSP): Knowledge and Attitudes of United States Advanced Out-of-

Training in Mechanical Ventilation. Ann Emerg Med. 1999;34(4):S2. 2. Dent AW, Weiland TJ, Paltridge D. Australasian emergency

Hospital Care Providers. Prehosp Disaster Med. 2013;28(2):104-6. 19. Veronese C, Richards JB, Pernar L, et al. A randomized pilot study of

physicians: a learning and educational needs analysis. Part Four:

the use of concept maps to enhance problem-based learning among

CPD topics desired by emergency physicians. Emerg Med Australas. 2008;20(3):260-6.

first-year medical students. Med Teach. 2013;35(9):e1478-84. 20. Goligher EC, Ferguson ND, Kenny LP. Core competency in

3. Varon J, Fromm RE,Jr, Levine RL. Emergency department

mechanical ventilation: development of educational objectives using

procedures and length of stay for critically ill medical patients. Ann Emerg Med. 1994;23(3):546-9.

the Delphi technique. Crit Care Med. 2012;40(10):2828-32. 21. Brenner B, Corbridge T, Kazzi A. Intubation and mechanical

4. Svenson J, Besinger B, Stapczynski JS. Critical care of medical and

ventilation of the asthmatic patient in respiratory failure. J Emerg

surgical patients in the ED: length of stay and initiation of intensive care procedures. Am J Emerg Med. 1997;15(7):654-7.

Med. 2009;37(2 Suppl):S23-34. 22. Manthous CA. Avoiding circulatory complications during endotracheal

5. Wilcox SR, Seigel TA, Strout TD, et al. Emergency medicine

intubation and initiation of positive pressure ventilation. J Emerg Med.

residents’ knowledge of mechanical ventilation. J Emerg Med. 2015;48(4):481-91.

2010;38(5):622-31. 23. Ventilation with lower tidal volumes as compared with traditional

6. Wood S, Winters ME. Care of the intubated emergency department

tidal volumes for acute lung injury and the acute respiratory distress

patient. J Emerg Med. 2011;40(4):419-27.

syndrome. The Acute Respiratory Distress Syndrome Network. N

7. Gajic O, Frutos-Vivar F, Esteban A, et al. Ventilator settings as a risk factor for acute respiratory distress syndrome in mechanically

Western Journal of Emergency Medicine

Engl J Med. 2000;342(18):1301-8. 24. Davis DP, Dunford JV, Poste JC, et al. The impact of hypoxia

278

Volume XVII, no. 3 : May 2016

Wilcox et al. and hyperventilation on outcome after paramedic rapid

Academic Physiciansâ&#x20AC;&#x2122; Knowledge of Mechanical Ventilation 30. Rose L and Gerdtz MF. Use of invasive mechanical ventilation

sequence intubation of severely head-injured patients. J Trauma.

in Australian emergency departments. Emerg Med Australas.

2004;57(1):1,8;discussion 8-10. 25. Davis DP, Idris AH, Sise MJ, et al. Early ventilation and outcome in

2009;21(2):108-16. 31. Hockberger RS, Binder LS, Chisholm CD, et al. The model of the

patients with moderate to severe traumatic brain injury. Crit Care

clinical practice of emergency medicine: a 2-year update. Ann Emerg

Med. 2006;34(4):1202-8. 26. Davis DP, Meade W, Sise MJ, et al. Both hypoxemia and extreme

Med. 2005;45(6):659-74. 32. Cox CE, Carson SS, Ely EW, et al. Effectiveness of medical resident

hyperoxemia may be detrimental in patients with severe traumatic

education in mechanical ventilation. Am J Respir Crit Care Med.

brain injury. J Neurotrauma. 2009;26(12):2217-23. 27. Dumont TM, Visioni AJ, Rughani AI, et al. Inappropriate prehospital

2003;167(1):32-8. 33. Fuller BM, Mohr NM, Dettmer M, et al. Mechanical ventilation and

ventilation in severe traumatic brain injury increases in-hospital

acute lung injury in emergency department patients with severe

mortality. J Neurotrauma. 2010;27(7):1233-41.

sepsis and septic shock: an observational study. Acad Emerg Med.

28. von Elm E, Schoettker P, Henzi I, et al. Pre-hospital tracheal intubation in patients with traumatic brain injury: systematic review of

2013;20:659-69. 34. Fuller BM, Mohr NM, Miller CN, et al. Mechanical ventilation and

current evidence. Br J Anaesth. 2009;103(3):371-86.

ARDS in the ED: A multicenter, observational, prospective, cross-

29. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap) - A metadata-driven methodology and workflow

sectional study. Chest. 2015;148(2):365-74. 35. Fuller BM and Mohr NM. Limiting acute respiratory distress syndrome

process for providing translational research informatics support. J

in the emergency department: a survey of US academic emergency

Biomed Inform. 2009;42(2):377-81.

medicine physicians. Eur J Emerg Med. 2014;21(5):387-8.

Volume XVII, no. 3 : May 2016

279

Western Journal of Emergency Medicine

Case Report

Anti-N-Methyl-D-Aspartate Receptor Encephalitis, an Underappreciated Disease in the Emergency Department Daniel R. Lasoff, MD*† Jimmy Corbett-Detig, MD‡ Rebecca Sell, MD§ Matthew Nolan, MD‡ Gabriel Wardi MD, MPH§

*VA Medical Center, San Diego, California † UCSD Medical Center, Department of Emergency Medicine, Division of Medical Toxicology, San Diego, California ‡ UCSD Medical Center, Department of Emergency Medicine, San Diego, California § UCSD Medical Center, Department of Internal Medicine, Division of Pulmonary, Critical Care, and Sleep Medicine, San Diego, California

Section Editor: Rick A. McPheeters, DO Submission history: Submitted December 16, 2015; Revision received February 10, 2016; Accepted February 22, 2016 Electronically published May 2, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29554

Anti-N-Methyl-D-Aspartate Receptor (NMDAR) Encephalitis is a novel disease discovered within the past 10 years. Antibodies directed at the NMDAR cause the patient to develop a characteristic syndrome of neuropsychiatric symptoms. Patients go on to develop autonomic dysregulation and often have prolonged hospitalizations and intensive care unit stays. There is little literature in the emergency medicine community regarding this disease process, so we report on a case we encountered in our emergency department to help raise awareness of this disease process. [West J Emerg Med. 2016;17(3):280–282.]

CASE REPORT A 23-year-old man presented to the emergency department (ED) after a witnessed tonic-clonic seizure. He was previously healthy with no prior seizures and had no recent trauma, fevers, vomiting, or history of substance abuse except for marijuana. His family reported over the preceding weeks he had been trying to lose weight and recently had started using three weight-loss supplements: Erratic, Thermovex, and Prozein. A review of these supplements revealed they were a mixture of various amino acids, proteins, vitamins, and caffeine. He had reportedly been agitated and increasingly manic over the previous few days, and co-workers stated that he had seemed confused at work earlier in the day. A review of systems was otherwise negative. On physical examination, the patient was a young, athletic male who was somnolent but arousable. His temperature was 37.1°C, blood pressure 132/71, heart rate 62bpm, and respiratory rate 16 breaths per minute. Pupils were 4mm, equal, round, and reactive. His face was symmetric and tongue was midline on protrusion. He had 5/5 strength in both upper and lower extremities, and sensation was intact throughout to fine touch. Patellar and ankle reflexes were 2+, symmetric and without clonus. Laboratory analysis was remarkable for a glucose level of 232mg/dL, a urine drug screen that was positive for THC and benzodiazepines Western Journal of Emergency Medicine

(the latter of which had been given by the field medics and the ED for seizures). Following an unremarkable computed tomography (CT) of his brain, he had a lumbar puncture, which showed 370 white blood cells/mm3, 300 red blood cells/ mm3, and a protein level of 147mg/dL. The gram stain of his cerebrospinal fluid (CSF) was negative. Empiric ceftriaxone, vancomycin, and acyclovir were started for presumed infectious meningitis. While in the ED, the patient had an additional tonic-clonic seizure and was intubated for airway protection. The patient was then admitted to the intensive care unit (ICU) for further care. During his course in the ICU he failed to improve, remained intermittently agitated and was unable to be extubated. An electroencephalogram (EEG) shortly after admission revealed ongoing epileptiform activity, and he received aggressive anti-seizure therapy. Consultations from infectious disease and rheumatology services were unable to provide a diagnosis. Autoimmune panels and several viral, fungal, and bacterial assays were all negative. On hospital day 15, the patient’s CSF was sent for an anti-NMethyl-D-Aspartate receptor (NMDAR) antibody assay, and the test returned two days later with a titer of 1:5120 (normal <1:10) consistent with anti-NMDAR encephalitis. The patient was started on intravenous immunoglobulin (IVIG) for treatment, followed by plasmapheresis, cyclophosphamide,

280

Volume XVII, no. 3 : May 2016

Lasoff et al.

Anti-N-Methyl-D-Aspartate Receptor Encephalitis

and eventually rituximab. His course has been complicated by episodes of autonomic instability, delirium, and hospitalassociated infections. He remains in the ICU six months after admission for management of severe autonomic instability and remains dependent upon a tracheostomy and gastrostomy tube. DISCUSSION Anti-NMDAR encephalitis is an autoimmune encephalitis syndrome that is underappreciated and frequently missed in the ED due to lack of awareness. It was initially described in 2007 by Dalmau et al., and to our knowledge, it has not received any attention in the US emergency medicine literature to date. It is one of a growing family of neuronal surface antibody syndromes (NSAS) with auto-antibodies directed against the NR1 subunit of the NMDA-receptor.1,2 As awareness grows, it seems that anti-NMDAR encephalitis is likely to be four times more common than HSV encephalitis.3 Our goal is to increase awareness of anti-NMDAR encephalitis as many patients initially present to the ED with classic histories for this condition, but the diagnosis is not considered until much later in the clinical course. The diagnosis is often difficult to make due to the nonspecific nature of symptoms.4 Anti-NMDAR encephalitis classically presents with a prodromal syndrome of malaise, headache, and fever followed by psychiatric symptoms such as irritability, agitation, hallucinations, memory loss, mania, or frank psychosis.5-8 Neurological symptoms such as aphasia, seizures, dyskinesias, catatonia, or coma distinguish this syndrome from a pure psychiatric illness.2,5-7 Patients frequently develop autonomic dysregulation as well, which can manifest as tachycardia, hyperthermia, hypothermia, blood pressure abnormalities, or hypoventilation which frequently necessitates mechanical ventilation.2,5-7 Due to the nonspecific presenting symptoms, patients will often undergo lengthy workups, repeated imaging and blood work, and several consultations from specialists, without a clear diagnosis. Symptoms are frequently attributed incorrectly to a toxicological or psychiatric cause.5,9 In Dalmau et al.â&#x20AC;&#x2122;s 2008 case series of 100 patients, 77 were initially seen by a psychiatrist.5 Due to the often-varied presenting symptoms, diagnoses on average were delayed 21 days in children and 28 days in adults from the time of symptom onset.10 This is quite concerning as patient outcome seems to worsen when treatment is delayed.10 Anti-NMDAR encephalitis is more prevalent in women.3,11 Patients tend to be young with ages ranging from 2-40.3,4 The autoantibodies can be associated with paraneoplastic syndromes in 20-59% of cases, most commonly ovarian teratomas.2,5-7,10 Males and young children are less likely to have an underlying tumor that is responsible for their encephalitis.5,11 In cases associated with a tumor, patients may improve with removal of the tumor.13 Patients with undifferentiated encephalitis typically Volume XVII, no. 3 : May 2016

undergo evaluation with neuroimaging, lumbar puncture, and EEG. Patients with anti-NMDAR encephalitis may show nonspecific abnormalities on MRI, although the majority are normal.5,7 Almost all patients will have nonspecific EEG changes such as delta waves, theta waves, or slowing, and about half of patients may show epileptiform activity.5 CSF analysis is also non-specific, but common findings include lymphocytic pleocytosis, increased protein, and increased opening pressures; however, a significant number of patients present without any of these findings.7,14 The hallmark of the disease is the presence of anti-NMDAR antibodies that can be found in the both the serum and the CSF. The CSF appears to be more sensitive, as one study showed all 43 patients were positive for antibodies in the CSF, but only 27 patients tested positive for antibodies in their serum.14 First-line therapy for anti-NMDAR encephalitis includes steroids, IVIG, and plasma exchange.10,11 Patients who do not respond to first-line therapy have been treated with immunomodulators such as cyclophosphamide and rituximab.10,15 If there is evidence of ovarian teratomas, operative removal may be beneficial, and complete cessation of symptoms has been reported.10,16 While the therapy for antiNMDAR encephalitis will almost certainly not be started in the ED, the consideration of and appropriate diagnostic testing for this condition will greatly aid these patients. Unfortunately, many patients with anti-NMDAR encephalitis tend to undergo prolonged hospitalizations and require lengthy ICU stays. The mortality seems to be lower than initially thought and is estimated at approximately 10%; the majority of patients make a meaningful neurologic recovery. 1,10 It also appears that early treatment results in a better neurologic outcome for patients.18 CONCLUSION Anti-NMDAR encephalitis carries a significant morbidity and mortality that is worsened by delays in diagnosis; it is underappreciated and unrecognized in the ED This case illustrates some common features of anti-NMDAR encephalitis: a patient with recent psychiatric symptoms who presents with a neurological complaint, a delay in diagnosis and a lengthy course of treatment. Although our patient fits the usual age range and had common presenting symptoms of behavioral changes and seizures, anti-NMDAR encephalitis is more common in women. Emergency physicians who encounter patients with new onset neurologic complaints preceded by psychiatric symptoms should consider antiNMDAR encephalitis in the differential diagnosis as it may promote earlier treatment and improve outcomes.

Address for Correspondence: Daniel Lasoff, MD, VA Medical Center, San Diego, California and UCSD Medical Center, Department of Emergency Medicine, Division of Medical Toxicology, 200 W Arbor Dr #8676, San Diego, CA 92103. Email: dlasoff@ucsd.edu.

281

Western Journal of Emergency Medicine

Anti-N-Methyl-D-Aspartate Receptor Encephalitis

Lasoff et al.

receptor (NMDAR) encephalitis in children and adolescents. Ann Neurol. 2009;66(1):11-8. 8. Hughes EG, Peng X, Gleichman AJ, et al. Cellular and synaptic mechanisms of anti-NMDA receptor encephalitis. J Neurosci. 2010;30(17):5866-75.

Copyright: © 2016 Lasoff et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

9. Punja M, Pomerleau AC, Devlin JJ, et al. Anti-N-methyl-Daspartate receptor (anti-NMDAR) encephalitis: an etiology worth considering in the differential diagnosis of delirium. Clin Toxicol (Phila). 2013;51(8):794-7. 10. Titulaer MJ, McCracken L, Gabilondo I, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA

REFERENCES

receptor encephalitis: an observational cohort study. Lancet Neurol. 2013;12(2):157-65.

1. Zuliani L, Graus F, Giometto B, et al. Central nervous system neuronal surface antibody associated syndromes: review

11. Ramanathan S, Mohammad SS, Brilot F, et al. Autoimmune encephalitis: Recent updates and emerging challenges. J Clin

and guidlines for recognition. J Neurol Neurosurg Psychiatry.

Neurosci. 2014;21(5):722-30.

2012;83(6):638-45. 2. Dalmau J, Gleichman AJ, Hughes EG, et al. Anti-NMDA-receptor

12. Dalmau J, Lancaster E, Martinez-Hernandez E, et al. Clinical experience and laboratory investigations in patients with anti-NMDAR

encephalitis: case series and analysis of the effects of antibodies.

encephalitis. Lancet Neurol. 2011;10(1):63-74.

Lancet Neurol. 2008;7(12):1091-8. 3. Gable MS, Sheriff H, Dalmau J, et al. The frequency of

13. Seki M, Suzuki S, Iizuka T, et al. Neurological response to early removal of ovarian teratoma in anti-NMDAR encephalitis. J Neurol

autoimmune N-Methyl-D-Aspartate receptor encephalitis

Neurosurg Psychiatry. 2008;79(3):324-6.

surpasses that of individual viral etiologies in young individuals enrolled in the California Encephalitis Project. Clin Infect Dis.

14. Wang R, Guan HZ, Ren HT, et al. CSF findings in patients with anti-Nmethyl-D-aspartate receptor-encephalitis. Seizure. 2015;29:137-42.

2012;54(7):899-904. 4. Barry H, Byrne S, Barrett E, et al. Anti-N-methyl-D-aspartate receptor

15. Liba Z, Sebronova V, Komarek V, et al. Prevalence and treatment of anti-NMDA receptor encephalitis. Lancet Neurol. 2013;12(5):424-5.

encephalitis: review of clinical presentation, diagnosis and treatment. BJPsych Bull. 2015;39(1):19-23.

16. Iizuka T, Sakai F, Ide T, et al. Anti-NMDA receptor encephalitis in Japan Long-term outcome wtihout tumor removal. Neurology.

5. Irani SR, Bera K, Waters P, et al. N-methyl-D-aspartate antibody

2008:70(7):504-11.

encephalitis: temporal progression of clinical and paraclinical observations in a predominantly non-paraneoplastic disorder of both

17. Peery HE, Day GS, Dunn S, et al. Anti-NMDA receptor encephalitis. The disorder, the diagnosis, and the immunobiology. Autoimmun Rev.

sexes. Brain. 2010;133(Pt6):1655-67.

2012;11(12):863-72.

6. Dalmau J, Tuzun E, Wu HY, et al. Paraneoplastic anti-N-methyl-Daspartate receptor encephalitis associated with ovarian teratoma.

18. Finke C, Kopp UA, Pruss H, et al. Cognitive Deficits following anti-NMDA receptor Encephalitis. J Neurol Neurosurg Psychiatry.

Ann Neurol. 2007;61(1):25-36.

2012;83(2):195-8.

7. Florance NR, Davis BA, Dalmau J, et al. Anti-N-Methyl-D-Asparate

Western Journal of Emergency Medicine

282

Volume XVII, no. 3 : May 2016

Original Research

ACE-I Angioedema: Accurate Clinical Diagnosis May Prevent Epinephrine-Induced Harm R. Mason Curtis, MD*† Sarah Felder, MD‡ Rozita Borici-Mazi, MD‡ Ian Ball, MD†§¶

* Western University, Division of Emergency Medicine, Department of Medicine, London, Ontario, Canada † Queen’s University, Departments of Emergency Medicine and Biomedical and Molecular Sciences, Kingston, Ontario, Canada ‡ Queen’s University, Division of Allergy and Immunology, Kingston, Ontario, Canada § Queen’s University, Program in Critical Care Medicine, Kingston, Ontario, Canada ¶ Western University, Division of Critical Care of Medicine, Department of Medicine, London, Ontario, Canada

Section Editor: Kenneth S. Whitlow, DO Submission history: Submitted November 11, 2015; Revision received January 17, 2016; Accepted February 8, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29224

Introduction: Upper airway angioedema is a life-threatening emergency department (ED) presentation with increasing incidence. Angiotensin-converting enzyme inhibitor induced angioedema (AAE) is a non-mast cell mediated etiology of angioedema. Accurate diagnosis by clinical examination can optimize patient management and reduce morbidity from inappropriate treatment with epinephrine. The aim of this study is to describe the incidence of angioedema subtypes and the management of AAE. We evaluate the appropriateness of treatments and highlight preventable iatrogenic morbidity. Methods: We conducted a retrospective chart review of consecutive angioedema patients presenting to two tertiary care EDs between July 2007 and March 2012. Results: Of 1,702 medical records screened, 527 were included. The cause of angioedema was identified in 48.8% (n=257) of cases. The most common identifiable etiology was AAE (33.1%, n=85), with a 60.0% male predominance. The most common AAE management strategies included diphenhydramine (63.5%, n=54), corticosteroids (50.6%, n=43) and ranitidine (31.8%, n=27). Epinephrine was administered in 21.2% (n=18) of AAE patients, five of whom received repeated doses. Four AAE patients required admission (4.7%) and one required endotracheal intubation. Epinephrine induced morbidity in two patients, causing myocardial ischemia or dysrhythmia shortly after administration. Conclusion: AAE is the most common identifiable etiology of angioedema and can be accurately diagnosed by physical examination. It is easily confused with anaphylaxis and mismanaged with antihistamines, corticosteroids and epinephrine. There is little physiologic rationale for epinephrine use in AAE and much risk. Improved clinical differentiation of mast cell and non-mast cell mediated angioedema can optimize patient management. [West J Emerg Med. 2016;17(3):283–289.]

INTRODUCTION Background Angioedema, defined as a transient, localized nonpitting edema of skin or mucous membranes is a potentially Volume XVII, no. 3 : May 2016

life-threatening emergency department (ED) presentation.1,2 Prompt identification of the cause of angioedema is essential, as it will guide management strategies and prevent the unnecessary and potentially harmful use of ineffective 283

Western Journal of Emergency Medicine

Accurate ACE-I Angioedema Diagnosis

Curtis et al.

interventions. Etiologies may be broadly classified into mast cell mediated and non-mast cell mediated processes, the latter of which are thought to be mediated through accumulation of bradykinin.3 Bradykinin increases vascular permeability leading to plasma leakage into the dermis and resultant angioedema. This condition may be mediated by a hereditary or acquired etiology, the former of which is due to a deficiency in C-1 esterase inhibitor and presents with isolated angioedema without urticaria. Contrasting this, histamine release from mast cell mediated etiologies, such as an allergic response, leads to increased vascular permeability of more superficial layers of the dermis resulting in urticaria, and less commonly involves deeper dermal structures leading to angioedema. Angiotensin-converting enzyme inhibitor (ACE-I) medications are reported to be responsible for approximately 30% of patients presenting to tertiary care and community EDs with angioedema.4,5 ACE-I induced angioedema (AAE), an example of an acquired form of non-mast cell mediated etiology of angioedema, is estimated to affect approximately 1 in 200 individuals taking an ACE-I and typically causes swelling of the lips, face and tongue, which has led to asphyxiation and death in severe cases.6,7 Although angioedema is considered a rare side effect of ACE-Is, the widespread use of this medication class, over 35-40 million individuals worldwide, means that a large number of individuals may be affected by this potentially life-threatening adverse effect.8 Importance ACE-Is are widely used for the management of hypertension, heart failure, remodeling after myocardial infarction, and for the prevention of diabetic nephropathy, cardiovascular events and secondary stroke. The incidence of AAE will continue to grow as the prevalence of cardiac disease and our average population age continue to rise.9 There are limited data on the optimal method for treating AAE, and the evidence and physiologic rationale supporting strategies such as epinephrine, corticosteroids and antihistamines is poor. Unnecessary exposure to these medications may be harmful, in particular epinephrine, as many of these patients are elderly and have ischemic heart disease. A greater awareness of appropriate therapeutic approaches to treat angioedema is required.10 Goals of This Investigation We sought to describe the presentation and management of angioedema in the ED, and highlight any epinephrineinduced harm. We highlight the need to differentiate mast cell and non-mast cell mediated etiologies of angioedema in order to optimize patient care and reduce iatrogenic morbidity. METHODS Study Population We performed a retrospective chart review of all patients Western Journal of Emergency Medicine

with angioedema presenting to two tertiary North American EDs between July 2007 and May 2012. Every patient visit with an International Classification of Disease, 10th Revision discharge diagnosis of T782 (anaphylactic shock, unspecified), T783 (angioneurotic edema), T784 (allergy unspecified), D841 (defects in the complement system), T886 (anaphylactic shock due to adverse effect of correct drug or medicament) or T887 (unspecified adverse effect of drug or medicament) was eligible for review. Patient visits were excluded from the study if visible swelling was not documented on the medical record, if the medical evaluation was incomplete secondary to patient leaving against medical advice, or if the swelling was from a nonsystemic reaction to an insect sting, trauma or irritant exposure. Methodology Data abstraction was performed using a structured data extraction form. Training with the data extraction form was conducted on the first 20 medical records reviewed, which were then re-assessed during the formal review. Due to the binary nature of the abstracted data, such as the presence or absence of a documented clinical sign or symptom, or extraction of objective measures such as vitals, double data entry and interrater reliability were not assessed. This project was approved by the institutional review board at our academic center. Etiology We assessed the most responsible etiology for angioedema based on the following a priori study definitions: 1. Food allergy: consumption of a known allergen in the two hours before presentation 2. Drug allergy: any change in drug regimen in the past 24 hours, except ACE-I or angiotensin receptor blocker (ARB) 3. Environmental allergy: exposure to a known or likely allergen (seasonal allergy, pets etc.) 4. Stinging insect: presentation of systemic response to known insect sting 5. ACE-I/ ARB induced: patients taking an ACE-I or ARB with no other known cause of angioedema. 6. Nonsteroidal anti-inflammatory drug (NSAID) or acetylsalicylic acid (ASA) induced: patients taking chronic NSAIDs or ASA with no other known cause of angioedema, who are not taking an ACE-I or ARB 7. C1-Esterase Inhibitor Deficiency: patients with known C1-esterase deficiency 8. Contact allergy: patients who present with a systemic reaction to contact with an allergen (i.e. cosmetic products, latex etc.) 9. Etiology unknown: patients whose presentation is not explained by any of the above definitions These definitions have been previously used in similar studies.4,11 Patients using an ACE-I or an ARB were both defined as having AAE, as the mechanism behind ARBinduced angioedema has yet to be fully elucidated. This practice is consistent with previous studies.12,13

284

Volume XVII, no. 3 : May 2016

Curtis et al.

Accurate ACE-I Angioedema Diagnosis

RESULTS Study Population Figure 1 describes the study population, representing a continuous sample of all patients presenting to the ED with angioedema during the five-year study period. Patient demographics and clinical variables are shown in Table. Figure 2 displays the likely presenting etiology, based on the a priori definitions. Notably, approximately 51% (n=270) had an unknown etiology, and the most common identifiable etiology was AAE (n=85, 33% of identifiable causes). Of the 85 AAE patients, ramipril was the most common ACE-I being used at the time of ED presentation (n=45, 53%). Seven of these patients were on an ARB at the time of ED presentation (8.2%). Figure 3 depicts the management strategies employed for patients presenting to the ED with AAE. (This figure does not include other causes of angioedema.) Antihistamines and corticosteroids were the most common strategies employed for AAE management. Of particular interest, epinephrine was administered through the subcutaneous or intramuscular route in 21% (n=18) of patients, five of whom received repeated doses.

Records identified for n=1702

Lack of ED record (ie. left while awaiting physician assessment) n=65

Secondary to local response (ie. trauma, non-systemic insect sting) n=176

Lack of swelling on physical exam n=934

Underwent full review n = 527

Figure 1. Flow chart of study population. ED, emergency department

Epinephrine-Induced Morbidity Epinephrine caused harm in two of 18 AAE patients who received it as management for AAE. In both cases, patients presented with isolated lingual edema with no signs of a mast cell mediated process (urticaria/pruritus, respiratory compromise, hypotension, nausea/vomiting). Both patients had been started on ramipril for cardioprotection following myocardial infarctions and were therefore particularly vulnerable to complications of epinephrine administration. One patient developed angina five minutes after epinephrine administration. The second patient developed runs of premature ventricular contractions after epinephrine administration, shown in Figure 4. There was no immediate improvement of the lingual edema after intramuscular epinephrine in either case. Angioedema symptoms slowly subsided over approximately five hours in both cases. DISCUSSION We report that AAE is the most common identifiable cause of angioedema. Concerningly, our study found epinephrine, corticosteroids and diphenhydramine to be commonly used for the treatment of AAE. Of the 18 AAE patients who received epinephrine (21% of all AAE patients), five received multiple doses. This finding is not a local phenomenon, as other studies have identified that approximately 10-33% of AAE patients are being managed with epinephrine, some of whom receive repeated administrations.4,11,12 Cleary, epinephrine is still being used to manage AAE, despite its recognition as a non-mast cell mediated process. In our study, two of the 18 AAE patients (11%) given epinephrine developed morbidity following intramuscular epinephrine administration. Given the high prevalence of Volume XVII, no. 3 : May 2016

Figure 2. Graphical representation of angioedema etiologies in the emergency department. ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; NSAID, non-steroidal anti-inflammatory drug; ASA, acetylsalicylic acid

epinephrine use in other studies for management of AAE, the two cases are unlikely to be unique.4,11,12 These patients may be some of the most vulnerable to complications from epinephrine administration, given their cardiac co-morbidities. As illustrated in the cases presented, the administration of epinephrine in patients with cardiac histories may cause ischemia, dysrhythmias, or complications thereof.

285

Western Journal of Emergency Medicine

Accurate ACE-I Angioedema Diagnosis

Curtis et al.

Table. Demographics, clinical variables and pre-emergency department (ED) management of patients presenting to the ED with angioedema. AAE (n=85) Mean age (SD) Male sex, %

Non-AAE (n=442)

65 (13)

33 (21)

Peri-orbital

2.4

Lips

Tongue

Cheek/face

Pharynx

Glottis

1.2

Extremities

3.5

Genitalia

1.2

1.1

Multiple sites

<1 hour

7.1

1-6 hours

6-24 hours

>24 hours

2.4

Swelling location, %

Time from onset (%)

Pre-ED treatment (%) Antihistamine

Self Epi-PenÂŽ

4.3

3.5

3.4

EMS epinephrine EMS H1-blocker Mean ED stay in minutes (SD)

5.9 238 (192)

Figure 3. Management strategies employed in the emergency department for treatment of angiotensin-converting enzyme inhibitor induced angioedema (n=85). IM, intramuscular; SC, subcutaneous

2.5 198 (152)

Admission (%)

4.7

4.0

Intubation (%)

1.2

0.7

AAE, Angiotensin-converting enzyme inhibitor induced angioedema; EMS, emergency medical services

In the context of mast cell mediated angioedema, prompt administration of epinephrine is appropriate. It is well recognized that early administration of epinephrine for this form of angioedema, related to anaphylaxis, is the treatment of choice.14,15 The benefit is attributed to improved peripheral vascular tone, bronchodilation, reduction in mast cell mediated histamine release and positive inotropic and chronotropic effects on the myocardium. In the acutely ill patient, when differentiation between a mast cell and non mast cell mediated etiology is difficult, we feel that clinicians should err on the side of giving epinephrine. In clear cases of AAE, however, clinicians should not administer epinephrine out of a need to do something. Removal of the offending medication, supportive care, and airway management are the appropriate approach in these scenarios. Western Journal of Emergency Medicine

Figure 4. Rhythm strip of patient who developed ventricular bigeminy following epinephrine administration for treatment of angiotensin-converting enzyme inhibitor induced angioedema. Baseline rhythm shown in top line, post epinephrine dysrhythmia shown in middle and bottom line.

To aid clinicians in the diagnosis of the systemic mastcell mediated response that occurs in anaphylaxis, clinical guidelines have been established to guide physicians in its recognition and management.16 Sampson and colleagues have produced well known and widely accepted criteria for the diagnosis of anaphylaxis. These guidelines are based on clinical signs and stratified by potential exposures; if any of 286

Volume XVII, no. 3 : May 2016

Curtis et al.

Accurate ACE-I Angioedema Diagnosis

the following three criteria are met, anaphylaxis is likely and epinephrine should be administered promptly: 1) Acute onset of illness (minutes to hours) with dermal or mucosal involvement without known exposure, and either respiratory compromise, or reduced blood pressure 2) Exposure of a likely allergen with acute onset symptoms or signs in two organ systems (skin/mucosal tissue, respiratory compromise, reduced blood pressure, persistent gastrointestinal systems) 3) Exposure to a known allergen with reduced blood pressure (<90 mmHg systolic in adults, or >30% decrease in age specific systolic blood pressure). When a patientâ&#x20AC;&#x2122;s presentation falls outside of these clinical criteria, it does not exclude anaphylaxis as a cause. However, in patients presenting with isolated lingual angioedema concurrently taking an ACE-I who do not meet these guidelines and have no urticaria, it is likely the etiology of their presentation is not driven by a mast-cell mediated mechanism. Epinephrine administration for angioedema management outside the realm of anaphylaxis may cause significant morbidity and mortality. Previous case reports of the administration of epinephrine for non-anaphylaxis angioedema describe both cardiac dysrhythmias and ischemia, similar to what was observed in our study.17,18 Even when epinephrine is indicated in the setting of anaphylaxis, in a review of 166 admissions for management of anaphylaxis, Kanwar et al. found errors in dose and route of administration in 2.4% of patients, leading to coronary artery dissection, cardiogenic shock, coronary vasospasm and ventricular dysrhythmias requiring admission to critical care.19 This case series underscores the inherent risk of administering epinephrine and potentially lethal adverse outcomes that may arise. Given this risk of administration, combined with the unproven benefit of epinephrine in the setting of non-mast cell mediated angioedema, such as AAE, clinicians should not consider this an appropriate therapy for non-anaphylaxis angioedema. Avoiding the use of epinephrine and corticosteroids in the setting of hereditary and acquired angioedema is supported by consensus guidelines.20-22 How Should We Manage AAE? A recent review by Jaiganesh and colleagues highlights the need to differentiate mast cell and non-mast cell etiologies.10 They highlight that a focused assessment for the presence or absence of urticaria may aid clinicians in determining the etiology and appropriate therapeutic interventions. This distinction is key, as many mast cell mediated processes will present with urticaria secondary to histamine release. In contrast, the bradykinin mediated forms of angioedema, such as AAE and hereditary angioedema, are not associated with histamine release and urticaria; therefore, interventions such as antihistamines, corticosteroids and epinephrine offer limited therapeutic value based on underlying physiology.6 Volume XVII, no. 3 : May 2016

When a patient presents with angioedema, a prompt airway assessment is essential. Jaiganesh and colleagues describe a logical and effective step-wise approach to the patient with acute angioedema.10 Beyond airway management, treatment of AAE consists of supportive care (including potential endotracheal intubation) and discontinuation of the offending ACE-I. While seemingly obvious, discontinuation is necessary, for the recurrence of AAE is 10 times more likely when ACE-I are not discontinued.23 It is important to note that the duration of time between initiation of ACE-I therapy and onset of angioedema may be quite variable, reported to range from one day, up to eight years of therapy, with a median of six months.24 In the current study, no AAE patients received fresh frozen plasma or the bradykinin receptor antagonist icatibant, both of which have demonstrated potential benefit in small case series.25,26 In theory, the use of fresh frozen plasma would replenish angiotensin converting enzyme, and begin the breakdown of accumulated bradykinin; however, no prospective randomized trials have evaluated this and physicians must be prepared to deal with paradoxical exacerbation of symptoms.21 Contrasting this, icatibant is a bradykinin-receptor antagonist that has been used with good effect in the hereditary angioedema population, and recently been proposed for acquired cases of angioedema (such as in AAE). A recent randomized controlled trial of icatibant, for management of AAE demonstrated reduced time to symptom relief. 27 However, the small sample size, elevated cost of treatment and lack of impact on prognostic outcomes (need for intubation, need for surgical airway, prevention of admission), limit the generalizability of these results. A thorough review of these potential treatment modalities is nicely reviewed by Jaiganesh, and is found in many recent consensus guidelines and position statements, for the interested reader.10, 20-22 LIMITATIONS The retrospective nature of this investigation limits our data to the existing medical record. â&#x20AC;&#x153;Unknown Etiologyâ&#x20AC;? was grouped with all other causes of angioedema not due to an ACE-I. We did not perform double data entry and kappa analysis during data extraction. However, the binary nature (presence/absence) of extracted clinical variables limits the potential for bias. CONCLUSION This descriptive analysis of angioedema in the ED identified AAE as the most common identifiable etiology of angioedema. Antihistamines and corticosteroids were the most frequently used ED therapies. Concerningly, epinephrine was often used despite a lack of evidence or physiologic rationale, and was responsible for documented morbidity in these high-risk patients. It is essential for physicians to distinguish between mast cell and bradykinin-mediated etiologies of

287

Western Journal of Emergency Medicine

Accurate ACE-I Angioedema Diagnosis

Curtis et al.

angioedema, because their treatments differ. We do accept that epinephrine use is appropriate in undifferentiated cases. Unless new evidence of benefit from epinephrine use in AAE arises, epinephrine should be avoided for clear cases of nonmast cell mediated angioedema.

2008;100:327–32. 6. Bas M, Adams V, Suvorava T, et al. Nonallergic angioedema: role of bradykinin. Allergy. 2007;62(8):842–56. 7. Tiemensma M, Burger EH, Dempers JJ, et al. Fatal angioedema induced by angiotensin conversion enzyme (ACE) inhibitors. SA Fam

ACKNOWLEDGMENTS The authors have no direct conflicts of interest to declare regarding this publication. Relevant interests outside of this publication include: author RB has received funding from CSL Behring for support of summer research students regarding a case review. RB also participates on advisory board meetings for CSL Behring and Viropharma. Neither of these companies had involvement in this publication. A Queen’s University Research Initiation Grant from author IB was used to fund summer research students in this investigation.

Pract. 2010;52(3). 8. Messerli FH and Nussberger J. Vasopeptidase inhibition and angioedema. Lancet. 2000;356:608. 9. Statistics Canada [Internet], Ottawa: Statistics Canada 2013 [cited 2013 September 30]. Available at: http://www.statcan.gc.ca/tablestableaux/sum-som/l01/cst01/health70a-eng.htm 10. Jaiganesh T, Wiese M, Hollingsworth J, et al. Acute angioedema. Eur J Emerg Med. 2013;20(1):10–7. 11. Sondhi D. Airway compromise due to angiotensin-converting enzyme inhibitor-induced angioedema clinical experience at a large community teaching hospital. Chest. 2004;126(2):400–4. 12. Malde B, Regalado J, Greenberger PA. Investigation of angioedema associated with the use of angiotensin-converting enzyme inhibitorsand angiotensin receptor blockers. Ann Allergy Asthma

Address for Correspondence: R. Mason Curtis, MD, London Health Sciences Center, Division of Emergency Medicine, 800 Commissioners Road East, London, ON, Canada, N6C 2R6. Email: rcurtis3@uwo.ca.

Immunol. 2007;98(1):57–63. 13. Grant N, Deeb Z, Chia S. Clinical experience with angiotensinconverting enzyme inhibitor–induced angioedema. Otolaryngol Head Neck Surg. 2007;137(6):931–5.

14. Lieberman P, Kemp S, Oppenheimer, et al. The diagnosis and management of anaphylaxis: An updated practice parameter. J Allergy Clin Immunol. 2005;115(3):S483–S523. 15. Lieberman P, Nicklas RA, Oppenheimer J, et al. The diagnosis and

Copyright: © 2016 Curtis et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

management of anaphylaxis practice parameter: 2010 Update. J Allergy Clin Immunol. 2010;126(3):477–480. 16. Sampson HA, Muñoz-Furlong A, Campbell RL, et al. Second symposium on the definition and management of anaphylaxis: Summary report—Second National Institute of Allergy and Infectious Disease/Food Allergy and Anaphylaxis Network symposium. J Allergy Clin Immunol. 2006;117(2):391–7. 17. Johnston SL. Lesson of the week: Adrenaline given outside

REFERENCES

the context of life threatening allergic reactions. BMJ

1. Grigoriadou S and Longhurst HJ. Clinical immunology review series:

2003;326(7389):589–90.

An approach to the patient with angio-oedema. Clinical Exp Immunol.

18. Sood JD, Empson M, Fitzharris P, et al. Too much of a good thing, is

2009;155(3):367–77. 2. Lin RY and Shah SN. Increasing hospitalizations due to angioedema

it bad? Adrenaline on trial. NZ Med J. 2007. 19. Kanwar M, Irvin CB, Frank JJ, et al. Confusion About Epinephrine

in the United States. Ann Allergy Asthma Immunol. 2008;101(2):185-

Dosing Leading to Iatrogenic Overdose: A Life-Threatening Problem

92. 3. Cugno M, Nussberger J, Cicardi M, et al. Bradykinin and

With a Potential Solution. Ann Emerg Med. 2010;55(4):341–4. 20. Lang DM, Aberer W, Bernstein J, et al. International consensus on

the pathophysiology of angioedema. Int Immunopharmacol.

hereditary and acquired angioedema. Ann Allergy Asthma Immunol.

2003;3(3):311–7. 4. Bluestein HM, Hoover TA, Banerji AS, et al. Angiotensin-

2012;109:395-402. 21. Zuraw B, Bernstein J, Lang D et al. A focused parameter update:

converting enzyme inhibitor-induced angioedema in a community

Hereditary angioedema, acquired C1-inhibitor deficiency, and

hospital emergency department. Ann Allergy Asthma Immunol.

angiotensin-converting enzyme inhibitor-associated angioedema. J

2009;103(6):502. 5. Banerji A, Clark S, Blanda M, et al. Multicenter study of patients

Allergy Clin Immunol. 2013;131:1491-3. 22. Cicardi M, Aberer W, Bas M et al. Classification, diagnosis, and

with angiotensin-converting enzyme inhibitor–induced angioedema

approach to treatment for angioedema: consensus report from

who present to the emergency. Ann Allergy Asthma Immunol.

the Hereditary Angioedema International Working Group. Allergy.

Western Journal of Emergency Medicine

288

Volume XVII, no. 3 : May 2016

Curtis et al.

Accurate ACE-I Angioedema Diagnosis

2014;69:602-16.

progressive and refractory angiotensin-converting enzyme inhibitor-

23. Brown NJ, Snowden M, Griffin MR. Recurrent angiotensin-converting

induced angioedema. J Emerg Med. 2013;44(4):764–72.

enzyme inhibitor—associated angioedema. JAMA. 1997;278(3):232–3.

26. Bas M, Greve J, Stelter K, et al. Therapeutic efficacy of icatibant in

24. Beltrami L, Zingale L, Carugo S, et al. Angiotensin-converting

angioedema induced by angiotensin-converting enzyme inhibitors: A

enzyme inhibitor-related angioedema: how to deal with it. Expert Opin. Drug Saf. 2006;5(5):643-9.

case series. YMEM. 2010;56(3):278–82. 27. Bas M, Greve J, Stelter K, et al. A randomized trial of icatibant in

25. Hassen GW, Kalantari H, Parraga M, et al. Fresh frozen plasma for

Volume XVII, no. 3 : May 2016

ACE-inhibitor-induced angioedema. N Engl J Med. 2015;372:418-25.

289

Western Journal of Emergency Medicine

Case Report

Case Series of Synthetic Cannabinoid Intoxication from One Toxicology Center Kenneth D. Katz, MD* Adam L. Leonetti, DO* Blake C. Bailey, DO* Ryan M. Surmaitis, DO* Eric R. Eustice, DO* Sherri Kacinko, PhD† Scott M. Wheatley, MD‡

*Department of Emergency Medicine, Lehigh Valley Health Network, Allentown, Pennsylvania † NMS Labs, Willow Grove, Pennsylvania ‡ Pediatric Critical Care, Lehigh Valley Health Network, Allentown, Pennsylvania

Section Editor: Rick A. McPheeters, DO Submission history: Submitted December 17, 2015; Accepted February 22, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29519

Synthetic cannabinoid use has risen at alarming rates. This case series describes 11 patients exposed to the synthetic cannabinoid, MAB-CHMINACA who presented to an emergency department with life-threatening toxicity including obtundation, severe agitation, seizures and death. All patients required sedatives for agitation, nine required endotracheal intubation, three experienced seizures, and one developed hyperthermia. One developed anoxic brain injury, rhabdomyolysis and died. A significant number were pediatric patients. The mainstay of treatment was aggressive sedation and respiratory support. Synthetic cannabinoids pose a major public health risk. Emergency physicians must be aware of their clinical presentation, diagnosis and treatment. [West J Emerg Med. 2016;17(3):290–294.]

INTRODUCTION Synthetic cannabinoids (SCs) were synthesized to mimic the effects of Δ-9 Tetrahydrocannabinol (THC), the psychoactive component of the Cannabis sativa plant. Due to excessive cannabinoid receptor 1 (CB1) and cannabinoid receptor 2 (CB2) receptor agonism, use of SCs has been associated with unexpected and significant toxicity, including lethargy, agitation, tachycardia, hyperthermia, acute tubular necrosis, myocardial infarction, seizure and even death.1 Marketed as “Spice” and “K2,” its popularity among adolescents and adults has risen substantially and is the second most common drug of abuse (after marijuana) among high school students.2 In April 2015, an unprecedented, massive nationwide surge in SC cases was reported to regional poison control centers. The American Association of Poison Control Centers reported a 562% increase in April compared to March (1,512 versus 269).3 The state of New York issued a health alert after more than 160 citizens were hospitalized between April 8-17 for suspected SC toxicity. Health departments in Alabama, Mississippi, Connecticut, Maryland and the National Institute on Drug Abuse Western Journal of Emergency Medicine

issued similar warnings.4 Tyndall et al. recently reported an outbreak of MAB (or ADB)-CHMINACA toxicity in multiple patients at the University of Florida Health Medical Center.5 Northeastern Pennsylvania simultaneously witnessed an epidemic of SC exposures. The following cases describe 11 patients presenting to a tertiary care medical facility between April 20 and June 6 of 2015, who had serologic confirmed exposure to a novel, carboxamide indazole SC, MAB-CHMINACA (N-(1-amino-3,3dimethyl-1-oxobutan-2-yl)-1-(cyclohexylmethyl)-1 H-indazole3-carboxamide).6 This case series received expedited approval from the hospital’s institutional review board. CASE REPORTS Case #1: An 18-year-old man without prior medical history was found by police, unresponsive in a parking lot after smoking “K2.” The patient became agitated and was brought into the emergency department (ED). He was tachycardic and admitted to the tertiary care intensive care unit (TCICU). His EKG showed sinus tachycardia and rightward axis. Pupil examination was 4mm bilaterally with sluggish reactivity. The patient recovered uneventfully and

290

Volume XVII, no. 3 : May 2016

Katz et al.

Case Series of Synthetic Cannabinoid Intoxication

was discharged later that day. Case #2: A 28-year-old man with a history of substance abuse and hepatitis presented to an outside healthcare facility (OHF) unresponsive, hallucinating and tachycardic. The patient was endotracheally intubated for airway protection and then transferred to a TCICU. Although he developed aspiration pneumonia, he was treated and discharged uneventfully on hospital day (HD) six. Case #3: A 17-year-old woman without medical history was transferred to the TCICU from an OHF for agitation, delirium and tachycardia after exposure to SC. She was treated with benzodiazepines, but did not require intubation. he patient was discharged uneventfully on HD two. Case #4: A 14-year-old boy without medical history was transferred from an OHF to the tertiary care pediatric intensive care unit (TCPICU) after being found unresponsive on the street. The patient became agitated and combative at the OHF, and was endotracheally intubated. The patient admitted to using “K2” and was eventually extubated and discharged on HD two. Case #5: A 13-year-old girl with a history of marijuana abuse was transferred from an OHF to a TCPICU after ingesting SC. The patient was found intermittently responsive and combative at home. Her pupils were 3mm bilaterally and sluggish, but reactive. At the OHF, she was tachycardic and-after being administered benzodiazepines--became obtunded with hypoventilation. She was endotracheally intubated and transferred to the TCPICU. The patient was extubated and discharged on HD two. Case #6: A 13-year-old boy without medical history was found unresponsive in a park. On examination, his pupils were 3mm bilaterally and reactive. At the OHF, the patient was endotracheally intubated and transferred to a TCPICU. He required significant amounts of sedatives due to periods of agitation and combativeness. During his hospitalization the patient developed aspiration pneumonia, but was discharged uneventfully on HD three. Case #7: A 50-year-old man with a history of polysubstance abuse was transported to the ED after using SCs and being found unresponsive by his roommate. The patient was apneic and cyanotic, and--after endotracheal intubation--was transferred to the TCICU. He was extubated, and discharged to an inpatient drug and rehabilitation facility on HD eight. Case #8: A 40-year-old woman with a history of bipolar disorder presented to the ED with a witnessed seizure after smoking SC. Her EKG demonstrated sinus tachycardia with an incomplete right bundle branch block. The patient was tachycardic, administered benzodiazepines and endotracheally intubated. In the TCICU, the patient was extubated, had an unremarkable electroencephalogram and discharged on HD two. Case #9: A 19-year-old woman with a history of epilepsy, bipolar disorder and substance abuse presented Volume XVII, no. 3 : May 2016

to an OHF after suffering a seizure after smoking SC. Her EKG showed sinus rhythm with sinus arrhythmia and nonspecific T-wave abnormality. The patient was found unresponsive; she was endotracheally intubated and transferred to a TCICU. She was also treated for a wound infection secondary to intravenous drug abuse. The patient was extubated and discharged on HD four. Case #10: A 14-year-old boy with a history of substance abuse was found by his mother, agitated after exposure to SC and transported to an OHF. His pupils were 4mm bilaterally and sluggishly reactive on examination. The patient’s EKG demonstrated sinus bradycardia; early repolarization was noted. The patient demonstrated periods of unresponsiveness followed by severe agitation and combativeness requiring sedation and endotracheal intubation. He was transferred to the TCICU and eventually extubated and discharged on HD two. Case #11: A 20-year-old man without medical history was found unresponsive by family members after SC exposure and transported to an OHF. The patient had been without medical care for approximately 24 to 36 hours. He was hyperthermic, tachycardic and demonstrated decorticate posturing along with areas of trunk and extremity edema on examination. He was endotracheally intubated. Laboratory analysis demonstrated significant rhabdomyolysis and acute renal failure. He was transferred to the TCICU where clinical examination was consistent with anoxic brain injury. A brain magnetic resonance imaging (MRI) scan confirmed these findings. Care was eventually withdrawn due to the grim prognosis, and the patient died on HD seven. These 11 patients were all exposed to the SC, MABCHMINACA, and presented to the ED with significant, life-threatening toxicity, ranging from obtundation to severe agitation, seizures and death. All had serum samples drawn upon arrival to the hospital that were sent to a contracted laboratory for further identification of specific SCs. All patients had expanded toxicologic analysis (liquid chromatography/mass spectroscopy, LC/MS) to determine any potential adulterants or other substances to which each patient was exposed. All patients required benzodiazepines or other sedatives for agitation. Nine of 11 patients required endotracheal intubation for either control of severe agitation or decreased responsiveness. Three experienced seizures. One patient presented with hyperthermia, and six of the 11 had tachycardia. One patient developed rhabdomyolysis and died. The majority suffered from mental illness or substance abuse. A significant number were pediatric patients, the youngest of whom was only 13. Common chemistries and laboratory values were unremarkable for these patients; pupil size examination and/or EKG results were not available for five of the cases. The results are summarized in Table. Any drug material recovered from patients was analyzed by the county criminal investigative laboratories to determine the presence of SCs (see Figure). Results confirmed MABCHMINACA in all samples tested.

291

Western Journal of Emergency Medicine

Katz et al.

Case Series of Synthetic Cannabinoid Intoxication

Table. Summary of case reports of 11 patients presenting to the emergency department with life-threatening toxicity after exposure to synthetic cannabinoids. Qualitative serum level, SC screen*

Comprehensive urine drug screen findings (LC/MS)

Hospital day

Clinical presentation

#1. 18yo male

Day 1

Altered mental status, combative, agitated

Discharged same day.

HR 113bpm, BP 95/53mmHg, temp 99.9°F, RR 20rpm, pulse ox 100%, room air

#2. 28yo male

Day 3

Combative and hallucinating

Yes

Developed aspiration pneumonia. Discharged hospital day 6.

HR 102bpm, BP 119/88mmHg, temp 96°F, RR 15rpm, pulse ox 100%, vent

X10

#3. 17yo female

Day 2

Agitated and delirious

Discharged following day.

HR 110bpm, BP 112/63mmHg, temp 97.3°F, RR 16rpm, pulse ox 98%, room air

Lorazepam

#4. 14yo male

Day 2

Intermittent unresponsiveness followed by agitation

Yes

Discharged following day.

HR 96bpm, BP 123/58mmHg, temp 96.7°F, RR 19rpm, pulse ox 98%, vent

X45

Norfentanyl

#5. 13yo female

Day 2

Altered mental status with unresponsiveness

Yes

Discharged HR 115bpm, BP hospital day 2. 98/46mmHg, temp 98.9°F, RR 17rpm, Ppulse ox 100%, vent

X20

Phenylephrine, midazolam, fentanyl, norfentanyl, diphenhydramine, cotinine

#6. 13yo male

Day 0

Found unresponsive

Yes

Developed aspiration pneumonia. Discharged on hospital day 3.

HR 81bpm, BP 131/80mmHg, temp 97.1°F, RR 16rpm, pulse ox 99%, room air

X30

Lorazapam, hydroxymidazolam

#7. 50yo male

Day 1

Found unresponsive

Yes

Discharged to inpatient drug rehabilitation facility.

HR 59bpm, BP 93/57mmHg (patient prescribed Lopressor), temp 98.7°F, RR 14rpm, pulse ox 95%, vent

X36

Ethanol, naloxone, metoprolol, caffeine

#8. 40yo female

Day 1

Combative, delirious, seizures

Yes

Case

Intubated

Outcome

Vital signs

Caffeine

Caffeine, morphine, midazolam, lorazepam

Discharged HR 114bpm, BP X32 Acetaminophen hospital day 2. 120/62mmHg, temp 99.4°F, RR 18rpm, pulse ox 84%, room air (improved after intubation) SC, synthetic cannabinoids; LC, liquid chromatography; MS, mass spectroscopy; HR, heart rate; BP, blood pressure; RR, respiratory rate; RPM, respirations per minute; yo, year old *Value represents the number of times higher each case was in comparison to Case #1, which had the lowest relative concentration.

DISCUSSION To date, this is the largest SC case series describing MABCHMINACA toxicity. In the fall of 2014, MAB-CHMINACA Western Journal of Emergency Medicine

was responsible for more than 125 patients seeking hospital care in Baton Rouge, LA.7 It is a highly potent SC, which was just recently added to the Schedule 1 Controlled Substance Act in 292

Volume XVII, no. 3 : May 2016

Katz et al.

Case Series of Synthetic Cannabinoid Intoxication

Table. Continued. Case

Hospital day

Clinical presentation

Intubated

Outcome

Vital signs

Qualitative serum level, SC screen*

Comprehensive urine drug screen findings (LC/MS)

#9. 19yo female Day 3

Found unresponsive; seizure-like activity

Yes

Treated for infected right forearm wound secondary to IV drug use. Discharged hospital day 4.

HR 74bpm, BP 118/85mmHg, temp 97.6°F, RR 14rpm, pulse ox 100%, 40% oxygen

X55

Morphine, norfentanyl, cocaine, amphetamine, methamphetamine, codeine, midazolam, lorazepam

#10. 14yo male

Day 2

Agitated and combative

Yes

Discharged following day.

HR 95bpm, BP 114/66mmHg, temp 99.9°F, RR 22rpm, pulse ox 100% vent

X13

Sertraline

#11. 20yo male

Day 3

Found unresponsive and posturing; approximate down-time was 24-36 hours

Yes

Expired. Anoxic brain injury. Family withdrew care on hospital day 7

HR 146bpm, BP 200/74mmHg, temp 104.1°F, RR 33rpm, pulse ox 100%, vent (outside facility)

X1.5

Sertraline

HR 75bpm, BP 138/55mmHg, temp 99.5°F, RR 17rpm, pulse ox 100%, vent (upon ICU arrival) SC, synthetic cannabinoids; LC, liquid chromatography; MS, mass spectroscopy; IV, intravenous; HR, heart rate; BP, blood pressure; RR, respiratory rate; RPM, respirations per minute; ICU, intensive care unit *Value represents the number of times higher each case was in comparison to Case #1, which had the lowest relative concentration.

late 2015.8 The clinical manifestations observed in these patients appear similar to those described with other SC toxicities.5,9 The exact mechanism of MAB-CHMINACA toxicity is unknown. Affinities of SCs and their metabolites are multifold higher at the CB1 and CB2 receptors than that of Δ-9 THC and are thought responsible for such severe clinical manifestations.9 CB1 receptors are mostly located in the brain and regulate the central nervous system effects of Δ-9 THC and other cannabinoids; they are also expressed peripherally in adipocytes and skeletal muscle.1 The identification of CB1 receptors presynaptically on GABA and glutamatergic terminals with increased excitatory and decreased inhibitory tone may be responsible.10 Other postulated mechanisms include activation of non-cannabinoid receptors and drug-drug synergistic effects.1 Some patients demonstrated stimulant and serotonergic agent use as detected by comprehensive urine drug LC/ MS testing that could have contributed to their clinical presentation, such as sympathomimetic or serotonergic toxidromes. However, this was not universal and would not explain the fairly consistent clinical presentation of all of these patients. Further, the relative concentrations of MABVolume XVII, no. 3 : May 2016

CHMINACA do not necessarily correlate with toxicity and may demonstrate the potential lipophilic distribution of the drug to tissue at the time of blood draw. For example, patient #11 had a relatively low concentration, yet expired. Interestingly, post-mortem analysis of a patient who died from MAB-CHMINACA toxicity demonstrated a relatively low amount in the adipose tissue, atypical of other SCs. The authors hypothesized the compound may require time to distribute to that tissue, and that may have been the salient issue regarding patient #11 given the length of time from medical attention.11 Limitations in real-time detection by routine toxicologic immunoassay screening appear to be a factor in SC use.12 The reasons for the abrupt, recent increase in exposures, however, remain unclear. The celebration of “4/20” (a counterculture holiday celebrating Cannabis) correlated with this surge and might have influenced users. Management of the SC-intoxicated patient centers on meticulous supportive care, as no specific antidote currently exists. Decontamination is of little value. Given their beneficial pharmacokinetic profile, administration of benzodiazepines titrated to clinical effect can control agitation, delirium,

293

Western Journal of Emergency Medicine

Katz et al.

Case Series of Synthetic Cannabinoid Intoxication REFERENCES

1. Zhao A, Tan M, Maung A, et al. Rhabdomyolysis and acute kidney injury requiring dialysis as a result of concomitant use of atypical neuroleptics and synthetic cannabinoids. Case Rep Nephrol. 2015;2015:235982. 2. NIH, National Institute on Drug Abuse. Media Guide: Most Commonly Used Addictive Drugs. Available at: http://www.drugabuse.gov/ publications/media-guide/most-commonly-used-addictive-drugs. Accessed on Jan 23, 2016. 3. Synthetic Marijuana Data. American Association of Poison Control Centers Web site. Available at: https://aapcc.s3.amazonaws.com/ files/library/Syn_Marijuana_Web_Data_through_7.6.15.pdf. Accessed on Nov 17, 2015. 4. Schwartz A. Potent “Spice” drug fuels rise in visits to emergency room. New York Times (New York Ed.). April 25, 2015. Available at:

Figure. Sample synthetic cannabinoid packaging.

http://www.nytimes.com/2015/04/25/health/surge-in-hospital-visitslinked-to-a-drug-called-spice-alarms-health-officials.html?_r=0. Accessed on Nov 17, 2015.

hyperthermia and seizures. Cooling measures without antipyretic administration can also be implemented for hyperthermic patients. Administration of intravenous fluids is helpful for associated rhabdomyolysis. Endotracheal intubation may be required for the severely intoxicated patient.9

5. Tyndall JA, Gerona R, De Portu G, et al. An outbreak of acute delirium from exposure to the synthetic cannabinoid AB-CHMINACA. Clin Toxicol (Phila). 2015;53(10):950-6. 6. Debruyne D and Le Boissilier R. Emerging drugs of abuse: Current perspectives on synthetic cannabinoids. Subst Abuse Rehabil. 2015;6:113–29.

CONCLUSION SCs pose a substantial and dangerous public health risk. Emergency physicians must be aware of their clinical presentation, difficulties in identification and treatment. Further studies are required to elucidate the exact mechanism of SCs and, more specifically, MAB-CHMINACA toxicity.

7. DHH Bans Another Synthetic Drug. WAFB, Baton Rouge, Louisiana Web site. Available at: http://www.wafb.com/story/27157019/dhhbans-another-synthetic-drug. Accessed on Nov 19, 2015. 8. Schedules of Controlled Substances: Temporary Placement of the Synthetic Cannabinoid MAB-CHMINACA into Schedule I (A Proposed Rule). Drug Enforcement Administration, Washington DC Web site. Available at: https://www.federalregister.gov/ articles/2015/09/16/2015-23198/schedules-of-controlled-substances-

Address for Correspondence: Kenneth D. Katz, MD, Lehigh Valley Health Network, Department of Emergency Medicine, 2545 Schoenersville Road, 4th Floor, South Wing, Bethlehem, PA 18017. Email: katzkd1@gmail.com.

temporary-placement-of-the-synthetic-cannabinoid-mab-chminaca. Accessed on Nov 19, 2015. 9. Nelson ME, Bryant SM, Aks SE. Emerging drugs of abuse. Emerg Med Clin North Am. 2014;32(1):1-28. 10. Lapoint J, James LP, Moran CL, et al. Severe toxicity

following synthetic cannabinoid ingestion. Clin Toxicol (Phila). 2011;49(8):760-4. 11. Hasegawa K, Wurita A, Minakata K, et al. Postmortem distribution of MAB-CHMINACA in body ﬂuids and solid tissues of a human

Copyright: © 2016 Katz et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

Western Journal of Emergency Medicine

cadaver. Forensic Toxicol. 2015;33(2):380-7. 12. Tait RJ, Caldicott D, Mountain D, et al. A systematic review of adverse

294

events arising from the use of synthetic cannabinoids and their associated treatment. Clin Toxicol (Phila). 2015;2015:1-13.

Volume XVII, no. 3 : May 2016

Original Research

Gender Differences in Emergency Department Visits and Detox Referrals for Illicit Drug Use and Nonmedical Use of Opioids Hyeon-Ju Ryoo, BA* Esther K. Choo, MD, MPH†

*The Warren Alpert Medical School of Brown University, Providence, Rhode Island † The Warren Alpert Medical School of Brown University, Department of Emergency Medicine, Providence, Rhode Island

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted December 3, 2015; Revision received January 27, 2016; Accepted February 6, 2016 Electronically published April 28, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29425

Introduction: Visits to the emergency department (ED) for use of illicit drugs and opioids have increased in the past decade. In the ED, little is known about how gender may play a role in drugrelated visits and referrals to treatment. This study performs gender-based comparison analyses of drug-related ED visits nationwide. Methods: We performed a cross-sectional analysis with data collected from 2004 to 2011 by the Drug Abuse Warning Network (DAWN). All data were coded to capture major drug categories and opioids. We used logistic regression models to find associations between gender and odds of referral to treatment programs. A second set of models were controlled for patient “seeking detox,” or patient explicitly requesting for detox referral. Results: Of the 27.9 million ED visits related to drug use in the DAWN database, visits by men were 2.69 times more likely to involve illicit drugs than visits by women (95% CI [2.56, 2.80]). Men were more likely than women to be referred to detox programs for any illicit drugs (OR 1.12, 95% CI [1.02-1.22]), for each of the major illicit drugs (e.g., cocaine: OR 1.27, 95% CI [1.15-1.40]), and for prescription opioids (OR 1.30, 95% CI [1.17-1.43]). This significant association prevailed after controlling for “seeking detox.” Conclusion: Women are less likely to receive referrals to detox programs than men when presenting to the ED regardless of whether they are “seeking detox.” Future research may help determine the cause for this gender-based difference and its significance for healthcare costs and health outcomes. [West J Emerg Med. 2016;17(3):295–301.]

INTRODUCTION Visits to the emergency department (ED) for use of illicit drug and misuse of opioid analgesic have increased disproportionately among certain sub-populations during the past decade.1,2 The number of ED visits for nonmedical use of opioid analgesics increased by 111% between 2004 and 20083 in the setting of marked increase in opioid prescriptions from both primary care and ED.4 Interestingly, a clear gender-specific pattern has emerged from the epidemiologic data on opioid overdoses: the rate of drug overdose deaths Volume XVII, no. 3 : May 2016

increased 500% in women over the past decade, compared to a 360% increase in men.3 Drug use in the ED population is associated with high prevalence of comorbid conditions, injuries, high recidivism, morbidity, and mortality.5 Thus, the ED is an important point of contact for providing standardized screening, brief interventions and referrals to treatment (SBIRT) for drug use.6 In the face of modest outcomes from existing ED SBIRT studies, investigators have been turning to more targeted programs, reasoning that subgroups of patients might have 295

Western Journal of Emergency Medicine

Ryoo et al.

Gender Differences in Emergency Department Utilization different motivations for and barriers to changing drug use, responsiveness to specific intervention elements, and referral needs. Designing interventions based on gender is one promising avenue: gender is one of the most influential factors in determining trajectory of drug use disorders, accessing treatments, and achieving recovery.7-10 Outside the ED setting, a number of successful programs have been developed specifically targeting subgroups of men and women.11 Developing gender-specific interventions for the ED will require a better understanding of how men and women differ on various aspects of ED care. Frequently, studies include gender as a covariate but do not stratify samples or examine interactions between gender and other demographic or clinical variables; many studies explicitly studying gender are likely underpowered to detect differences between genders. Overall, there are few data on ED utilization by men and women with drug misuse and clinical outcomes of visits. As the opioid epidemic has demonstrated, following gender-related trends over time can shine a light on clinical problems that are manifesting in gender-specific or gender-sensitive ways, which in turn may represent gender-related vulnerabilities to adverse clinical consequences or special needs for prevention or treatment. Although there have been gender-based comparisons of utilization of substance-use treatment facilities and treatment outcomes,12,13 we know little about how treatment entry is facilitated and any gender disparity therein. The importance of emergency care research focused on gender disparities in patterns of referrals to substance-abuse treatments has been identified as a national priority.14 Our objective was to perform a gender-based comparative analysis of drug-related ED visits using a nationally representative database, Substance Abuse and Mental Health Administration (SAMHSA)’s Drug Abuse Warning Network (DAWN). Objectives were to examine, by gender the following: 1) rates of presentations and dispositions of patients presenting with use of major illicit drugs and nonmedical use of prescription opioids; 2) trends in drug-related visits over time; 3) patterns of referral to outpatient therapy upon discharge referred to as “detox referrals.” METHODS DAWN is a nationally representative sample of ED visits to hospitals throughout the 50 states of the U.S.15 DAWN includes non-Federal, short-stay, general surgical and medical hospitals with a 24-hour ED.15 A retrospective chart review was performed by a trained DAWN reporter at each hospital. ED visits in which drug ingestion was the direct cause or contributing factor of the visit were identified.15 ED visits were codified based on a standardized algorithm – a “DAWN Decision Tree” – available in the DAWN Methodology Report.16 ED visits reportable to DAWN involve not only all forms of drug misuse and abuse but also adverse reactions, accidental ingestions, and visits where patients were seeking detox services.15 All drugs are Western Journal of Emergency Medicine

classified based on DAWN Drug Reference Vocabulary, a drug coding system based on Multum Lexicon, © 2012.15 “Illicit” drugs are defined as cocaine, heroin, marijuana, synthetic cannabinoids, amphetamines, methamphetamine, Ecstasy (MDMA), gamma-hydroxybutyric acid (GHB), flunitrazepam (Rohypnol®), ketamine, lysergic acid diethylamide (LSD), phencyclidine (PCP), hallucinogens, or substances inhaled for psychotropic properties (e.g., sniffing model airplane glue). Nonmedical use of pharmaceuticals include misuse or abuse of prescription medications, over-the-counter medications, or dietary supplements that result from taking higher-thanprescribed dose, taking pharmaceuticals prescribed to another individuals, or malicious poisoning by another individual.15 This analysis used all available DAWN data, from 2004 to 2011, when data collection was terminated due to limited funding. Across the years, 2.6 million drug-related ED visits were identified: Applying post-stratified weights, these cases extrapolated to an estimate of 27.9 million drug-related ED visits out of an estimated 937 million total visits. DAWN captures limited data on individual visits, including age, gender, race, and disposition categories. It does not capture admission diagnoses, measures of illness severity, procedures performed, or length of stay. Our analysis was restricted to adult patients (≥18 years of age). Individual drug types reported in the DAWN database were reviewed by the authors and further coded into categories of the following: 1) cocaine, 2) hallucinogens, 3) heroin, 4) marijuana; 5) methamphetamines; and 6) any illicit drug. We coded separately recreational drugs not clearly falling into one of these categories but included in DAWN’s definition of “illicit drugs” (e.g., cathinone and GHB). Drug combinations (such as cocaine/heroin) were placed in both categories. A single code for “any illicit drug use” was created that was positive if any of the previous drug categories were present and reflect “illicit drugs” defined by DAWN. We also created a variable for all prescription opioids that were included in the database for nonmedical use. All drug names under each category were reviewed by the authors, and any discrepancies between the two authors’ lists were resolved by using a toxicology database. Alcohol was a pre-defined variable in the DAWN database and only included underage drinking. Other variables of interest extracted for this study included gender, age, race, and clinical disposition, including 1) hospital admission or transfer; 2) intensive care unit (ICU) admission; 3) referral to outpatient detoxification; 4) admission for inpatient detoxification; and 5) psychiatric care. As this study used only existing, publically available, de-identified data, it was exempt from institutional review board review. Data Analysis We calculated proportions for demographic and drug use variables. The proportions for men and women were compared using unadjusted odds ratios.

296

Volume XVII, no. 3 : May 2016

Ryoo et al.

Gender Differences in Emergency Department Utilization

Trends over time were presented graphically and examined using Stata 12.0 (StataCorp LP, College Station, TX). We developed logistic regression models to test for associations between gender and the binary outcome of referral to substance-use treatment programs among patients discharged from the ED. The first model adjusted for age, race, number of involved substances, and the time of day of the visit. These variables were determined a priori based on previous studies as characteristics available in the DAWN database that could potentially confound the effect of gender.17-23 The second model additionally included the chief complaint of “seeking detox.” This variable was selected after noting a difference between proportions of men and women “seeking detox” in the univariate analysis. Model variables were examined for evidence of collinearity with variance inflation factors (VIF), as high multicollinearity may lead to increased variance of model coefficients. Adjusted odds ratios (aOR) for which the 95% CI did not cross the null value of aOR=1.0 were considered statistically significant. For all analyses, we used “svy” commands in Stata to account for weights and clustering and to obtain accurate point estimates, standard errors, confidence intervals and tests of hypothesis. These have been described in previous studies.24 Model fit was evaluated using HosmerLemeshow goodness of fit statistics.

drug use, cocaine, heroin, and marijuana, as compared to women (Figure 1). Men and women had similar rates of visits involving hallucinogens, methamphetamines, and prescription opioids across all years studied (Figure 1). No significant changes in numbers of visits were found across the years for visits involving either male or female patients (Figure 1). In logistic regression analysis, men were more likely than women to be referred to detox programs for any illicit drugs, each of the illicit drug categories, as well as prescription opioids (Table 2). This significant gender-based difference in detox referrals remained after adjusting for “seeking detox” for all categories of drugs, except for the combined category of “all illicit drugs” (Table 2). The VIFs were near unity, supporting lack of collinearity (Table 2).

RESULTS Of the 27.9 million ED visits related to drug use in the DAWN database 2004 to 2011, visits by men were 2.69 times more likely to involve illicit drugs than visits by women (95% CI [2.56, 2.80]). For every major category of illicit drugs, visits by men were more likely than visits by women to involve the illicit drug (Table 1).Visits by men were less likely to involve patients that were White (OR=0.75; 95% CI [0.70, 0.81]) and ages >55 years old (OR=0.75; 95% CI [0.72, 0.79]) as compared to women (Table 1). Gender differences were observed in the ultimate disposition of ED patients presenting with drug use. ED visits by men were more likely than visits by women to result in hospital admissions with odds ratio of 1.19 (95% CI [1.14, 1.24]) (Table 1). Among discharged patients, men were 1.90 times more likely than women to receive detox referrals (95% CI [1.72, 2.09]) (Table 1). Among those seeking detox, 23.1% (95% CI [19.0, 27.2]) of women actually received referrals to detox compared to 24.5% (95% CI [20.7, 28.3]) of men. Conversely, among those discharged with detox referrals, 33.2% (95% CI [29.4, 37.1]) of women were “seeking detox” compared 41.4% (95% CI [38.2, 44.6]) of men. In other words, although a similar proportion of male and female patients “seeking detox” received detox referrals, “seeking detox” constituted a smaller proportion of women receiving referrals as compared to men. For most years from 2004 to 2011, visits by male patients were significantly more likely to involve any illicit Volume XVII, no. 3 : May 2016

DISCUSSION Approximately 23.9 million Americans aged 12 or older are current users of illicit drugs and the proportion of women using illicit drugs is growing faster than men.25 Women face many gender-specific barriers to treatment such as financial dependence, family responsibilities, and more frequent self-reports of shame and stigma.7 Given the potentially high impact of substance-abuse treatments, improving access to treatment services–particularly for women–is a public health priority. Visits to the ED by women involving substance abuse present an opportunity to connect them to substance-use treatment and other mental health resources and address any existing barriers to accessing these resources. The DAWN database, however, demonstrates that nationwide, women are less likely than men to receive referrals to treatment for use of illicit drugs or misuse of prescription opioids. There are a number of potential causes for gender-based difference in referrals, including severity of drug use, comorbid conditions, patient motivation, and physician biases. The current literature, however, suggests that many of these factors should result in higher rate of treatment referrals for female than male patients. In particular, past studies have shown that women have greater severity of substance abuse upon presentation to treatment,7 and faster progression to dependence, also known as “telescoping.”8-10 The addiction literature also demonstrates that women are more likely to suffer from comorbid psychiatric conditions such as depression and anxiety and other health-related consequences than men; it is likely that women have greater needs for dual substance-use and mental healthcare facilities, which would address these interrelated problems simultaneously. Overall, these findings suggest that severity of drug use and comorbid conditions should have prompted more treatment referrals for women than men. Our analysis indicates that patient motivation could have played a role in the fact that referrals were given more often to men than women. Although similar proportion of men and women “seeking detox” actually received detox referrals,

297

Western Journal of Emergency Medicine

Ryoo et al.

Gender Differences in Emergency Department Utilization

Table 1. Characteristics of emergency department visits related to drug use, by gender, from the Drug Abuse Warning Network (DAWN) database (N=27,865,483). Female proportion, %

Male proportion, %

Male:Female unadjusted OR (95% CI)

Age category 18-29

25.3

30.0

1.15 (1.11, 1.18)

30-44

26.5

28.0

1.08 (1.04, 1.11)

45-54

16.7

18.2

1.11 (1.08, 1.15)

55 or older*

31.5

25.6

0.75 (0.72, 0.79)

White*

72.5

66.5

0.75 (0.70, 0.81)

Black/African-American*

17.4

20.9

1.25 (1.16, 1.34)

Other*

10.1

12.7

1.29 (1.17, 1.43)

18.5

37.8

2.69 (2.56, 2.80)

Cocaine*

9.6

18.9

2.18 (2.06, 2.32)

Marijuana*

5.5

12.4

2.41 (2.31, 2.52)

Heroin*

3.6

8.6

2.59 (2.35, 2.64)

Methamphetamines*

1.7

3.3

1.95 (1.78, 2.14)

Race

Drug category Any illicit drug*

Hallucinogens*

1.6

2.5

1.54 (1.35, 1.76)

17.9

17.2

0.95 (0.90, 1.00)

Hospital admission*

30.5

34.3

1.19 (1.14, 1.24)

ICU admission*

17.0

15.6

0.90 (0.85, 0.96)

4.1

6.5

1.62 (1.44, 1.83)

56.5

54.4

0.76 (0.72, 0.79)

Prescription opioids Disposition

†

Psychiatric admission* Discharged*

Discharged with detox referral* 3.2 5.9 1.90 (1.72, 2.09) ICU, intensive care unit *Statistically significant difference between genders. † Disposition categories do not add up to 100% because several disposition categories were omitted (e.g. transferred, deceased, left against medical advice). Admissions to the ICU and psychiatry are subset of the total hospital admissions. Omitted admissions include those to inpatient, surgery, and inpatient detox unit. Similarly, discharged with detox referral is a subset of detox. Discharged home and released to police/jail is not included in this table.

Table 2. Male:Female adjusted odds ratios (aORs) for referral to detox programs. Second column includes models adjusted for covariates used in first column in addition to an additional variable indicating if the patient presented to the emergency department with a complaint of “seeking detox”. Model 1 Male:Female aOR (95% CI)

Mean VIF

Model 2: “Seeking detox” Male:Female aOR (95%CI)

Any illicit drug

1.12 (1.02, 1.22)

1.05

1.06 (0.96, 1.17)

1.10

Cocaine

1.27 (1.15, 1.40)

1.06

1.13 (1.02, 1.26)

1.09

Hallucinogens

1.31 (1.19, 1.45)

1.02

1.14 (1.02, 1.27)

1.06

Heroin

1.23 (1.12, 1.35)

1.02

1.12 (1.01, 1.06)

1.07

Marijuana

1.30 (1.17, 1.44)

1.06

1.13 (1.01, 1.26)

1.09

Methamphetamines

1.31 (1.18, 1.45)

1.02

1.14 (1.02 1.27)

1.06

1.30 (1.17, 1.43)

1.04

1.13 (1.02, 1.26)

1.08

Mean VIF

Discharge with detox referral

Prescription opioids VIF, variance inflation factors

Western Journal of Emergency Medicine

298

Volume XVII, no. 3 : May 2016

Ryoo et al.

Figure 1. Trends in drug use: 2004-2011. ED, emergency department

more men than women with detox referrals were “seeking detox.” This descriptive finding suggests that motivation may play a larger role in obtaining referrals to treatments for men than for women. The regression model, however, demonstrates a greater rate of treatment referrals for men even after controlling for the chief complaint “seeking detox” for all drug categories. This finding suggests that lower rate of explicit request for treatment services by women may not entirely account for their lower rate of detox referrals for most categories of illicit drugs. On the other hand, men and women had similar rate of detox referral for “all illicit drugs” after controlling for “seeking detox.” One potential explanation is that “seeking detox” was an important reason for receiving detox referrals for patients using drugs that are part of “all illicit drugs” but not coded as major categories of illicit drugs, such as GHB, LSD, and PCP. Prior studies have demonstrated physician bias in screening for substance use in the emergency setting.10,26 Similarly, past literature regarding alcohol treatment suggests that women are less likely than men to receive physician referrals to treatment centers.7,27 The higher prevalence of ED visits involving drug use among men as well as the higher admission rates found in DAWN database suggest that emergency physicians may be more aware of significant substance use among their male patients, and thus more vigilant about providing detox referrals for them. Physicianand healthcare-related barriers to detox referrals for women Volume XVII, no. 3 : May 2016

Gender Differences in Emergency Department Utilization and interventions to overcome these barriers are important areas for future research. Overall, the rate of referrals was very low: only 3.2% in women and 5.9% in men of those presenting with drug use problems. Although our analysis indicates that women are less likely than men to receive referrals, it is important to note that both genders have low rates of referrals and may therefore both be likely to face significant barriers to treatment. The low rate of referrals is consistent with past studies that have found that 27% of ED patients had unmet substance-abuse treatment needs28. The low rate of referrals reflects, in part, that the baseline population includes those who clearly did not present with drug abuse, such as those with accidental ingestions or adverse reactions. However, it may also demonstrate the limited availability of and referral to substance-use treatment programs, particularly for patients who are uninsured. Other barriers may include cost, lack of transportation, other family or work responsibilities, lack of information, functional differences arising from psychiatric illness, or domestic violence.7,26 It is also plausible that many of the patients presented with low amounts of drug abuse and were deemed not at risk enough to receive a referral. However, there is no “safe” or “low risk” amount of illicit drug use or nonmedical use of prescription drugs ,and virtually all illicit drug users should receive some type of referral. Over the past 40 years, there has been a dramatic improvement in the recognition of the specific needs of women with substance use disorders and the development of women-focused substance-abuse treatment programs7. Data from the National Treatment Improvement Evaluation Survey (NTIES), a study of publically-funded substance-use treatment units nationwide, demonstrated that women are more likely than men to receive individualized counseling and access-related services such as transportation and child care.29 Compared to men, women have comparable treatment retention, similar reductions in post-treatment substance abuse,20,29-32 and are less likely to experience relapse.33,34 Despite these advances, our analysis suggests that referrals in the ED may lag in recognition of substance use in women as suggested by our analysis and serve as a potential barrier to accessing substance-use treatments that have otherwise been shown to be efficacious for women. LIMITATIONS Although the DAWN database is a useful resource for nationally representative drug-related ED visits, there are several limitations. The DAWN database relies on diagnosis of illicit drug use or misuse of opioids by ED physicians. Future studies with patients’ reports of illicit drug use and nonmedical use of opioids are needed to determine whether the gender disparity reflects the actual prevalence of drug use by patients in the ED or differences in the recognition of drug use by emergency physicians. The DAWN database was created as 299

Western Journal of Emergency Medicine

Ryoo et al.

Gender Differences in Emergency Department Utilization a surveillance system to monitor trends of drug use over time and thus poses several challenges in answering our clinical research question of gender differences in ED visits and detox referrals. Because the cases reportable to the DAWN database include only visits involving drug misuse, proportions are expressed as a percentage of DAWN cases rather than total ED visits. This does not allow estimation of the proportion of total visits with illicit drug use and may inaccurately reflect the differences between genders; specifically, it is vulnerable to underestimating visits by female patients, who are less likely to be identified as using illicit drugs10. The inherent limitations of data captured by the DAWN database also do not allow more in-depth inquiry regarding the severity of illness, patient level factors such as past drug abuse, socioeconomic status, or physical or mental health comorbidities, or physician-level reasons for failing to provide detox referrals. DAWN is not clear whether a referral to a dual diagnosis treatment center – a resource potentially more relevant to women than men, as discussed above – would have been captured as a referral to a substance-use treatment center; if dual diagnosis referrals were not categorized as substance-use treatment, this could lead to a systematic bias underestimating resources provided to women. It is unclear whether physicians did not recommend detox or whether the patient refused detox despite being offered a referral. A study with prospective data collection focused on capturing these details that were omitted in DAWN could better identify contributing factors to a detox referral or lack thereof. “Seeking detox” in the DAWN database is also a heterogeneous definition and includes a wide range of causes, such as experiencing withdrawal and request for medical clearance before entering jail or a detox program. More research is needed to investigate each cause for seeking detox and the impact of patient motivation in receiving referrals to detox. Specifically, future studies should stratify specific motivations for seeking detox (e.g., medical clearance for jail or for withdrawal symptoms) and their impact on detox referrals.

Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Ryoo et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES 1. Cherpitel CJ and Ye Y. Trends in alcohol- and drug-related ED and primary care visits: data from three US National Surveys (19952005). Am J Drug Alcohol Abuse. 2008;34(5):576-583. 2. Center for Disease Control. Vital Signs: Overdoses of Prescription Opioid Pain Relievers and Other Drugs Among Women-United States, 1999–2010. MMWR Morb Mortal Wkly Rep. 2013;62(26):529-532. 3. Center for Disease Control. Emergency Department Visits Involving Nonmedical Use of Selected Prescription Drugs. MMWR Morb Mortal Wkly Rep. 2010;59(23):705-709. 4. Mazer-Amirshahi M, Mullins PM, Rasooly I, et al. Rising opioid prescribing in adult U.S. emergency department visits: 2001-2010. Acad Emerg Med. 2014;21(3):236-243. 5. Dischinger PC, Mitchell KA, Kufera JA, et al. A Longitudinal Study of Former Trauma Center Patients : Subsequent Injury Mortality. J Trauma. 1995;51(5):877-886. 6. Bogenschutz M and Donovan DM. Screening Motivational Assessment and Referral to Treatment in Emergency Departments (SMART-ED): Evaluation of Screening, Brief Intervention, Referral to Treatment (SBIRT) and Booster Session for Drug Use Patients Presenting for Treatment in the Emergency. Available at: http:// ctndisseminationlibrary.org/protocols/ctn0047.htm. Accessed Aug 1, 2014. 7. Green C. Gender and use of substance abuse treatment services.

CONCLUSION The DAWN database suggests that nationwide, women are less likely to present to the ED and receive detox referrals for illicit drug use or nonmedical use of prescription opioids compared to men. Future research is needed to determine the cause for this disparity, including more in-depth investigation of patient- and physician-level factors leading to referral, its significance to clinical outcomes, and whether increasing referrals to detox programs for women in the ED may improve substance-abuse treatment use and other patient-centered patient outcomes.

Alcohol Res Health. 2006;29(1):55-62. Available at: http://www. pubmedcentral.nih.gov/articlerender.fcgi?artid=3091618&tool=pmcen trez&rendertype=abstract. 8. McCance-Katz EF, Carroll KM, Rounsaville BJ. Gender differences in treatment-seeking cocaine abusers--implications for treatment and prognosis. Am J Addict. 1999;8(4):300-311. 9. Allegri F, Belvederi Murri M, Paparelli A, et al. Current cannabis use and age of psychosis onset: a gender-mediated relationship? Results from an 8-year FEP incidence study in Bologna. Psychiatry Res. 2013;210(1):368-370. 10. Beasley GM, Ostbye T, Muhlbaier LH, et al. Age and gender differences in substance screening may underestimate injury severity: a study of 9793 patients at level 1 trauma center from 2006

Address for Correspondence: Hyeon-Ju Ryoo, BA, 222 Richmond Street, Providence, RI 02903 Box G-9999. Email: hyeon_ju_ ryoo@brown.edu.

Western Journal of Emergency Medicine

to 2010. J Surg Res. 2014;188(1):190-7. 11. Greenfield SF, Rosa C, Putnins SI, et al. Gender research in the National Institute on Drug Abuse National Treatment Clinical

300

Volume XVII, no. 3 : May 2016

Ryoo et al.

Gender Differences in Emergency Department Utilization

Trials Network: a summary of findings. Am J Drug Alcohol Abuse.

23. Otiniano Verissimo AD, Grella CE, Amaro H, et al. Discrimination

2011;37(5):301-312.

and Substance Use Disorders Among Latinos: The Role of Gender,

12. Grella CE. From generic to gender-responsive treatment:

Nativity, and Ethnicity. Am J Public Health. 2014;104(8):1421-8.

changes in social policies, treatment services, and outcomes of

24. Choo EK, Douriez C, Green T. Gender and Prescription Opioid

women in substance abuse treatment. J Psychoactive Drugs.

Misuse in the Emergency Department. Acad Emerg Med.

2008;Suppl 5:327-343. Available at: http://www.ncbi.nlm.nih.gov/

2014;21(12):1493-1498.

pubmed/19256044. Accessed Jul 4, 2014.

25. U.S. Department of Health and Human Services Substance Abuse

13. Brady KT and Randall CL. Gender differences in substance use

and Mental Health Services. Results from the 2012 NSDUH:

disorders. Psychiatr Clin North Am. 1999;22(2):241-252. Available at:

Summary of National Findings, SAMHSA, CBHSQ. Available at:

http://www.ncbi.nlm.nih.gov/pubmed/10385931. Accessed Jan 24,

http://www.samhsa.gov/data/nsduh/2012summnatfinddettables/

2012.

nationalfindings/nsduhresults2012.htm#ch2. Accessed Aug 1, 2014.

14. Choo EK, Beauchamp G, Beaudoin FL, et al. A Research Agenda for

26. Beckman LJ. Treatment Needs for Women With Alcohol Problems.

Gender and Substance Use Disorders in the Emergency Department. Acad Emerg Med. 2014;21(12):1438-1446.

Alcohol Heal Res World. 1994;18(3):206-211. 27. Schober R and Annis HM. Barriers to Help-Seeking for Change in

15. SAMHSA. Drug Abuse Warning Network (DAWN): Data, Outcomes,

Drinking: A Gender-Focused Review of the Literature. Addict Behav.

and Quality. Substance Abuse and Mental Health Services

1996;21(I):81-92.

Administration. 2011.

28. Rockett IRH, Putnam SL, Jia H, et al. Unmet substance

16. Center for Behavioral Health Statistics and Quality. Drug Abuse

abuse treatment need, health services utilization, and cost: A

Warning Network Methodology Report, 2011 Update. Rockville, MD;

population-based emergency department study. Ann Emerg Med.

2013. Available at: http://store.samhsa.gov.

2005;45(2):118-127.

17. Mertens JR and Weisner CM. Predictors of substance abuse

29. Guerrero EG, Marsh JC, Cao D, et al. Gender disparities in utilization

treatment retention among women and men in an HMO. Alcohol Clin

and outcome of comprehensive substance abuse treatment among

Exp Res. 2000;24(10):1525-1533.

racial/ethnic groups. J Subst Abuse Treat. 2014;46(5):584-591.

18. Almog YJ, Anglin MD, Fisher DG. Alcohol and heroin use patterns of

30. Marsh JC, Cao D, Dâ&#x20AC;&#x2122;Aunno T. Gender differences in the impact of

narcotics addicts: gender and ethnic differences. Am J Drug Alcohol

comprehensive services in substance abuse treatment. J Subst

Abuse. 1993;19(2):219-238.

Abuse Treat. 2004;27(4):289-300.

19. Alvanzo A a H, Storr CL, Mojtabai R, et al. Gender and race/ethnicity

31. McHugh RK, Devito EE, Dodd D, et al. Gender differences in a

differences for initiation of alcohol-related service use among persons

clinical trial for prescription opioid dependence. J Subst Abuse Treat.

with alcohol dependence. Drug Alcohol Depend. 2014;140:48-55.

2013;45(1):38-43.

20. Callaghan RC and Cunningham JA. Gender differences in

32. Greenfield SF, Brooks AJ, Gordon SM, et al. Substance abuse

detoxification: predictors of completion and re-admission. J Subst

treatment entry, retention, and outcome in women: a review of the

Abuse Treat. 2002;23(4):399-407.

literature. Drug Alcohol Depend. 2007;86(1):1-21.

21. Dawson DA. Gender differences in the probability of alcohol

33. Walitzer KS and Dearing RL. Gender differences in alcohol and

treatment. J Subst Abuse. 1996;8(2):211-225.

substance use relapse. Clin Psychol Rev. 2006;26(2):128-148.

22. Miller NS and Greenfeld A. Patient characteristics and risks factors

34. Grella CE, Scott CK, Foss MA, et al. Gender similarities and

for development of dependence on hydrocodone and oxycodone. Am

differences in the treatment, relapse, and recovery cycle. Eval Rev.

J Ther. 2004;11(1):26-32.

2008;32(1):113-137.

Volume XVII, no. 3 : May 2016

301

Western Journal of Emergency Medicine

Original Research

Variations in Substance Use Prevalence Estimates and Need for Interventions among Adult Emergency Department Patients Based on Different Screening Strategies Using the ASSIST Roland C. Merchant, MD, MPH, ScD*† Tao Liu, PhD‡ Janette R. Baird, PhD*

*Brown University, Alpert Medical School, Department of Emergency Medicine, Providence, Rhode Island † Brown University, School of Public Health, Department of Epidemiology, Providence, Rhode Island ‡ Brown University, School of Public Health, Center for Statistical Sciences, Department of Biostatistics, Providence, Rhode Island

Section Editor: Chadd K. Kraus, DO, DrPH, MPH Submission history: Submitted January 8, 2016; Revision received March 3, 2016; Accepted March 11, 2016 Electronically published May 10, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29723

Introduction: Among adult emergency department (ED) patients, we sought to examine how estimates of substance use prevalence and the need for interventions can differ, based on the type of screening and assessment strategies employed. Methods: We estimated the prevalence of substance use and the need for interventions using the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) in a secondary analysis of data from two cross-sectional studies using random samples of English- or Spanish-speaking 18-64-year-old ED patients. In addition, the test performance characteristics of three simplified screening strategies consisting of selected questions from the ASSIST (lifetime use, past threemonth use, and past three-month frequency of use) to identify patients in need of a possible intervention were compared against using the full ASSIST. Results: Of 6,432 adult ED patients, the median age was 37 years-old, 56.6% were female, and 61.6% were white. Estimated substance use prevalence among this population differed by how it was measured (lifetime use, past three-month use, past three-month frequency of use, or need for interventions). As compared to using the full ASSIST, the predictive value and accuracy to identify patients in need of any intervention was best for a simplified strategy asking about past three-month substance use. A strategy asking about daily/near-daily use was better in identifying patients needing intensive interventions. However, some patients needing interventions were missed when using these simplified strategies. Conclusion: Substance use prevalence estimates and identification of ED patients needing interventions differ by screening strategies used. EDs should carefully select strategies to identify patients in need of substance use interventions. [West J Emerg Med. 2016;17(3):302–314.]

INTRODUCTION Recent research indicates high prevalences of alcohol, smoking and drug use among United States (U.S.) emergency department (ED) adult patients.1-6 However, estimated ED patient substance-use prevalence and anticipated need for Western Journal of Emergency Medicine

interventions likely are impacted by the screening strategies used to measure them, such as who is screened; where, when, how and by whom screening is conducted; the simplicity or complexity of these strategies; the types of screening instruments used; and the domains these instruments measure. 302

Volume XVII, no. 3 : May 2016

Merchant et al.

Emergency Department Patient Substance Use Screening

Variability across screening strategies may lead to disparate or contradictory recommendations for addressing substance use among ED patients, which in turn could have consequent deleterious effects on ED and public health priorities chosen for substance use prevention, treatment efforts and the allocation of extramural funding to evaluate interventions. Although ED-based screening and initiation of consequent interventions are recommended for alcohol and smoking,7-9 this practice is not yet routine.10-18 Comprehensive assessments of substance use and intervention need can be a time-consuming process, which can discourage their use. An enticing way to assess ED patients substance use is to use simplified screening-question strategies involving one or two screening questions that might identify individuals in possible need of an intervention (e.g., “In the past 3 months, did you drink alcohol?”).3,4 This strategy might be applied by ED staff at triage or incorporated into the electronic medical record (EMR), and the results could prompt or obviate the need for an intervention during the clinical care encounter. Before such simplified strategies can be recommended, their yield and accuracy against relatively more comprehensive screening and assessment strategies need to be evaluated. Our primary aim was to examine how estimates about the prevalence of substance use among adult ED patients differ when these estimates are based on the following: any lifetime, any past three-month, or past three-month frequency of substance use; or according to the need for any, a brief, or an intensive intervention per the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST).19 Our secondary goal was to compare the ability of three simplified screening-question strategies to identify adult ED patients in need of any intervention vs. no intervention, and in need of an intensive intervention vs. no intensive intervention (i.e., no intervention needed or only a brief intervention [BI] needed), to the full ASSIST as the “gold standard.” The three simplified strategies constituted screening based on any lifetime use, any past three-month use, or past three-month frequency of use. METHODS Study design and setting This investigation was a secondary analysis of two concurrent studies on substance use at two EDs affiliated with a medical school in the same hospital system and city from July 2010 to December 2012. The data were based on two cross-sectional studies that involved surveying random samples of adult ED patients. The hospital institutional review board approved the study. Selection of participants For both studies, bilingual (English- and Spanishspeaking) research assistants (RAs) randomly selected ED patients for possible study inclusion, reviewed their EMR for exclusion criteria, and confirmed study eligibility through a brief interview. A random sample of patients present in the ED Volume XVII, no. 3 : May 2016

during study collection periods was approached and evaluated for study inclusion through random selection to their patient care rooms. If the ED EMR indicated that a patient potentially was study eligible, a RA would confirm study eligibility through a brief interview. Data collection for the study was performed from 8 AM to midnight seven days/week when bilingual (English- and Spanish-speaking) RAs were available to conduct the study. Patients were study eligible if they were 18-64 years-old; English- or Spanish-speaking; not critically ill or injured; not prison inmates, under arrest, nor undergoing home confinement; not presenting for an acute psychiatric illness; not requesting treatment for substance use; not intoxicated; and did not have a physical or mental impairment that prevented them from providing consent or participating in the study. The study population aimed to reflect the general adult ED population that would be included in a screening, brief intervention, and referral to treatment (SBIRT) program, i.e., excluding those presenting for evaluation of their substance use, those acutely intoxicated, and those undergoing a formal substance use or psychiatric evaluation. Methods and measurements Participants completed the ASSIST, which we adapted for these studies. (See supplemental material for an Englishlanguage copy of the study instrument.)19 The cognitive-based assessments, pilot testing and evaluations of the adapted ASSIST have been described in detail previously.20 In brief, this adaption involved preparing it for audio computer selfadministered interviewing (ACASI) to increase veracity of responses of sensitive or stigmatizing information (i.e., substance use/misuse)21-25 and improve the flow of the instrument; clarifying questions, responses, and instructions; distinguishing misuse from the use of prescription drugs; and adding or expanding drug categories (e.g., barbiturates, benzodiazepines, prescription opioid analgesics). Cronbach’s α ranged from 0.86 to 0.95 for the drug categories assessed in our adapted ASSIST. As shown in Figure 1, participants first were asked in the ASSIST about any lifetime substance use by substance category, and if they indicated lifetime use of a given substance, they were asked about any past three-month use and frequency of use during the past three months (ASSIST Questions 1 and 2). Following these initial questions, the ASSIST proceeded through an evaluation of substance-use severity. Substancespecific scores are calculated for those who have used a given substance; if an individual has not used that substance, no score is calculated. Per World Health Organization (WHO) recommendations, an ASSIST score of ≥4 points for smoking or any drug category or a score of ≥11 points for alcohol suggests a need for BI, and a score of ≥27 points for any substance suggests a need for a more intensive intervention.19 Those who report use in the past three months are assessed for the need of a BI or an intensive intervention, while those who report ever

303

Western Journal of Emergency Medicine

Merchant et al.

Emergency Department Patient Substance Use Screening ASSIST Question 1 Lifetime use Q1: “At any time in your life have you used [substance]?”

If no to any substance use, screening ends ASSIST Question 2 Past 3 month use and frequency of use of substances reported for lifetime use Q2: “In the past 3 months, how often have you used [substance]?”

ASSIST Questions 6 & 7 Reported lifetime, but no past 3 month use of a substance Q6: “Has a friend, relative or anyone else ever said they were concerned about your use of [substance]?” Q7: “Have you ever tried to control, cut down or stop using [substance]?”

Used in the past 3 months

Not used in the past 3 months

ASSIST score calculation

No intervention

ASSIST Questions 3, 4, 5, 6A, & 7A Reported lifetime and past 3 month use of a substance Q3: “In the past 3 months, how often have you had a strong urge or craving to use tobacco [substance]?” Q4: “In the past 3 months, how often has your use of [substance] caused health, legal or financial problems or problems with friends, relatives, coworkers or any other person?” Q5: “In the past 3 month, how often have you failed to do the things that are usually expected of you because of your use of [substance]?” Q6: “Has a friend, relative or anyone else ever said they were concerned about your use of [substance]?” Q6A: “In the past 3 months, has a friend, relative or anyone else ever said they were concerned about your use of [substance]?” Q7: “Have you ever tried to control, cut down or stop using [substance]?” Q7A: “In the past 3 months, have you ever tried to control, cut down or stop using [substance]?”

ASSIST score calculation

No intervention

Brief intervention

Intensive intervention Brief intervention

Figure 1. ASSIST screening and substance misuse intervention algorithm. ASSIST, Alcohol, Smoking and Substance Involvement Screening Test

using a given substance (lifetime use) but deny past 3-month use are assessed for the need for a BI only. In addition to the ASSIST, we queried participants about the specific drugs that they had used within the past three months and whether or not these drugs had been injected or prescribed. (See supplemental material [Appendix A].) The reading level of study questionnaires in English was at a FleschKincaid grade level of 6.6 (Microsoft Word; Microsoft Corp., Redmond, WA) and in Spanish was at a Huerta Reading Ease score of 80, both indicating an easy reading level.26 Participants completed the questionnaires in approximately 10-15 minutes. The RAs received 40 hours of training on the study protocol, including mock interviews with the study investigators and pilot testing of the study protocol prior to collecting data for the study. The RAs met with the study investigators throughout the study to discuss procedural issues arising from the conduct of the study. To ensure fidelity to Western Journal of Emergency Medicine

the study protocol, study investigators directly observed RAs during participant encounters. Deviations from the study protocol were addressed and suggestions for improvement were provided. Analysis Analysis of the study and presentation of study findings followed current Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations for cross-sectional studies (www.strobe-statement.org). We summarized study eligibility assessments and enrollment using current recommendations,27 participant demographic characteristics, and responses to the ASSIST. ASSIST scores were calculated for each participant. The need for any, a brief, or more intensive intervention was calculated according to WHO recommendations for all participants, and as stratified by those reporting no past three-month or any past three-

304

Volume XVII, no. 3 : May 2016

Merchant et al.

Emergency Department Patient Substance Use Screening

month use of a given substance.19 For the primary aim, substance use prevalence was estimated for each substance use category (e.g., smoking, alcohol, benzodiazepines) as stratified by responses to Questions 1 and 2 of the ASSIST (any lifetime use, any past three-month use, and past three-month frequency of use), and by the need for interventions using ASSIST scores. Substance use was ranked in order of decreasing magnitude of prevalence or frequency, respectively, across all substance categories according to these strata. We estimated accompanying 95% confidence intervals (CIs) for the rankings. The ranks were provided to assist in distinguishing differences among proportions within each category. For the secondary aim, we calculated the test performance characteristics for the ability of the three simplified screening question strategies to identify individuals in need of any intervention vs. no intervention, and in need of an intensive intervention vs. no intensive intervention (i.e., no intervention needed or only a BI needed), as compared to the full ASSIST as the â&#x20AC;&#x153;gold standard.â&#x20AC;? The three simplified strategies constituted screening based on any lifetime use, any past three-month use,

or past three-month frequency of use (per ASSIST Questions 1 and 2). Sensitivity, specificity, negative and positive predictive values, and accuracy with corresponding 95% CIs were estimated. We performed all analyses using STATA 13 (Stata Corporation, College Station, TX). RESULTS Participant eligibility assessment, enrollment, and demographic characteristics Participant eligibility assessment and enrollment results are depicted in Figure 2. As shown, of the 9,813 randomly selected 18- to 64-year-old English- or Spanish-speaking ED patients, 78% (7,643) were assessed in person for study eligibility. Of these, 97% (7,409) were eligible to complete the ASSIST, and 87% (6,432) completed it and comprised the final study sample. The majority of participants were female, white/non-Hispanic, most had 12 or fewer years of formal education, had private healthcare insurance, were married or part of an unmarried couple, were employed, not homeless, and received their medical care from a private clinic/practice (Table 1).

18 to 64 year-old English- or Spanish-speaking randomly selected ED patients n=9813 Ineligible for study n=234 (2.4%) Reasons: Physically disabled Mentally disabled Previously in study Not English or Spanish-speaking HIV/HBV/HCV Infected Intoxicated Prisoner/Home confinement

Refused or dropped out during in-person eligibility assessment n=2170 (22.1%) Reasons:

% 38.5 20.1 12.8 11.5 9.8 6.8 0.4

Completed in-person eligibility assessment n=7643 (77.9%)

I am in too much pain I am feeling sick I am leaving the ED I do not like being in surveys I am too tired I am uncomfortable with the subject I have no time My family is present Other reason No reason provided

% 26.8 18.8 16.9 9.2 6.4 2.8 1.5

1.6 10.6 5.5

*Percentages will not total 100% since patients could be ineligible for multiple reasons

4.3

Eligible for ASSIST n=7409 (96.9%) Declined to take the ASSIST n=667 (9.0%)

Dropped out during ASSIST n=310 (4.2%)

Reasons: I am leaving the ED I am in too much pain I am feeling sick I do not like being in surveys I am too tired I am uncomfortable with the subject I have no time My family/friends are present Other reason No reason provided

Reasons:

% 26.8 11.6 10.3 6.8 5.2 3.5 1.0 1.0 13.5 20.3

Completed ASSIST n=6432 (86.8%)

I am in too much pain I am feeling sick I do not like being in surveys I am too tired I have no time I am leaving the ED I am uncomfortable with the subject My family/friends are present Other reason No reason provided

% 20.2 19.5 12.7 9.7 8.4 7.5 4.5 1.6 12.9 2.8

Figure 2. Eligibility assessment and enrollment. ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; HBV, Hepatitis B Virus; HCV, Hepatitis C Virus; HIV, human immunodeficiency virus; ED, emergency department

Volume XVII, no. 3 : May 2016

305

Western Journal of Emergency Medicine

Merchant et al.

Emergency Department Patient Substance Use Screening In regards to specific substances used among these ED patients, the highest past three-month use prevalences for illicit drugs were cocaine (4.4%), crack (2.6%), and ecstasy/3,4-methylenedioxy-N-methylamphetamine (MDMA) (1.0%) (Supplemental Table 1). Although among all participants the highest past three-month use prevalences for

Table 1. Demographic characteristics of participants. Demographic characteristics Median age, years (IQR) Gender Female Male Ethnicity/race White, non-Hispanic White, Hispanic Black/African-American, non-Hispanic Black/African-American, Hispanic Other Years of formal education <12 years Grade 12 College 1-3 years College 4 years (college graduate)/ >College Health insurance status Private Governmental None Don’t know/refuse to answer Partner status Married Divorced/widowed/separated Never married Unmarried couple Homeless status Currently homeless Past 12 months homeless Never/not homeless past 12 months Employment status Employed Disability Student Unemployed Don’t know/refuse to answer Usual source of medical care Private clinic/practice Hospital or community health clinics Emergency department Urgent care center Don’t know/refuse to answer IQR, interquartile range

Western Journal of Emergency Medicine

n=6432 37 (26-48) % 56.6 43.4 61.6 11.2 17.3 6.5 3.3 25.6 29.8 26.9 17.6 40.8 34.0 25.1 0.1 27.9 18.0 38.8 15.4 5.5 3.1 91.4 48.8 18.0 8.3 24.8 0.1 46.4 26.2 24.8 2.5 0.2

prescription opioids were acetaminophen and hydrocodone (4.0%), acetaminophen and oxycodone (3.6%), and oxycodone (1.1%), only 2.2% of all participants stated that had been prescribed acetaminophen and hydrocodone, 1.9% acetaminophen and oxycodone, and 0.7% oxycodone within the past three months. Lifetime prevalence of injectiondrug use was 4.4% and was 1.7% for past three-month use, predominately heroin (0.93%) and cocaine (0.56%). Differences in substance use prevalence estimates among ED patients When estimating the prevalence of substance use based on lifetime use or past 3-month use, the rank order was the same for the first three substances (alcohol, smoking, and then marijuana) (Table 2). When estimating substance use prevalence according to daily/near-daily use, smoking had the highest rank, followed by marijuana, alcohol, and then prescription opioids. There were subtle differences in the ranks for the remaining drugs across these three ways of estimating the prevalence of substance use. When estimating substance use prevalence based on the need for interventions (Table 3a and 3b), the need for any intervention or at least a BI among these patients was greatest for smoking, marijuana, and then alcohol; however, the need for an intensive intervention was greatest for smoking, alcohol, and then marijuana. The need for any, a brief, or an intensive intervention differed slightly for the remaining substances. The relative rank order of the need for a BI also differed when ASSIST score results were stratified by lifetime only vs. past three-month use. Among those who reported lifetime-only use, the need for a BI was greatest for opioids, methadone or buprenorphine, prescription opioid analgesics, barbiturates and then cocaine or crack. However, among those who reported past three-month substance use, the need for a BI was greatest for smoking, marijuana, amphetamines, prescription opioid analgesics, and then methamphetamines. The need for an intensive intervention based on ASSIST scores among those who reported past three-month substance use was greatest for opioids, gamma hydroxybutyrate (GHB), methadone or buprenorphine, cocaine or crack, and then barbiturates. The relative need for an intervention differed for specific substances when comparing past or no past threemonth use of these substances. For example, the need for a BI for illicit opioids was greater among those who reported no past three-month use of this substance; whereas for smoking, the need for a BI was greater for those who reported smoking in the past three months. Performance of simplified screening question strategies in identifying need for substance use interventions The performance of using simplified screening question strategies (any lifetime use, any past three-month use, or daily/ near-daily use) as compared to the full ASSIST as the “gold 306

Volume XVII, no. 3 : May 2016

Volume XVII, no. 3 : May 2016 90.1 94.5 84.0 91.6 98.4 93.8 90.3

Methamphetamines

Inhalants

Hallucinogens

Illicit opioids

Gamma-hydroxybutyrate (GHB)

Amphetamines

Benzodiazepines 13 (13,14)

6 (6,6)

8 (8,9)

14 (13,14)

7 (7,7)

11 (10,11)

12 (12,12)

9 (8,10)

5 (4,5)

3 (3,3)

1 (1,1)

2 (2,2)

Rank of % (95% CI)

4 (4,5)

0.3

3.6

1.7

0.2

2.3

1.2

0.6

1.6

5.8

26.6

61.6

43.7

10 (9,11)

12 (12,12)

14 (13,14)

8 (7,8)

10 (10,11)

13 (13,14)

9 (9,9)

5 (5,5)

11 (10,11)

7 (7,8)

4 (4,4)

3 (3,3)

1 (1,1)

2 (2,2)

Rank of % (95% CI)

Past 3-month use

1.3

3.2

1.5

9.7

6.2

1.6

8.4

16.0

5.5

9.9

24.5

60.2

86.9

66.3

Lifetime use

Prescription opioid analgesics 85.1 14.9 6 (6,6) 6.3 n.b. Rows do not total to 100% because participants could refuse to answer.

96.8

75.5

Cocaine or crack

Methadone or buprenorphine

39.8

Marijuana

98.5

13.1

Alcoholic beverages

Barbiturates

33.7

Smoking

Never used or misused

Any lifetime or any past 3-month use

307 1.2

0.3

0.1

0.7

0.2

0.1

0.7

0.1

0.2

1.0

9.5

6.5

32.8

4 (4,5)

8 (8,9)

14 (12,14)

7 (5,7)

10 (9,13)

13 (12,14)

6 (6,7)

12 (9,12)

11 (9,13)

9 (9,12)

5 (4,6)

2 (2,2)

3 (2,3)

1 (1,1)

Rank of % (95% CI)

5 to 7 days a week

1.8

0.3

0.1

0.9

0.4

0.0

0.6

0.1

0.2

1.0

6.9

19.3

5.3

1 to 4 days a week

1.6

0.3

0.0

0.8

0.4

0.0

0.4

0.3

0.1

0.4

1.7

4.5

18.5

2.5

1 to 2 days a month

1.7

0.3

0.1

1.2

0.7

0.1

0.6

0.8

0.3

0.8

2.2

5.7

17.3

3.1

1 to 2 days in the past 3 months

Past 3-month frequency of use

Table 2. Extent of substance use as assessed by any lifetime use, any past 3-month use, or past 3-month frequency of use (n=6432).

93.1

97.9

99.2

95.9

97.6

99.1

97.2

98.1

99.1

98.0

93.7

72.9

38.0

56.1

Not used in the past 3 months

Merchant et al. Emergency Department Patient Substance Use Screening

Western Journal of Emergency Medicine

308

0.2

1.5

3.6

0.4

1.5

6.8

GHB

Amphetamines

Benzodiazepines

Barbiturates

Methadone or buprenorphine

Prescription opioid analgesics 5 (4,5)

9 (8,11)

13(12,13)

6 (6,7)

10 (8,11)

14(14,14)

7 (6,7)

11 (9,11)

12 (12,13)

8 (8,11)

4 (4,5)

2 (2,2)

3 (3,3)

1 (1,1)

Rank of % (95% CI)

5.0

1.1

0.3

2.6

1.2

0.1

2.3

1.2

0.4

1.4

5.2

19.5

11.6

34.8

5 (4,5)

11 (8,11)

12 (12,13)

6 (6,7)

9 (8,11)

14 (14,14)

7 (6,7)

10 (8,11)

13 (12,13)

8 (8,10)

4 (4,5)

2 (2,2)

3 (3,3)

1 (1,1)

Rank of % (95% CI)

Brief intervention

Entire population (n=6432)

1.8

0.5

0.1

1.0

0.3

0.1

1.1

0.2

0.1

0.2

1.9

3.1

4.6

9.1

5 (4,5)

8 (8,8)

13(12,14)

7 (6,7)

9 (9,11)

14 (13,4)

6 (6,7)

11 (9,12)

12 (10,14)

10 (9,12)

4 (4,5)

3 (3,3)

2 (2,2)

1 (1,1)

Rank of % (95% CI)

Intensive intervention

550

121

394

288

390

950

315

535

1203

2162

1628

1449

16.4

19.8

16.3

10.4

5.6

6.5

20.8

3.7

1.6

5.6

12.1

3.5

0.0

10.4

3 (2,4)

2 (1,4)

4 (1,7)

6 (4,8)

10 (8,12)

8 (5,12)

1 (1,3)

11 (10,12)

13 (12,13)

9 (8,12)

5 (4,7)

12 (10,12)

14 (14,14)

7 (5,8)

Rank of % (95% CI)

Brief intervention

No past 3-month substance use

WHO recommendations for interventions

407

232

109

150

101

372

1708

3959

2812

56.5

56.1

43.8

53.0

56.9

30.0

42.7

48.8

46.3

56.4

49.7

69.1

18.8

74.3

4 (3,8)

6 (3,11)

11 (2,13)

7 (4,10)

3 (3,10)

13 (4,14)

12 (8,13)

9 (4,12)

10 (3,13)

5 (3,10)

8 (6,11)

2 (2,3)

14 (13,14)

1 (1,1)

Rank of % (95% CI)

Brief intervention

29.0

35.4

31.3

28.0

15.6

40.0

48.0

13.8

19.5

12.9

33.1

11.6

7.6

13.7

6 (3,7)

3 (2,7)

5 (1,13)

7 (3,8)

9 (8,13)

2 (1,12)

1 (1,3)

10 (9,14)

8 (4,13)

12 (9,14)

4 (2,6)

13 (9,13)

14 (13,14)

11 (6,10)

Rank of % (95% CI)

Intensive intervention

Any past 3-month substance use

x̅ , mean; SE, standard error; ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; WHO, World Health Organization; GHB, gamma-hydroxybutyrate

53.8

3.4

Illicit opioids

Total ASSIST score

1.3

Hallucinogens

28.6

0.5

Inhalants

Total ASSIST drug score

1.6

Methamphetamines

22.6

Marijuana

7.1

16.2

Alcoholic beverages

Cocaine or crack

43.9

Any intervention

Smoking

Substances

Table 3a. Extent of substance use according to the need for substance misuse interventions per WHO recommendations based on ASSIST scores.

Emergency Department Patient Substance Use Screening Merchant et al.

Volume XVII, no. 3 : May 2016

Merchant et al.

Emergency Department Patient Substance Use Screening was better for intensive intervention need.

Table 3b. Continued. Substances

ASSIST score for reporting any lifetime use n

x̅ (SE)

Smoking

4261

13.3 (0.17)

Alcoholic beverages

5587

6.5 (0.11)

Marijuana

3870

5.9 (0.14)

Cocaine or crack

1575

5.4 (0.24)

Methamphetamines

636

2.6 (0.24)

Inhalants

356

1.9 (0.30)

Hallucinogens

1030

1.4 (0.13)

Illicit opioids

540

7.8 (0.50)

GHB

103

2.5 (0.71)

Amphetamines

397

4.0 (0.37)

Benzodiazepines

626

7.1 (0.43)

Barbiturates Methadone or buprenorphine Prescription opioid analgesics

4.3 (0.88)

203

9.2 (0.82)

957

8.6 (0.36)

Total ASSIST drug score

4146

13.6 (0.44)

Total ASSIST score

5898

25.3 (0.45)

x̅ , mean; SE, Standard Error; ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; GHB, gammahydroxybutyrate

standard” to identify need for WHO-recommended substance use interventions is shown in Table 4a-4d. Across all substances, sensitivity decreased while specificity increased when using querying about lifetime use vs. past three-month vs. daily/near-daily use. Also across substances, querying about past three-month use generally performed better in regards to predictive value and accuracy for any intervention need, but querying about daily/near-daily use was generally better for identifying intensive intervention need. The test performance characteristics of these three simplified screening strategies varied by substance category. For any smoking intervention, predictive values and accuracy of querying about past three-month use was higher than screening for lifetime alone or frequency of use (daily/neardaily use). However, frequency of use performed better for identifying need for smoking intensive interventions, although the positive predictive value was low. For any or alcohol intensive interventions, querying about frequency of use performed better than the other screening strategies. For any marijuana interventions, querying about past three-month use performed better, but querying about frequency of use was better for intensive interventions. For all other drugs, querying about past three-month use was better than other queries for any intervention need, yet querying about frequency of use Volume XVII, no. 3 : May 2016

DISCUSSION Our investigation demonstrates how estimated substance use prevalence among ED patients can differ depending on how it is derived. For example, prescription opioid analgesics is only sixth when estimating its prevalence based on lifetime use, but rises to fourth in prevalence when considering past three-month use, daily/near-daily use, or the need for a BI based on past three-month use. Likewise, smoking becomes the number one substance use problem when considering the need for interventions, rather than measuring lifetime or past threemonth use prevalence. Subtle small differences in estimates such as these can have large public health implications on which substances are prioritized for interventions in United States EDs. This investigation adds a unique perspective to the small number of large sample studies estimating the extent of adult ED patient substance use. The results from these prior studies also illustrate how differences in who is screened, how they are screened, and how prevalence is estimated can change perspectives on the extent of substance use in this setting. Wu, et al. conducted a secondary analysis of past year prevalence of alcohol and drug use disorders using ACASI-collected data from persons ≥18 years-old in the U.S. who participated in the 2007-2009 National Survey on Drug Use and Health and who had visited an ED at least once within the past year.2 Approximately 66% of those who had visited an ED drank alcohol, 12% had used at least one of nine different drugs within the past year, 10% met criteria for alcohol abuse or dependence and 4% for drug abuse or dependence per Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria. Wu, et al. observed that the prevalence of substance use disorders was greatest for heroin, sedatives, cocaine, and then opioids. As noted, Wu, et al. reported a lower prevalence of drug use and markedly different order of drugs of concern than for our study. These differences likely are due to the study methodology, drug categories used, their inability to measure recency of substance use, and use of DSM-IV criteria for abuse and dependence. Four recent U.S. ED-based studies screened patients concurrently with their ED visit, as we did in our investigation. Two employed a brief screening strategy administered aloud by nurses at ED triage. From 2009 to 2012 among ≥18-year-old ED patients in Macon, Georgia, Johnson, et al. evaluated three one-question screening tools for past year tobacco product use, alcohol use (≥4 drinks in a day for women, ≥5 drinks in a day for men), and drug use (“pot [marijuana], use of another street drug, or use of a prescription painkiller, stimulant or sedative for a non-medical reason”), which had been adapted from primary care settings.3 Approximately 22% screened positive for at-risk alcohol or drug use. Hankin, et al. from 2009 to 2010 employed a screening approach similar to that by Johnson, et al. among a predominately Black, non-Hispanic population at the Grady 309

Western Journal of Emergency Medicine

Merchant et al.

Emergency Department Patient Substance Use Screening

Table 4a. Test performance characteristics of simplified screening strategies for substance misuse interventions as compared to the full ASSIST. Need for any intervention vs. no intervention Marijuana

Any other drug

Lifetime use only

Past 3-month use

Daily to near daily use

Lifetime use only

Past 3-month use

Daily to near daily use

% (95% CI)

Sensitivity

100

94.8 (93.6, 95.9)

42.0 (39.4, 44.6)

100

78.2 (75.3, 80.9)

22.5 (19.8, 25.5)

Specificity

51.1 (49.7, 52.5)

93.3 (92.6, 94.0)

100

71.0 (69.8, 72.2)

97.3 (96.9, 97.7)

100

PPV

37.5 (36.0, 39.1)

80.7 (78.7, 82.5)

100

35.3 (33.4,37.2)

82.2 (79.4,84.7)

100

NPV

100

98.4 (98.0, 98.7)

85.4 (84.5, 86.3)

100

96.6 (96.1, 97.0)

89.1 (88.3, 89.9)

62.2 (61.0, 63.4)

93.7 (93.1, 94.3)

86.8 (86.0, 87.6)

75.0 (73.9, 76.1)

94.7 (94.2, 95.3)

89.4 (88.7, 90.2)

Accuracy

ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; PPV, positive predictive value; NPV, negative predictive value

Table 4b. Continued. Need for any intervention vs. no intervention Smoking

Alcohol

Lifetime use only

Past 3-month use

Daily to near daily use

Lifetime use only

Past 3-month use

Daily to near daily use

% (95% CI)

Sensitivity

100

94.7 (93.8,95.5)

74.7 (73.1,76.3)

100

30.9 (28.1, 33.8)

Specificity

60.0 (58.4,61.6)

96.2 (95.5,96.8)

100

15.3 (14.3, 16.2)

45.6 (44.3, 47.0)

98.2 (97.8, 98.5)

PPV

66.3 (64.9,67.7)

95.1 (94.2,95.9)

100

18.7 (17.7, 19.7)

26.3 (25.0, 27.7)

77.0 (72.7, 81.0)

NPV

100

95.8 (95.1,96.4)

83.4 (82.3,84.5)

100

88.0 (87.1, 88.8)

77.6 (79.6,78.6)

95.5 (95.0,96.0)

88.9 (88.1,89.6)

29.1 (27.9, 30.2)

54.5 (53.3, 55.7)

87.2 (86.4, 88.1)

Accuracy

ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; PPV, positive predictive value; NPV, negative predictive value

Western Journal of Emergency Medicine

310

Volume XVII, no. 3 : May 2016

Merchant et al.

Emergency Department Patient Substance Use Screening

Table 4c. Continued. Need for an intensive intervention vs. no intensive intervention (no intervention or a brief intervention) Smoking

Alcohol

Lifetime use only

Past 3-month use

Daily to near daily use

Lifetime use only

Past 3-month use

Daily to near daily use

% (95% CI)

94.7 53.8 100 100 (92.6, 96.4) (48.0, 59.6) 37.0 61.8 73.3 13.4 40.1 95.8 Specificity (35.7,38.2) (60.6, 63.1) (72.2, 74.4) (12.6, 14.3) (38.8, 41.3) (95.3, 96.3) 13.7 20.8 26.2 5.4 7.6 38.5 PPV (12.7,14.8) (19.3, 22.4) (24.4, 28.2) (4.8, 6.0) (6.7, 8.4) (33.8, 43.4) 99.3 97.7 NPV 100 100 100 100 (99.0, 99.5) (97.3, 98.1) 42.7 65.3 75.2 17.4 42.8 93.8 Accuracy (45.1,43.9) (64.1, 66.5) (74.2, 76.3) (16.5, 18.4) (41.6, 44.1) (93.2, 94.4) ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; PPV, positive predictive value; NPV, negative predictive value Sensitivity

100

Table 4d. Continued. Need for an intensive intervention vs. no intensive intervention (no intervention or a brief intervention) Marijuana

Sensitivity

Any other drug

Lifetime use only

Past 3-month use

Daily to near daily use

Lifetime use only

Past 3-month use

Daily to near daily use

% (95% CI)

100

73.2 (66.5, 79.3)

100

40.7 (39.5, 42.0) 5.1 (4.4, 5.9)

75.6 (74.5, 76.7) 11.6 (10.1, 13.2)

51.0 (44.7, 57.3)

92.5 63.9 90.7 98.9 (91.8, 93.1) (62.7, 65.1) (89.9, 91.4) (98.7, 99.2) 23.8 10.5 31.1 67.0 PPV (20.4, 27.4) (9.3, 11.7) (28.0, 34.4) (59.9, 73.6) 99.1 98.0 NPV 100 100 100 100 (98.8, 99.3) (97.6, 98.3) 42.6 76.4 91.8 65.4 51.0 97.0 Accuracy (41.4, 43.8) (75.3, 77.4) (91.2, 92.6) (64.2, 66.6) (44.7, 57.3) (96.6, 97.4) ASSIST, Alcohol, Smoking and Substance Involvement Screening Test; PPV, positive predictive value; NPV, negative predictive value Specificity

Volume XVII, no. 3 : May 2016

311

Western Journal of Emergency Medicine

Merchant et al.

Emergency Department Patient Substance Use Screening Hospital ED in Atlanta, Georgia.4 Patients with a positive screen were administered the ASSIST through an interview with a health educator. Of those screened, 27.6% had a positive one-question screen for alcohol or drug use within the past year; marijuana and cocaine were the most commonly used drugs. The lower yields of screening for these studies as compared to our study might be due to the type of screening instruments used and mode of screening (nurse-administered aloud at ED triage). Employing three different screening instruments at EDs in six states among â&#x2030;Ľ18-year-old Englishspeaking patients, Sanjuan, et al.5 and Konstantopoulos, et al.6 found that 39% reported daily tobacco use, 45% risky alcohol use, 22% any past 30-day drug use, and 17% moderate to severe drug problems. The most frequently reported drugs used were marijuana, cocaine, â&#x20AC;&#x153;street opioidsâ&#x20AC;? and prescription opioids. Although Spanish-speaking ED patients were excluded and different instruments were used, the prevalence of use from these two analyses is relatively similar to ours. In their practical application, our study findings also indicate how screening strategies can affect adult ED patients identified (or missed) as possibly needing an intervention. For example, a simplified screening question strategy that queries about past three-month use to identify adult ED patients needing any intervention would perform well for smoking, marijuana, and other drugs, but would perform poorly for alcohol. Likewise, querying about daily/ near-daily use to identify need for an intensive intervention would perform well for marijuana, other drugs, and alcohol, but less well for smoking. As the ASSIST assessments that stratify by lifetime or past three-month use remind us, remarkably high proportions of ED patients who deny current or recent substance use still qualify for a BI per WHO recommendations. This finding indicates that asking about lifetime use and not just past three-month or frequency of use remains an important facet in identifying those who might benefit from substance use interventions. This stratification also demonstrates that the relative need for an intervention can vary substantially when considering past three-month use, as the example of illicit opioids and smoking illustrate. One implication for ED clinical practice is to recognize that a need for an intervention is based on more than current substance use. Querying about lifetime and current use and frequency of use can help to identify those who need assistance. The challenge for the future is to learn how to screen efficiently yet effectively in EDs to identify those who might benefit from assistance because of their substance use. Unfortunately, we do not yet have a single perfect instrument to screen adult ED patients quickly and easily for substance abuse in a manner that identifies all of those in need of an intervention, without overburdening providers and patients, and minimizes false positive and negative results. One caveat is that effective interventions for adult ED Western Journal of Emergency Medicine

patients to help them reduce or eliminate their substance use continue to elude us. Although BIs have been recommended28 for alcohol among adolescent and adult ED patients, studies evaluating them have had mixed results.29-42 Three different recently published studies also observed the following: (1) no short-term (three-month follow-up) benefit from a BI in reducing drug use or increasing uptake of drug treatment services;43 (2) no differences in past 30day drug use abstinence six months post-enrollment44; and no advantage to a BI with telephone boosters as compared with screening, assessment and referral to treatment or minimal screening only in terms of drug use at three, six, and 12 months post-enrollment.45 As such, although instruments such as the ASSIST might indicate the need for an intervention, the optimal type of intervention needed is not yet known. LIMITATIONS This study had several limitations. Although care was taken to reach a representative sample of adult patients at these two EDs not presenting for a substance use or psychiatric evaluation, those who were excluded might have different substance use profiles and need for interventions. An advantage of this study is that random sampling of patients was used and Spanish-speaking patients were included, which should increase the internal and external validity of the study findings. Of course, patients at EDs in other locales might have a different spectrum of substance use. Although we provide estimates of predictive value and accuracy for the simplified screening strategies, such estimates are based on the prevalence at these EDs, and likely would be different at other EDs. Nevertheless, these parameters provide an illustration of the application of the screening strategies. Even though confidentiality and fidelity of responses likely was enhanced by the use of the ACASI approach,21-25 we cannot be assured that all participants answered the questions truthfully. As would be expected, other screening strategies and substance-use evaluation instruments other than the ASSIST might yield different results. In addition, because we only used questions taken from the ASSIST and not any other instrument, we cannot determine how the three simplified screening strategies would compare to other screening instruments. Comparisons to different instruments might have led to different conclusions. CONCLUSION The results of this investigation indicate that estimated substance use prevalence among adult ED patients differs according to how it is measured. In addition, the yield from simplified screening question strategies can vary by substance, and risk missing patients who might benefit from an intervention. Public policy makers and EDs contemplating substance use screening programs should be cognizant that the

312

Volume XVII, no. 3 : May 2016

Merchant et al.

Emergency Department Patient Substance Use Screening

priorities for what substances need intervention and the types of interventions needed are highly dependent on how adult ED patients are assessed for substance use.

9. American College of Emergency Physicians. Alcohol screening in the emergency department. 2011. July 24, 2014. 10. Rockett IR, Putnam SL, Jia H, et al. Assessing substance abuse treatment need: a statewide hospital emergency department study. Ann Emerg Med. 2003;41(6):802-13. 11. Tong EK, Strouse R, Hall J, et al. National survey of U.S. health professionals’ smoking prevalence, cessation practices, and beliefs.

Address for Correspondence: Roland C. Merchant, MD, MPH, ScD, Rhode Island Hospital, 593 Eddy Street, Claverick Building, Providence, RI 02903. Email: rmerchant@lifespan.org.

Nicotine Tob Res. 2010;12(7):724-33. 12. Greenberg MR, Weinstock M, Fenimore DG, et al. Emergency department tobacco cessation program: staff participation and

intervention success among patients. J Am Osteopath Assoc. 2008;108(8):391-6. 13. Vokes NI, Bailey JM, Rhodes KV. “Should I give you my smoking lecture now or later?” Characterizing emergency physician

Copyright: © 2016 Merchant et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

smoking discussions and cessation counseling. Ann Emerg Med. 2006;48(4):406-14, 14 e1-7. 14. Prochazka A, Koziol-McLain J, Tomlinson D, et al. Smoking cessation counseling by emergency physicians: opinions, knowledge, and training needs. Acad Emerg Med. 1995;2(3):211-6. 15. Cunningham RM, Harrison SR, McKay MP, et al. National survey of emergency department alcohol screening and intervention practices. Ann Emerg Med. 2010;55(6):556-62.

REFERENCES

16. Yokell MA, Camargo CA, Wang NE, et al. Characteristics of United States Emergency Departments that Routinely Perform Alcohol Risk

1. Blow FC, Walton MA, Barry KL, et al. Alcohol and drug use among patients presenting to an inner-city emergency department: a latent

Screening and Counseling for Patients Presenting with Drinking-

class analysis. Addict Behav. 2011;36(8):793-800.

related Complaints. West J Emerg Med. 2014;15(4):438-45. 17. Delgado MK, Acosta CD, Ginde AA, et al. National survey of

2. Wu LT, Swartz MS, Wu Z, et al. Alcohol and drug use disorders

preventive health services in US emergency departments. Ann

among adults in emergency department settings in the United States.

Emerg Med. 2011;57(2):104-8 e2.

Ann Emerg Med. 2012;60(2):172-80 e5. 3. Johnson JA, Woychek A, Vaughan D, et al. Screening for at-risk

18. O’Rourke M, Richardson LD, Wilets I, et al. Alcohol-related problems: emergency physicians’ current practice and attitudes. J Emerg Med.

alcohol use and drug use in an emergency department: integration

2006;30(3):263-8.

of screening questions into electronic triage forms achieves high screening rates. Ann Emerg Med. 2013;62(3):262-6.

19. Humeniuk R and Ali R. Validation of the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) and pilot brief

4. Hankin A, Daugherty M, Bethea A, et al. The Emergency Department as a prevention site: a demographic analysis of substance use

intervention [electronic resource] : a technical report of phase II

among ED patients. Drug Alcohol Depend. 2013;130(1-3):230-3.

findings of the WHO ASSIST Project. Geneva, Switzerland: World Health Organization; 2006.

5. Sanjuan PM, Rice SL, Witkiewitz K, et al. Alcohol, tobacco, and drug use among emergency department patients. Drug Alcohol Depend.

20. Merchant RC, Baird JR, Liu T, et al. Brief intervention to increase emergency department uptake of combined rapid human

2014;138:32-8.

immunodeficiency virus and hepatitis C screening among a drug

6. Konstantopoulos WLM, Dreifuss JA, McDermott KA, et al. Identifying

misusing population. Acad Emerg Med. 2014;21(7):752-67.

patients with problematic drug use in the emergency department: results of a multisite study. Annals Emerg Med. (in press).

21. Perlis TE, Des Jarlais DC, Friedman SR, et al. Audio-computerized self-interviewing versus face-to-face interviewing for research

7. Babcock-Irvin C, Wyer PC, Gerson LW. Preventive care in the

data collection at drug abuse treatment programs. Addiction.

emergency department, Part II: Clinical preventive services-

2004;99(7):885-96.

-an emergency medicine evidence-based review. Society for Academic Emergency Medicine Public Health and Education

22. Caldwell DH and Jan G. Computerized assessment facilitates

Task Force Preventive Services Work Group. Acad Emerg Med.

disclosure of sensitive HIV risk behaviors among African Americans

2000;7(9):1042-54.

entering substance abuse treatment. Am J Drug Alcohol Abuse. 2012;38(4):365-9.

8. Bernstein SL, Boudreaux ED, Cydulka RK, et al. Tobacco control interventions in the emergency department: a joint statement

23. Islam MM, Topp L, Conigrave KM, et al. The reliability of sensitive

of emergency medicine organizations. Ann Emerg Med.

information provided by injecting drug users in a clinical setting:

2006;48(4):e417-26.

clinician-administered versus audio computer-assisted self-

Volume XVII, no. 3 : May 2016

313

Western Journal of Emergency Medicine

Merchant et al.

Emergency Department Patient Substance Use Screening interviewing (ACASI). AIDS Care. 2012;24(12):1496-503.

35. D’Onofrio G and Degutis LC. Preventive care in the emergency

24. Newman JC, Des Jarlais DC, Turner CF, et al. The differential effects

department: screening and brief intervention for alcohol problems in

of face-to-face and computer interview modes. Am J Public Health.

the emergency department: a systematic review. Acad Emerg Med.

2002;92(2):294-7.

2002;9(6):627-38.

25. Turner CF, Villarroel MA, Rogers SM, et al. Reducing bias

36. Taggart IH, Ranney ML, Howland J, et al. A systematic review of

in telephone survey estimates of the prevalence of drug

emergency department interventions for college drinkers. J Emerg

use: a randomized trial of telephone audio-CASI. Addiction.

Med. 2013;45(6):962-8.

2005;100(10):1432-44.

37. D’Onofrio G, Fiellin DA, Pantalon MV, et al. A brief intervention

26. Fernandez Huerta JM. Medidas sencillas de lecturabilidad (Simple

reduces hazardous and harmful drinking in emergency department

readability measures). Consigna. 1959;214:29-32. 27. von Elm E, Altman DG, Egger M, et al. The Strengthening the

patients. Ann Emerg Med. 2012;60(2):181-92. 38. D’Onofrio G, Pantalon MV, Degutis LC, et al. Brief intervention for

Reporting of Observational Studies in Epidemiology (STROBE)

hazardous and harmful drinkers in the emergency department. Ann

statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-9.

Emerg Med. 2008;51(6):742-50 e2. 39. Sommers MS, Lyons MS, Fargo JD, et al. Emergency department-

28. Cunningham RM, Bernstein SL, Walton M, et al. Alcohol, tobacco,

based brief intervention to reduce risky driving and hazardous/

and other drugs: future directions for screening and intervention in

harmful drinking in young adults: a randomized controlled trial.

the emergency department. Acad Emerg Med. 2009;16(11):1078-

Alcohol Clin Exp Res. 2013;37(10):1753-62.

88.

40. Bernstein J, Heeren T, Edward E, et al. A brief motivational interview

29. Nilsen P, Baird J, Mello MJ, et al. A systematic review of emergency

in a pediatric emergency department, plus 10-day telephone follow-

care brief alcohol interventions for injury patients. J Subst Abuse

up, increases attempts to quit drinking among youth and young

Treat. 2008;35(2):184-201.

adults who screen positive for problematic drinking. Acad Emerg

30. Havard A, Shakeshaft A, Sanson-Fisher R. Systematic review and meta-analyses of strategies targeting alcohol problems in emergency

Med. 2010;17(8):890-902. 41. Academic ED SBIRT Research Collaborative. The impact of

departments: interventions reduce alcohol-related injuries. Addiction.

screening, brief intervention and referral for treatment in emergency

2008;103(3):368-76; discussion 77-8.

department patients’ alcohol use: a 3-, 6- and 12-month follow-up.

31. Yuma-Guerrero PJ, Lawson KA, Velasquez MM, et al. Screening, brief intervention, and referral for alcohol use in adolescents: a

Alcohol Alcohol. 2010;45(6):514-9. 42. Newton AS, Dong K, Mabood N, et al. Brief emergency department

systematic review. Pediatrics. 2012;130(1):115-22.

interventions for youth who use alcohol and other drugs: a systematic

32. Cochran G, Field C, Caetano R. Injury-related consequences of alcohol misuse among injured patients who received screening and

review. Pediatric emerg care. 2013;29(5):673-84. 43. Merchant RC, Baird JR, Liu T. Short-term efficacy of a brief

brief intervention for alcohol: a latent class analysis. Substance

intervention to reduce drug misuse and increase drug treatment

abuse. 2014;35(2):153-62.

utilization among adult emergency department patients. Acad Emerg

33. Dent AW, Weiland TJ, Phillips GA, et al. Opportunistic screening and clinician-delivered brief intervention for high-risk alcohol use among

Med. 2015;22(10):1172-1180. 44. Woodruff SI, Clapp JD, Eisenberg K, et al. Randomized clinical trial

emergency department attendees: a randomized controlled trial.

of the effects of screening and brief intervention for illicit drug use:

Emerg Med Australasia. 2008;20(2):121-8. 34. Wojnar M and Jakubczyk A. Brief interventions for hazardous

the Life Shift/Shift Gears study. Addict Sci Clin Pract. 2014;9:8. 45. Bogenschutz MP, Donovan DM, Mandler RN, et al. Brief intervention

and harmful alcohol consumption in accident and emergency

for patients with problematic drug use presenting in emergency

departments. Front Psychiatry. 2014;5:152.

departments: a randomized clinical trial. JAMA. 2014;174(11):1736-45.

Western Journal of Emergency Medicine

314

Volume XVII, no. 3 : May 2016

Original Research

Undertriage of Trauma-Related Deaths in U.S. Emergency Departments Jenelle A. Holst, MD*† Sarah M. Perman, MD, MSCE* Roberta Capp, MD, MHS* Jason S. Haukoos, MD, MSc*† Adit A. Ginde, MD, MPH*

*University of Colorado School of Medicine, Department of Emergency Medicine, Aurora, Colorado † Denver Health Medical Center, Department of Emergency Medicine, Denver, Colorado

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted November 19, 2015; Revision received February 14, 2016; Accepted February 17, 2016 Electronically published May 2, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29327

Introduction: Accurate field triage of critically injured patients to trauma centers is vital for improving survival. We sought to estimate the national degree of undertriage of trauma patients who die in emergency departments (EDs) by evaluating the frequency and characteristics associated with triage to non-trauma centers. Methods: This was a retrospective cross-sectional analysis of adult ED trauma deaths in the 2010 National Emergency Department Sample (NEDS). The primary outcome was appropriate triage to a trauma center (Level I, II or III) or undertriage to a non-trauma center. We subsequently focused on urban areas given improved access to trauma centers. We evaluated the associations of patient demographics, hospital region and mechanism of injury with triage to a trauma versus non-trauma center using multivariable logistic regression. Results: We analyzed 3,971 included visits, representing 18,464 adult ED trauma-related deaths nationally. Of all trauma deaths, nearly half (44.5%, 95% CI [43.0-46.0]) of patients were triaged to non-trauma centers. In a subgroup analysis, over a third of urban ED visits (35.6%, 95% CI [34.137.1]) and most rural ED visits (86.4%, 95% CI [81.5-90.1]) were triaged to non-trauma centers. In urban EDs, female patients were less likely to be triaged to trauma centers versus non-trauma centers (adjusted odds ratio [OR] 0.83, 95% CI [0.70-0.99]). Highest median household income zip codes (≥$67,000) were less likely to be triaged to trauma centers than lowest median income ($1-40,999) (OR 0.54, 95% CI [0.43-0.69]). Compared to motor vehicle trauma, firearm trauma had similar odds of being triaged to a trauma center (OR 0.90, 95% CI [0.71-1.14]); however, falls were less likely to be triaged to a trauma center (OR 0.50, 95 %CI [0.38-0.66]). Conclusion: We found that nearly half of all trauma patients nationally and one-third of urban trauma patients, who died in the ED, were triaged to non-trauma centers, and thus undertriaged. Sex and other demographic disparities associated with this triage decision represent targeted opportunities to improve our trauma systems and reduce undertriage. [West J Emerg Med. 2016;17(3):315–323.]

INTRODUCTION Regionalized trauma systems have been developed to improve outcomes after injury by preferentially triaging injured patients to designated trauma centers.1 Survival Volume XVII, no. 3 : May 2016

of trauma patients is higher at trauma centers, due to immediate and ongoing access to highly skilled clinicians and resources.2-4 Prehospital emergency medical services (EMS) trauma triage protocols have been developed to aid 315

Western Journal of Emergency Medicine

Holst et al.

Undertriage of Trauma-Related Deaths in identifying and triaging the most critically ill patients.5-13 These protocols are used to minimize both under- and overtriage of patients, in an attempt to match the acuity of the patient with the appropriate level of hospital care.10 The most severely injured patients should be transported to major trauma centers (Level I or II), if available at a reasonable distance, whereas patients with minor injuries may be taken to lower level trauma centers (Level III) or non-traumadesignated hospitals. Inaccurate triage that results in a patient requiring a higher level of care not being transported to a trauma center is termed undertriage,13 and can lead to avoidable morbidity and mortality. If a patient with minor trauma is taken to a trauma center this is considered overtriage and can lead to unnecessary cost and burden on the limited number of trauma centers, as cost of maintaining a trauma center is considerable.14 Injured patients who ultimately die in the emergency department (ED) are the most critically ill subset of trauma patients and need to be appropriately triaged to a trauma center where they have the best chance of survival. Previous studies have assessed undertriage of trauma patients with varying degrees of injury severity, but have not individually assessed the most severely injured subset of patients who die in the ED.15-17 Level I and II trauma centers are most often centered in urban areas,18,19 while rural areas frequently have lower level or no nearby trauma center. With readily available access to tertiary care trauma centers, urban areas should have very low rates of undertriage if field triage criteria are used accurately and appropriately. When assessing undertriage, previous studies did not differentiate between urban and rural areas.15-17 This geographic distinction when assessing undertriage is important given the anticipated differences in trauma center availability in urban versus rural areas. Triage of injured patients is one of the most important components of an effective regionalized trauma system, yet little is currently known about national rates of undertriage of severely injured patients who die in the ED. We hypothesized that since ED death is a marker of the most critically injured trauma patient, undertriage of this patient population would be rare, especially in urban areas with readily accessible trauma centers. However, rural areas, with their inherent limited access to trauma centers, were hypothesized to have higher rates of undertriage. Additionally, we sought to identify patient and hospital characteristics associated with undertriage and determine targeted opportunities to improve EMS triage decisions. METHODS Study Design and Population This study was a retrospective, cross-sectional analysis of the 2010 National Emergency Department Sample (NEDS), Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality.20 The NEDS is the Western Journal of Emergency Medicine

largest all-payer ED database in the United States. The NEDS provides patient-level data on a 20% stratified sample of ED visits from 969 hospitals in 29 states, of which 164 (17%) were designated trauma centers. Hospitals are selected using a stratified probability sample based on hospital characteristics to provide weighted national estimates of ED visits, which were approximately 129 million in 2010. Analysis of this publically available dataset was approved by the institutional review board. We included all adult, age â&#x2030;Ľ18 years, trauma-related ED visit patients. Trauma-related ED visits were defined by the injury variable available in the NEDS, which uses injuryrelated International Classification of Diseases, 9th edition, (ICD-9) codes in any diagnosis field as previously defined.21 We then restricted our cohort to patients whose ED visit resulted in death during the index ED visit (prior to hospital admission or transfer). The primary outcome was trauma center designation of the hospital (i.e., trauma center level I, II or III versus non-trauma center). We defined undertriage as a patient visit with a traumatic injury ending in death in the ED of a non-trauma center. NEDS obtains trauma center status from the Trauma Information Exchange Program database,20 which includes state designation or American College of Surgeons verification. After initial overall descriptive analyses of deaths in EDs located in metropolitan (urban) and rural hospitals combined, we further restricted our cohort to ED deaths in urban hospitals to focus on a population with improved access to trauma centers. Metropolitan areas were defined in NEDS based on the county-based Urban Influence Codes (UIC)20: â&#x2030;Ľ50,000 people (metropolitan), non-metropolitan regions with <50,000 and >10,000 people (micropolitan locations), and rural locations contain <10,000 people (micropolitan and rural categories combined for this analysis). Patient socio-demographic characteristics included age, sex, median household income based on the patientâ&#x20AC;&#x2122;s zip code, and primary insurance/payer. Clinical characteristics included whether the ED visit occurred during a weekday or weekend day, month, and mechanism of injury, as defined by ICD-9 External-Cause-of-Injury-Codes.22 Additionally, we examined hospital characteristics including census region (Northeast, Midwest, South and West) ownership (public, private and combined, which were hospitals in strata too small to stratify based on control), teaching status, annual ED volume, and safety net status. Safety net status was defined as an ED with >30% ED visits with Medicaid, or self-pay/no charge (uninsured), or >40% ED visits combined Medicaid or uninsured.23 Statistical Analysis We used descriptive statistics to summarize the data. Among urban hospitals, we used multivariable logistic regression to estimate the associations between sociodemographic and clinical characteristics of ED trauma 316

Volume XVII, no. 3 : May 2016

Holst et al.

Undertriage of Trauma-Related Deaths

deaths with triage to trauma versus non-trauma centers. Characteristics were removed from the multivariate model if they were collinear with hospital trauma or urban status (i.e., teaching hospital and safety net status). Because there was <5% missing for each variable of interest, missing observations were dropped rather than imputed (final multivariable model with <10% missing data). We used survey commands to account for the complex survey design and provide national estimates, per NEDS guidelines. Analyses were conducted in SAS 9.3 (SAS Inc, Cary, NC) and Stata 12.1 (Stata Corp, LP, College Station, TX). RESULTS The 2010 NEDS contained 42,614 observations of adult ED deaths, of which 3,971 (9.3%) were traumarelated, representing 18,464 ED trauma deaths nationally. Patient, visit and hospital characteristics of these deaths are presented in Table 1. The largest demographic groups among trauma deaths were young (age 18-34 years), male, low median household income, and self-pay. The mechanisms of injury for ED deaths are displayed in Figure 1. The four most common mechanisms were motor vehicle trauma (MVT), including occupant of or person struck by an automobile or motorcycle (30.7%, 95% CI [29.2-32.2]), injury by firearm (19.0% , 95% CI [17.8-20.3]), other (18.9%, 95% CI [17.7-20.2]), and falls (11.1%, 95% CI [10.1-12.1]). Of all combined rural and urban U.S. trauma deaths, nearly half (44.5%, 95% CI [43.0-46.0]) of patients were triaged to nontrauma centers, and thus undertriaged. Figure 2 displays trauma vs. non-trauma center status of ED deaths in urban and rural locations. For patients taken to EDs in urban areas, most patients were triaged to trauma centers; however, still over a third (35.6%, 95% CI [34.1-37.1]) were triaged to non-trauma centers, and thus undertriaged. Most ED trauma deaths were triaged to nontrauma centers in rural areas (86.4%, 95% CI [81.5-90.1]). Next our analysis focused on urban areas to further explore the characteristics associated with undertriage of ED trauma deaths triaged to non-trauma centers (descriptive results in Table 2). The multivariable logistic regression results for characteristics associated with triage to an urban trauma vs. urban non-trauma center are shown in Table 3. Female patients were less likely to be triaged to trauma centers versus non-trauma centers (adjusted odds ratio [OR] 0.83, 95% CI [0.70-0.99]). Highest median household income zip codes were less likely to be triaged to trauma centers than lowest median income (OR 0.54, 95% CI [0.43-0.69]). Compared to MVT, firearm trauma had similar odds of being triaged to a trauma center (OR 0.90, 95% CI [0.71-1.14]); however, falls were less likely to be triaged to a trauma center (OR 0.50, 95% CI [0.38-0.66]). DISCUSSION In this study we found that nationwide, nearly half Volume XVII, no. 3 : May 2016

(44.5%) of trauma patients who died in the ED died in a nontrauma center. This is the most concerning form of undertriage because this subset of trauma patients who die in the ED are the most gravely injured and triage to a trauma center is crucial for improving their chances of survival. To our knowledge this is the first national study focusing solely on evaluating the destination and outcomes of the most critically ill trauma patients, those who die in the ED. Based on our results, the extent of undertriage of U.S. trauma patients who ultimately die in the ED is remarkably high. Previous studies estimated undertriage ranging from 34% to 69%; however, they studied less severely injured patients,15,16 or grouped ED deaths in non-trauma centers with other forms of undertriage such as patients treated and released from the ED of a nontrauma center or admitted to a non-trauma center.17 Identifying these critically injured patients in the prehospital setting and finding ways to ensure triage to a trauma center will potentially be the most impactful way to reduce undertriage and prevent mortality. Our results also identified a large burden of undertriage in urban areas, where over a third (35.6%) of trauma patients who died in the ED died in a non-trauma center. This is the first national estimate of urban trauma undertriage as previous studies did not differentiate between urban and rural.15-17 This differentiation is important because accessibility to trauma centers is very different in urban versus rural areas. Trauma systems should be the most advanced in the urban setting given the closer proximity that Americans living in cities have to trauma centers. One study indicated that 84.1% of Americans have access to level I or II trauma centers within one-hour transport time, and these people live mostly in urban areas.18,19 This leaves 46.7 million living in mostly rural areas without trauma center access within an hour. Accordingly, we anticipated the observed difference in undertriage between rural and urban ED trauma deaths in our study. However, the large amount of trauma deaths in urban areas that were undertriaged to nontrauma centers was higher than anticipated. Prehospital EMS trauma triage protocols have been developed to aid in identifying the most critically ill patients.5 Generally trauma triage protocols incorporate physiologic criteria, anatomic criteria, mechanism of injury and special considerations (age, comorbidities, etc).7 The 2011 Guidelines for Field Triage of Injured Patients from the Centers for Disease Control and Prevention (CDC) recommend a stepwise approach designed to identify serious injuries as early as possible during the prehospital assessment. Step one assesses vital signs and Glasgow Coma Scale (physiologic); step two assesses visible injuries such as penetrating injuries, crush injuries or long bone fractures (anatomic); step three assesses for high risk mechanisms including auto versus pedestrian or falls from a significant height (mechanism); and step four assesses for other complicating factors such as anticoagulated, pediatric, elderly, burned or pregnant patients among other

317

Western Journal of Emergency Medicine

Holst et al.

Undertriage of Trauma-Related Deaths

Table 1. Baseline characteristics of trauma-related emergency department (ED) visits in the 2010 National Emergency Department Sample. ED deaths (n=3,971) Characteristics

Survived ED visits (n=4,975,715)

Weighted % (95% CI)

18-34

1331

33.9 (32.4-35.5)

1,882,605

37.9 (37.8-37.9)

35-49

805

19.9 (18.7-21.3)

1,257,935

25.1 (25.1-25.1)

50-64

772

19.5 (18.2-20.9)

911,985

18.3 (18.3-18.4)

1063

26.6 (25.2-28.1)

937,768

18.7 (18.7-18.8)

1126

28.1 (26.6-29.6)

2,512,067

50.1 (50.1-50.2)

$1-40,999

1177

30.1 (28.6-31.6)

1,474,500

29.7 (29.7-29.8)

$41,000-50,999

1072

27.1 (25.7-28.6)

1,345,638

27.1 (27.1-27.2)

$51,000-66,999

831

20.6 (19.3-22.0)

1,112,678

22.1 (22.0-22.1)

â&#x2030;Ľ$67,000

686

17.1 (15.9-18.4)

941,503

18.8 (18.7-18.8)

Medicare

901

22.6 (21.2-24.0)

1,069,394

21.4 (21.4-21.4)

Medicaid

341

8.7 (7.8-9.7)

716,103

14.4 (14.4-14.4)

Private

1086

28.4 (26.9-29.9)

1,634,268

33.4 (33.3-33.4)

Self-pay

1358

34.1 (32.5-35.6)

1,067,111

21.1 (21.1-21.1)

255

6.3 (5.6-7.1)

476,157

9.7 (9.7-9.8)

1316

33.1 (31.5-34.6)

1,506,155

30.4 (30.3-30.4)

January-March

571

18.4 (17.2-19.7)

934,814

22.1 (22.1-22.2)

April-June

739

22.0 (20.6-23.4)

1,120,229

26.5 (26.5-26.6)

July-September

877

22.3 (20.9-23.7)

1,165,350

27.7 (27.5-27.6)

October-December

900

22.4 (21.0-23.7)

1,008,723

23.8 (23.8-23.9)

Northeast

678

16.7 (15.6-17.8)

965,508

20.2 (20.2-20.2)

Midwest

784

22.0 (20.7-23.3)

1,092,194

24.6 (24.6-24.6)

1752

42.1 (40.6-43.6)

2,120,934

38.4 (38.3-38.4)

Demographics Age, years

â&#x2030;Ľ65 Female sex Median household income

Primary payer

No charge/other Weekend arrival Month of arrival

Hospital characteristics Region

South West

757

19.3 (18.0-20.5)

811,657

16.8 (16.7-16.8)

2021

55.5 (54.0-57.0)

1,532,579

35.0 (35.0-35.0)

Metropolitan non-teaching

1437

32.5 (31.1-33.9)

2,365,176

43.4 (43.3-43.4)

Metropolitan teaching

1864

50.1 (48.5-51.6)

1,742,525

37.0 (37.0-37.0)

670

17.5 (16.3-18.8)

882,592

19.6 (19.6-19.6)

14222

21.3 (20.1-22.5)

1,536,808

26.4 (26.4-26.5)

2735

73.9 (72.6-75.1)

3,059,866

67.4 (67.3-67.4)

4.9 (4.3-5.5)

393,619

6.2 (6.2-6.2)

Trauma center Teaching/urban-rural status

Micropolitan/rural (teaching and non-teaching) Ownership Private Collapsed (public or private)

Public- government, non-federal 3321 Note: n represents raw (unweighted) number of observations.

Western Journal of Emergency Medicine

318

Volume XVII, no. 3 : May 2016

Holst et al.

Undertriage of Trauma-Related Deaths

Table 1. Continued. ED deaths (n=3,971) Characteristics

Weighted % (95% CI)

Survived ED visits (n=4,975,715) n

Weighted % (95% CI)

Emergency department (ED) volume <10,000

208

10,000-19,999

283

20,000-39,999

875

40,000-59,999

1059

60,000-79,999 >80,000

255,915

5.7 (5.7-5.7)

7.3 (6.4-8.2)

482,572

10.2 (10.1-10.2)

21.1 (19.9-22.5)

1,396,101

27.8 (27.8-27.9)

28.4 (27.0-30.0)

1,141,757

23.0 (23.0-23.0)

631

14.8 (13.8-15.9)

804,751

15.3 (15.2-15.3)

915

22.6 (21.3-23.9)

909,197

18.0 (18.0-18.1)

2705

67.5 (66.0-69.0)

3,189,345

62.3 (62.2-62.3)

Cutting or piercing

147

3.7 (3.1-4.3)

400,718

8.1 (8.1-8.1)

Drown

106

2.7 (2.2-3.3)

1,437

0.03 (0.03-0.03)

Fall

446

11.1 (10.1-12.1)

1,407,804

28.2 (28.2-28.2)

Safety net status

5.8 (5.0-6.8)

Mechanism of injury

Fire Firearm Machinery Motor vehicle trauma Natural/environmental Poison Struck by or against Suffocation

1.3 (1.0-1.8)

66,530

1.4 (1.3-1.4)

745

19.0 (17.8-20.3)

13,518

0.3 (0.3-0.3)

0.4 (0.2-0.7)

26,011

0.5 (0.5-0.5)

1209

30.7 (29.2-32.2)

566,551

11.3 (11.3-11.4)

1.8 (1.4-2.2)

192,578

3.8 (3.8-3.9)

182

4.7 (4.0-5.6)

153,324

3.1 (3.0-3.1)

2.0 (1.5-2.6)

575,861

11.6 (11.6-11.6)

186

4.5 (3.9-5.2)

7,356

0.2 (0.2-0.2)

18.9 (17.7-20.2)

1,592,205

32.1 (32.1-32.1)

Other 756 Note: n represents raw (unweighted) number of observations.

unique situations (special considerations). As soon as one of these criteria is met, the EMS provider should make the decision to transport to a trauma center.13 These guidelines include core elements meant to be adapted to the needs of each individual EMS system, and thus protocols across the country differ slightly. Multiple studies have assessed the sensitivity and specificity of various triage protocols for determining if an injured patient needed transportation to a trauma center and have shown variability among different protocols.6,8-12 Additionally, there is variability in uptake of these guidelines across EMS systems following guideline revisions.24 The accuracies of each piece or the sum of these trauma triage protocols are relatively unknown. Although the NEDS does not contain data regarding the mode of transportation to the ED we can estimate the rate of arrival by privately owned vehicle (POV) versus EMS from other studies. In one study using the National Trauma Data Bank, 12.6% of patients with gunshot wounds were transported to 182 trauma centers by POV.25 In a statewide study, 9.6% of all injured patients were transported to any trauma center by POV.26 In a regional study, 6.5% of patients with cervical spine injuries were transported to three trauma centers by POV.27 Although Volume XVII, no. 3 : May 2016

the mode of arrival of adult injured patients to non-trauma centers has not yet been assessed, these studies provide a starting point, estimating >85% of injured patients arrive to trauma centers via EMS. A high proportion of arrival via EMS reinforces the influence of pre-hospital trauma triage protocols on rates of under- and overtriage. Our results highlight an opportunity to improve prehospital trauma triage protocols, particularly with prehospital provider perception of the severity of mechanism of injury, as undertriage was found to be associated with falls. EMS protocols and prehospital providers may be more likely to underestimate the severity of injury from falls relative to more visually obvious mechanisms of injury due to firearms and MVT. It has also been shown that older adults with falls who die soon after hospital arrival are often transported to non-trauma centers because the severity of their injuries is not recognized in the field.28 In addition to potential unrecognized injury secondary to the trauma from the fall, other potentially lethal medical causes of the fall such as syncope and associated high risk cardiac events, spontaneous intracranial hemorrhage or other severe metabolic derangements may be under-recognized causes of ED deaths. Recognition of these subtle presentations of severe trauma and potential serious 319

Western Journal of Emergency Medicine

Holst et al.

Undertriage of Trauma-Related Deaths

Table 2. Patient and hospital characteristics for trauma patients who died in non-trauma vs trauma hospitals among urban emergency departments (EDs). Non-trauma hospital Characteristics

Trauma hospital

% (95% CI)

1,320

35.6 (34.1-37.1)

1,981

64.4 (62.9-65.9)

18-34

371

28.1 (25.7-30.6)

757

38.2 (36.0-40.4)

35-49

258

19.4 (17.3-21.6)

389

19.5 (17.7-21.4)

50-64

272

20.6 (18.4-22.9)

348

17.6 (15.9-19.4)

â&#x2030;Ľ65

419

32.0 (29.5-34.6)

487

24.8 (22.9-26.9)

424

32.7 (30.2-35.3)

520

25.9 (23.9-27.9)

$1-40,999

274

20.3 (18.2-22.6)

610

31.5 (29.4-33.7)

$41,000-50,999

349

25.9 (23.6-28.4)

495

25.0 (23.1-27.1)

$51,000-66,999

348

26.5 (24.1-29.0)

399

20.2 (18.3-22.1)

â&#x2030;Ľ$67,000

294

23.2 (21.0-25.7)

350

17.0 (15.4-18.7)

Medicare, private and other

704

54.1 (51.3-56.8)

923

47.4 (45.1-49.7)

Medicaid, self-pay and no charge

610

45.9 (43.2-48.7)

1,036

52.6 (50.3-54.9)

398

30.2 (27.8-32.8)

688

34.4 (32.2-36.7)

January-March

265

20.7 (18.6-23.0)

358

17.8 (16.2-19.7)

April-June

257

19.8 (17.7-22.1)

435

21.4 (19.6-23.4)

July-September

302

23.3 (21.1-25.7)

438

21.3 (19.5-23.2)

October-December

304

23.7 (21.4-26.1)

396

19.4 (17.7-21.2)

Northeast

223

20.0 (17.8-22.3)

395

17.2 (15.8-18.7)

Midwest

186

16.6 (14.6-18.8)

462

23.4 (21.8-25.2)

South

620

42.2 (39.7-44.8)

742

39.8 (37.8-41.9)

West

291

21.3 (19.2-23.5)

382

19.5 (18.0-21.2)

Motor vehicle trauma

310

23.1 (20.9-25.5)

709

35.6 (33.4-37.9)

Firearm

182

13.6 (11.9-15.6)

467

23.9 (22.0-25.9)

Fall

171

13.0 (11.3-14.9)

196

9.6 (8.4-11.1)

Other

657

50.3 (47.6-53.0)

612

30.8 (28.7-33.0)

Total Demographics Age, years

Female sex Median household income

Primary payer

Weekend arrival Month of arrival

Hospital characteristics Region

Mechanism of injury

medical causes of falls in older adults should be a target for improvement in prehospital trauma triage protocols. Some geographic and socioeconomic differences in rates of undertriage were not anticipated. For example, people living in the Northeast have been shown to have the closest proximity to trauma centers,18,29 yet in our study were more likely to be undertriaged compared to the Midwest. This suggests that factors other than distance from a trauma Western Journal of Emergency Medicine

center may account for this degree of undertriage. One variable that potentially affects prehospital adherence to trauma triage protocols is patient preference. For example, one study showed in approximately half of injured patients, EMS providers indicated patient preference as the reason for selecting the destination hospital.30 Perhaps this geographic difference in undertriage is a reflection of patient/family/EMS preference, lower utilization of EMS, or need for improvement 320

Volume XVII, no. 3 : May 2016

Holst et al.

Undertriage of Trauma-Related Deaths

Table 3. Multivariable logistic regression for characteristics of emergency department trauma deaths associated with triage to an urban trauma center vs. an urban non-trauma center. Characteristics

Adjusted odds ratio

Cutting or piercing 3.7%

95% CI

Demographics 18-34

Referent

35-49

0.86

0.69-1.07

50-64

0.74

0.59-0.92

â&#x2030;Ľ65

0.78

0.62-1.00

0.83

0.70-0.99

Female sex Median household income $1-40,999

0.53-0.81

$51,000-66,999

0.52

0.42-0.65

â&#x2030;Ľ$67,000

0.54

0.43-0.69

Referent

0.97

0.81-1.17

80%

1.13

0.97-1.33

70%

Referent

April-June

1.23

0.97-1.33

July-September

1.02

0.81-1.29

October-December

0.89

0.70-1.13

Trauma center

Non-trauma center

90%

Month of arrival

44.5%

35.6%

60%

86.4%

50% 40% 30%

Hospital characteristics

55.5%

64.4%

20%

Region Northeast

Firearm 19.0%

100%

Primary payer

January-March

Motor vehicle traffic 30.7%

Figure 1. Mechanism of injury of US emergency department trauma-related deaths.

0.65

Weekend arrival

Machinery 0.4%

Other 18.9%

Referent

Medicaid, self-pay and no charge

Fire 1.3%

Suffocation 4.5%

Fall 11.1%

$41,000-50,999

Medicare, private and other

Drown 2.7%

Poison 4.7%

Age, years

Natural/ environment 1.8%

Struck by or against 2.0%

10%

Referent

Midwest

1.88

1.47-2.40

South

0.80

0.65-0.99

West

1.17

0.93-1.48

13.6%

0% Overall

Urban

Rural

Figure 2. Trauma deaths triaged to trauma vs. non-trauma centers in urban, rural and overall emergency departments.

Mechanism of injury Motor vehicle trauma

Referent

Firearm

0.90

0.71-1.14

Fall

0.50

0.38-0.66

Other

0.37

0.31-0.45

of regionalized trauma systems in this area. Other factors such as hospital density, road conditions and ED divert status likely impact EMS triage decisions and would be an interesting direction for further study. Trauma patient deaths with higher median household income were more likely to die in a non-trauma center than poorer patients. This finding was unanticipated given the general assumption that higher socioeconomic status leads to better access to medical care. One possible explanation Volume XVII, no. 3 : May 2016

could be that many trauma centers are located in inner cities and thus lower socioeconomic status populations, which are frequently in the same location, may have better access to these trauma centers. Further research is needed to determine potential reasons for this disparity. We also found that female sex was associated with undertriage, consistent with a prior state-level study that reported female moderate to severely injured trauma patients at non-trauma centers were less likely to be properly transferred to a trauma center than men.15 While the cause of this disparity is unknown, we speculate that since women account for fewer trauma deaths than men overall, the severity of injury may be underestimated. Sex disparities have been noted in other areas of acute care including cardiac emergencies where men generally receive more aggressive early management than women.31 321

Western Journal of Emergency Medicine

Holst et al.

Undertriage of Trauma-Related Deaths LIMITATIONS This study is a secondary analysis and like any similarly conducted study cannot determine causality. In addition, the database relies on administrative data, potentially leading to coding errors or missing data. Some data elements are not available in the NEDS, such as race/ethnicity or physiologic/anatomic prehospital triage criteria. Additionally, some EMS trauma triage protocols may dictate that if a patient has an emergent airway threat or other serious form of instability, he should be transported to the nearest hospital even if it is not a trauma center. This could account for an unknown portion of our patient sample categorized as undertriaged. Furthermore, we were unable to determine the precise reason for death, and whether the trauma/injury diagnosis was truly the cause of death. Though we assume that most gravely injured patients would be transported to the hospital by EMS, data are not available in the NEDS for mode of transport to the ED. Data were also not available for the geographic location of the injury or distance to closest hospital or trauma center, limiting our ability to fully address prehospital triage decisions. We also considered using the Injury Severity Score to study moderate and severely injured patients, but the use of this score is limited and can be inaccurate because often patients are too unstable to complete imaging needed to assess all injury categories prior to hospital transfer; the full extent of the ISS is not determined until the end of hospitalization and these data after transfer would not be available in the NEDS. CONCLUSION High numbers of trauma patients who died in EDs were undertriaged to non-trauma centers, even in urban areas, where trauma centers are more accessible. Sex and other demographic disparities may impact trauma triage decisions in urban areas. These differences represent targeted opportunities to improve triage of specific populations to trauma centers.

REFERENCES 1. Eastman AB, Lewis FR Jr, Champion HR, et al. Regional trauma system design: critical concepts. Am J Surg. 1987;154:79–87. 2. MacKenzie EJ, Rivara FP, Jurkovich GJ, et al. A national evaluation of the effect of trauma-center care on mortality. N Engl J Med. 2006;354:366-78. 3. Celso B, Tepas J, Langland-Orban B, et al. A systematic review and meta-analysis comparing outcome of severely injured patients treated in trauma centers following the establishment of trauma systems. J Trauma. 2006;60:371-8. 4. Nathens AB, Jurkovich GJ, Rivara FP, et al. Effectiveness of state trauma systems in reducing injury-related mortality: a national evaluation. J Trauma. 2000;48:25-30. 5. Kane G, Engelhardt R, Celentano J, et al. Empirical development and evaluation of prehospital trauma triage instruments. J Trauma. 1985;25:482-9. 6. Henry MC, Hollander JE, Alicandro JM, et al. Incremental benefit of individual American College of Surgeons trauma triage criteria. Acad Emerg Med. 1996;3:992–1000. 7. Brown JB, Stassen NA, Bankey PE, et al. Mechanism of injury and special consideration criteria still matter: an evaluation of the National Trauma Triage Protocol. J Trauma. 2011;70:38-44. 8. West JG, Murdock MA, Baldwin LC, et al. A method for evaluating field triage criteria. J Trauma. 1986;26:655–9. 9. Boyle MJ, Smith EC, Archer F. Is mechanism of injury alone a useful predictor of major trauma? Injury. 2008;39:986–92. 10. Cook CH, Muscarella P, Praba AC, et al. Reducing overtriage without compromising outcomes in trauma patients. Arch Surg. 2001;136:752–6. 11. Newgard CD, Hsia RY, Mann NC, et al. Western Emergency Services Translational Research Network (WESTRN) Investigators. The tradeoffs in field trauma triage: a multiregion assessment of accuracy metrics and volume shifts associated with different triage strategies. J Trauma Acute Care Surg. 2013;74:1298-306. 12. Newgard CD, Zive D, Holmes JF, et al. WESTRN investigators. A multisite assessment of the American College of Surgeons Committee on Trauma field triage decision scheme for identifying seriously injured children and adults. J Am Coll Surg. 2011;213:709-21. 13. Sasser SM, Hunt RC, Faul M, et al. Centers for Disease Control

Address for Correspondence: Adit A. Ginde, MD, MPH, Anschutz Medical Campus, 12401 East 17th Avenue, Leprino Building, B215, Aurora, CO 80045. Email: adit.ginde@ucdenver.edu.

and Prevention (CDC). Guidelines for field triage of injured patients: recommendations of the National Expert Panel on Field Triage, 2011. MMWR Recomm Rep. 2012;61(RR-1):1-20.

14. Taheri PA, Butz DA, Lottenberg L, et al. The cost of trauma center

Copyright: © 2016 Holst et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

16. Nathens AB, Jurkovich GJ, MacKenzie EJ, et al. A resource-

Western Journal of Emergency Medicine

readiness. Am J Surg. 2004;187:7-13. 15. Mohan D, Rosengart MR, Farris C, et al. Assessing the feasibility of the American College of Surgeons’ benchmarks for the triage of trauma patients. Arch Surg. 2011;146:786-92. based assessment of trauma care in the United States. J Trauma. 2004;56:173-8. 17. Xiang H, Wheeler KK, Groner JI, et al. Undertriage of major trauma

322

Volume XVII, no. 3 : May 2016

Holst et al.

Undertriage of Trauma-Related Deaths

patients in the US emergency departments. Am J Emerg Med.

Variation in prehospital use and uptake of the national Field Triage

2014;32:997-1004. 18. Branas CC, MacKenzie EJ, Williams JC, et al. Access to trauma

Decision Scheme. Prehosp Emerg Care. 2013;17:135-48. 25. Zafar SN, Haider AH, Stevens KA, et al. Increased mortality

centers in the United States. JAMA. 2005;293:2626-33.

associated with EMS transport of gunshot wound victims when

19. Myers SR, Branas CC, French BC, et al. Safety in numbers: are major cities the safest places in the United States? Ann Emerg Med.

compared to private vehicle transport. Injury. 2014;45(9):1320-6. 26. Johnson NJ, Carr BG, Salhi R, et al. Characteristics and outcomes

2013;62:408-18.

of injured patients presenting by private vehicle in a state trauma

20. HCUP Nationwide Emergency Department Sample (NEDS). Healthcare Cost and Utilization Project (HCUP). 2010. Agency for

system. Am J Emerg Med. 2013;31(2):275-81. 27. Urdaneta AE, Stroh G, Teng J, Snowden B, Barrett TW, Hendey GW.

Healthcare Research and Quality, Rockville, MD. Available at: www.

Cervical spine injury: analysis and comparison of patients by mode of

hcup-us.ahrq.gov/nedsoverview.jsp. Accessed Oct 15, 2015. 21. Injury Surveillance Workgroup, State and Territorial Injury Prevention

transportation. J Emerg Med. 2013;44(2):287-91. 28. 28. Dean D, Wetzel B, White N, et al. WESTRN Investigators. From

Directors Association, 2003. Consensus Recommendations for

9-1-1 call to death: evaluating traumatic deaths in seven regions for

Using Hospital Discharge Data for Injury Surveillance. Marietta,

early recognition of high-risk patients. J Trauma Acute Care Surg.

GA. Available at: http://www.safestates.org/associations/5805/files/ HospitalDischargeData.pdf. Accessed Oct 15, 2015.

2014;76:846-53. 29. Carr BG, Branas CC, Metlay JP, et al. Access to emergency care in

22. Injury Prevention and Control, Centers for Disease Control and Prevention, 2011. Recommended framework of E-code groupings

the United States. Ann Emerg Med. 2009;54:261-9. 30. Newgard CD, Mann NC, Hsia RY, et al. Western Emergency

for presenting injury mortality and morbidity data. Available at: http://

Services Translational Research Network (WESTRN) Investigators.

www.cdc.gov/injury/wisqars/ecode_matrix.html. Accessed Oct 15,

Patient choice in the selection of hospitals by 9-1-1 emergency

2015.

medical services providers in trauma systems. Acad Emerg Med.

23. Fee C, Burstin H, Maselli JH, et al. Association of emergency department length of stay with safety-net status. JAMA.

2013;20:911-9. 31. Gan SC, Beaver SK, Houck PM, et al. Treatment of acute myocardial

2012;307:476-82.

infarction and 30-day mortality among women and men. N Engl J

24. Barnett AS, Wang NE, Sahni R, et al. WESTRN Investigators.

Volume XVII, no. 3 : May 2016

Med. 2000;343:8-15.

323

Western Journal of Emergency Medicine

Original Research

Vital Signs Predict Rapid-Response Team Activation Within Twelve Hours of Emergency Department Admission James M. Walston* Daniel Cabrera, MD† Shawna D. Bellew, MD† Marc N. Olive, BA† Christine M. Lohse, MS‡ M. Fernanda Bellolio, MD, MS†§

*Mayo Medical School, Rochester, Minnesota † Department of Emergency Medicine, Mayo Clinic, Rochester, Minnesota ‡ Department of Health Sciences Research, Division of Biostatistics, Mayo Clinic, Rochester, Minnesota § Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota

Section Editor: Michael Gottlieb, MD Submission history: Submitted August 24, 2015; Revision received February 2, 2016; Accepted February 3, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.28501

Introduction: Rapid-response teams (RRTs) are interdisciplinary groups created to rapidly assess and treat patients with unexpected clinical deterioration marked by decline in vital signs. Traditionally emergency department (ED) disposition is partially based on the patients’ vital signs (VS) at the time of hospital admission. We aimed to identify which patients will have RRT activation within 12 hours of admission based on their ED VS, and if their outcomes differed. Methods: We conducted a case-control study of patients presenting from January 2009 to December 2012 to a tertiary ED who subsequently had RRT activations within 12 hours of admission (early RRT activations). The medical records of patients 18 years and older admitted to a non-intensive care unit (ICU) setting were reviewed to obtain VS at the time of ED arrival and departure, age, gender and diagnoses. Controls were matched 1:1 on age, gender, and diagnosis. We evaluated VS using cut points (lowest 10%, middle 80% and highest 10%) based on the distribution of VS for all patients. Our study adheres to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for reporting observational studies. Results: A total of 948 patients were included (474 cases and 474 controls). Patients who had RRT activations were more likely to be tachycardic (odds ratio [OR] 2.02, 95% CI [1.25-3.27]), tachypneic (OR 2.92, 95% CI [1.73-4.92]), and had lower oxygen saturations (OR 2.25, 95% CI [1.42-3.56]) upon arrival to the ED. Patients who had RRT activations were more likely to be tachycardic at the time of disposition from the ED (OR 2.76, 95% CI [1.65-4.60]), more likely to have extremes of systolic blood pressure (BP) (OR 1.72, 95% CI [1.08-2.72] for low BP and OR 1.82, 95% CI [1.192.80] for high BP), higher respiratory rate (OR 4.15, 95% CI [2.44-7.07]) and lower oxygen saturation (OR 2.29, 95% CI [1.43-3.67]). Early RRT activation was associated with increased healthcare utilization and worse outcomes including increased rates of ICU admission within 72 hours (OR 38.49, 95%CI [19.03-77.87]), invasive interventions (OR 5.49, 95%CI [3.82-7.89]), mortality at 72 hours (OR 4.24, 95%CI [1.60-11.24]), and mortality at one month (OR 4.02, 95%CI [2.44-6.62]). Conclusion: After matching for age, gender and ED diagnosis, we found that patients with an abnormal heart rate, respiratory rate or oxygen saturation at the time of ED arrival or departure are more likely to trigger RRT activation within 12 hours of admission. Early RRT activation was associated with higher mortality at 72 hours and one month, increased rates of invasive intervention and ICU admission. Determining risk factors of early RRT activation is of clinical, operational, and financial importance, as improved medical decision-making regarding disposition would maximize allocation of resources while potentially limiting morbidity and mortality. [West J Emerg Med. 2016;17(3):324–330.] Western Journal of Emergency Medicine

324

Volume XVII, no. 3 : May 2016

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations

INTRODUCTION Rapid-response teams (RRTs) (also known as emergency response systems or medical emergency teams) have become an important component of American healthcare and are an integral part of the Institute for Healthcare Improvement’s “Saving 100,000 Lives” campaign.1 Hospitals have implemented this concept with the goal of improving clinical outcomes by targeting interventions at patients at risk for clinical deterioration and cardiac arrest.1,2 These systems have been introduced in many hospitals throughout the world, including our own institution in 2006. RRTs are interdisciplinary groups that intervene in the care of patients with early warning signs of impending deterioration. The goal of these teams is rapid activation, evaluation and intervention to prevent further clinical deterioration in patients in non-intensive-care settings.2 In addition to the general concern of care providers, RRT triggers commonly consist of changes in mental status, chest pain, bleeding, and vital sign (VS) deviations.3 Despite wide adoption, early data shed some doubt about the effectiveness of RRTs in preventing negative outcomes.4 However, recent studies indicate that RRT implementation decreases inhospital mortality and cardiac arrests,5 while the effect of these systems on overall mortality remains unclear. Safe disposition from the emergency department (ED) is one of the core tasks that emergency physicians are clinically, operationally and socially charged with.6 The medical decision-making in selecting admission to a regular floor setting versus an intensive care unit (ICU) is both difficult and high stakes,7 particularly considering demonstrated increases in morbidity and mortality in patients admitted to floor settings who deteriorate shortly thereafter.8,9 Patients with RRT activation within 24 hours of ED admission have been found to have a fourfold increased risk of in-hospital mortality.10 It has been suggested that RRT activations, such as these, are often the result of disposition errors, in addition to the scarcity of ICU beds.11 Previously, VS based clinical scores have been used to determine the need for rapid evaluation and eventually resuscitative interventions in inpatient settings. The ability of these scoring systems to predict intensive care admission and mortality is controversial,12-15 and their performance in anticipating deterioration appears to be poor.14,15 Recently, it has been suggested that the use of RRT criteria as an aid for disposition decisions may be helpful in identifying patients likely to trigger RRT activation and therefore at high risk for morbidity and mortality.10 We aim to describe patients having RRT activations within 12 hours of admission from the ED, including clinical characteristics, categorical diagnoses, and outcomes associated to RRT activations and to compare this group with matched controls admitted from the ED who did not activate RRTs. Determining risk factors of early RRT activation is of clinical, operational, and financial importance, as improved

medical decision-making regarding disposition would improve allocation of resources while potentially limiting morbidity and mortality.2,7

Volume XVII, no. 3 : May 2016

METHODS Study Design, Setting and Participants We obtained institutional review board approval at our institution, and we adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for reporting observational studies (Appendix A).16 We performed a case-control study of patients who presented to an academic tertiary care ED with 73,000 annual patient visits and were admitted to a non-ICU setting. Cases were obtained from an existing prospective, qualityimprovement database that collects information regarding all RRT events at our institution, including triggers and outcomes. We included all of the consecutive RRT events recorded from January 2009 to December 2012. Information included in the quality-improvement database was derived from incident reports made by the RRT leader, as well as from the electronic activation system (i.e., pager records). The RRT is composed of a critical care medicine fellow and attending in critical care medicine, senior level residents, ICU nurses, ICU pharmacists and respiratory therapists. At our institution a RRT is activated if the patient met one of the criteria in Figure. RRT activations were further restricted to only those that occurred within 12 hours of admission from the ED, where the time of admission was the electronic timestamp of the patient physically leaving the physical plant of the ED (a metric routinely tracked in our ED). Furthermore, in our institution boarding of patients is rare, and time from disposition-decision to the patient leaving the ED is consistently less than two hours. Controls were patients who presented from January 2011 to December 2012, that were recorded in a routinely maintained, comprehensive patient data warehouse that includes financial, clinical and operations data. From this data warehouse, a randomly selected group of patients who did not trigger a RRT activation within 12 hours of admission was matched to the case group. Patients included were admitted to the medical or surgical floors. We excluded patients admitted to the ICU. Patients were excluded if they did not consent to having their medical records reviewed for research purposes. Additionally, four patients’ datasets were incomplete and were excluded from the analyses. Cases and control identification and data analysis Patients who had a RRT within 12 hours of admission from the ED from January 2009 to December 2012 were identified from an existing prospective operations and qualityimprovement database. The dataset had information on time to RRT activation, reason for activating RRT (primary and secondary criterion), disposition after RRT was called (ICU, no location changed, expired), interventions performed by RRT

325

Western Journal of Emergency Medicine

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations A staff member is worried about the patient Acute and persistent declining oxygen saturations <90% Acute and persistent change in heart rate: <40 or >130 Acute and persistent change in systolic blood pressure <90mmHg Acute and persistent change in respiratory rate <10 or >28 per minute Acute chest pain suggestive of ischemia Acute and persistent change in conscious state New onset of symptoms suggestive of stroke Figure 1. Current clinical criteria used in the study institution for rapid-response team activation.

(intravenous fluids administration, blood transfusion, laboratory work, intubation, etc.), and healthcare provider who activated the RRT (nurse, resident physician, attending physician, other). Electronic medical records were extracted by a data quality analyst who provided patient demographic information (age, gender), date of ED visit, ED initial complaint, ED final diagnosis, first and last set of VS from the ED visit, disposition from the ED, and time of ED departure. Subsequently, the medical records were reviewed by three of the authors. The following data were extracted into a Microsoft Excel (Microsoft Corporation, Redmond, WA) spreadsheet: mortality within 72 hours and within one month; interventions including dialysis; need for interventions defined as interventional radiology, central line placement, chest tube placement or any operating room procedure. VS and ED diagnoses were manually reviewed for accuracy and consistency. A convenience sample of 10 charts were abstracted in duplicate by the abstractors, and after standardization of the methods and protocols for the data management each abstractor continued independently on different data sets. Data abstractors met frequently to guarantee quality assurance, maintaining the same methods and protocols. Data Analysis Each control was matched 1:1 to each of the 474 RRT cases by age (+/- 5 years), sex, and ED diagnosis, defined by the final primary ED diagnosis in the record. No secondary diagnosis or comorbidity matching was performed. VS collected included heart rate, respiratory rate, blood pressure, and oxygen saturation. These were measured and manually entered to the medical record by registered nurses (RN) at the time of the patientsâ&#x20AC;&#x2122; visits. VS were categorized as the lowest 10%, middle 80%, and highest 10% based on the distribution of all 948 patients included. These cut-offs were selected based on clinical significance of the higher extremes. Comparisons of categorized VS (first and last heart rate [HR], systolic blood pressure [SBP], O2 Saturation, and respiratory rate [RR] in the ED) between the cases and controls were evaluated using conditional logistic regression Western Journal of Emergency Medicine

models to account for the paired nature of the data. In terms of the timing of the VS records, we defined the first set was defined as the initial group in the record, while the last set was defined consequently as the last group of VS in the record. As described before, in our institution boarding of patients is non-existent; additionally, RNs are mandated to measure and record a set of VS prior to departure from the ED and notify the attending physician if VS are within RRT-activation criteria. Comparisons are reported as odds ratios (OR) with 95% confidence intervals (CI). We did not plan to perform sensitivity or secondary data analyses. Statistical analyses were performed using version 9.3 of the SAS software package (SAS Institute, Cary, NC). All tests were two-sided and p-values <0.05 were considered statistically significant. RESULTS During the study period (January 2009 to December 2012) there were 13,036 RRT activations at our institution. Of these, 1,259 were within 24 hours of admission from the ED (9.7%), and 522 within 12 hours of admission (4.0%). After removing patients with more than one RRT activation and patients who did not consent for medical record review (<5%), the final studied group comprised 474 cases that had RRT activations within 12 hours of ED admission (Table 1) and 474 age, gender and diagnosis matched controls. RRT patients The mean age of the 474 cases was 65.4 years old, 51.1% males. The most common reasons for activating a RRT were hypotension (21.5%), respiratory distress (18.4%), altered level of consciousness (17.7%), tachycardia (11.2%), oxygen saturation below 90% (9.9%), hypertension (4.2%), chest pain (3.4%), staff concern (2.7%), and seizure (2.3%). The remaining reasons included bradycardia, tachycardia, and arrhythmia and accounted for <9% combined. The care team member calling the RRT was a RN in 70.9%, a physician in 12.7%, and a combination of both a RN/MD in 9.7% of the calls. The caller was unknown 6.8% of the time. Regarding disposition of the patient following RRT activation, 62.5% were transferred to a higher level of care immediately as an intervention of the RRT and 71.4% were transferred to the ICU within 72 hours. The mortality rate was 4.5% at 72 hours and 17.1% at one month. The percentage of patients requiring an invasive intervention (operating room procedure, interventional radiology, central line placement, chest tube placement, intubation, dialysis, etc.) within 72 hours was 47.6%. Comparison RRT cases vs non-RRT matched controls Odds ratios for the comparison between patients with vs without RRT activation within 12 hours are in Table 2. We also compared disposition between the cases and controls. VS were evaluated using cut points (lowest 10%, middle 80% and highest 10%) based on the distribution of each VS

326

Volume XVII, no. 3 : May 2016

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations

Table 1. Characteristics of 474 patients with rapid-response team (RRT) activation within 12 hours of admission from the emergency department. Characteristic

Percentages (n=474)

Age +/- standard deviation

65.4 +/- 18.9

Female gender %

48.9%

Reason for activating RRT Hypotension

21.5%

Respiratory distress

18.4%

Altered level of consciousness

17.7%

Tachycardia

11.2%

Oxygen saturation below 90%

9.9%

Hypertension

4.2%

Chest pain

3.4%

Staff concern

2.7%

Seizure

2.3%

Other (bradycardia, tachypnea, arrhythmia, etc.)

8.7%

Care team member calling RRT Registered nurse (RN)

70.9%

Physician (MD)

12.7%

Combination of RN/MD

9.7%

Unknown

6.8%

Disposition after RRT Transferred to higher level of care

62.5%

Transferred to intensive care unit within 72 hours of RRT

71.4%

Mortality Mortality within 72 hours of RRT

4.5%

Mortality within 1 month of RRT

17.1%

Intervention (within 72 hours) Invasive intervention (operating room, interventional radiology, dialysis, intubation, central line, chest tube, etc.)

for all 948 patients in the matched analysis (474 cases and 474 controls combined). Patients who had RRT were more likely to have a first HR recorded in the highest 10% (OR 2.02, 95% CI [1.25-3.27]; were more likely to have a first oxygen saturation recorded in the lowest 10% (OR 2.25, 95% CI [1.42-3.56]); more likely to have a first respiratory rate recorded in the highest 10% (OR 2.92, 95% CI [1.73-4.92]). and had lower oxygen saturations (OR 2.25, 95% CI [1.423.56]) upon arrival to the ED. Additionally, at the time of disposition from the ED to the hospital floor, patients who activated RRT were more likely to have their last HR in the highest 10% percentile (OR 2.76, 95% CI [1.65-4.60]); last systolic blood pressure in the lowest 10% (OR 1.72, 95% CI [1.08-2.72]) and the highest 10% percentile (OR 1.82, 95% CI [1.19-2.80]), last respiratory rate in the highest 10% (OR 4.15, 95% CI [2.44-7.07]); and last oxygen Volume XVII, no. 3 : May 2016

47.6%

saturation in the lowest 10% percentile (OR 2.29, 95% CI [1.433.67]). A detailed description of the VS is depicted in Table 2. Using VS recorded in the ED and comparing the most extreme 10% measurements on both sides of the distribution (footnote of Table 2), we were able to calculate the values where VS in the ED had a higher likelihood to result in RRT activation within 12 hours of admission. These results showed that the most predictive VS were first and last ED heart rates in the highest 10% (OR 2.02 and 2.76), first and last ED RR in the highest 10% (OR 2.92 and 4.15), and first and last oxygen saturation in the lowest 10% (2.25 and 2.29). Outcomes Patients who triggered RRT activation were more likely to have an ICU admission within 72 hours (OR 38.49, 95% CI [19.03-77.87]), an invasive intervention within 72 hours (OR 327

Western Journal of Emergency Medicine

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations Table 2. Comparisons between patients with and without rapid-response team activations. Variables Disposition

Controls N(%)

Cases N(%)

Odds ratio (95% CI)

P-value*

ICU admission (<72h)

38.49 (19.03-77.87)

<0.001

Intervention (<72h)

5.49 (3.82-7.89)

<0.001

Mortality (<72h)

4.24 (1.60-11.24)

0.004

Mortality (<1 month)

4.02 (2.44-6.62)

<0.001

Lowest 10%

52 (11)

38 (8)

0.79 (0.50-1.24)

0.11

Middle 80%

391 (83)

373 (79)

reference

29 (6)

60 (13)

2.02 (1.25-3.27)

0.004

Vital sign First heart rate

Highest 10% First systolic blood pressure Lowest 10%

39 (8)

56 (12)

1.53 (0.98-2.40)

0.062

Middle 80%

389 (83)

365 (77)

reference

42 (9)

50 (11)

1.25 (0.81-1.92)

0.31

Lowest 10%

39 (8)

53 (11)

1.44 (0.91-2.27)

0.12

Middle 80%

405 (86)

353 (75)

reference

26 (6)

65 (14)

2.92 (1.73-4.92)

<0.001

Lowest 10%

33 (7)

65 (14)

2.25 (1.42-3.56)

<0.001

Middle 80%

388 (83)

352 (75)

reference

Highest 10%

49 (10)

54 (12)

1.19 (0.78-1.81)

0.43

Lowest 10%

59 (13)

36 (8)

0.61 (0.39-0.96)

0.031

Middle 80%

388 (82)

371 (79)

reference

25 (5)

65 (14)

2.76 (1.65-4.60)

<0.001

Highest 10% First respiratory rate

Highest 10% First oxygen saturation

Last heart rate

Highest 10% Last systolic blood pressure Lowest 10%

38 (8)

58 (12)

1.72 (1.08-2.72)

0.022

Middle 80%

393 (83)

351 (75)

reference

40 (8)

62 (13)

1.82 (1.19-2.80)

0.006

Lowest 10%

51 (11)

44 (9)

0.99 (0.63-1.56)

0.98

Middle 80%

395 (84)

350 (75)

reference

24 (5)

75 (16)

4.15 (2.44-7.07)

<0.001

Lowest 10%

31 (7)

61 (13)

2.29 (1.43-3.67)

<0.001

Middle 80%

398 (85)

358 (76)

reference

Highest 10% Last respiratory rate

Highest 10% Last oxygen saturation

Highest 10% 42 (9) 50 (11) 1.27 (0.82-1.95) 0.29 ICU, intensive care unit *P-value from conditional logistic regression model adjusted for age and sex. Footnote: Vital signs cut points for vital signs in the bottom 10, middle 80 and top 10% of the distribution are: First heart rate <66, 66120, >120 per minute; first systolic blood pressure <101, 101-167, >167mmHg, first respiratory rate <16, 16-26, >26 per minute, first oxygen saturation <92, 92-99, >99%, last heart rate <63, 63-112, >112 per minute, last systolic blood pressure <101, 101-157, >157 mmHg, last respiratory rate <16,16-24, >24 per minute, last oxygen saturation <93, 93-99, >99%. These cut points were chosen based on the distribution of each vital sign for all 948 patients.

Western Journal of Emergency Medicine

328

Volume XVII, no. 3 : May 2016

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations

5.49, 95% CI [3.82-7.89]), higher mortality at 72 hours (OR 4.24, 95% CI [1.60-11.24]), and higher mortality at one month (OR 4.02, 95% CI [2.44-6.62]).

underwent invasive treatment (dialysis, surgery, central line, chest tube, intubation or interventional radiology procedures) within 72 hours, with an odds ratio of 38.5 for ICU admission and 5.5 for invasive interventions. RRTs are currently being extensively studied as an integral component of overall patient management systems and procedures.2,10 Lovett et al. used 24-hour RRTs as a performance improvement tool for emergency physicians in a retrospective analysis fashion.17 Similarly, our study adds data to aid in the operational decision of the final disposition of the patient from the ED, providing a cognitive aid for identifying patients who may benefit from a higher level of care. Our results are congruent, albeit with different magnitudes, with recent data from Mora et al.10

DISCUSSION VS-based clinical scores have been used to predict risk of deterioration in the inpatient setting and the need to escalate interventions,4 but until recently the performance of these scores to predict risk of deterioration and RRT activation was not described. Mora et al.10 demonstrated recently that ED patients with abnormal VS, particularly heart rate and respiratory rate, are more likely to require RRT activation and had a subsequent higher mortality. This study adds to the recently published evidence, supporting the notion that using a set of predetermined discrete criteria (instead of continuous values as described by Mora et al.) can help identify patients who are at a high risk of RRT activation within 12 hours of admission from an ED. Furthermore, this is one of the largest single-center, case-control datasets of RRT activation published in the literature. Overall, the proportion of patients admitted through the ED that activate a RRT within 12 hours of admission is very small (4% of all RRT activations at our institution). Therefore, it can be inferred that most patients are admitted to the appropriate level of care and do not decompensate within the first 12 hours. Currently, there is no standard quality metric review analyzing these events, ranging from six (as in our institution) to 24 hours;17 therefore, we decided to adopt a 12hour framework as a pragmatic middle point. In our population, hypotension and respiratory distress were the most common causes of RRT activation. The mortality rate of patients who triggered early RRT was high, at nearly 5% at 72 hours from admission and 17.5% within 30 days. When compared to non-RRT patients, the odds ratios for mortality were 4.2 at 72 hours and 4.0 at one month, meaning higher mortality in those who activated an RRT; likely this difference in mortality is secondary to the severity of the underlying disease, and despite matching for age, gender and ED diagnosis, it is likely that more variables need to be considered when evaluating this outcome. Our data on mortality is similar to Mora et al.,10 in which they found four times greater mortality in patients with RRT activations within 24 hours for ED admission. Mora et al. report a higher respiratory rate and heart rate as the cause for RRT activation and an ICU transfer rate of 20%.10 We suspect the difference seen in ICU admissions compared to our study might be secondary to the availability of ICU beds. Supporting this notion, Stelfox et. al12 found an association between the number of ICU beds available and ICU admission but no change hospital mortality when the medical emergency team was activated. There was a significantly increased use of healthcare resources and utilization in the RRT group, as patients who triggered an RRT often required an escalation of care and 42% Volume XVII, no. 3 : May 2016

LIMITATIONS First, our study used a prospectively recorded quality dataset for the identification of study subjects, which is susceptible to recording and data-abstractions errors. To mitigate this limitation, we meticulously followed standardized guidelines for observational studies.16 Second, when working with medical records, we used data that were recorded for clinical purposes, and therefore some information was missing. Four patients had incomplete data and were excluded from the analyses. Third, our data abstractors were not blinded, and we did not perform statistical measurements of agreement to assess interrater reliability. Fourth, the data of this study were obtained from a single academic center with a complicated patient population and with particular disposition practices. This may limit the generalizability of our findings to other hospitals. Fifth, in matching cases with controls for diagnoses, we were unable to match for secondary diagnoses and comorbidities that may have been helpful in determining potential morbidity and mortality. Sixth, in terms of the RRT activation criteria (Figure), some of its components are not objectively defined (i.e., â&#x20AC;&#x153;A staff member is worried about the patientâ&#x20AC;?); therefore, they depend heavily on training, experience and institutional culture, which further limits generalizability. CONCLUSION After matching for age, gender and ED diagnosis, we found that abnormal heart rate, respiratory rate and oxygen saturation at the time of ED arrival and departure are predictive of RRT activation within 12 hours of admission. Patients who had a RRT activation had higher healthcare utilization with increased ICU admissions, invasive interventions, and mortality at 72 hours and one month. Among the reasons for activation, tachycardia, hypotension and respiratory distress were the most common. When compared to a population of admitted patients who did not trigger early RRT activation, the early RRT group was more likely to have a diagnosis of hemorrhage, hypotension

329

Western Journal of Emergency Medicine

Walston et al.

Vital Signs as Predictors of Rapid-Response Team Activations and sepsis and to have persistent tachycardia, low blood pressure and tachypnea while in the ED. These clinical findings may serve as aids for the identification of patients at risk of deterioration and who therefore may benefit from admission to a higher level of care. Further research is necessary to validate these findings in other practice settings. Determining risk factors of early RRT activation is of clinical, operational, and financial importance, as improved medical decision-making regarding disposition would improve allocation of resources while potentially limiting morbidity and mortality.

wards. Cochrane Database Syst Rev. 2007;(3):CD005529. 5. Maharaj R, Raffaele I, Wendon J. Rapid response systems: a systematic review and meta-analysis. Crit Care. 2015;19(1):254. 6. Asplin BR, Magid DJ, Rhodes KV, et al. A conceptual model of emergency department crowding. Ann Emerg Med. 2003;42(2):173-80. 7. Wiswell J, Tsao K, Bellolio MF, et al. “Sick” or “not-sick”: accuracy of System 1 diagnostic reasoning for the prediction of disposition and acuity in patients presenting to an academic ED. Am J Emerg Med. 2013;31(10):1448-52. 8. Considine J, Charlesworth D, Currey J. Characteristics and outcomes of patients requiring rapid response system activation within hours of emergency admission. Crit Care Resusc. 2014;16(3):184-9. 9. Liu V, Kipnis P, Rizk NW, et al. Adverse outcomes associated with delayed intensive care unit transfers in an integrated healthcare

Address for Correspondence: M. Fernanda Bellolio, MD, MS, Department of Emergency Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Email: Bellolio.fernanda@mayo.edu.

system. J Hosp Med. 2012;7(3):224-30. 10. Mora JC, Schneider A, Robbins R, et al. Epidemiology of early Rapid Response Team activation after Emergency Department admission. Australas Emerg Nurs J. 2016;19(1):54-61.

11. Litvak E, Pronovost PJ. Rethinking rapid response teams. JAMA. 2010;304(12):1375-6. 12. Dundar ZD, Ergin M, Karamercan MA, et al. Modified Early Warning Score and VitalPac Early Warning Score in geriatric patients admitted

Copyright: © 2016 Walston et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

to emergency department. Eur J Emerg Med. April 2015. 13. Armagan E, Yilmaz Y, Olmez OF, et al. Predictive value of the modified Early Warning Score in a Turkish emergency department. Eur J Emerg Med. 2008;15(6):338-340. 14. Gu M, Fu Y, Li C, et al. [The value of modified early warning score in predicting early mortality of critically ill patients admitted to emergency department]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue.

REFERENCES 1. Berwick DM, Calkins DR, McCannon CJ, et al. The 100,000 lives

2015;27(8):687-90. 15. Ho LO, Li H, Shahidah N, et al. Poor performance of the modified

campaign: setting a goal and a deadline for improving health care

early warning score for predicting mortality in critically ill patients

quality. JAMA. 2006;295(3):324-7.

presenting to an emergency department. World J Emerg Med. 2013;4(4):273-278.

2. Jones DA, DeVita MA, Bellomo R. Rapid-response teams. N Engl J Med. 2011;365(2):139-46.

16. Von Elm E, Altman DG, Egger M, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)

3. Rapid Response Team. Available at : https://www.icsi.org/

statement: guidelines for reporting observational studies. BMJ.

guidelines__more/catalog_guidelines_and_more/catalog_guidelines/

2007;335(7624):806-8.

catalog_patient_safetyreliability_guidelines/rrt/. Accessed on Nov 12, 2015.

17. Lovett PB, Massone RJ, Holmes MN, et al. Rapid response team activations within 24 hours of admission from the emergency

4. McGaughey J, Alderdice F, Fowler R, et al. Outreach and Early Warning Systems (EWS) for the prevention of intensive care

department: an innovative approach for performance improvement.

admission and death of critically ill adult patients on general hospital

Acad Emerg Med. 2014;21(6):667-72.

Western Journal of Emergency Medicine

330

Volume XVII, no. 3 : May 2016

Commentary

Trauma Triage and Trauma System Performance Gary Johnson, MD

University Hospital, SUNY Upstate Medical University, Department of Emergency Medicine, Syracuse, New York

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted January 26, 2016; Accepted February 3, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29900

[West J Emerg Med. 2016;17(3):331â&#x20AC;&#x201C;332.]

Trauma systems seek to provide complex medical care at the correct time and the correct place. During the past four decades numerous articles have been published that validate trauma systems from many points of view. Achievements of trauma systems include improvement in mortality and morbidity, efficiencies of care, and economic outcomes. Prehospital policy execution is intrinsic to trauma system performance. Trauma system criteria are relatively standardized. However, flexibility in emergency medical service (EMS) decision-making is commonly allowed. These decisions have major impacts on resource allocation, trauma center utilization, and patient outcome. In this edition, Holst, et al1 reviewed adult emergency department (ED) trauma deaths as reported in the 2010 National Emergency Department Sample. They recorded the association of these deaths to trauma or non-trauma center designation, as well as geographic and patient demographics including rural vs urban site, gender, and patient income data. They found that one half of all trauma ED deaths nationally and one third of ED urban trauma patients died in non-trauma centers. Both elderly trauma deaths and deaths due to falls more frequently occurred in non-trauma centers. Like most studies describing trauma system performance, this is a retrospective review taken from a large database. Therefore, causation of outcome cannot be directly attributed to undertriage. However, the magnitude of the non-trauma center death rate merits further investigation. Trauma system literature often describes the undertriage of trauma patients with regard to trauma center designation. The American College of Surgeons has a goal for trauma systems to achieve less than 5% undertriage. However, studies frequently estimate a much higher rate. Like the study published in this journal, undertriage rates are higher with elderly patients.2 Centers for Medicare/Medicaid Services claims also identify a high rate of undertriage, and these non-trauma center visits are associated with worse outcomes. Staudenmayer, et al3 found that in California undertriage varied substantially by region. Patient factors such as age greater than 55 years, female sex,

Volume XVII, no. 3 : May 2016

greater number of co-morbidities and a fall mechanism led to a higher rate of undertriage. Overtriage to trauma centers commonly occurs as well and contributes to resource mismatch. Tang, et al4 found that a significant number of trauma transfers were discharged within 24 hours. Tarima5 found that non-trauma centers were transferring 98% of their trauma patients including those with low Injury Severity Scores (ISSs) and high Glasgow Coma Scale (GCS) scores. Multiple factors contribute to trauma system formation. Population distribution and medical resource allocation strongly affect system performance. Commonly, injuries happen close to the home of the trauma victim. However, motor vehicle accidents have a higher rate of occurring farther from home. Trauma system volume is associated with improved trauma patient mortality and length of stay. Minei, et al6 found that increasing trauma center volume was also associated with more ventilator-free days and less severe organ failure outcomes. This triage process predictably leads to a large number of EDs without significant trauma experience, and hence makes them more likely to transfer out the mild to moderately injured patients they do see. As the cycle repeats, the distribution of trauma care skews even more toward trauma centers. Mohan, et al7 confirmed that emergency physicians working in non-trauma centers rarely encounter trauma patients with moderate to severe injuries. Trauma system improvement processes may initially realize gains that are hard to sustain. Winchell, et al8 found that trauma system evaluation planning committee consultations commonly improved outcomes. However, the gains were not self-sustaining after consultation. Trauma system literature is rich with retrospective analyses of trauma system performance. Holst, et al1 published in this edition contributes to this literature in a meaningful fashion by looking at all ED trauma deaths rather than trauma center mortality/morbidity alone. The study supports the body of literature that shows undertriage has a significant impact on system performance. Particular attention needs to be paid

331

Western Journal of Emergency Medicine

Trauma Triage and Trauma System Performance Johnson to the consistent theme of undertriage of elderly trauma and fall-related trauma. Moving forward, each system needs to examine its performance to the community it serves beyond the bounds of the trauma centers themselves. Regional trauma systems need to turn toward prospective validation of outcomes so that system structure and protocols can best serve our populations.

REFERENCES 1. Holst JA, Perman SM, Capp R, et al. Undertriage of Trauma-Related Deaths in U.S. Emergency Departments. West J Emerg Med. 2016;17(3). 2. Mohan D, Barnato AE, Rosengart MR, et al. Triage patterns for Medicare patients presenting to nontrauma hospitals with moderate or severe injuries. Ann Surg. 2015;261(2):383-9. 3. Staudenmayer K, Lin F, Mackersie R, et al. Variability in California triage from 2005 to 2009: A population-based longitudinal study of severely injured patients. J Trauma Acute Care Surg. 2014;76(4):1041-7. 4. Tang A, Hashmi A, Pandit V, et al. A critical analysis of secondary overtriage to a Level 1 trauma center. J Trauma Acute Care Surg. 2014;77(6):969-73. 5. Tarima S, Ertl A, Groner JI, et al. Factors associated with patients

Address for Correspondence: Gary Johnson, MD, SUNY Upstate University Hospital, 2nd Fl, 550 East Genesee Street, Syracuse, NY 13202. Email: johnsong@upstate.edu.

transferred from undesignated trauma centers to trauma centers. J Trauma Acute Care Surg. 2015;79(3):378-85. 6. Minei JP, Fabian TC, Guffey DM, et al. Increased trauma center

volume is associated with improved survival after severe injury. Ann Surg. 2014;260(3):456-65. 7. Mohan D, Barnato AE, Rosengart MR, et al. Trauma triage in the emergency departments of nontrauma centers: An analysis of individual physician caseload on triage patterns. J Trauma Acute

Copyright: ÂŠ 2016 Johnson. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

Western Journal of Emergency Medicine

332

Care Surg. 2013;74(6):1541-7. 8. Winchell RJ, Sanddal N, Ball J, et al. A reassessment of the impact of trauma system consultation on regional trauma system development. J Trauma Acute Care Surg. 2015;78(6):1102-10.

Volume XVII, no. 3 : May 2016

Infectious Disease

Inadequate Sensitivity of Laboratory Risk Indicator to Rule Out Necrotizing Fasciitis in the Emergency Department Elizabeth Burner, MD, MPH* Sean O. Henderson, MD* Guenevere Burke, MD† Jeffrey Nakashioya, BS* Jerome R. Hoffman, MD, MA‡

*Keck School of Medicine of the University of Southern California, Department of Emergency Medicine, Los Angeles, California † George Washington University, Department of Emergency Medicine, Washington, DC ‡ University of California Los Angeles, Department of Emergency Medicine, Los Angeles, California

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted October 28, 2015; Revision received January 12, 2016; Accepted February 1, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29069

Introduction: Necrotizing fasciitis (NF) is a life-threatening illness, particularly when surgical debridement is delayed. The Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score was developed to identify patients at higher risk for NF. Despite limited information in this regard, the LRINEC score is often used to “rule out” NF if negative. We describe the sensitivity of the LRINEC score in emergency department (ED) patients for the diagnosis of NF. Methods: We conducted a chart review of ED patients in whom coding of hospital discharge diagnoses included NF. We employed standard methods to minimize bias. We used laboratory data to calculate the LRINEC score, and confirmed the diagnosis of NF via explicit chart review. We then calculated the sensitivity of a positive LRINEC score (standardly defined as six or greater) in our cohort. We examined the role of patient characteristics in the performance of the LRINEC score. Finally, we performed sensitivity analyses to estimate whether missing data for c-reactive protein (CRP) results were likely to impact our results. Results: Of 266 ED patients coded as having a discharge diagnosis of NF, we were able to confirm the diagnosis, by chart review, in 167. We were able to calculate a LRINEC score in only 80 patients (due to absence of an initial CRP value); an LRINEC score of 6 or greater had a sensitivity of 77%. Sensitivity analyses of missing data supported our finding of inadequate sensitivity to rule out NF. In sub-analysis, NF patients with concurrent diabetes were more likely to be accurately categorized by the LRINEC score. Conclusion: Used in isolation, the LRINEC score is not sufficiently sensitive to rule out NF in a general ED population. [West J Emerg Med. 2016;17(3):333–336.]

INTRODUCTION Necrotizing fasciitis (NF) is a life-threatening infection with high mortality. Because NF can be misdiagnosed as a less lethal mimic, such as cellulitis and abscess, efforts have been made to identify clinical features that could help clinicians accurately diagnose NF and avoid delays to surgical debridement.1 Prior retrospective studies have shown certain Volume XVII, no. 3 : May 2016

laboratory values, particularly an extremely elevated leukocyte count and a low sodium concentration, are associated with NF.2 These abnormal values might help clinicians distinguish NF from less severe soft-tissue infections. The Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score was developed in a large cohort of admitted patients to identify patients at higher risk for NF.3 Patients are assigned a LRINEC 333

Western Journal of Emergency Medicine

Burner et al.

Inadequate Sensitivity of the Laboratory Risk Indicator score based on serum sodium, glucose, creatinine, c-reactive protein (CRP), leukocyte count and hemoglobin. Scores range from 0 to 13; a score 6 or greater was associated with a high risk of NF, and a score of 8 or greater with a very high risk. The LRINEC score can be easily misapplied, however. The score was not designed to exclude NF in patients with a low-risk score, and case reports and small studies externally validating the score failed to replicate the high sensitivity and negative predictive value reported in the initial paper.4-8 Additionally, the sensitivity of the score has not been addressed specifically among emergency department (ED) patients. We conducted this retrospective study to determine the sensitivity of the LRINEC score in ED patients with a confirmed diagnosis of NF and examine the role of patient factors in the score’s sensitivity. We also performed secondary analyses to estimate how missing laboratory values impacted our results.

met, were classified as “unconfirmed” and were excluded. LRINEC Scores During laboratory value review, abstractors were blinded to the final “confirmation” of NF in any individual case. Only the first value for each laboratory test was collected, eliminating conflicting data. Abstractors coded a result as “missing” if there was no result in the first 48 hours of the hospital course; this occurred only with CRP; all other laboratory tests were present for every patient. These cases were excluded as no score could be calculated. Patient Characteristics Patient age, gender and inpatient mortality were collected from administrative data. Past medical history was abstracted from medical history and final diagnoses on operative reports, discharge summaries and death notes.

Primary Study: Sensitivity of the LRINEC Score in ED Patients Chart Abstraction Methods We used standard abstraction methods to minimize bias.9 All abstractors received training. Medical students blinded to the study hypothesis calculated the LRINEC score after two training sessions; the lead author also reviewed their first results to ensure reliability and accuracy, and a coding guide was made available to abstractors (Appendix 1). The lead author reviewed 10% of cases, and with a Kappa calculated for abstracted variables. Diagnosis Confirmation The ICD-9 diagnosis of NF was confirmed if any of the following criteria were met: 1) NF was a diagnosis on the hospital discharge or death summary; 2) NF was confirmed at surgery, as documented by operative report; or 3) fascial necrosis was documented on an anatomic pathology specimen. Patients with ICD-9 coding of “NF” but in whom none of these criteria were Western Journal of Emergency Medicine

Primary Study: Sensitivity of LRINEC score, and Association with Patient Factors A LRINEC score was calculated for each patient (Figure), and the overall sensitivity of the score was calculated using STATA version13, with a binomial model for confidence intervals. We included only patients with complete data. We used chi-squared tests and t-tests to examine the sensitivity of the LRINEC score when stratifying by patient factors of age, gender, inpatient mortality and history of cirrhosis, intravenous drug use or diabetes, known risk factors for developing NF. Secondary Study: Analysis of Missing Data All cases without laboratory results to calculate a LRINEC score were missing a CRP, so we performed sensitivity analyses to determine how this impacted our results. We calculated a LRINEC score for each patient with missing data based on the assumption that the missing CRP value “would have been” positive in 50%, 77% (the sensitivity of the LRINEC score found in our cohort), or 100% of cases. Because of the large contribution of CRP to the LRINEC score, we assumed that if the CRP was positive, the LRINEC score would also be positive. We

LRINEC Score

METHODS The study was approved by the institutional review board prior to initiation. The study cohort consisted of patients evaluated in the ED of Los Angeles County+University of Southern California Medical Center (LAC+USC) with NF. The LAC+USC ED is an urban, academic, tertiary care hospital. Patients were identified by search of all International Statistical Classification of Diseases and Related Health Problems, Ninth Revision (ICD-9) hospital discharge codes between April 2003 and April 2013; charts of all patients coded as having NF (728.86) were then reviewed, and further categorized as either “confirmed” or “unconfirmed” (see below.) Patients who were not initially evaluated in the ED or who developed NF postoperatively after admission were excluded.

Count of cases at each LRINEC score

16 14 12 10 8 6 4 2 0 0

LRINEC Score

Figure. Count of cases at each Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score.

334

Volume XVII, no. 3 : May 2016

Burner et al.

Inadequate Sensitivity of the Laboratory Risk Indicator

calculated the sensitivity of the score under at each of these assumptions for CRP. RESULTS Primary Study A computerized records search identified 316 patients between April 2003 and April 2013 with ICD-9 discharge diagnosis of NF. We excluded 47 cases that were not admitted through the ED and three cases that developed NF in the post-operative period while inpatient. NF was “confirmed” by chart review in 167 cases, with 100% interrater agreement (Kappa=1.0), but only 80 of these patients had a CRP documented in the first 48 hours of presentation. The interrater reliability for a positive LRINEC score was excellent (Kappa=0.91), as was that for history of diabetes (K=0.84). Interrater reliability for history of cirrhosis and intravenous drug use was good (K=0.78 for both). Demographic characteristics and percentage of cases with each LRINEC score are shown in Table. In the final study cohort, the overall sensitivity of the LRINEC score was 77% (CI [66-85]). Patients with diabetes were more correctly categorized by the score than patients without diabetes (85% vs 67%, p=0.04). There was no difference in the score’s sensitivity when patients were stratified by age, gender, inpatient mortality nor history of cirrhosis or intravenous drug use. Secondary Analyses If the CRP and resultant LRINEC score are assumed to be “positive” in 50%, 77% or 100% of cases missing data, the sensitivity would be 63% (CI [55-70]), 76% (CI [69-82]) or 89% (CI [83-93]), respectively. Our analysis of missing data gives a range of sensitivities of as low as 55% to as high as 93%. DISCUSSION In our population of ED patients with NF, the LRINEC score had a measured sensitivity of 77%, substantially lower than the 91% reported by Wong.3 Our population differs from the Wong cohort in that our patients came exclusively from the ED and were younger, more frequently male and had a higher mortality rate. Our finding that the LRINEC score, applied in isolation, would miss over 20% of cases of NF is consistent with reports of sensitivities between 68% and 80% in smaller studies based in surgical referral centers.6-8,10 While our sensitivity analyses, based on realistic possibilities regarding missing CRP data, suggest that the sensitivity of the LRINEC score could range between 55% and 93%, it is likely that the true value is less than 80%. Our calculated sensitivity for LRINEC may be artificially high, since in clinical practice this score is often used to decide whether a patient needs further work up or surgical management. Patients with NF and a falsely Volume XVII, no. 3 : May 2016

negative LRINEC score are less likely to have the diagnosis confirmed through pathologic specimens (and would be missed by our study methodology). LIMITATIONS The focus of this study is the sensitivity of the LRINEC score in ED patients of one hospital; the findings may have limited generalizability. The use of a single ICD-9 code may miss cases of NF due to miscoding. The validity of any chart review is threatened to the extent that it relies on data that is frequently missing, internally inconsistent, and/or poorly gathered. To minimize this, we employed standard retrospective chart review methodology. Reviewers were trained and were largely blinded to our study hypothesis and outcome data. We conducted duplicate review to assess reliability, used precise definition of both independent and dependent variables, and relied on only initial laboratory data to decrease inconsistency. We minimized issues with missing data, except with regard to CRP. To address the high number of missing CRP values, we performed sensitivity analyses covering reasonable assumptions about how the missing values might have affected our results. As no possible values for the missing data could have produced a high sensitivity of the LRINEC, our primary conclusion – that a normal LRINEC score should not by itself be used to rule out NF – remains qualitatively unchallenged regardless of what these missing CRP values might have been. Additionally, it is possible that spectrum bias is present, and that the LRINEC score performs better in the most severe cases; however, an ideal diagnostic adjunct would aid a clinician in identifying the correct diagnosis in subtle cases. CONCLUSION In this cohort, the LRINEC score with the standard cut-off of six would miss over 20% of cases of NF. Our results suggest that clinicians must maintain a high index of

Table. Characteristics of patients with confirmed necrotizing fasciitis. n=81 % Male

80%

Inpatient mortality

35%

Age (mean, SD)

47.5 (1.4)

History of diabetes

49%

History IVDU

18%

History of cirrhosis

LRINEC score positive

76%

LRINEC score negative 34% IVDU, intravenous drug user; LRINEC, laboratory risk indicator for necrotizing fasciitis

335

Western Journal of Emergency Medicine

Burner et al.

Inadequate Sensitivity of the Laboratory Risk Indicator suspicion, and avoid the trap of using a “normal” LRINEC score, in isolation, to dismiss the diagnosis. While a lower cutoff might improve the sensitivity, the accompanying cost to specificity is not knowable in this study.

REFERENCES 1. Alayed KA, Tan C, Daneman N. Red flags for necrotizing fasciitis: a case control study. Int J Infect Dis. 2015;36:15-20. 2. Chan T, Yaghoubian A, Rosing D, et al. Low sensitivity of physical examination findings in necrotizing soft tissue infection is

ACKNOWLEDGEMENTS EB lead the study design, data collection, data analysis, wrote the first draft of the manuscript, and oversaw editing. She takes full responsibility for accuracy of data and findings. SOH participated in study design, manuscript preparation and editing. GB participated in study design, data collection, data analysis and editing. JN participated in data collection, data analysis and editing. JH participated in study design, manuscript preparation and editing. Dr. Burner’s time to work on this project was funded in part by grants NIDDK 1F32DK094547 and SC-CTSI KL2TR000131.

improved with laboratory values: a prospective study. Am J Surg. 2008;196:926-30. 3. Wong CH, Khin LW, Heng KS, et al. The LRINEC (Laboratory Risk Indicator for Necrotizing Fasciitis) score: a tool for distinguishing necrotizing fasciitis from other soft tissue infections. Crit Care Med. 2004;32:1535-41. 4. Gausepohl JS and Wagner JG. Survival from cervical necrotizing fasciitis. West J Emerg Med. 2015;16:172-4. 5. Wilson MP and Schneir AB. A case of necrotizing fasciitis with a LRINEC score of zero: clinical suspicion should trump scoring systems. J Emerg Med. 2013;44:928-31. 6. Hodgins N, Damkat-Thomas L, Shamsian N, et al. Analysis of the increasing prevalence of necrotising fasciitis referrals to a regional plastic surgery unit: a retrospective case series. J Plast Reconstr Aesthet Surg. 2015;68:304-11.

Address for Correspondence: Elizabeth Burner, MD, MPH, 1200 N State St, Room 1011, Los Angeles, CA 90033. Email: eburner@ usc.edu.

7. Holland MJ. Application of the Laboratory Risk Indicator in

8. Swain RA, Hatcher JC, Azadian BS, et al. A five-year review of

referral centre. Anaesth Intensive Care. 2009;37:588-92. necrotising fasciitis in a tertiary referral unit. Ann R Coll Surg Engl. 2013;95:57-60. 9. Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg

Copyright: © 2016 Burner et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

Western Journal of Emergency Medicine

Necrotising Fasciitis (LRINEC) score to patients in a tropical tertiary

Med. 1996;27:305-8. 10. Thomas AJ and Meyer TK. Retrospective evaluation of laboratory-

336

based diagnostic tools for cervical necrotizing fasciitis. Laryngoscope. 2012;122:2683-7.

Volume XVII, no. 3 : May 2016

Review Article

Ten Tips for Engaging the Millennial Learner and Moving an Emergency Medicine Residency Curriculum into the 21st Century Shannon L. Toohey, MD* Alisa Wray, MD* Warren Wiechmann, MD, MBA* Michelle Lin, MD† Megan Boysen-Osborn, MD, MHPE*

*University of California Irvine, Department of Emergency Medicine, Orange, California † University of California San Francisco, Department of Emergency Medicine, San Francisco, California

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted January 21, 2016; Accepted March 18, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.29863

Introduction: Millennial learners are changing the face of residency education because they place emphasis on technology with new styles and means of learning. While research on the most effective way to teach the millennial learner is lacking, programs should consider incorporating educational theories and multimedia design principles to update the curriculum for these new learners. The purpose of the study is to discuss strategies for updating an emergency medicine (EM) residency program’s curriculum to accommodate the modern learner. Discussion: These 10 tips provide detailed examples and approaches to incorporate technology and learning theories into an EM curriculum to potentially enhance learning and engagement by residents. Conclusion: While it is unclear whether technologies actually promote or enhance learning, millennials use these technologies. Identifying best practice, grounded by theory and active learning principles, may help learners receive quality, high-yield education. Future studies will need to evaluate the efficacy of these techniques to fully delineate best practices. [West J Emerg Med. 2016;17(3):337–343.]

INTRODUCTION Medical education is rapidly changing as millennial learners place priority on self-directed learning, using a digital library of videos, podcasts, social media, and other online resources.1-3 The status quo of routine, classroom lectures does not meet millennial learner needs. We propose that a learner-centered approach with active engagement optimizes knowledge retention. While research on effective teaching strategies in graduate medical education is still lacking, many aspects of educational theory are well established. These 10 tips for engaging the millennial learner and moving an emergency medicine (EM) residency into the 21st century incorporate these principles and discuss working examples to advance one’s residency curriculum to better engage today’s learners. Volume XVII, no. 3 : May 2016

Tip #1: “Flip” your curriculum While textbooks and traditional references provide residents with a foundation of basic concepts and knowledge,4,5 classroom time should avoid regurgitating textbook facts and figures. Instead, these didactic sessions should focus on clarifying points of confusion, explaining high-order concepts, or challenging learners to apply concepts to clinical scenarios.6,7 A “flipped classroom” requires residents to complete basic learning at home. Then, they attend didactic sessions that promote active learning. This model ensures that residents are taught both basic and advanced concepts, while remaining engaged. Studies on the “flipped classroom” model in undergraduate and graduate health professions education have been small, but have demonstrated improved learner satisfaction and either

337

Western Journal of Emergency Medicine

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century improved or similar knowledge acquisition when compared to traditional methods of instruction.8-16 We see an additional benefit of the flipped classroom in promoting lifelong learning, but this has not been studied. Flipping the residency program curriculum is multifaceted and incorporates many existing curricular practices. Residency programs must first outline the content they expect their residents to complete asynchronously. We call this learner responsible content (LRC). Traditionally, programs recommend a textbook reading schedule; however, as digital media such as blogs, podcasts, and reference websites permeate the Internet17 and become increasingly popular with EM learners,18,19 these sources should be included in the LRC. Because some of these digital resources are not written by experts, peer-reviewed, or consistently reliable,20 it is imperative that programs guide their learners to dependable sources or use quality indicators to evaluate educational blogs and podcasts.21,22 Life In The Fast Lane Review and the Academic Life in Emergency Medicine (ALiEM) Approved Instructional Resources (AIR) Series are resources that curate high-quality content.22,23 Once the curriculum of LRC is established, in-classroom didactics should then build on the foundation of LRC through such application exercises as problem-based learning (PBL), audience response systems (ARS), and team-based learning (TBL). This allows residents to pursue higher order Bloom taxonomy levels24 of learning. By aligning didactic content with LRC, the learner has the impetus to complete the assignments and may benefit from improved knowledge retention, as material is repeated, built upon, and applied.25 Tip #2: Incorporate individualized interactive instruction into a learning management system The flipped curriculum blends nicely with asynchronous learning. In 2008, the Accreditation Council for Graduate Medical Education’s (ACGME) EM Residency Review Committee (RRC) and Council of Residency Directors (CORD) recommended that a program’s ideal didactic curriculum should include a mixture of asynchronous and synchronous learning.7 The EM RRC began allowing up to 20% of the didactic curriculum to be replaced by asynchronous learning, termed individualized interactive instruction (III). Resident participation in III requires that learning be monitored by the program director, evaluated, overseen by faculty, and monitored for efficacy.26,27 A learning management system (LMS), such as Schoology, Canvas, or BlackBoard, can organize a program’s LRC, providing a platform for learners to access a menu of videos, interactive educational modules, articles, quizzes, recorded didactics, and other assignments. Furthermore, an LMS can help meet the above four III criteria. Costs for implementing an LMS at a residency program can vary from free (Schoology) to a per-learner cost (Blackboard, Canvas) that may be covered by a university’s contract. The benefits of III include Western Journal of Emergency Medicine

identifying and addressing an individual learner’s needs; encouraging practice-based learning and improvement and lifelong learning; and the ability to use expert content for asynchronous materials.7 Tip #3: Incorporate modern didactic approaches Small group learning (SGL) is an ideal format for highyield, interactive didactic sessions. Its benefits include providing opportunities for assessment, clarification of knowledge gaps, improved learner engagement, teamwork, and a varied didactic format. Anecdotally, SGL uses more resources than are feasible for routine use; however, in our experience it requires similar preparation to a standard, high-quality lecture. Between faculty, fellows, local alumni, and chief residents, programs should have a sufficient number of instructors. Instructors should have clear objectives for the learning sessions, as well as information on how to best facilitate the session. The ideal SGL session is selfdirected by learners, with faculty available to provide additional information, clarify any points of confusion, ask open-ended questions, and ensure the group remains on track.28-32 Small group learning sessions can be carried out through a TBL or PBL approach with application exercises or may focus on areas such as board review or visual diagnosis. While instructors can create content de novo, there are many sources to acquire existing learning materials, such as MedEdPORTAL,33 CORD Teaching Cases,34 as well as many other textbook and Internet sources. In undergraduate medical education, PBL has been shown to be more effective than standard instruction in improving clinical performance, knowledge and reasoning,35 while TBL showed improvement of examination scores.36 Its efficacy has not been as well studied in graduate medical education. Problem-based learning: A group works through a clinical vignette in a small group setting,37 encouraging self-directed learning and self-evaluation. Facilitators are available as expert “consultants” to answer questions.38 Team-based learning: Previous research suggests that millennial learners prefer team-oriented projects and learning.2 Team-based learning exercises require only a single instructor. After completing LRC, learners participate in an individual readiness assurance test (iRAT), followed by a group readiness assurance test (gRAT), ensuring learners “know” the content.39 Groups then apply their knowledge during a group application exercise where the learners “show how”.39 A facilitator reviews learning points and clarifies any confusion. Teambased learning encourages teamwork and communication, improves learning outcomes and examination scores, and develops lifelong learning skills.6 Tip #4: Improve lecture efficacy by keeping it high yield and brief While lectures have a negative reputation in modern learning theory, they should not be eliminated entirely from residency didactics. Lectures are practical, as they require only

338

Volume XVII, no. 3 : May 2016

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century

one instructor for an almost unlimited number of learners. However, rather than regurgitate textbook information, lecturers should use the time to build on concepts that have been covered in textbooks or other LRC. Because experts suggest that learners only absorb 5% of lecture material and have an attention span of only 15-20 minutes,40,41 decreased lecture time, coupled with methods that promote active learning (pause procedures, simulation, small group discussion), should improve retention.6 For example, our residency program frequently divides the 45-minute “lectures” into a 15-minute lecture, followed by a 30-minute SGL exercise. Alternatively, a short lecture, clarifying commonly asked questions or reiterating concepts, could follow a TBL exercise. Instructors can better engage learners through commitment activities, such as audience response software (ARS) and pause procedures, such as the “one-minute paper” approach.6 Wolff and colleagues provide an outstanding overview of multiple methods for engaging learners in didactic lectures (Table).6 Further methods that ensure high-quality lectures include using a manageable scope of content, clear objectives, and case-based scenarios.42 Not all topics, however, are best taught using SGL. For instance, more dense material can be covered in a required reading or multimedia assignment, followed by a high-yield board review lecture with ARS or a game format, both of which have been shown to increase engagement, motivation, and retention.43-49

Simulation has been shown to be an effective teaching strategy, with studies demonstrating improved quality of care, knowledge retention, comfort in performing procedures, and performance in repeated simulated medical scenarios.54,55 While the initial investment of a simulation center can be quite high, once the physical space and resources exist, the creation of cases and content can be developed by faculty and residents, or found in any number of online databases. Sharing a simulation center with other departments or local institutions may be an option to help finance it. Tip #7: Think of any moment as a potential teaching moment Traditionally, teaching is restricted to on-shift (clinical) learning and in-classroom didactics, with LRC serving as supplemental learning material. With modern technology (smartphones, tablets, and laptops), educators have the opportunity to create educational opportunities throughout the day beyond these traditional teaching venues. For example, web services can deliver pre-scheduled or automated announcements via email,56 text messages, tweets, or Facebook posts. These might include images, key facts, or even brief quizzes that residents can access while on the go. Tablets that display board review questions or visual diagnosis quizzes can be placed in charting rooms (i.e., a “fact board”), so that learners can engage in a steady stream of information throughout their shift, although we have found that medical students, nurses, and technicians seemed to engage in the fact board more often than our residents. Publishing periodic educational content on a residency-run blog, which can be likened to an online magazine or journal, can be used to post interesting cases, or discuss diseases or local relevant medical news. These resources should be updated frequently so that residents consistently have the most useful information available.

Tip #5: Have a central, cohesive technology plan for resident education Establishing a central agreed-upon technology, such as iPads or other tablet devices, can be essential to ensuring consistent access to online resources and electronic textbooks.50 The tablet provides access to the program’s LMS, which provides an organized, real-time updated database of the LRC and monitoring of modules, projects, and quizzes. Mobile access to LRC can encourage completion of assigned reading by providing more convenient, portable access.51 Our residency program uses Schoology, a free LMS, which is functionally advanced and user friendly for both administrators and learners. Tip #6: Use simulation to its full potential Simulation, broadly defined, includes a number of interactive tools, such as human simulation with “standardized patients,” oral board cases, games, computer-based training exercises, or the more realistic high fidelity, mannequinbased simulators. Simulation provides team-based, engaging, relevant, active learning with emotional connection. It is an ideal supplement for residency curriculum because it allows educators to expose residents to uncommon or critical situations in a more realistic form.52,53 Furthermore, simulation exercises can be another tool in the flipped classroom model for residents to demonstrate and solidify lessons from their LRC. Volume XVII, no. 3 : May 2016

Tip #8: Use technology for more effective, formative feedback Traditional feedback exists via Likert scale-based competency assessment, with comment boxes for formative feedback. The implementation of the ACGME Milestones was intended to provide a more consistent assessment process, to accurately describe a residents’ performance and allow tracking of improvement.26,57 However, it still provides rote, non-personalized, summative feedback that does not necessarily help the resident improve specific skills. Furthermore, a traditional semi-annual format of evaluations may diminish faculty recall of resident performance. Millennial learners value prompt, “real-time” feedback.2 Evaluation apps and programs can be used to encourage more frequent assessment, which may lead to real-time feedback and a more accurate assessment of a resident’s overall performance. Real-time feedback may result in more formative feedback, which can drive further learning.58 We use

339

Western Journal of Emergency Medicine

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century Table. Techniques to engage learners in didactive lectures.* Technique

Description

Commitment activities

Learners are posed a question and must “commit” to a response. Examples: audience response system, iRATs, multiple choice questions, visual diagnosis.

Pause procedures

A brief “pause” in a didactic session to allow learners to clarify and assimilate learning points. Examples: One-minute paper where learners spend one-minute writing a response to a question posed by the instructor (also a commitment activity) and the Muddiest Point where learners reflect on and share on primary points of confusion

Jigsaw

Learners are tasked to become an expert on one of many small parts of a particular topic. Each expert teaches his part of the topic to the other learners. In the end, the topics all come together like a jigsaw puzzle.

Role-play

Specific to case scenarios, learners play a part (for example patient, physician or family) to work through a case and understand concepts and theories.

Think-pair-share

Instructors ask learners a question. Learners are encouraged to “think” about their response. Then, they “pair” with a neighbor and “share” their answers with each other. iRAT, individual readiness assurance test *Adapted from reference 6.

Instant Eval, a program built for milestone assessment during or after a shift. (Other options include New Innovations and MyEvaluations.com.) It can be especially challenging for supervisors to provide feedback to learners during procedures and resuscitations. Providing specific, real-time, formative feedback can be difficult for faculty, as their attention is often split between managing the patient and observing the resident. Faculty members also may be hesitant to verbalize feedback in front of patients in order to preserve confidence in their providers. Furthermore, busy EM faculty may not observe the entirety of a procedure, missing subtle technical errors. Because of these reasons, procedural feedback is typically provided afterwards, often resulting in recall bias and focusing on the success or failure of the procedure as the major outcome. Meanwhile technical difficulty and other factors are forgotten. One potential solution is first-person video recording using optical head-mounted displays, such as Google Glass or Go Pro cameras. These devices can capture the entire procedure and are non-obstructive to patient care. Faculty can review these recordings, and annotate them using an application such as Coach’s Eye or notes with associated timestamp marks. This allows more formative, substantial evaluation which could lead to resident improvement on procedural skills,59,60 communication during resuscitations, and general approach to an acutely ill patient. While only a few EM residency programs use video recordings for feedback, these programs find that it the most effective method for providing feedback to residents.61 First-person recording has associated HIPAA compliance issues, as many of the patients undergoing difficult procedures and resuscitations cannot consent to recording. Many programs already record traumas and resuscitations for quality improvement, and this first-person recording can be used in the same way. However, it will require approval from hospital Western Journal of Emergency Medicine

compliance committees, as well as appropriate software to ensure encrypted recordings. Tip #9: Teach educators to be proficient in a variety of educational technologies through professional development opportunities Educators should be proficient in a variety of educational approaches both involving and not involving technologies in order to best accomplish particular educational objectives.26 Professional development sessions can help faculty members use effective bedside teaching and lecturing, audience response systems, and small group learning to enhance resident learning experiences.62 Active experimentation and application of various techniques in these sessions allows educators to experience, troubleshoot, and observe the operational details. The culture of technology can be further promoted through educational fellowship programs. The University of California, Irvine is the first to offer a fellowship in Multimedia Design and Educational Technologies (MDEdTech). This program offers fellows the opportunity to participate in residency curriculum design, creation and curation of educational technologies for residents and medical students, and the use of educational technologies for patient education. Tip #10: Identify an instructional technology champion in the department With the ongoing growth and evolution of new instructional technologies and digital resources, it can be challenging for any residency program to stay current and abreast of these developments. A program should identify a faculty or resident (such as a chief resident of technology) to monitor the medical education literature, educational technology websites (e.g. EducatorsTechnology.com, Edudemic.com, Edutopic.org), technology websites (e.g.

340

Volume XVII, no. 3 : May 2016

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century

TechCrunch.com, iMedicalApps.com), and key Twitter accounts, (e.g. @Educ_Technology and @EduTweetTech). Furthermore, s/he could periodically check for new blogs or podcasts that demonstrate high educational value for EM residents to incorporate into the LRC curriculum. Effectively applying these 10 tips to your residency In our own residency program, we have incorporated each of these “tips” over the past three years. In the first phase of these curricular modifications, we replaced one hour of conference time with asynchronous curricula. Despite decreasing conference time, we saw no associated change of in-training exam scores pre- and post-implementation.63 While it is reassuring that in-training exam scores did not decrease, we felt that we were not properly organizing and integrating this asynchronous curriculum into our synchronous curriculum and we were not meeting the requirements of III. Now, we have better aligned our asynchronous and synchronous curriculum and added back 30 minutes of conference (total conference time 4.5 hours) in order to review asynchronous content. Our chief resident of technology works with our residency leadership team to identify, create, and post III content onto our LMS. In the second phase of our curricular restructuring, we revised our didactic (synchronous) curriculum. To implement these changes, we formed a curriculum planning committee (CPC), which includes the program director, assistant program directors, and MDEdTech fellows. We held a two-hour faculty development session to introduce TBL, SGL, and other effective didactic methods to our faculty members. We reminded faculty members about these concepts at monthly faculty meetings and educated our chief residents during a “chief dinner” at the beginning of the academic year. We assigned a faculty member, academic chief resident, and member of the CPC to each block of didactic curriculum. The faculty member and chief resident do the majority of planning for the conference block, but the CPC member ensures that the appropriate content is covered and that the didactic curriculum incorporates effective learning strategies. Time- and/or feedback-based modifications have helped to overcome many of the challenges we have faced in implementing these curricular changes. For example, we found that senior level residents had the lowest participation in our asynchronous curriculum when we first implemented it. Now, since our current senior residents have used our LMS since the start of their residency, we have equal participation between the senior and junior learners in III. We have an 80% completion rate among all classes. The chief residents have been integral to ensuring the success of this culture change. They have encouraged faculty members of all experience levels to incorporate these strategies into their didactics and have helped create or find much of the content for our SGL exercises.

Volume XVII, no. 3 : May 2016

CONCLUSION Education is changing rapidly with the current instructional technologies of the 21st century, and medical education is no different. Although it is still unclear whether these technologies are actually more effective than previous, less technologically focused approaches, millennial learners are using social media and other technologies for their learning. Residencies will be well served by becoming proficient in these digital tools in order to provide an optimal resident learning experience. When any new digital innovation or technology is incorporated in the curriculum, it should be evaluated in terms of efficacy, logistical burden, accuracy, and user experience. These experiences, whether positive or negative, should be shared with others in a public forum to help the greater education community more quickly identify best practices in modern medical education.

Address for Correspondence: Shannon L. Toohey, MD, University of California Irvine, Department of Emergency Medicine, 333 City Boulevard West, Suite 640, Route 128-01, Orange, CA 92868. Email: stoohey@uci.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. Dr. Michelle Lin is the Editor-in-Chief of the Academic Life in Emergency Medicine website. Copyright: © 2016 Toohey et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

341

1. Howe N and Strauss W. Milennials Rising: the Next Generation. New York: Vintage Books; 2009. 2. Monaco M and Martin M. The millennial student: A new generation of learners. ATEJ. 2007;2(2):42-46. 3. Deiorio NM, Fitch MT, Jung J, et al. Evaluating Educational Interventions in Emergency Medicine. Kowalenko T, ed. Acad Emerg Med. 2012;19(12):1442-1453. 4. Young T, Bailey C, Guptill M, et al. The flipped classroom: A modality for mixed asynchronous and synchronous learning in a residency program. West J Emerg Med. 2014;15(7):938-944. 5. Pearson D, Cooney R, Bond M. Recommendations from the Council of Residency Directors (CORD) social media committee on the role of social media in residency education and strategies on implementation. West J Emerg Med. 2015;16(4):510-515. 6. Wolff M, Wagner MJ, Poznanski S, et al. Not Another Boring Lecture: Engaging Learners with Active Learning Techniques. J Emerg Med.

Western Journal of Emergency Medicine

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century 2012;48(1):85-93.

ALiEM. Available at: http://www.aliem.com/new-air-series-aliem-

7. Sadosty AT, Goyal DG, Gene Hern H Jr, et al. Alternatives to the

approved-instructional-resources/. Accessed Nov 15, 2015.

conference status quo: Summary recommendations from the 2008

23. Cadogan M and Nickson CP. LITFL Review - Life in the fast lane

CORD Academic Assembly conference alternatives workgroup. Acad

Medical Blog. Available at: http://lifeinthefastlane.com/collections/

Emerg Med. 2009;16:S25-S31.

litfl-review/.

8. Tan E, Brainard A, Larkin GL. Acceptability of the flipped classroom

24. Bloom BS and Krathwohl DR. Taxonomy of Educational Objectives:

approach for in-house teaching in emergency medicine. Emerg Med

the Classification of Educational Goals, by a Committee of College

Australas. 2015;27(5):453-459.

and University Examiners. Handbook 1: Cognitive Domain. New

9. Boysen-Osborn M, Anderson C, Navarro R, et al. Better

York: Longmans.

performance on written tests for medical students in a flipped

25. Augustin M. How to learn effectively in medical school: test

classroom model with team based learning for advanced cardiac life

yourself, learn actively, and repeat in intervals. Yale J Biol Med.

support compared with lecture at the University of California, Irvine,

2014;87(2):207-12.

United States. J Educ Eval Health Prof. 2016;13:11-20.

26. Accreditation Council for Graduate Medical Education. ACGME

10. Periyakoil VS and Basaviah P. The flipped classroom paradigm for

Program Requirements for Graduate Medical Educationin

teaching palliative care skills. Virtual Mentor. 2013;15(12):1034-1037.

Emergency Medicine.2013:1-33. Available at: http://www.acgme.

11. O’Flaherty J and Phillips C. The use of flipped classrooms

org/acgmeweb/portals/0/pfassets/2013-pr-faq-pif/110_emergency_

in higher education: A scoping review. Internet High Educ. 2015;25(C):85-95.

medicine_07012013.pdf. 27. Accreditation Council for Graduate Medical Education. Frequently

12. Morgan H, Chapman C, Fitzgerald T, et al. A Flipped Classroom Curriculum for Milestones Preparation. Obstetrics & Gynecology.

Asked Questions: Emergency Medicine. 2012:1-11. 28. Wood DF. ABC of learning and teaching in medicine: Problem based

2015;126:46S–47S. 13. Liebert CA, Lin DT, Mazer LM, et al. Effectiveness of the Surgery

learning. BMJ. 2003;326(7384):328-330. 29. Donner RS and Bickley H. Problem-based learning in

Core Clerkship Flipped Classroom: a prospective cohort trial. Am J

American medical education: an overview. Bull Med Libr Assoc.

Surg. 2015;0(0).

1993;81(3):294-298.

14. Pierce R and Fox J. Vodcasts and active-learning exercises in a

30. Davis MH. AMEE Medical Education Guide No. 15: Problem-based

“flipped classroom” model of a renal pharmacotherapy module. Am J Pharm Educ. 2012;76(10):196.

learning: a practical guide. Med Teach. 1999;21(2):130-140. 31. Maudsley G. Roles and responsibilities of the problem based

15. Belfi LM, Bartolotta RJ, Giambrone AE, et al. “Flipping” the

learning tutor in the undergraduate medical curriculum. BMJ.

introductory clerkship in radiology: Impact on medical student performance and perceptions. Acad Radiol. 2015;22(6):794-801.

1999;318(7184):657-661. 32. Barrows HS. A taxonomy of problem-based learning methods. Med

16. Ramar K, Hale CW, Dankbar EC. Innovative model of delivering quality improvement education for trainees – a pilot project. Med

Edu. 1986;20(6):481-486. 33. Reynolds RJ and Candler CS. MedEdPORTAL: educational

Educ Online. 2015;20:28764. 17. Cadogan M, Thoma B, Chan TM, et al. Free Open Access

scholarship for teaching. J Contin Educ Health Prof. 2008;28(2):91-94. 34. Council of Emergency Medicine Residency Directors. CORD

Meducation (FOAM): The rise of emergency medicine and

Teaching Cases: Oral Board & Simulation Cases. Available at: http://

critical care blogs and podcasts (2002-2013). Emerg Med J.

www.cordem.org/i4a/pages/index.cfm?pageid=3403. Accessed Nov

2014;31(e1):e76-e77.

15, 2015.

18. Mallin M, Schlein S, Doctor S, et al. A survey of the current utilization

35. Distlehorst LH, Dawson E, Robbs RS, et al. Problem-based learning

of asynchronous education among emergency medicine residents in the United States. Acad Med. 2014;89(4):598-601.

outcomes: the glass half-full. Acad Med. 2005;80(3):294-299. 36. Koles PG, Stolfi A, Borges NJ, et al. The Impact of Team-Based

19. Purdy E, Thoma B, Bednarczyk J, et al. The use of free online

Learning on Medical Studentsʼ Academic Performance. Acad Med.

educational resources by Canadian emergency medicine residents and program directors. CJEM. 2015;17(02):101-106.

2010;85(11):1739-1745. 37. Wilson BG. Constructivist Learning Environments: Case Studies

20. Nickson CP and Cadogan MD. Free Open Access Medical education

in Instructional Design. (Wilson BG, ed.). Englewood Cliffs, NJ:

(FOAM) for the emergency physician. Emerg Med Australas. 2014;26(1):76-83.

Educational Technology Publications; 1996:135-148. 38. Barrows HS. Problem‐based learning in medicine and beyond:

21. Lin M, Thoma B, Trueger NS, et al. Quality indicators for blogs and

A brief overview. New Directions for Teaching and Learning.

podcasts used in medical education: Modified Delphi consensus recommendations by an international cohort of health professions

1996;1996(68):3-12. 39. Miller GE. The assessment of clinical skills/competence/performance.

educators. Postgrad Med J. 2015;91(1080):546-550. 22. Grock A. New AIR series: ALiEM approved instructional resources.

Western Journal of Emergency Medicine

Acad Med. 1990;65(9 Suppl):S63-S67. 40. The National Teaching and Learning Forum. Blackwell Publishing Ltd

342

Volume XVII, no. 3 : May 2016

Toohey et al.

Ten Tips for Moving Residency Curriculum into 21st Century Simulation Education Curriculum for Emergency Medicine Residents.

41. Middendorf J and Kalish A. The “Change-Up” in Lectures. The

Ann Emerg Med. 2007;49(4):495–504.e11.

National Teaching and Learning Forum. 1996;5(2):1-12.

53. McLaughlin SA, Doezema D, Sklar DP. Human simulation in

42. Kessler CS, Dharmapuri S, Marcolini EG. Qualitative Analysis

emergency medicine training: A model curriculum. Acad Emerg Med.

of Effective Lecture Strategies in Emergency Medicine. YMEM.

2002;9(11):1310-1318.

2011;58(5):482-489.e487.

54. Wayne DB, Didwania A, Feinglass J, et al. Simulation-Based

43. Shiroma PR, Massa AA, Alarcon RD. Using game format to teach psychopharmacology to medical students. Med Teach.

Education Improves Quality of Care During Cardiac Arrest

2011;33(2):156-160.

Team Responses at an Academic Teaching Hospital. Chest. 2008;133(1):56-61.

44. O’Leary S, Diepenhorst L, Churley-Strom R, et al. Educational

55. Okuda Y, Bryson EO, DeMaria S Jr, et al. The Utility of Simulation

games in an obstetrics and gynecology core curriculum. Am J Obstet

in Medical Education: What Is the Evidence? Mt Sinai J Med.

Gynecol. 2005;193(5):1848-1851.

2009;76(4):330-343.

45. Khan MN, Telmesani A, Alkhotani A, et al. Comparison of jeopardy game format versus traditional lecture format as a teaching methodology in

56. Kerfoot BP and Baker H. An Online Spaced-Education Game to

medical education. Saudi Med J. 2011;32(11):1172-1176.

Teach and Assess Residents: A Multi-Institutional Prospective Trial.

46. Schackow TE, Chavez M, Loya L, et al. Audience response

ACS. 2012;214(3):367-373.

system: effect on learning in family medicine residents. Fam Med.

57. The Accreditation Council for Graduate Medical Education, The

2004;36(7):496-504.

American Board of Emergency Medicine. The Emergency Medicine

47. Latessa R and Mouw D. Use of an audience response system to

Milestone Project. October 2013:1-29.

augment interactive learning. Fam Med. 2005;37(1):12-14.

58. Cox M, Irby DM, Epstein RM. Assessment in medical education. N

48. Pradhan A, Sparano D, Ananth CV. The influence of an audience

Engl J Med. 2007;356(4):387-396.

response system on knowledge retention: An application to resident

59. Aggarwal R, Grantcharov T, Moorthy K, et al. Toward Feasible, Valid,

education. Am J Obstet Gynecol. 2005;193(5):1827-1830.

and Reliable Video-Based Assessments of Technical Surgical Skills

49. Kay RH and LeSage A. Examining the benefits and challenges

in the Operating Room. Ann Surg. 2008;247(2):372-379.

of using audience response systems: A review of the literature.

60. Chen AC-Y, Lee MS, Chen W-J, et al. Assessment in orthopedic

Computers & Education. 2009;53(3):819-827.

training-an analysis of rating consistency by using an objective

50. Surry DW, Ensminger DC, Haab M. A model for integrating

structured examination video. J Surg Educ. 2013;70(2):189-192.

instructional technology into higher education. Br J Educ Technol.

61. Hile D. Unpublished Data.

2005;36(2):327-329.

62. Ogunyemi D, Fung E, Alexander C, et al. A Faculty and Resident

51. Hanover Reseach. Usage of E-Books in Higher Education. Academic

Development Program to Improve Learning and Teaching Skills. J

Administration Practice. January 2013:1-24. 52. Binstadt ES, Walls RM, White BA, et al. A Comprehensive Medical

Volume XVII, no. 3 : May 2016

Grad Med Educ. 2009;1(1):127-131. 63. Wray A. Unpublished Data.

343

Western Journal of Emergency Medicine

Original Research

Survey of Individual and Institutional Risk Associated with the Use of Social Media Manish Garg, MD* David A. Pearson, MD, MS† Michael C. Bond, MD‡ Michael Runyon, MD† M. Tyson Pillow, MD, MEd§ Laura Hopson, MD¶ Robert R. Cooney, MD|| Jay Khadpe, MD# Jason T. Nomura, MD** Pholaphat C. Inboriboon, MD, MPH††

* Temple University Hospital, Department of Emergency Medicine, Philadelphia, Pennsylvania † Carolinas Medical Center, Department of Emergency Medicine, Charlotte, North Carolina ‡ University of Maryland School of Medicine, Department of Emergency Medicine, Baltimore, Maryland § Baylor College of Medicine, Department of Emergency Medicine, Houston, Texas ¶ University of Michigan, Department of Emergency Medicine, Ann Arbor, Michigan || Conemaugh Memorial Medical Center, Department of Emergency Medicine, Johnstown, Pennsylvania # SUNY Downstate Medical Center, Department of Emergency Medicine, Brooklyn, New York ** Christiana Care Health System, Department of Emergency Medicine, Newark, Delaware †† University of Missouri-Kansas City, Department of Emergency Medicine, Kansas City, Missouri

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted November 26, 2015; Revision received February 12, 2016; Accepted February 26, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.28451

Introduction: Residents and faculty in emergency medicine (EM) residency programs might be unaware of the professional and legal risks associated with the use of social media (SM). The objective of this study was to identify and characterize the types and reported incidence of unprofessional SM behavior by EM residents, faculty, and nurses and the concomitant personal and institutional risks. Methods: This multi-site study used an 18-question survey tool that was distributed electronically to the leaders of multiple EM residency programs, members of the Council of Emergency Medicine Residency Directors (CORD), and the residents of 14 EM programs during the study period May to June 2013. Results: We received 1,314 responses: 772 from residents and 542 from faculty. Both groups reported encountering high-risk-to-professionalism events (HRTPE) related to SM use by residents and non-resident providers (NRPs), i.e., faculty members and nurses. Residents reported posting of one of the following by a resident peer or nursing colleague: identifiable patient information (26%); or a radiograph, clinical picture or other image (52%). Residents reported posting of images of intoxicated colleagues (84%), inappropriate photographs (66%), and inappropriate posts (73%). Program directors (PDs) reported posting one of the following by NRPs and residents respectively: identifiable patient information (46% and 45%); a radiograph, clinical picture or other image (63% and 58%). PDs reported that NRPs and residents posted images of intoxicated colleagues (64% and 57%), inappropriate photographs (63% and 57%), or inappropriate posts (76% and 67%). The directors also reported that they were aware of or issued reprimands or terminations at least once a year (30% NRPs and 22% residents). Residents were more likely to post photos of their resident peers or nursing colleagues in an intoxicated state than were NRPs (p=0.0004). NRPs were more likely to post inappropriate content (p=0.04) and identifiable patient information (p=0.0004) than were residents. Conclusion: EM residents and faculty members cause and encounter HRTPE frequently while using SM; these events present significant risks to the individuals responsible and their associated institution. Awareness of these risks should prompt responsible SM use and consideration of CORD’s Social Media Task Force recommendations. [West J Emerg Med. 2016;17(3):344–349.] Western Journal of Emergency Medicine

344

Volume XVII, no. 3 : May 2016

Garg et al.

Assesment of Risk Associated with Social Media Use

INTRODUCTION The use of social media (SM) continues to increase among both emergency medicine (EM) residents and attending physicians. In a recent investigation covering a one-year period, the number of EM and critical care blogs increased from 130 to 201 and EM, and critical care information dissemination increased 10% via TwitterTM and 7% via FacebookTM.1 SM utilization in EM residency training programs is also on the rise. A recent survey of asynchronous education among 226 EM residents in 12 EM residency programs in the United States found that almost 70% of the residents endorsed EM podcasts as the most beneficial learning modality compared with textbooks, journals, and GoogleTM.2 In addition to asynchronous learning and educational resources, EM residency training programs use SM for professional marketing and for personal communication, socializing, and editorials. The personal use of SM is of particular concern, as it could have substantial crossover professional implications. Healthcare workers in emergency departments (EDs) might share protected patient information on SM inappropriately or inadvertently, putting employees and institutions at risk. Furthermore, because the Accreditation Council for Graduate Medical Education (ACGME) has not yet established a standard for regulating the use of SM, EM residency programs have limited oversight of their departmental SM utilization. The combination of increased SM utilization, a lack of structured guidelines, and a lack of awareness of risk on the part of ED personnel presents substantial personal and institutional risk, as demonstrated by a number of recent highly publicized cases. For example, an EM physician was fired for a FacebookTM public overshare, in which community members were able to identify a patient from a post.3 We thus sought to characterize observed SM behavior by EM residents, faculty, and nurses that carries potential personal and institutional risks. To our knowledge, this study by the Social Media Committee of the Council of Emergency Medicine Residency Directors (CORD) is the first investigation of its kind. METHODS CORD is a national organization of faculty and program leaders for EM residencies accredited by the ACGME and the American Osteopathic Association. The CORD Social Media Committee developed an 18-question electronic survey tool (SurveyMonkey™, Palo Alto, CA), which was distributed in May 2013; reminders were sent two and four weeks later. The protocol and survey tool were approved by the institutional review board. The electronic survey was distributed to the committee members’ home institutions, which created a diverse geographic sample of 432 residents. Additional residents were contacted, using the CORD listserv as a distribution method. The listserv consists of all CORD member program Volume XVII, no. 3 : May 2016

leaders, staff, and key faculty; these individuals were asked to distribute the electronic survey link to their programs’ residents. Faculty respondents consisted of the program directors (PDs), associate/assistant program directors, and core faculty members who were members of CORD and received the survey via the CORD listserv. The survey tool is presented in Appendix A. Key measurements included the use of SM by residents, their knowledge of institutional policies regarding SM, and a comparison of SM use by residents and faculty members. Responses were voluntary and the study period was from May to June 2013. “Inappropriate” posts were left to the individual survey respondent to define, as a universal definition to cover everything is extremely difficult. We calculated descriptive statistics using the chi-squared test or Fisher’s exact test. StatsDirect software (v2.8.0, StatsDirect, Cheshire, UK) was used for all analyses. RESULTS We received 1,314 responses: 772 from residents and 542 from faculty members. At the end of May 2013, the listserv had 841 faculty participants. According to the American Board of Emergency Medicine Report on residency training information for the academic year 2012 to 2013, there were 5,734 residents in accredited US categorical emergency medicine programs. Therefore, the survey response rate was 13% for residents and 44% for faculty. The participants’ demographics are summarized in Table 1. The faculty respondents’ geographic distribution was as follows: Northeast, 32%; South, 31%; Midwest, 27%; and West, 8%. The residents had a similar geographic distribution: Northeast, 33%; South, 33%; Midwest, 24%; and West, 8% (p=0.58). Residents reported high-risk-to-professionalism events (HRTPE) resulting from their resident peers or nursing colleagues posting one of the following: identifiable patient information (177/680 [26%]); and a radiograph, clinical picture or other image (352/679 [52%]). Residents reported HRTPE created by residents and NRPs who posted images of intoxicated colleagues (564/674 [84%]), inappropriate photographs (445/675 [66%]), or inappropriate posts (490/672 [73%]). Some residents did not respond to all questions, which resulted in a lower denominator for some questions. PDs reported HRTPE associated with non-resident providers (NRPs) and residents posting one of the following respectively: identifiable patient information (33/71 [46%] and 30/67 [45%]); and a radiograph, clinical picture or other image (44/70 [63%] and 39/67 [58%]) (Figure 1). PDs reported HRTPE related to posted images of intoxicated colleagues (44/69 [64%] and 38/67 [57%]); inappropriate photographs (45/71 [63%] and 38/67 [57%]); or inappropriate posts (54/71 [76%] and 44/66 [67%]) (Figure 2). PDs also reported being aware of or issuing reprimands or termination at least once a year (21/71 [30%] for NRPs and 15/67 [22%] for residents. Residents were more likely to post photos of other residents

345

Western Journal of Emergency Medicine

Garg et al.

Assesment of Risk Associated with Social Media Use

Table 1. Demographic information for respondents (N=1314) in survey regarding inappropriate social media postings by emergency medicine residents and faculty. Gender Male 63%

Female 37%

Age (years)

Prefer not to answer 1%

Residents

Faculty (n=542)

<30

31–39

>40

772

15% PDs; 85% other faculty

40%

39%

21%

PDs, program directors

or their nursing colleagues in an intoxicated state than were NRPs (p=0.0004). NRPs were more likely to post inappropriate content (p=0.04) and identifiable patient information (p=0.0004) than were residents. There was no difference for posting of inappropriate photographs (p=0.28) or radiographs, clinical pictures or other images (p=0.12) (Table 2). DISCUSSION The practice of medicine mandates a high level of professionalism to ensure continued societal confidence in the profession and the physicians who deliver care. The EM Milestone project created by the ACGME includes two professionalism milestones. The first is “Professional Values,” which asks that the EM physician “demonstrates compassion,

integrity, and respect for others as well as adherence to the ethical principles relevant to the practice of medicine.” The second is “Accountability,” which asks that the EM physician “demonstrates accountability to patients, society, profession and self.” The basic level of this second milestone requires that the EM resident “maintains patient confidentiality” and “uses social media ethically and responsibly.”4 Although these milestones are intended for the clinical care of patients, we believe it is equally important to ensure resident education regarding behavior outside the hospital setting. Due to the high number of inappropriate posts reported in this study, it is important that residents understand the potential pitfalls associated with the use of SM. Once SM posts are in the public sphere, it is often difficult, if not impossible, to remove them.

Figure 1. Reported high-risk-to-professionalism events by residents and program directors (PD) involving the posting of (1) identifiable patient information and (2) radiographs, clinical pictures or other images. Group 1=Resident responses regarding postings by nonresident providers (NRP). Group 2=PD responses regarding postings by NRPs (other faculty/nurses). Group 3=PD responses regarding postings by residents.

Western Journal of Emergency Medicine

346

Volume XVII, no. 3 : May 2016

Garg et al.

Assesment of Risk Associated with Social Media Use

Figure 2. Reported high-risk-to-professionalism events by residents and program directors (PD) involving the posting of (1) images of intoxicated colleagues, (2) inappropriate photographs, or (3) inappropriate posts. NRP, non-resident providers Table 2. Comparison of reported high-risk-to-professionalism events (HRTPE) in program director (PD) responses regarding postings by other faculty and nurses versus resident responses regarding postings by other residents and nurses. HRTPE

PD, non-resident, peers/colleagues

Resident, peers/colleagues

P-value 0.0004

Intoxicated state

44/69, 64%

564/674, 84%

Inappropriate content

54/71, 76%

490/672, 73%

0.04

Identifiable patient information

33/71, 46%

177/680, 26%

0.0004

Inappropriate photographs

45/71, 63%

445/675, 66%

0.28

Radiographs, clinical pictures or other images

44/70, 63%

352/679, 52%

0.12

Institutional risk introduced by the personal use of SM has not been investigated in the literature. Half of the PDs who completed our survey reported being aware of at least one occurrence of the sharing of identifiable patient information by NRPs or residents during the last year. Interestingly, 25% of participating residents knew about similar postings by their peers and nurse colleagues of information that allowed a patient to be identified. Collectively, 207 survey respondents reported awareness of postings that allowed patients to be identified. It Volume XVII, no. 3 : May 2016

is impossible to know if these are distinct events, but it is clear that breaches of patient information on SM occur. Confidentiality and privacy are fundamental expectations of the patientâ&#x20AC;&#x2019;physician relationship. The Health Insurance Portability and Accountability Act of 1996 (HIPAA) has privacy, security, and breach notification rules that are enforced by the Office for Civil Rights within the U.S. Department of Health and Human Services.5 Some of the observations and occurrences documented in our study fall within the scope of 347

Western Journal of Emergency Medicine

Garg et al.

Assesment of Risk Associated with Social Media Use HIPAA and put individuals and institutions at risk. In addition to federal ramifications for medical institutions in regard to unprofessional conduct on SM by employees, the individuals responsible for the HRTPE face state licensing consequences. In 2012, a national survey of state medical boards investigated physician violations of online professionalism and documented the subsequent disciplinary actions. For example, approximately 15% of state medical boards reported known physician violations resulting from online depiction of intoxication. Disciplinary consequences for all types of violations included letters of reprimand or probation; restriction, suspension or revocation of licensure; mandated education or community service; or monetary fines.6 Our results are more serious than those reported in this investigation. From the viewpoint of a residency PD, having nearly a quarter of the residents reporting seeing SM misuse should be cause for program-wide education and policy. Consequences for individuals were reported by the participants in our study, but institutional consequences were never mentioned in the free-text responses. A few lawsuits have been filed against teaching institutions based on inappropriate SM utilization. The first is based on a posting, by a non-EM physician, of private information on FacebookTM and InstagramTM after evaluating an intoxicated female patient in the ED.7 The lawsuit names the physician, the hospital, and the affiliated medical school and is seeking $1.5 million in damages. The second suit occurred after non-physician hospital employees (allegedly including one nurse) posted a diagnosis of maternal syphilis on a FacebookTM group with a large distribution.8 The lawsuit names the hospital employees and the medical center and is seeking more than $25,000 in damages. Though the number of serious violations is low, the incidence is likely to increase with larger payouts unless the responsible use of SM becomes a part of medical education either at the undergraduate or graduate level. Regulation of the use of SM in residency training programs is hampered by the lack of clear guidelines from the accrediting body. The ACGME has not yet issued formal SM standards or guidelines for residency training programs, but the American Medical Association (AMA) has SM guidelines for physicians.9 Guidance tailored to the residency training environment is available from the CORD’s Social Media Task Force.10 The authors believe the AMA and CORD recommendations can help residency training programs use SM professionally and minimize personal and institutional risk. Professional use would include using SM for educational purposes, announcement of accolades and professional accomplishments while ensuring nothing unprofessional (e.g.: images of colleagues intoxicated, protected patient information, or posts that could be considered offensive by distinct groups) is posted. Responsible SM use has many advantages for resident education. SM is interactive, brings together a geographically diverse distribution of information, and has Western Journal of Emergency Medicine

near real-time peer review. If it is used unprofessionally or irresponsibly, SM can lead to serious personal, professional, and institutional consequences, including employment termination, fines, and federal violations. Our survey revealed violations of patient privacy and other inappropriate use of SM, and these represent risks to the residents and institutions that cannot be ignored as SM expands. LIMITATIONS Our study has several limitations. First, the survey methodology could have led to inaccurate entry of the respondent data (i.e., if a respondent could have incorrectly selected a choice that did not reflect a true observation). Second, we asked respondents to self-report the frequency of observations, which is subject to recall bias (individuals who have knowledge of adverse events could be more likely to respond). Third, we left the definition of an inappropriate photograph up to the survey respondent. A finite definition was not provided knowing that individuals often make a value judgment of inappropriateness based on their own perceptions and life experiences. Finally, none of respondents actually reported any institutional consequence from the improper SM postings, despite the reports of individual consequences. CONCLUSION Emergency medicine residents and faculty members who use SM frequently encounter HRTPE that hold potential personal and institutional risk. Awareness of these risks, and the risk of their own behaviors and potential liability should encourage responsible SM use and consideration of using the SM recommendations from the CORD’s Social Media Task Force.

Address for Correspondence: Manish Garg, MD, 3401 N Broad St. Temple University Hospital, Philadelphia, PA 19140. Email: manish.garg@tuhs.temple.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Garg et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

348

1. Cadogan M. EMCC Blog update. LITFL. Available at: http://

Volume XVII, no. 3 : May 2016

Garg et al.

Assesment of Risk Associated with Social Media Use

lifeinthefastlane.com/emcc-blog-update-2013. Accessed Aug 22,

6. Greysen SR, Chretien KC, Kind T, et al. Physician violations of online

2015.

professionalism and disciplinary actions: a national survey of state

2. Mallin M, Schlein S, Doctor S, et al. A survey of the current utilization

medical boards. JAMA. 2012;307(11):1141-1142.

of asynchronous education among emergency medicine residents in

7. Lawsuit: Dr Posted Pics of Drunk ER Patient to Social media.

the United States. Acad Med. 2014;89(4):598–601.

Available at: http://lawmedconsultant.com/6297/lawsuit-dr-posted-

3. Danzig C. ER Doc Forgets Patient Info is Private, Gets Fired for

pics-drunk-er-patient-social-media-video/. Accessed Aug 22, 2015.

Facebook Overshare. Available at: http://abovethelaw.com/2011/04/

8. Lawsuit: UC Med Center employees leaked patient’s history of

er-doc-forgets-patient-info-is-private-gets-fired-for-facebook-

Facebook. Available at: www.wlwt.com/news/lawsuit-uc-med-center-

overshare/. Accessed Aug 22, 2015.

employees-leaked-patients-history-on-facebook/26316492. Accessed

4. The Emergency Medicine Milestone Project. Available at:

Aug 22, 2015.

https://wwwacgme.org/acgmeweb/Portals/0/PDFs/Milestones/

9. Report of the Council of Ethical and Judicial Affairs. American

EmergencyMedicineMilestones.pdf. Accessed Aug 22, 2015. 5. Health Information Privacy. U.S. Department of Health & Human

Medical Association, CEJA Report 8-I-10, November 2010, Page 6. 10. Pillow MT, Hopson L, Bond M, et al. Social media guidelines and best

Services. Available at: www.hhs.gov/ocr/privacy/. Accessed Aug 22,

practices: recommendations from the Council of Residency Directors

2015.

Social Media Task Force. West J Emerg Med. 2014;15(1):26–30.

Volume XVII, no. 3 : May 2016

349

Western Journal of Emergency Medicine

Original Research

Impact of Doximity Residency Rankings on Emergency Medicine Applicant Rank Lists William J. Peterson, MD* Laura R. Hopson, MD* Sorabh Khandelwal, MD† Melissa White, MD, MPH‡ Fiona E. Gallahue, MD§ John Burkhardt, MD* Aimee M. Rolston, MD¶ Sally A. Santen, MD, PhD*

* University of Michigan, Department of Emergency Medicine, Ann Arbor, Michigan † The Ohio State University College of Medicine, Department of Emergency Medicine, Columbus, Ohio ‡ Emory University, Department of Emergency Medicine, Atlanta, Georgia § University of Washington, Department of Emergency Medicine, Seattle, Washington ¶ University of Michigan, Department of Obstetrics and Gynecology, Ann Arbor, Michigan

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted January 12, 2016; Revision received March 15, 2016; Accepted April 1, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.4.29750

Introduction: This study investigates the impact of the Doximity rankings on the rank list choices made by residency applicants in emergency medicine (EM). Methods: We sent an 11-item survey by email to all students who applied to EM residency programs at four different institutions representing diverse geographical regions. Students were asked questions about their perception of Doximity rankings and how it may have impacted their rank list decisions. Results: Response rate was 58% of 1,372 opened electronic surveys. This study found that a majority of medical students applying to residency in EM were aware of the Doximity rankings prior to submitting rank lists (67%). One-quarter of these applicants changed the number of programs and ranks of those programs when completing their rank list based on the Doximity rankings (26%). Though the absolute number of programs changed on the rank lists was small, the results demonstrate that the EM Doximity rankings impact applicant decision-making in ranking residency programs. Conclusion: While applicants do not find the Doximity rankings to be important compared to other factors in the application process, the Doximity rankings result in a small change in residency applicant ranking behavior. This unvalidated ranking, based principally on reputational data rather than objective outcome criteria, thus has the potential to be detrimental to students, programs, and the public. We feel it important for specialties to develop consensus around measurable training outcomes and provide freely accessible metrics for candidate education. [West J Emerg Med. 2016;17(3):350–354.]

INTRODUCTION Background Influences on applicant rank lists have been well studied; however, the advent of the new Doximity ranking system may have introduced new considerations. Studies have shown that applicants base their decisions on a combination of personal Western Journal of Emergency Medicine

factors including geographic location and quality of life, as well as program-specific factors including expected clinical experience, curriculum quality, interview day, experience with residents and faculty, and reputation of program.1-5 This process leads to an important decision that will impact the applicant’s future practice and location.6 350

Volume XVII, no. 3 : May 2016

Peterson et al.

Doximity Residency Rankings Impact on Applicant Rank Lists

In the 2014 application cycle, Doximity released residency program rankings by specialty in collaboration with U.S. News and World Report.7 Doximity is a free, HIPAA-compliant online platform for physicians’ social networking, collaboration and education. To create the residency rankings, Doximity administered a survey to their physician network in which they asked practicing physicians to “nominate up to 5 residency programs in your medical specialty that offer the best clinical training.8” More than 17,000 Doximity members responded to the survey, which resulted in a list of programs ranked by number based on majority vote. No independent consensus exists as to metrics for quality of training. The creation of a ranking system based on potentially biased responses from a selected group of physicians is controversial and has raised significant concerns, especially in the emergency medicine (EM) community.9

Methods, Measurements, and Outcomes We assembled a research team consisting of two assistant deans, four residency program directors, a clerkship director, a resident, and a fourth-year medical student. For content validity, we modeled the questions after previous studies.4,11 The survey was reviewed by all authors with attention to response process and revised. For additional content validity, the questions were modeled after previous studies on a similar population (fourth-year students applying to residency). That survey was piloted by 20 residents and faculty and revised for response process validity. All authors reviewed this survey with attention to response process. The survey was emailed to students using a web-based platform, QualtricsTM, and student responses were anonymous. The survey was initially distributed at the beginning of March 2015 after rank lists were submitted by applicants and closed before Match Day. Three repeated requests were sent weekly to non-responders. The survey asked first whether the student applicant was aware of or had looked at the Doximity rankings prior to submitting their rank list (Figure). Students who had looked at the rankings were eligible to complete the questions that assessed whether the Doximity rankings impacted their rank list construction. Students were also asked basic demographic information, how accurate they perceived the rankings to be on a 100-point scale (0 being not accurate at all and 100 being very accurate), and whether they increased or decreased the rank of programs based on the Doximity rankings. Additionally, space was provided for students to comment about the Doximity rankings. The comments were qualitatively reviewed and categorized into three groups: negative impression of rankings, neutral, and positive impressions of rankings. The negative category contained statements about how Doximity rankings were perceived as inaccurate or biased. The neutral category contained comments where students were unsure or did not care about the rankings. The positive category contained comments about how Doximity rankings were helpful or perceived as accurate. Finally, respondents were asked what factors affected their choice of programs.

Importance The impact of U.S. News and World Report rankings on the undergraduate college application process has been well studied and has been shown to affect applicant decisionmaking, as well as public perception of universities and the funding that universities receive, regardless of debates about its accuracy.10 Prior to the Doximity emergency medicine residency list being published, no central ranking system existed for residency applicants to refer to and to potentially impact their rank list. Goals of this Investigation The effects of the Doximity findings, which have both reputational and ranking implications, are not well studied. This new ranking system may result in changes to applicants’ selections of residency programs. An initial study indicated that applicants are using Doximity in their choice of program applications.11 This impact is potentially concerning, since there has been significant resistance to Doximity rankings in the EM community due to concerns about lack of objective criteria, inaccurate portrayal of residency programs, bias towards programs with larger alumni networks and provision of potentially misleading information to students as well as patients in the community.12 The objective of this study is to investigate the impact of the Doximity rankings on the rank list choices made by residency applicants in EM. METHODS Study Design, Participants and Setting Design was an 11-item survey emailed to all students who applied to EM residency programs at four different geographically diverse institutions: University of Michigan (Midwest), Ohio State University (Midwest), Emory University (South), and University of Washington (West). These email addresses were obtained through ERAS with the permission of ERAS. Volume XVII, no. 3 : May 2016

Analysis Data analysis included descriptive statistics using SPSS 22. This study was determined to be exempt from institutional review board review at all four participating sites. RESULTS Characteristics of Study Subjects We sent 1,641 emails to individual applicants for EM resident positions; 1,372 people opened the email, 850 started the survey, and 793 students completed the survey across the sample (overall response rate of 93% of people who started the survey, 58% of people who opened the email and 48% of total emails sent). The demographics of this sample of applicants who looked at the rankings were as follows: 63% male, 73% self-identified as White, 11% Asian, 5% Hispanic, 351

Western Journal of Emergency Medicine

Peterson et al.

Doximity Residency Rankings Impact on Applicant Rank Lists

decrease 1.46 (SD 1.14). Students’ relative value of factors affecting residency preference are noted in Tables 1 and 2. They included preference for a particular geographic location, listed interview experience and experience with residents.

793 Respondents 260 not aware of rankings; 2 skipped this question 531 aware of rankings 166 did not look at rankings 359 looked at rankings

Mean number programs added=1.2 (n=350)

Mean number programs dropped=0.3 (n=347)

Mean number programs increased in rank = 0.6 (n=351)

Mean number programs decreased in rank = 0.5 (n=351)

Figure. Applicants who looked at rankings.

4% Black or African American, 1% American Indian, and 7% other. The regions of the institutions from which the applicants applied included 25% Northeast, 29% South, 27% Midwest, and 18% West. From the National Residency Matching Program for 2015, there were 1,613 U. S. senior applicants and 2,352 total applicants to EM. Main Results Among the respondents, 531 students (67%) were aware of the Doximity rankings prior to submitting their rank lists. Among the students who were aware of the rankings, 359 (68%) looked at the rankings (Figure). Respondents found the Doximity rankings to be somewhat accurate with the mean score for accuracy of 41 (SD 23, range 0-100). Students were asked to “explain your assessment of the accuracy of the Doximity rankings.” Comments varied widely from “worthless” and “completely subjective” to “seems accurate” to “I don’t know.” Of the comments, 65% fell into the negative impressions category, 35% were neutral, and 10% were positive. Of the students who looked at the rankings, 26% added programs to their rank list and 9.8% dropped programs from their rank list (Figure). The mean number of programs added per applicant who looked at the rankings was 1.2 (range 0 to >10) and the mean number of programs dropped was 0.3 (range 0 to 7). However, for those students who did add or drop programs to their rank list based on the rankings, the mean number added was 2.15 (SD 2.40) and dropped 1.31 (SD 1.07). Similarly, 26% of students increased the rank of programs on their rank list based on the Doximity rankings, and 19% decreased the rank of programs on their rank list. The mean number of programs that an applicant who looked at the Doximity rankings increased in rank was 0.6 programs (range 0 to 10), and the mean number of programs the applicant decreased in rank was 0.5 programs (range 0 to 8). For those students who changed their rank list based on the Doximity rankings, the mean increase in rank was 1.60 (SD 1.34) and Western Journal of Emergency Medicine

DISCUSSION This study found that a majority of medical students applying to residency in EM were aware of the Doximity rankings prior to submitting rank lists. A substantial number of applicants looked at the rankings and about a quarter of these applicants changed the number of programs and ranks of those programs when completing their rank list. Notably, the Doximity rankings were the least important factor compared to the other factors assessed in this study (Table 1). While these rankings were the least important, applicants did make changes in their rankings because of Doximity, demonstrating that the Doximity rankings may have some impact in applicant decision-making in ranking residency programs. A previous study similarly found that Doximity rankings affected the number of programs to which students applied.13 We did not assess final match position of applicants, and without that information we cannot comment on how the Doximity rankings may have impacted final match position of applicants. There has been significant resistance to the Doximity rankings in the EM community due to concerns about lack of objective criteria and inaccurate portrayal of residency programs. A consensus statement against the Doximity rankings endorsed by all major EM organizations was recently released in response to the rankings. The letter highlighted the significant threats to the validity of Doximity’s polling methods including the risk of sampling bias since EM physician survey responses were generated from Doximity members recruited through social media.14 It further emphasized to applicants the importance of looking at programs for fit versus an arbitrary ranking system. Despite the concerns expressed by the EM community and by students directly through their comments in our study, the existence of Doximity rankings allows students to make inferences about the reputation and value of programs based solely on these rankings and allows institutions to lay claim to reputation as well. It is well documented that

Table 1. Factors of importance affecting choice of residency programs to which medical students applied. Number (%) n=772 Geographical preference

693 (90%)

Interview experience

636 (82%)

Experience with residents

596 (77%)

Proximity to spouse/significant other/family

439 (57%)

Doximity rankings

352

41 (5%)

Volume XVII, no. 3 : May 2016

Peterson et al.

Doximity Residency Rankings Impact on Applicant Rank Lists

Table 2. Factors important in making a rank list (rank those selected in question above). Factor

Number of respondents placing factors in certain ranking 1st

2nd

3rd

4th

5th

Interview experience

183

174

134

Geographical preference

175

168

161

Experience with residents

109

158

141

Proximity to spouse/significant other/family

133

Doximity rankings

reputation affects decision-making, and although the effect size is small, our study supports that applicant perception of reputation through rankings may impact their decision-making with residency rank lists.15,16 Medical students believe their programâ&#x20AC;&#x2122;s reputation will impact their future career prospects and the medical school faculty consider a schoolâ&#x20AC;&#x2122;s reputation in terms of their own visibility and the opportunities for career advancement, resources and research. There is a strong interest on the part of students and EM programs for accurate, objective data about training programs. The inclusion of objective data could help guide applicants in selection of the best training environment for each learner. However, objective data for residency programs is limited and varies, and there is also the question of what data to include for a ranking system and more importantly, whether programs are willing to be transparent with certain information. Board passage rates were included in the Residency Navigator by Doximity for programs in internal medicine, family medicine, surgery and pediatrics. While markers like these are often considered by trainees and programs as indicators of successful training, these data speak to a single facet of training. Our study also confirms the results of a previous study by Love and colleagues4 that students are choosing programs based on personal factors such as geography and experiences. This may be due to the absence of objective data to assist decision-making. While it would be preferable to focus on objective data in lieu of rankings, we know from previous research (including this study) that rankings impact decisionmaking, and now Doximity has introduced an Internetsearchable residency ranking system that most applicants are aware exists. Perhaps efforts can be made to shift the way these rankings are generated and to promote searchable objective data about programs so that applicants can better identify the characteristics of programs that fit their individual interests and needs rather than creating an artificial roster of program superiority. Efforts to identify useful objective data, collect that data, and disseminate it in an easily navigable and Internet-searchable form could tremendously benefit student applicants by providing a set of metrics to evaluate and characterize programs in a transparent way. Other specialties have previously initiated work on this.17 There is currently a study in progress looking to build consensus around a set of reportable metrics that may allow applicants to rank programs Volume XVII, no. 3 : May 2016

according to their needs and expectations of a program, such as percentage of grads who go on to academic practice. LIMITATIONS Response rate is a limitation of this study, with only 793 students completing the survey out of an initial 1,641 students who were emailed, and may limit generalizability and provide response bias. It is possible that students chose to complete or not complete the survey based on preconceived perception of Doximity. Recall bias is another limitation in any survey-based study. Students may not remember exactly how Doximity affected their rank lists, as this survey was distributed after students had already submitted their lists. Another limitation is that students may not be able to fairly measure the impact of Doximity on their list choices since they do not have a personal comparison of applying prior to the release of Doximity. CONCLUSION In conclusion, while applicants do not find the Doximity rankings to be important compared to other factors in the application process, the Doximity rankings result in a small change in residency applicant ranking behavior. This unvalidated ranking, based principally on reputational data rather than objective outcome criteria, thus has the potential to be detrimental to students, programs, and the public. We feel it important for specialties to develop consensus around measurable training outcomes and provide freely accessible metrics for candidate education. In addition, there should be a greater emphasis on student advising and matching to a bestfit program rather than to the most highly ranked one.

Address for Correspondence: William J. Peterson, MD, University of Michigan, Department of Emergency Medicine, TC B1382, 1500 E Medical Center Dr, Ann Arbor, MI 48109-5002. Email: wpet@med.umich.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none.

353

Western Journal of Emergency Medicine

Peterson et al.

Doximity Residency Rankings Impact on Applicant Rank Lists Copyright: © 2016 Peterson et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

health-news/blogs/second-opinion/2014/09/10/doximitys-residencynavigator-injects-transparency-into-gme. Accessed Oct 25, 2014. 9. Joint EM Organization Letters Regarding Doximity Medical Student Survey. 2014. Available at: http://www.cordem.org/i4a/pages/index. cfm?pageid=3359. Accessed Nov 1, 2014. 10. Meredith M. Why Do Universities Compete in the Ratings Game? An Empirical Analysis of the Effects of the U.S. News and World Report

REFERENCES

College Rankings. Research in Higher Education. 2004;45(5):443461.

1. Davydow D, Bienvenu OJ, Lipsey J, et al. Factors Influencing the Choice of a Psychiatric Residency Program: A Survey of Applicants

11. Rolston AM, Hartley SE, Khandelwal S, et al. Effect of Doximity Residency Rankings on Residency Applicants’ Program Choices

to the Johns Hopkins Residency Program in Psychiatry. Acad

West J Emerg Med. 2015;16(g):889-893.

Psychiatry. 2008;32(2):143-146. 2. DeSantis M and Marco CA. Emergency Medicine Residency

12. House H and Gore L. Emergency Medicine Objects to Methodology

Selection: Factors Influencing Candidate Decisions. Acad Emerg

of Residency Rankings Published by Doximity and U.S. News &

Med. 2005;12(6):559-561.

World Report. ACEP Now 2014. Available at: http://www.acepnow. com/article/emergency-medicine-objects-methodology-residency-

3. Lee J, Alfieri M, Patel T, et al. Choosing family medicine residency

rankings-published-doximity-u-s-news-world-report/. Accessed Sep 1,

programs: What factors influence residents’ decisions? Can Fam

2015.

Physicians. 2011;57(3):e113-e121. 4. Love JN, Howell JM, Hegarty CB, et al. Factors That Influence

13. Bassin BS, Sikavitsas A, Benner, CA, et al. Development and evaluation of a multidisciplinary simulation based crisis management

Medical Student Selection of an Emergency Medicine Residency

curriculum. Acad Emerg Med. 2013;20:S301.

Program: Implications for Training Programs. Acad Emerg Med. 2012;19(4):455-460.

14. Rosenau AM, Mercer M, Mitchell M, et al. Letter to Mr. Ben Harder, USNWR. 2014; Available at: http://saem.org/docs/default-source/

5. Stefanidis D, Miles WS, Greene FL. Factors Influencing Residency

default-document-library/finalusnewsletter.pdf?sfvrsn=2. Accessed

Choice of General Surgery Applicants—How Important is the

Oct 25, 2014.

Availability of a Skills Curriculum? J Surg Educ. 2009;66(6):325-329. 6. American Association of Medical Colleges. Physician Specialty

15. Harvey N and Fischer I. Taking Advice: Accepting Help, Improving Judgment, and Sharing Responsibility. Org Behav Human Decision

Data Book. 2012. Available at: https://http://www.aamc.org/

Proc. 1997;70(2):117-133.

download/313228/data/2012physicianspecialtydatabook.pdf. Accessed Oct 25, 2014.

16. Meshi D, Biele G, Korn CW, et al. How Expert Advice Influences Decision Making. PLoS One. 2012;7(11):e49748.

7. Doximity. About Doximity. Available at: https://http://www.doximity. com/about/company. Accessed Oct 25, 2014.

17. Association of Professors of Gynecology and Obstetrics. Residency Directory. Available at: https://http://www.apgo.org/student/residency.

8. Harder B and Heilbrunn E. Doximity’s ‘Residency Navigator’ Injects

html. Accessed Oct 25, 2014.

Transparency into GME. 2014. Available at: http://health.usnews.com/

Western Journal of Emergency Medicine

354

Volume XVII, no. 3 : May 2016

Original Research

Emergency Medicine Resident Perceptions of Medical Professionalism Joshua Jauregui, MD* Medley O. Gatewood, MD* Jonathan S. Ilgen, MD, MCR* Caitlin Schaninger, MD† Jared Strote, MD, MS†

*University of Washington, Department of Emergency Medicine, Seattle, Washington † University of Cincinnati, Department of Emergency Medicine, Cincinnati, Ohio

Section Editor: Andrew W. Phillips, MD, MEd Submission history: Submitted October 30, 2015; Revision received January 26, 2016; Accepted February 21, 2016 Electronically published May 2, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29102

Introduction: Medical professionalism is a core competency for emergency medicine (EM) trainees; but defining professionalism remains challenging, leading to difficulties creating objectives and performing assessment. Because professionalism is dynamic, culture-specific, and often taught by modeling, an exploration of trainees’ perceptions can highlight their educational baseline and elucidate the importance they place on general conventional professionalism domains. To this end, our objective was to assess the relative value EM residents place on traditional components of professionalism. Methods: We performed a cross-sectional, multi-institutional survey of incoming and graduating EM residents at four programs. The survey was developed using the American Board of Internal Medicine’s “Project Professionalism” and the Accreditation Council of Graduate Medical Education definition of professionalism competency. We identified 27 attributes within seven domains: clinical excellence, humanism, accountability, altruism, duty and service, honor and integrity, and respect for others. Residents were asked to rate each attribute on a 10-point scale. We analyzed data to assess variance across attributes as well as differences between residents at different training levels or different institutions. Results: Of the 114 residents eligible, 100 (88%) completed the survey. The relative value assigned to different professional attributes varied considerably, with those in the altruism domain valued significantly lower and those in the “respect for others” and “honor and integrity” valued significantly higher (p<0.001). Significant differences were found between interns and seniors for five attributes primarily in the “duty and service” domain (p<0.05). Among different residencies, significant differences were found with attributes within the “altruism” and “duty and service” domains (p<0.05). Conclusion: Residents perceive differences in the relative importance of traditionally defined professional attributes and this may be useful to educators. Explanations for these differences are hypothesized, as are the potential implications for professionalism education. Because teaching professional behavior is taught most effectively via behavior modeling, faculty awareness of resident values and faculty development to address potential gaps may improve professionalism education. [West J Emerg Med. 2016;17(3):355–361.]

INTRODUCTION Medical Professionalism is one of six core competencies required by the Accreditation Council for Graduate Medical Volume XVII, no. 3 : May 2016

Education (ACGME). Emergency medicine (EM) residents must demonstrate a commitment to carrying out professional responsibilities and an adherence to ethical principles.1 However, 355

Western Journal of Emergency Medicine

Jauregui et al.

Resident Perceptions of Medical Professionalism over a decade after the implementation of these standards, the teaching and assessment of professionalism remains a challenge, in great part due to a lack of consensus on its definition.2-6 One of the greatest challenges to defining professionalism is its dynamic nature: reflecting ever-evolving expectations of patients and physicians in society, particularly regarding attributes that reflect the core of the doctor-patient relationship. In this context, it is important to understand trainees’ perspectives on what constitutes professionalism. From a top-down viewpoint, these residents’ culture and core values will inform professionalism standards in the future for emergency medicine. From a bottom-up consideration, it is important to measure a baseline set of values to inform and prioritize educational goals. Despite its importance, there have been few studies examining the values that residents place on different aspects of professionalism and none focusing solely on EM trainees.7-10 The primary objective of this study was to explore current general conceptualizations of professionalism among EM residents by assessing the relative value these trainees place on various professionalism attributes. The secondary objectives were to compare interns’ and seniors’ responses as a proxy of how of clinical and training experiences in EM may shape these values and to compare resident responses across four different sites to explore potential site- or region-specific differences. METHODS Study Design and Population This cross-sectional study surveyed a convenience sample of incoming and graduating residents at four EM residency programs representing the South (A), West (B), Midwest (C) and Northeast (D) regions of the United States. In 2011, incoming residents were polled during the first two months of their internship and graduating residents were polled within two months of graduation. Survey Instrument Using the American Board of Internal Medicine’s (ABIM’s) “Project Professionalism” and the ACGME’s definition of professionalism competency as guiding frameworks, we identified seven domains of professionalism (clinical excellence, humanism, accountability, altruism, duty and service, honor and integrity, respect for others) to be represented in the survey. Project Professionalism was a comprehensive multi-year undertaking by the ABIM to provide a modern definition, raise awareness, and guide education and assessment.11 Similarly, the ACGME’s definition is a core element of an initiative created jointly with the American Board of Medical Specialties to identify key educational elements of physician competency.12 Each domain had several specific attributes and each attribute was represented by an individual item. All items were developed through an iterative process by EM faculty after a review of the literature and published standards of professionalism. Subjects were asked to rate, on a 10-point Western Journal of Emergency Medicine

scale, to what extent each of 27 attributes contributed to their concept of medical professionalism with “none” and “completely” used as anchors at each end of the scale. In addition, subjects were asked whether professionalism was teachable in medical school or residency and whether these attributes could be assessed. We collected additional demographics, including residency location and year of training. We pilot tested the survey with 10 intern and senior internal medicine and EM residents at a single site for response process and for clarity.13 Feedback was incorporated into the final draft, and adaptations of items for the finalized instrument were based upon group consensus of the authors. Both the draft and final instruments had a total of 27 items. The survey instrument is provided in Appendix A. After the study was completed, in order to assess the internal structure of the instrument, we calculated internal consistency (n=100),using Cronbach’s alpha for the entire 27-item survey and for each of its domains.14 These values— representing the degree to which the instrument or domains map to the construct of professionalism or its domains, respectively—were classified a priori as “suboptimal” (values <0.70), “good” (0.70–0.89) or “substantial” (>0.90).15 Across the survey in its entirety, internal consistency was substantial (0.91). Within domains, internal consistency was good for clinical excellence (0.75), humanism (0.75), altruism (0.76), duty and service (0.83) and honor and integrity (0.77); internal consistency was suboptimal within the domains of accountability (0.52) and respect for others (0.66). Also assessed post hoc, in order to examine whether each question added to the survey, we assessed differences in distribution among responses within each domain using Wilcoxon signed rank repeated measures (two items) and Friedman chi-squared test for repeated measures (three or more items). This analysis demonstrated statistically significant differences (p<0.02). Survey Protocol We recruited participants by in-person, phone, and email requests. The survey was administered via an anonymized, secure, web-based platform. All participation was voluntary and there was no compensation for taking the survey. We defined response rate as those who submitted the survey, regardless of the time or form of request they were responding to. The Human Subjects Division at the primary author’s institution study approved the study with a waiver of consent. Data Analysis Data were compiled and entered into SPSS Statistics ver. 22, IBM Corporation (Chicago, IL). We used descriptive statistics to measure the mean and median for each item. Differences in mean scores for each domain were compared using repeated measures analysis of variance with Bonferroni-corrected post hoc comparisons. We performed two-tailed t-tests for each item, comparing 356

Volume XVII, no. 3 : May 2016

Jauregui et al.

Resident Perceptions of Medical Professionalism

responses from 1) incoming and graduating residents and 2) males and females. A one-way analysis of variance (ANOVA) was used to compare responses from the different residencies. We considered a p-value less than 0.05 statistically significant. RESULTS Of the 114 residents eligible to complete the survey, 100 (88%) completed it, with 36 (100%), 22 (92%), 19 (79%), and 23 (77%) completed at the South, West, Midwest and Northeast residencies respectively. Interns represented 54% of the sample and 55% were male. Males represented 29% of the interns and 84% of the senior residents. Mean and median scores for each professionalism attribute for both incoming and graduating residents are shown in Table 1. Scores varied considerably, with means ranging from 4.6 to 9.6. Table 2 shows mean scores within each professionalism domain. A one-way ANOVA revealed a significant difference in the mean domain scores (F=63.3, p=<0.001), which is attributable to lower scores in items related to “altruism” (p=≤0.004, differences ranged from 0.72 to 2.54), and higher scores in “respect for others” (p=0.000, differences ranged from 0.89 to 2.33) as well as “honor and integrity” (p=0.000, differences ranged from 1.1 to 2.54), relative to the other domains. There was no significant difference between “respect for others” and “honor and integrity” (p=0.182). A significant difference (p<0.05) was found between incoming and graduating residents for five attributes, each corresponding to an individual item under a domain (Table 3). These included “Commitment to lifelong learning,” “I should be an active leader in my community,” “A portion of my care for patients should be for those without pay,” “Active involvement in teaching and/or a professional organization,” and “Compassion and empathy.” In each case, the graduating seniors assigned less value when compared to their incoming intern counterparts. Differences among residencies were significant for three attributes: “I should always be there for my patients” (p=0.04); “ In an emergency, putting the welfare of others over my own” (p=0.001); and “My patients’ welfare should come above my need for sleep” (p=0.02) (Table 4). Resident responses to specific professionalism teachability/testability questions are summarized in Table 5. Overall, 82% felt that professionalism was teachable to residents, but only 37% thought it could be assessed. DISCUSSION Our assessment of EM residents’ self-reported conception of professionalism revealed variance in value placed on the different domains of professionalism competency as defined by the ACGME. This study adds to a growing literature on resident perspectives7-10 and is the first to focus solely on EM residents. Given the unique training experiences, cultural environments, and work practices of each specialty, different Volume XVII, no. 3 : May 2016

concepts of professionalism may be emphasized; analyses between residents of different specialties have confirmed such differences.8 Prior studies that pooled residents from multiple disciplines may therefore demonstrate additional variability or diluted results that reflect these different populations, making their results less generalizable to EM trainees. Our study differs from others in an additional way. The majority of prior studies were performed in a structured interview or focus group format,7-10 which risk an “interview effect:” values that could be perceived as different from traditional norms (such as minimizing the placement of others before oneself) could be de-emphasized. Our anonymous survey allowed residents to clearly and safely appraise concepts of professionalism without risk of judgment. Our most notable finding, and different from prior studies, was that “altruism” was rated significantly lower than all other domains. Altruism can be defined as when a physician, “adheres to (the) best interest of the patient; (and) puts (the) best interest of the patient above self-interest and the interest of other parties.”16 As the ACGME states in its common program requirements, “Residents are expected to demonstrate responsiveness to patient needs that supersede self-interest.”17 It is likely that our residents’ responses reflect a current reconceptualization of the traditional concept of altruism. Wellness, including restrictions on patient care such as work-hour limits, has been a priority for the entirety of their medical training, creating a culture where higher value may be frequently placed on physician self-interest than patient needs.20 The increasing ACGME emphasis on wellness both reflects and drives this change. It should be noted that physician altruism is a complicated concept with widely varying interpretations.18 One prior study found that residents perceived a focus on work-life balance to enhance professionalism by promoting well-being and teamwork;21 others note a perceived conflict between altruism and self-interest.22 A common concern in the literature regarding the development of professionalism in our learners is the continual commercialization of medicine and ongoing evolution of the biological and technical aspects of practice.2 As physicians become regarded more as service “providers,” an unintended consequence may be that our millennial learners are becoming less “patient relationship-focused” and more “commodity-focused” learners.23 Although the effects of changes in the perceived role of altruism as part of medical professionalism remain unclear, the speed of change in this domain relative to others likely creates an increasingly difficult divide between educators and learners in teaching and assessment. There is evidence that professionalism is largely learned in an implicit and experiential manner, creating difficulties for both faculty development and role-modeling when significant differences in values exist.28 If, as a specialty, we are committed to creating physicians who place high value on all traditional professional concepts,

357

Western Journal of Emergency Medicine

Jauregui et al.

Resident Perceptions of Medical Professionalism Table 1. Value placed by residents on each medical professionalism attribute. To what extent do the following contribute to your concept of medical professionalism?

Domain Excellence

Humanism

Accountability

Altruism

Mean

95% CI (Mean)

Median

Excellence in communication and listening

9.0

8.8-9.3

Technical competence, skill, excellence

8.6

8.3-9.0

Hard work and discipline

8.7

8.4-9.0

Ability to make difficult decisions with limited information

7.5

7.0-8.0

Commitment to lifelong learning

8.7

8.4-9.0

Compassion and empathy

8.9

8.6-9.2

Emotional Intelligence

8.4

8.0-8.7

An artist as much as a scientist

6.7

6.3-7.2

Commitment to social justice

7.5

7.1-7.9

Self-reflection and insight

8.1

7.7-8.4

Taking responsibility for mistakes

9.3

9.0-9.5

Autonomy in my decision making

7.8

7.4-8.2

My patients’ welfare should come before my need for balance in my life

5.1

4.6-5.6

My patients’ welfare should come above my financial interests

8.6

8.3-8.9

My patients’ welfare should come above my need for sleep

4.6

4.1-5.1

In an emergency, putting the welfare of others over my own safety

4.8

4.2-5.3

8.1

7.8-8.5

I should be an active leader in my community

7.6

7.3-8.0

Active involvement in teaching and/or a professional organization

7.6

7.2-8.0

A portion of my care for patients should be for those without means to pay

7.6

7.2-8.0

I should volunteer my skill and expertise for the welfare of the community

7.5

7.1-7.9

Honesty

9.6

9.5-9.8

Commitment to one’s personal and professional codes

9.1

8.9-9.4

My behavior should be used as a model for the community

7.9

7.5-8.3

My behavior away from work should be respectable

8.5

8.2-8.8

All patients should be treated equally

8.9

8.6-9.2

Respect for co-workers

9.4

9.2-9.6

Duty and service I should always be there for my patients

Honor and integrity

Respect for others

Table 2. Mean resident responses for each domain of professionalism. Domain

Mean (SD)

Excellence

8.27 (2.02)

Humanism

7.56 (2.20)

Accountability

7.93 (1.94)

Altruism

6.83 (2.69)

Duty and service

7.79 (1.95)

Honor and integrity

9.37 (1.09)

Respect for others 9.16 (1.32) Note: Significance of mean domain differences: Differences in mean scores for each domain was compared using repeated measures analysis of variance with Bonferroni-corrected post hoc comparisons. F=63.3, P=<0.001. Altruism < all others; Respect for others, Honor and integrity > all others.

the path may be difficult. In our study, for example, the values of many attributes were rated significantly lower when evaluated in more experienced residents. Devaluing Western Journal of Emergency Medicine

“commitment to lifelong learning” is notable, given their recent immersion in focused learning and the early point in their educational journey. Lowering the value placed on “a portion of 358

Volume XVII, no. 3 : May 2016

Jauregui et al.

Resident Perceptions of Medical Professionalism

Table 3. Difference between intern and senior resident responses. Attribute

Incoming residents’ mean score (SD)

Graduating residents’ mean score (SD)

P valuea

Commitment to lifelong learning

9.02 (1.35)

8.39 (1.51)

0.031

I should be an active leader in my community

8.00 (1.79)

7.22 (1.99)

0.042

A portion of my care for patients should be for those without means to pay

8.04 (1.88)

7.13 (2.20)

0.029

Active involvement in teaching and/or a professional organization

8.11 (1.72)

7.00 (1.96)

0.003

8.54 (1.93)

0.047

Compassion and empathy 9.17 (1.11) Comparing mean scores between incoming and graduating residents using a two-tailed t test.

Table 4. Differences among residency programs. Attribute

South (SD)

West (SD)

Midwest (SD) Northeast (SD)

P valuea

I should always be there for my patients

7.78 (1.72)

8.82 (1.18)

7.47 (1.87)

8.50 (2.13)

0.042 F[3,96] = 2.83

In an emergency, putting the welfare of others over my own

4.14 (2.29)

6.41 (3.16)

5.00 (2.47)

4.05 (2.44)

0.006 F[3,96] = 4.37

My patients’ welfare should come above my need for sleep 4.43 (2.14)

5.77 (2.96)

3.37 (2.31)

4.77 (2.37)

0.02 F[3,96] =3.45 a Comparing attribute scores among four different residencies using one-way ANOVA. F, F statistic. Number of responses=100.

Table 5. Resident responses to specific professionalism questions. Professionalism questions

Yes

Is professionalism teachable through a residency curriculum?

82%

18%

Is professionalism testable?

37%

63%

my care for patients should be for those without means to pay” seems to be misaligned with EM’s commitment to being the safety net for a community’s healthcare needs. The decreased significance of “compassion and empathy” also seems out of sync with EM’s core values, and may reflect changes in rolemodeling or organizational priorities in our teaching hospitals, or a natural cynicism arising from experiences in patient care. Although there is no way to fully assess the multi-factorial causes of these changes we found, some of the differences seen are consistent with studies of medical students that show a similar progressive decrease in baseline humanistic and empathic qualities.24-27 Such changes have been postulated to be due at least in part to an informal curriculum (interpersonal experiences and work expectations) that devalues altruism as well as a hidden curriculum (organizational structure and culture) in academic medical centers that may place value on metrics such as efficiency or billing over altruism.28-30 Although it is clear to the teachers that professionalism is difficult to teach, our participants overwhelmingly believe that it can be taught effectively. And while not formally analyzed, comments from the residents entered as free text in the survey consistently agreed that role modeling was the best way for Volume XVII, no. 3 : May 2016

them to learn professionalism. In a recent “Best Evidence in Medical Education” review, role modeling and mentoring were considered to be the most effective techniques for developing professionalism.24 And when EM and surgical residents were asked about their perspectives on professionalism, learning professionalism through role modeling was the most common theme.7,31 Despite these findings, fewer than half of U.S. and Canadian medical schools report providing formal faculty development in mentoring and only 8% provide assistance in the development and nurturing of professionalism.32 In addition, role models are often unaware of their educational impact, making faculty development or a reliance on informal teaching a challenge.29 Improved and increased faculty development, therefore, may be the low-hanging fruit to improve our ability to reinforce and teach professional values in our residents. Such development may need to take into account local custom and culture if there are, in fact, differences at training sites. Although survey responses may not adequately assess actual ethical and cultural values, among the four different institutions included in our study, residents’ self-conception of professionalism differed significantly among three attributes, two in the altruism category and one in the duty category. Of

359

Western Journal of Emergency Medicine

Jauregui et al.

Resident Perceptions of Medical Professionalism note, one of the institutions studied was a military emergency medicine residency program where much of the statistical difference occurred. These physicians’ professional values as military officers may have impacted their responses to the different professional attributes,33 reflecting how social pressures and environmental constraints influence professional attitudes and behaviors.10 The power of institutional culture on individual learning of professionalism is significant and can also potentially inform education and evaluation reform.34 One final challenge to professionalism education is assessment, viewed pessimistically by the majority of our respondents. Consistent with that sentiment, a 2012 consensus conference working group on assessing professionalism in EM concluded that existing instruments demonstrate insufficient reliability and validity to provide psychometrically robust assessment.4 Because professionalism is a “complex construct” this group recommended that it be evaluated with multiple methods, including personal portfolios and narratives, simulation, and direct observation among others.4 A significant challenge in assessing professionalism is adequate faculty awareness and confidence,35 and further emphasis on teacher training could potentially provide substantial benefits in this regard as well. LIMITATIONS There are several limitations to our study. Only 86% of residents responded; those who did not respond may have represented a different population that could substantially change our findings. Furthermore, variability of response rate from different institutions could have skewed the differences among institutions. Areas where no difference was found should be viewed with caution, as the study was designed to be primarily descriptive in nature and was not prospectively powered. Furthermore, the study results are largely descriptive and limited by the methods in its ability to draw any comparative conclusions to current professionalism conceptualizations. As this was the first use of this survey instrument, its validity evidence is thus limited, and there is potential that scores do not adequately reflect resident perceptions of professionalism. Given the large number of items included in the survey, it is possible that significant differences exist purely due to sampling error. In addition, the domains of professionalism in our survey instrument may be interdependent, but we did not test this. Also, we attempted to study a representative sample of the population (EM residents) of interest by surveying residents from four different institutions from four different regions of the country; however, this extremely limited sample (including a military residency that may differ from non-military residencies in critical ways) may not properly reflect EM residencies as a whole. Although we found certain differences between residency programs this cannot be interpreted as differences between regions. When comparing interns and graduating residents, we only looked at one snapshot of time. We also did not compare individuals before and after training or examine personal, Western Journal of Emergency Medicine

educational, or cultural factors that could have had influenced changes. We present the data for consideration only and do not attempt to draw conclusions about the effect of experience or training. Each residency had professionalism training and assessment during the period in question that may have affected the results. None of these were formal programs and given the ambiguity surrounding them, we chose not to include them in the analysis. Given all these limitations, the data in this study are provided to guide further research and education programs, rather than to draw definitive conclusions. CONCLUSION The relative value assigned to different professional attributes by the residents we surveyed showed variance and was significantly lower in the altruism domain. Differences were also found comparing learners at different levels of training and location. These findings likely reflect, at least in part, multiple different challenges in defining and teaching professionalism. Because the concept of professionalism is dynamic and likely best taught through role-modeling, increased faculty development, including an understanding of the current generations’ perceptions, should probably play a significant role in professionalism education.

Address for Correspondence: Joshua Jauregui, MD, Division of Emergency Medicine. Harborview Medical Center, Box 359702, 1CT89, 325 Ninth Avenue, Seattle, WA 98104. Email: joshjaur@ uw.edu. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Jauregui et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

360

1. Larkin GL, Binder L, Houry D, et al. Defining and evaluating professionalism: a core competency for graduate emergency medicine education. Acad Emerg Med. 2002;9(11):1249-56. 2. Birden H, Glass N, Wilson I, et al. Defining professionalism in medical education: a systematic review. Med Teach. 2014;36(1):47-61. 3. Lurie SJ, Mooney CJ, Lyness JM. Measurement of the general competencies of the accreditation council for graduate medical education: a systematic review. Acad Med. 2009;84(3):301-9. 4. Rodriguez E, Siegelman J, Leone K, et al. Assessing professionalism:

Volume XVII, no. 3 : May 2016

Jauregui et al.

Resident Perceptions of Medical Professionalism

summary of the working group on assessment of observable learner

org/acgmeweb/Portals/0/PDFs/jgme-11-00-47-51%5B1%5D.pdf.

performance. Acad Emerg Med. 2012;19(12):1372-8. 5. Sullivan C, Murano T, Comes J, et al. Emergency medicine directors’

Accessed May 6, 2015. 21. Ratanawongsa N, Bolen S, Howell EE, et al. Residents’ perceptions

perceptions on professionalism: a Council of Emergency Medicine

of professionalism in training and practice: barriers, promoters, and

Residency Directors survey. Acad Emerg Med. 2011;18 Suppl 2:S97103.

duty hour requirements. J Gen Intern Med. 2006;21(7):758-63. 22. Eggly S, Brennan S, Wiese-Rometsch W. “Once when i was on

6. van Mook WN, van Luijk SJ, O’Sullivan H, et al. The concepts of

call...,” theory versus reality in training for professionalism. Acad

professionalism and professional behaviour: conflicts in both definition and learning outcomes. Eur J Intern Med. 2009;20(4):e85-9.

Med. 2005;80(4):371-5. 23. Moreno-Walton L, Brunett P, Akhtar S, et al. Teaching across the

7. Cho CS, Delgado EM, Barg FK, et al. Resident perspectives on

generation gap: a consensus from the Council of Emergency

professionalism lack common consensus. Annals Emerg Med.

Medicine Residency Directors 2009 academic assembly. Acad Emerg

2014;63(1):61-7. 8. Krain LP, Lavelle E. Residents’ perspectives on professionalism. J

Med. 2009;16 Suppl 2:S19-24. 24. Birden H, Glass N, Wilson I, et al. Teaching professionalism

Grad Med Educ. 2009;1(2):221-4.

in medical education: a Best Evidence Medical Education

9. Regis T, Steiner MJ, Ford CA, et al. Professionalism expectations

(BEME) systematic review. BEME Guide No. 25. Med Teacher.

seen through the eyes of resident physicians and patient families. Pediatrics. 2011;127(2):317-24.

2013;35(7):e1252-66. 25. Newton BW, Barber L, Clardy J, et al. Is there hardening of the heart

10. van Mook WN, De Grave WS, Gorter SL, et al. Intensive care medicine trainees’ perception of professionalism: a qualitative study.

during medical school? Acad Med. 2008;83(3):244-9. 26. Rabow MW, Remen RN, Parmelee DX, et al. Professional formation:

Anaesth Intensive Care. 2011;39(1):107-15.

extending medicine’s lineage of service into the next century. Acad

11. ABIM. American Board of Internal Medicine Project Professionalism. 2001. Available at: http://www.abimfoundation.org/~/media/

Med. 2010;85(2):310-7. 27. Stephenson AE, Adshead LE, Higgs RH. The teaching of professional

Foundation/Professionalism/Project professionalism.ashx?la=en.

attitudes within UK medical schools: Reported difficulties and good

Accessed May 6, 2015.

practice. Med Educ. 2006;40(11):1072-80.

12. Swing SR. The ACGME outcome project: retrospective and

28. Bernard AW, Malone M, Kman NE, et al. Medical student

prospective. Med Teacher. 2007;29(7):648-54.

professionalism narratives: a thematic analysis and interdisciplinary

13. Cook DA and Beckman TJ. Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med.

comparative investigation. BMC Emerg Med. 2011;11:11. 29. House JB, Theyyunni N, Barnosky AR, et al. Understanding ethical

2006;119(2):166 e167-16.

dilemmas in the emergency department: views from medical

14. Downing SM. Validity: on meaningful interpretation of assessment data. Med Educ. 2003;37(9):830-7.

students’ essays. J Emerg Med. 2015;48(4):492-8. 30. Kim JY, Kang SJ, Kim JW, et al. Survey of attitudes on

15. Nunnally J. Psychometric Theory. 2nd ed. New York, NY: McGraw-

professionalism in plastic and reconstructive surgery. Arch Plast

Hill; 1978. 16. Rider EA, Nawotniak RH, Smith G. A Practical Guide to Teaching

Surg. 2013;40(2):134-40. 31. Park J, Woodrow SI, Reznick RK, et al. Observation, reflection, and

and Assessing the ACGME Core Competencies. Marblehead, MA:

reinforcement: surgery faculty members’ and residents’ perceptions

HCPro, Inc.; 2007. 17. ACGME. Accreditation Council for Graduate Medical Education

of how they learned professionalism. Acad Med. 2010;85(1):134-9. 32. Lown BA, Chou CL, Clark WD, et al. Caring attitudes in medical

Common Program Requirements. Available at: https://http://www.

education: perceptions of deans and curriculum leaders. J Gen Intern

acgme.org/acgmeweb/Portals/0/PFAssets/ProgramRequirements/ CPRs2013.pdf. Accessed May 6, 2015.

Med. 2007;22(11):1514-22. 33. Nace MC, Dunlow S, Armstrong AY. Professionalism in medicine: we

18. Glannon W, Ross LF. Are doctors altruistic? J Med Ethics. 2002;28(2):68-9; discussion 74-66.

should set the standard. Mil Med. 2009;174(8):807-10. 34. Hafferty FW, Wynia MK, Papadakis MA, et al. In reply to

19. Roberts DH, Newman LR, Schwartzstein RM. Twelve tips for

Barnhoorn and Youngson and to Jones and Thaxton. Acad Med.

facilitating Millennials’ learning. Med Teacher. 2012;34(4):274-8. 20. Ludwig S and Day S. The ACGME 2011 Duty Hour Standards.

2014;89(12):1579-80. 35. Pauls MA. Teaching and evaluation of ethics and professionalism: in

Chapter 7 New Standards for Resident Professionalism: Discusion

Canadian family medicine residency programs. Can Fam Physician.

and Justification. 2011. Available at: https://http://www.acgme.

2012;58(12):e751-6.

Volume XVII, no. 3 : May 2016

361

Western Journal of Emergency Medicine

Educational Advances

Introducing a Fresh Cadaver Model for Ultrasound-guided Central Venous Access Training in Undergraduate Medical Education Ryan Miller, BS* Hang Ho, BS* Vivienne Ng, MPH, MD† Melissa Tran, BS* Douglas Rappaport, MD‡ William J.A. Rappaport, MD§ Stewart J. Dandorf, MPH, MS* James Dunleavy, BS* Rebecca Viscusi, MD§ Richard Amini, MD†

*College of Medicine, The University of Arizona, Tucson, Arizona † Department of Emergency Medicine, The University of Arizona, Tucson, Arizona ‡ Department of Emergency Medicine, Beth Israel Deaconess, Boston, Massachusetts § Department of Surgery, The University of Arizona, Tucson, Arizona

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted February 15, 2016; Accepted March 21, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.30069

Introduction: Over the past decade, medical students have witnessed a decline in the opportunities to perform technical skills during their clinical years. Ultrasound-guided central venous access (USGCVA) is a critical procedure commonly performed by emergency medicine, anesthesia, and general surgery residents, often during their first month of residency. However, the acquisition of skills required to safely perform this procedure is often deficient upon graduation from medical school. To ameliorate this lack of technical proficiency, ultrasound simulation models have been introduced into undergraduate medical education to train venous access skills. Criticisms of simulation models are the innate lack of realistic tactile qualities, as well as the lack of anatomical variances when compared to living patients. The purpose of our investigation was to design and evaluate a life-like and reproducible training model for USG-CVA using a fresh cadaver. Methods: This was a cross-sectional study at an urban academic medical center. An 18-point procedural knowledge tool and an 18-point procedural skill evaluation tool were administered during a cadaver lab at the beginning and end of the surgical clerkship. During the fresh cadaver lab, procedure naïve third-year medical students were trained on how to perform ultrasound-guided central venous access of the femoral and internal jugular vessels. Preparation of the fresh cadaver model involved placement of a thin-walled latex tubing in the anatomic location of the femoral and internal jugular vein respectively. Results: Fifty-six third-year medical students participated in this study during their surgical clerkship. The fresh cadaver model provided high quality and lifelike ultrasound images despite numerous cannulation attempts. Technical skill scores improved from an average score of 3 to 12 (p<0.001) and procedural knowledge scores improved from an average score of 4 to 8 (p<0.001). Conclusion: The use of this novel cadaver model prevented extravasation of fluid, maintained ultrasound-imaging quality, and proved to be an effective educational model allowing third-year medical students to improve and maintain their technical skills. [West J Emerg Med. 2016;17(3):362–366.]

Western Journal of Emergency Medicine

362

Volume XVII, no. 3 : May 2016

Miller et al.

Ultrasound-guided Central Venous Access Training

INTRODUCTION Over the past decade, medical students have witnessed a decline in the opportunities to perform technical skills during their clinical years.1,2 Although the reasons for this diminution in procedural training are complex, it has led to medical students matriculating into residency with a shortage of exposure to common technical skills.3,4 As a result, simulationbased medical education has come to the forefront to provide training opportunities in a safe environment prior to patient encounters.5,6 Despite the level of quality that simulation can provide, the tactile and anatomic differences between manikin, phantom models, and the authentic human body make the latter a more attractive training modality.7,8 In recent years, literature reports on cadaver models for training in central venous cannulation have emerged.9-11 Though cadaver models improve intravenous cannulation simulation, free flow of fluid into a cannulated post-mortem vessel leads to extravasation of fluid thereby eliminating the cadaver’s utility for other procedures.10 In addition, extravasation of fluid results in a distortion of normal anatomy and eventual deterioration of ultrasound image quality.11 The purpose of our investigation was to design and evaluate a lifelike and reproducible training model for ultrasound-guided central venous access (USG-CVA) using a fresh cadaver. METHODS This was a single-center cross-sectional study conducted at an academic medical center. The study was reviewed by the institutional review board and determined to be exempt. The study participants were 56 third-year medical students without training in USG CV placement. The population had an average age of 29 of which 47% were men and 53% were female. The hands-on fresh cadaver procedure-lab was conducted at the beginning of the surgical clerkship and again five weeks later, and at our institution, the cadaver procedure lab is the first formal training session for USG CV placement. Participation in this study was voluntary and data was gathered from January 2015 to August 2015. All bodies used in our fresh cadaver lab were provided by the College of Medicine Willed Body Program and selflessly donated by our donors and their families. Model Design Equipment necessary includes the following: a marking pen, number 10 scalpel, latex tubing, free ties of 0 silk suture, Toomey syringe, water, red dye, 18-gauge and 22-gauge needles, guide wires, and US with a linear probe. Latex tubing that measures 5/16” outer diameter, 1/4” inner diameter with 1/32” thickness is used to construct the femoral vein model; a smaller latex tubing that measures 7/32” outer diameter, 5/32” inner diameter with 1/32” thickness is used for the internal jugular vein. Femoral Vein (FV) Model First, cut a 12-inch length of latex tubing. Use the 0 Volume XVII, no. 3 : May 2016

suture to tie off one end of the tubing so that injected fluid is confined to the tubing. An “X” is placed halfway between the anterior superior iliac spine and the pubic tubercle. This represents the approximate location of the femoral artery (FA) with the femoral vein (FV) lying approximately one centimeter medial. (Six to seven cm above the “X” a threecm transverse incision, is made down to the external oblique fascia. A similar transverse incision is made about 5cm below the inguinal ligament and likewise carried down to fascia. A long clamp is used to tunnel from the superior to inferior incision (Figure 1A). Once the tip of the clamp is seen, the open end of the latex tubing is placed in between the jaws of the clamp and pulled out from the superior wound. The tubing lies immediately on top of the femoral vein. As seen in figure 1B, by using a longer length of tubing than necessary one can pull the tube either inferiorly or superiorly to obtain a portion of tubing free of perforation sites if leaking were to cause either deterioration of the venous image and/or if venous distention becomes more difficult. Internal Jugular Vein (IJV) model For the internal jugular vein (IJV) cannulation a similar technique is used. Make an incision, just above the clavicle between the two heads of the sternocleidomastoid muscle (SCM) and as close to the upper border of the clavicle as possible. (Figure 2A). The inferior incision is carried down to the level of the fascia just superficial to the IJV, which is visible in the incision site as a light blue structure. Care must be taken due to the fact that the IJV lies superficial in the space between the two heads of the SCM muscle. The superior incision is made just inferior to the mandible and in a straight line connecting the inferior and superior incision. A long clamp is inserted into the inferior incision and tunneled just superficial to the exiting the superior incision. The latex tubing is grasped at the upper incision and pulled so as to exit through the inferior incision (Figure 2B). As in the FV model, using a longer than necessary segment of latex allows the tubing to be pulled so that a fresh, non-penetrated segment can be used when necessary. Fluid Infusion for Both Models A Toomey syringe filled with the fluid solution is fastened to the exposed open end of the tubing. In order to maximize the quality of US imaging, one should attempt to aspirate all air from the tubing prior to injecting fluid into the latex tubing. After aspirating, inject a solution of water mixed with 1ml red dye (approximately 20cc of fluid). After fluid infusion, both exposed ends of tubing and the Toomey syringe are hidden under the drape so as not to provide students with needle guiding cues (Figure 3). The simulated vessels are easily imaged and provide realistic anechoic, fluid-filled structures that can be compressed and cannulated (Figure 4). Students may then attempt US-guided identification and venous cannulation. An instructor, well versed in the procedure, is

363

Western Journal of Emergency Medicine

Miller et al.

Ultrasound-guided Central Venous Access Training

Figure 1. 1A) Femoral line model: The clamp is tunneled into the deep subcutaneous tissue midway between the ASIS and pubic tubercle, the bony landmarks are annotated by X’s; 1B) The latex tube has been tunneled into the femoral tissues immediately overlying the femoral sheath to replicate the native vessel anatomy. ASIS, anterior superior iliac spine

Figure 3. A drape is used to hide the exposed latex tubing and Toomey syringe. The model is now ready for implementation.

Figure 2. 2A) Internal jugular vein model: The two heads of the sternocleidomastoid (SCM) are drawn (in purple) to demonstrate the anterior triangle of the neck. The first incision is made between the two heads of the SCM just above the supraclavicular line (blue arrow). The second incision is made at the submandibular line (red arrow) 2B) A long clamp is used to tunnel through the anterior triangle at the level of the fascia from inferior to superior incision points. A length of latex tubing is clamped and pulled through the tunnel to replicate the IJV. IJV, internal jugular vein

present to actively instruct and ensure proper technique. Once the vein is cannulated and fluid aspirated, guide wire insertion may be performed. If extravasation into the tissue space occurs after multiple cannulation attempts, the latex tubing can be pulled from the upper or lower prior incision providing a non-punctured region of latex tubing for cannulation. Educational Curriculum and Assessment During the third-year surgery clerkship at our institution, all students participate in a fresh cadaver lab designed to teach the ABCs of trauma. This is a technical skill lab Western Journal of Emergency Medicine

Figure 4. The latex tubing provides an anechoic, compressible vessel clone that is durable enough to withstand multiple “sticks” without extravasation.

where students actively participate in chest tube insertion, endotracheal intubation, and CVA placement. After learning general principles in groups of 15, students were divided into groups of three to practice USG-CVA cannulation of the internal jugular and femoral veins. Upon completion of the educational session, students’ procedure knowledge and technical skills were individually assessed (post-test). Five weeks later, students were asked to return to the cadaver lab and their knowledge and technical skills were assessed again (retention-test). Student procedural knowledge was assessed with an 18-item checklist (Appendix A) that included components regarding indications, contraindications, and complications related to the procedure. Student technical skill 364

Volume XVII, no. 3 : May 2016

Miller et al.

Ultrasound-guided Central Venous Access Training

was assessed using an 18-item checklist (Appendix B), which was developed by two of the authors experienced with USGCVA placement.

Table. Mean (range) of pre- and post-test scores (N=56). Item Central line placement*

Data Analysis Wilcoxon ranked-sum testing for non-parametric samples was used to assess differences in population means. All analyses were performed using SAS statistical software (Version 9.2, Cary, NC, USA).

DISCUSSION During the past decade, numerous authors have noted a marked decline in formal procedural skills training during medical school.7,12-14 This is especially concerning for a procedure-rich residency such as emergency medicine, anesthesia, and general surgery, where numerous procedures such as CVC are performed and can lead to complications.15,16 The use of US has been shown to significantly decrease CVC complication risk, which may be further mitigated by improving procedural familiarity and competence through simulation education.17,18 In fact, the Agency for Healthcare Research and Quality has identified US-guided central venous placement as one of the practices in which strong evidence supports its widespread implementation.19 To ensure proficiency of this commonly performed procedure, US simulation models have been introduced to train venous access skills.19 Simulation-based medical education has proven to be a viable tool in CVC training and assessment.18 The fresh cadaver model improves upon the sim-man model by supplementing both tactile realism and maintaining the varied anatomic relationships found in live patients.7,8,12 Unfortunately, the fresh cadaver model introduces its own unique challenges. In the early years of using the fresh cadaver as a venous cannulation model, we found that after infusing approximately one liter of intra-venous fluid our cadavers became “bloated” as fluid extravasates out of the small vessels and lymphatics and into the interstitial space making the body less useful as the session proceeded. The result is distortion of normal anatomy and eventual

Post-test

P-value

2.7 (0-13) 12.4 (8-18) <0.0001

Number of indications

1.5 (1-3)

3.2 (2-4) <0.0001

Number of contraindications

0.8 (0-2)

1.7 (0-3) <0.0001

1.6 (1-3)

2.7 (1-4) <0.0001

Number of complications *Out of 18 maximum points.

RESULTS A total of 56 students participated in this study. All 56 (100%) of the students completed a survey before and after the educational session. Mean post- and retention-test scores for technical skills performing USG-CVA (out of 18 maximum points) and number of correctly identified indications, contraindications, and complications for central venous lines are shown in Table. There was a significant difference between mean scores for pre- and post-test central venous line placement (2.7 vs 12.4, respectively, p<0.0001). There was also significant difference in the mean number of correct indications, contraindications, and complications of central venous line placement (1.5 vs 3.2, 0.8 vs 1.7, and 1.6 vs 2.7, respectively, all p<0.0001).

Volume XVII, no. 3 : May 2016

Pre-test

deterioration of US imaging. One other author also describes this finding as a difficulty in the utilization of the fresh cadaver as a cannulation model.10 Similarly, in our own attempts to infuse fluid directly into cadaver vessels, we noted local fluid extravasation and subsequent deterioration of US imaging after only a few cannulation attempts. Our model improves upon this limitation, preserves cadaver imaging, and facilitates continued use of the cadaver for other procedures such as intubation, chest tube insertion, pericardiocentesis, etc. The thin-walled latex tubing used our novel fresh cadaver model minimizes fluid extravasation and maintains venous distention despite numerous cannulation attempts. This tubing is a very close replica to the native vein, providing both a compressible structure and an US image accurately resembling the natural vein seen in the living patient. In our experience, depending upon the gauge of needle used, one can obtain up to 20-30 “sticks” with a 22-gauge needle and 5-10 “sticks” with an 18-gauge needle, which is used if guide wire insertion is required of the trainees. Tunneling a long segment of the tubing permits pulling the tube at one end providing a new segment for continued cannulation. (Figure 1B) In addition, the tube diameter can be varied so as to accommodate different cadaver body habitus. The objective is to have the tubing in its normal anatomic position without being able to palpate the tubing. This allows students to use landmarks without unrealistic visual or tactile clues to guide needle placement. Finally, tunneling the latex tubing directly on top of the vein gives the added value of retaining the anatomic relationships seen in the living patient. Incorporating latex tubing of variable diameters can enable trainees to experience difficult as well as simpler cannulations, thus grading the challenge according to experience. LIMITATIONS Although this model appears to create a lifelike training simulation for CVC cannulation, the transference and retention of skills has been tested in a small group of students from one institution. In addition, our institution has an active well-supported Willed Body Program that provides donor cadavers at a very reasonable cost. Not every medical center will have this program available to them, and as a result cadaver and material cost are variable. In our study, we did not compare our novel model for CVC simulation to the standard fresh cadaver model. Another limitation of our model was not

365

Western Journal of Emergency Medicine

Miller et al.

Ultrasound-guided Central Venous Access Training having a simulated carotid artery. This certainly would have added to the realism of our model and will be considered in future design modifications. Lastly, our cadaver model relies on instructors for successful implementation.

6. Britt RC, Reed SF, Britt LD. Central line simulation: a new training algorithm. Am Surg. 2007;73:680-3. 7. Kaplan SJ, Carroll JT, Nematollahi S, et al. Utilization of a nonpreserved cadaver to address deficiencies in technical skills during the of medical school: a cadaver model for teaching technical skills.

CONCLUSION The use of this novel cadaver model prevented extravasation of fluid, maintained ultrasound-imaging quality, and proved to be an effective educational model allowing third-year medical students to improve and maintain their technical skills.

World J Surg. 2013;37:953-5. 8. Yang JH, Kim YM, Chung HS, et al. Comparison of four manikins and fresh frozen cadaver models for direct laryngoscopic orotracheal intubation training. Emerg Med J. 2010;27:13-6. 9. Wadman MC, Lomneth CS, Hoffman LH, et al. Assessment of a new model for femoral ultrasound-guided central venous access procedural training: a pilot study. Acad Emerg Med. 2010;17:88-92. 10. Varga S, Smith J, Minneti M, et al. Central venous catheterization

Address for Correspondence: Richard Amini, MD, The University of Arizona, Department of Emergency Medicine, 1501 N Campbell Ave, Tucson, AZ 85724. Email: richardamini@gmail.com.

using a perfused human cadaveric model: application to surgical education. J Surg Educ. 2015;72:28-32. 11. Hoyer R, Means R, Robertson J, et al. Ultrasound-guided procedures in medical education: a fresh look at cadavers. Intern Emerg Med.

2015 [in press]. 12. DiMaggio PJ, Waer AL, Desmarais TJ, et al. The use of a lightly preserved cadaver and full thickness pig skin to teach technical skills on the surgery clerkship--a response to the

Copyright: © 2016 Miller et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

economic pressures facing academic medicine today. Am J Surg. 2010;200:162-6. 13. Remes V, Sinisaari I, Hargula A, et al. Emergency procedure skills of graduating medical doctors. Med Teach. 2003;25:149-54. 14. Niemi-Murola L, Helenius I, Turunen J, et al. Graduating medical students and emergency procedure skill teaching in Finland—does a clinical skills centre make the difference? Med Teach. 2007;29:821-6.

REFERENCES

15. Petrosoniak A, Herold J, Woolfrey K. Emergency medicine procedural

1. Morton J, Anderson L, Frame F, et al. Back to the future: teaching medical students clinical procedures. Med Teach. 2006;28:723-8.

skills: what are residents missing? CJEM. 2013;15:241-8. 16. Promes SB, Chudgar SM, Grochowski CO, et al. Gaps in procedural

2. Talbott VA, Marks JA, Bodzin AS, et al. Technical skills acquisition

experience and competency in medical school graduates. Acad

in surgery-bound senior medical students: an evaluation of student assertiveness. J Surg Educ. 2012;69:529-35.

Emerg Med. 2009;16:S58-62. 17. Shekelle PG, Wachter RM, Pronovost PJ, et al. Making health care

3. Fox RA, Ingham Clark CL, Scotland AD, et al. A study of pre-

safer II: an updated critical analysis of the evidence for patient safety

registration house officers’ clinical skills. Med Educ. 2000;34:1007-12. 4. Yudkowsky R, Loy G, York J. Ensuring medical student competency

practices. Evid Rep Technol Assess (Full Rep). 2013;(211):1-945. 18. Ma IW, Brindle ME, Ronksley PE, et al. Use of simulation-based

in basic procedural skills. Med Educ. 2005;39:515-6.

education to improve outcomes of central venous catheterization: a

5. Evans DE, Wood DF, Roberts CM. The effect of an extended hospital induction on perceived confidence and assessed clinical

systematic review and meta-analysis. Acad Med. 2011;86:1137-47. 19. Evans LV, Dodge KL, Shah TD, et al. Simulation training in central

skills of newly qualified pre-registration house officers. Med Educ.

venous catheter insertion: improved performance in clinical practice.

2004;38:998-1001.

Acad Med. 2010;85:1462-9.

Western Journal of Emergency Medicine

366

Volume XVII, no. 3 : May 2016

Original Research

Authorship Trends of Emergency Medicine Publications over the Last Two Decades Richard Lammers, DO† Thomas Simunich, MSc, MBA* John Ashurst, DO, MSc†

* Duke Lifepoint Conemaugh Memorial Medical Center, Department of Research, Johnstown, Pennsylvania † Duke Lifepoint Conemaugh Memorial Medical Center, Department of Emergency Medicine, Johnstown, Pennsylvania

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted January 15, 2016; Accepted February 24, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29779

Introduction: With the recent merger of the American Osteopathic Association (AOA) and the Accreditation Council for Graduate Medical Education (ACGME) a heightened pressure for publication may become evident. Our objective was to determine whether there was a gap in the type of both medical degree designation and advanced degree designation among authorship in three United States-based academic emergency medicine journals. Methods: We reviewed the Journal of Emergency Medicine, Academic Emergency Medicine and Annals of Emergency Medicine for the type of degree designation that the first and senior authors had obtained for the years 1995, 2000, 2005, 2010 and 2014. Results: A total of 2.48% of all authors held a degree in osteopathic medicine. Osteopathic physician first authors contributed to 3.26% of all publications while osteopathic physician senior authors contributed 1.53%. No statistical trend could be established for the years studied for osteopathic physicians. However, we noted an overall trend for increased publication for allopathic senior authors (p=0.001), allopathic first authors with a dual degree (p=0.003) and allopathic senior authors with a dual degree (p=0.005). For each journal studied, no statistical trend could be established for osteopathic first or senior authors but a trend was noted for allopathic first and senior authors in the Journal of Emergency Medicine (p-value=0.020 and 0.006). Of those with dual degrees, osteopathic physicians were in the minority with 1.85% of osteopathic first authors and 0.60% of osteopathic senior authors attaining a dual degree. No statistical trend could be established for increased dual degree publications for osteopathic physicians over the study period, nor could a statistical trend be established for any of the journals studied. Conclusion: Very few osteopathic physicians have published in the Journal of Emergency Medicine, Academic Emergency Medicine or Annals of Emergency Medicine over the last two decades. Despite a trend for increased publication by allopathic physicians in certain journals, there appears to be no trend for increased publication of osteopathic physicians in emergency medicine. [West J Emerg Med. 2016;17(3):367–371.]

INTRODUCTION With the recent merger of the American Osteopathic Association (AOA) and the Accreditation Council for Volume XVII, no. 3 : May 2016

Graduate Medical Education (ACGME) new standards will be established for scholarly activity criteria and designation for each specialty. The ACGME requires “all core faculty 367

Western Journal of Emergency Medicine

Lammers et al.

Authorship Trends of Emergency Medicine Publications members to be involved in scholarly projects” with “one scientific peer-reviewed publication for every five core faculty members per year.”1 The American College of Osteopathic Emergency Physicians (ACOEP) requirements state “scholarly activity should include a minimum of 2 major or 1 major and 2 minor scholarly activities” every four years.2 Some of these activities include serving on committee, participating in item writing or grant writing, manuscript publications, or serving as a judge/moderator for an academic meeting.2 Recent research has shown that 39% of allopathic emergency medicine (EM) residencies require their residents to participate in an “original research project,” but this research did not include osteopathic residencies or faculty members in these results.3 Several other studies have also highlighted that osteopathic EM residencies are underrepresented in top tier EM journal publications and very few editors of top tier academic journals are osteopathic physicians.4,5 In this study, we sought to determine if there was a difference between allopathic and osteopathic physician publications in EM journals over the last two decades. METHODS Study Design This was a retrospective study of graduate degree (M.D. vs D.O.) trends in publications from the Annals of Emergency Medicine (Annals), Academic Emergency Medicine (AEM), and the Journal of Emergency Medicine (JEM). We chose journals based upon impact factor and citation half-life. For each author, reviewers designated medical degree based on the names listed with each article. An author’s graduate degree was determined by initial inspection of his or her suffix. For those with more than one advanced degree, both the medical degree and advanced degree were tallied for analysis. In articles published by a single author, the author was considered the first author for statistical purposes. All original articles, review articles, case reports, and image reports were reviewed for the type of medical degree the first and senior author had obtained from 1995, 2000, 2005, 2010 and 2014 by searching each journal’s web portal. We excluded from analysis articles such as media reviews, editorials, book reviews, letters to the editor and correspondences. Data Analysis We obtained comparison of the proportions of allopathic and osteopathic physician authorship across the years by using simple descriptive statistics. The percentage of those holding both a medical degree and an advanced degree was also analyzed using descriptive statistics. We analyzed trends in authorship using simple linear regression. RESULTS We reviewed 3,189 articles and 5,805 authors for the years studied. A total of 3,189 authors were considered Western Journal of Emergency Medicine

first authors and 810 of these authors held a dual degree. A total of 573 manuscripts were authored by a single author as well. A total of 2,616 authors were considered the senior author on a manuscript, with 667 of these authors holding a dual degree. Overall, 2.48% (144/5805) of authors held a degree in osteopathic medicine. Of those authors considered to be the first author, 3.26% (104/3189) held a degree in osteopathic medicine and 1.53% (40/2616) of senior authors held a degree in osteopathic medicine. Similarly, 1.85% (15/812) of first authors who held a dual advanced degree and 0.60% (4/667) of senior authors who held a dual advanced degree also had a degree in osteopathic medicine. Further analysis into each journal shows that 2.71% (48/1771) of all authors in JEM, 2.57% (50/1943) in AEM and 2.20% (46/2091) in Annals were osteopathic physicians. In 1995 a total of 3.82% (21/550) of osteopathic physicians served in the first author role and 2.06% (9/436) of osteopathic physicians served in the senior author role, while in 2014 a total of 2.68% (19/708) and 1.46% (9/615) of osteopathic physicians served in this role (Table 1). Likewise, in 1995 a total of 1.27% (1/79) of osteopathic physicians served in the first author role and held an advanced degree as compared to 1.23% (3/246) in 2014 (Table 2). For those holding both an osteopathic degree and an advanced degree who served in the senior author role in 1995, 0% (0/80) were osteopathic physicians as compared to 0.93% (2/214) in 2014 (Table 2). Overall, no trend was noted for increased osteopathic physician publication in either the first or senior author spot over time (p=0.375 and p=0.882). Neither could we establish a statistically significant trend for first and senior osteopathic physician publication over the years studied in any journal (Table 1). In several journals, however, statistically significant trends were established for allopathic publication for both the senior author and first author over time (Table 1). DISCUSSION This study highlights the significant disparity in degrees held by those publishing in three American academic EM journals. Based upon the results, osteopathic physicians publish less frequently in the journals studied and hold fewer advanced degrees as compared to their allopathic counterparts. Previous literature has shown that authors affiliated with osteopathic EM residencies are under-represented in high impact academic EM journals, but that study was only conducted for the year 2011.4 The static nature of reviewing only a single year made that data difficult to extrapolate. Based upon this study, however, osteopathic EM physicians have been under-represented in several key EM journals over the last two decades. Even as time has progressed, no trend could be established for increased publication by both first and senior authors who hold an osteopathic medical degree.

368

Volume XVII, no. 3 : May 2016

Lammers et al.

Authorship Trends of Emergency Medicine Publications

Table 1. Percent representation of osteopathic and allopathic physicians among first and senior authors of published articles in three journals of emergency medicine. 1995

2000

2005

2010

2014

Adj R Sq

P-value

Overall DO First DO Senior

3.82 (21/550)

3.60 (21/583)

3.65 (23/631)

2.79 (20/717)

2.68(19/708)

0.020

0.375

2.06 (9/436)

1.36 (6/442)

1.72 (9/523)

1.17 (7/600)

1.46 (9/615)

-0.322

0.882

MD First

96.18 (529/550)

96.40 (562/583) 96.35 (608/631)

97.21 (697/717)

97.32 (689/708)

0.677

0.055

MD Senior

97.94 (427/436)

98.64 (436/442) 98.28 (514/523)

98.83 (593/600)

98.54 (606/615)

0.978

0.001

JEM DO First

3.91 (5/128)

5.63 (8/142)

2.89 (5/173)

3.36 (8/238)

2.85 (7/246)

-0.092

0.475

DO Senior

2.88 (3/104)

2.5 (3/120)

1.85 (3/162)

0.45 (1/223)

2.13 (5/235)

-0.287

0.764

MD First

96.09 (123/128)

94.37 (134/142) 97.11 (168/173)

96.64 (230/238)

97.15 (239/246)

0.833

0.020

MD Senior

97.12 (101/104)

97.5 (117/120) 98.15 (159/162)

99.55 (222/223)

97.87 (230/235)

0.927

0.006

Acad EM DO First

3.24 (6/185)

3.40 (8/235)

4.07 (9/221)

2.90 (7/241)

3.57 (7/196)

-0.301

0.802

DO Senior

1.53 (2/131)

1.18 (2/170)

1.56 (3/192)

1.94 (4/206)

1.20 (2/166)

-0.130

0.516

MD First

96.76 (179/185)

96.60 (227/235) 95.93 (212/221)

97.10 (234/241)

96.43 (189/196)

-0.333

0.972

MD Senior

98.47 (129/131)

98.82 (168/170) 98.44 (189/192)

98.06 (202/206)

98.80 (164/166)

0.412

0.146

2.10 (5/238)

1.88 (5/266)

0.208

0.248

Annals of EM DO First DO Senior MD First

4.22 (10/237) 1.99 (4/201) 95.78 (227/237)

2.43 (5/206)

3.80 (9/237)

0.66 (1/152)

1.78 (3/169)

1.17 (2/171)

0.93 (2/214)

-0.097

0.480

97.57 (201/206) 96.20 (228/237)

97.90 (233/238)

98.12 (261/266)

-0.136

0.522

MD Senior 98.01 (197/201) 99.34 (151/152) 98.22 (166/169) 98.83 (169/171) 99.07 (212/214) 0.390 0.156 DO, Doctor of Osteopathic Medicine; MD, Doctor of Medicine; Annals, Annals of Emergency Medicine; AEM, Academic Emergency Medicine; JEM, Journal of Emergency Medicine

No study to date has directly explored why osteopathic physicians are less likely to publish manuscripts in journals, but several limitations for those with the osteopathic medical degree may exist that their allopathic counterparts do not face. For instance, allopathic residencies are at large academic institutions with a trend for increased funding, protected academic time, and large collaborative trials.6 These added resources may allow allopathic physicians the opportunity to publish articles at an increased rate. Another barrier that may exist is that there is insufficient research training in graduate medical education. After a survey of residents, Riveria et al noted the largest barriers to completing a successful research project were the lack of time, having inadequate research skills and a lack of a formal research curriculum.7 According to our results, allopathic physicians who hold a dual degree have been increasingly publishing manuscripts in the EM literature. In 2004, the American Association of Colleges of Osteopathic Medicine noted that only 26 osteopathic medical students were enrolled in a dual DO/PhD degree program, while in 2012 there were only 29.8,9 Several factors may dissuade students from pursuing a dual DO/ PhD degree, among them emphasis on acquiring extramural funding, the length of time to acquire the dual degree and the Volume XVII, no. 3 : May 2016

possibility of decreased income as an academic physician.10-12 Over the past several years, however, the osteopathic profession has made adjustments to the curriculum to better train dual degreed physicians and now offers 91 combined programs highlighted by nine DO/PhD, 15 DO/MS, and 16 DO/MPH programs.13 Since the merger of the AOA and the ACGME, neither organization has commented on an updated research requirement for core faculty. As it stands however, very few core faculty osteopathic EM physicians may meet the requirement if the ACGME standards are adopted. One way to improve the academic output and lay the ground work of scholarly activity for osteopathic EM physicians may be to enroll in the American College of Emergency Physicians Emergency Medicine Basic Research Skills (EMBRS) workshop, the American College of Osteopathic Physicians Faculty Development Workshop, or the Medical Education Research Certification with the Council of Emergency Medicine Program Directors. These programs not only offer the fundamentals of research but also allow for mentoring and collaborative efforts amongst the participants. LIMITATIONS There are several limitations to this study. The authors

369

Western Journal of Emergency Medicine

Lammers et al.

Authorship Trends of Emergency Medicine Publications

Table 2. Percent representation of osteopathic physicians among first and senior authors of published articles in three journals of emergency medicine who held a dual degree. 1995

2000

2005

2010

2014

Adj R Sq

P-value

Overall DO First

1.27 (1/79)

3.48 (4/115)

2.10 (3/143)

1.32 (3/227)

1.22 (3/246)

0.052

0.350

0 (0/80)

1.19 (1/84)

0.78 (1/128)

0 (0/161)

0.93 (2/214)

0.060

0.344

98.73 (78/79)

96.52 (111/115)

97.90 (140/143)

98.69 (224/227)

98.78 (243/246)

0.955

0.003

100 (80/80)

98.81 (83/84)

99.22 (127/128)

100 (161/161)

99.07 (212/214)

0.934

0.005

DO First

0 (0/16)

5.56 (1/18)

0 (0/27)

3.64 (2/55)

0 (0/61)

-0.277

0.740

DO Senior

0 (0/25)

0 (0/15)

0 (0/22)

0 (0/50)

1.61 (1/62)

0.278

0.209

MD First

100 (16/16)

94.44 (17/18)

100 (27/27)

96.36 (53/55)

100 (61/61)

0.853

0.016

MD Senior

100 (25/25)

100 (15/15)

100 (22/22)

100 (50/50)

98.39 (61/62)

0.687

0.052

DO First

0 (0/35)

4.35 (2/46)

3.57 (2/56)

1.34 (1/88)

2.53 (2/79)

0.041

0.358

DO Senior

0 (0/24)

2.86 (1/35)

1.79 (1/56)

0 (0/64)

1.31 (1/76)

-0.231

0.653

MD First

100 (35/35)

95.65 (44/46)

96.43 (54/56)

98.86 (87/88)

97.47 (77/79)

0.864

0.014

MD Senior

100 (24/24)

97.14 (34/35)

98.21 (55/56)

100 (64/64)

98.68 (75/76)

0.998

0.000

3.57 (1/28)

1.96 (1/51)

1.67 (1/60)

1.19 (1/84)

0.94 (1/106)

DO Senior MD First MD Senior JEM

Acad EM

Annals of EM DO First DO Senior MD First

0 (0/31)

0 (0/34)

0 (0/50)

0 (0/47)

0 (0/76)

96.43 (27/28)

98.04 (50/51)

98.33 (59/60)

98.81 (83/84)

99.06 (105/106)

0.971

0.001

MD Senior 100 (31/31) 100 (34/34) 100 (50/50) 100 (47/47) 100 (76/76) 0.823 0.021 DO, Doctor of Osteopathic Medicine; MD, Doctor of Medicine; Annals, Annals of Emergency Medicine; AEM, Academic Emergency Medicine; JEM, Journal of Emergency Medicine

cannot comment on the years that were not studied, which may have led to other osteopathic physicians who served as either the first or senior author that were not included in the study. The authors also cannot comment on the number of manuscripts submitted to each journal by osteopathic physicians over the time period. Although unlikely based upon trend analysis, if more or less osteopathic physicians published in these years not surveyed, it may have altered the results. Also, several authors were listed as “unknown” due to no degree being established with their authorship. While we made every effort to remove all “Letters to the Editor,” during the early years several of the journals did not clearly note these publications as such. All efforts were made to remove these from the study by review of the abstracts when available. Lastly, several journals publish brief reports, case reports and clinical images in the “Letters to the Editor” section. These were not included in the totals due to publication status in the “Letters” section, and it was felt that little change would be present in the results because of this. CONCLUSION Over the last two decades, very few osteopathic Western Journal of Emergency Medicine

physicians published original research articles, review articles, case reports, or image reports in JEM, AEM and Annals. Over the years studied, we saw no trend for increased publication by either the osteopathic first author or osteopathic senior author for the journals studied. As compared to their allopathic counterparts, very few osteopathic first or senior authors hold an advanced degree.

Address for Correspondence: John Ashurst, DO, MSc, Duke Lifepoint Memorial Medical Center, Department of Emergency Medicine, 1086 Franklin Street, Johnstown, PA 15905. Email: ashurst.john.32.research@gmail.com. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Lammers et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

370

Volume XVII, no. 3 : May 2016

Lammers et al.

Authorship Trends of Emergency Medicine Publications

REFERENCES

2000;342:250–255. 7. Riveria JA, Levine RB and Wright SM. Completing a scholarly project

1. ACGME Program Requirements for Graduate Medical Education in Emergency Medicine. (2012, September 30). Available at:

during residency training: Prospectives of residents who have been

http://www.acgme.org/acgmeweb/portals/0/pfassets/2013-pr-faq-

successful. J Gen Intern Med. 2005;(4):366–369. 8. American Association of Colleges of Osteopathic Medicine. 2004

pif/110_emergency_medicine_07012013.pdf. Accessed Apr 22,

Annual report on Osteopathic medical education.

2015.

9. American Association of Colleges of Osteopathic Medicine. 2012-

2. Basic Standards for Residency Training in Emergency Medicine. (2012, June 1). Available at: http://www.osteopathic.org/inside-aoa/

2013 Annual Report on Allied Health and Graduate Programs.

accreditation/postdoctoral-training-approval/postdoctoral-training-

Available at: http://www.aacom.org/docs/default-source/data-and-

standards/Documents/Basic-Standards-Emergency-Medicine.pdf.

trends/2012-13-allied-health-and-GProg.pdf?sfvrsn=8. Accessed Dec 13, 2015.

Accessed Apr 22, 2015. 3. Geyer BC, Kaji AH, Katz ED et al. A national evaluation of the scholarly

10. Schrier RW. Ensuring the survival of the clinician-scientist. Acad Med. 1997;72:589 –594.

activity requirement in residency programs: A survey of emergency medicine program directors. Acad EM. 2015;22(11):1337–1344.

11. Weinberg AM. Lessons in financial health and debt management for young physicians: loan consolidation programs, loan deferments, and

4. Baskin S, Lin C and Carlson J. Osteopathic emergency medicine

tax savings. J Am Osteopath Assoc. 2002;102:545-554.

programs infrequently publish in high impact emergency medicine journals. WestJEM. 2014;15(7):908–912.

12. Ley TJ and Rosenburg LE. The physician-scientist career pipeline in 2005. Build it, and they will come. JAMA. 2005;294:1343-1351.

5. Galuska M and Ashurst JV. Osteopathic Physicians on the Editorial Boards of Major Medical Journals Over the Last 30 Years. Accepted

13. American Association of Colleges of Osteopathic Medicine. Osteopathic Medical College Information Book: Entering Class 2016.

for publication. J Am Osteopath Assoc.

Available at: http://www.aacom.org/docs/default-source/cib/2016_cib.

6. Moy E, et al. Distribution of research awards from the National

pdf. Accessed Dec 13, 2015.

Institutes of Health among medical schools. N Engl J Med.

Volume XVII, no. 3 : May 2016

371

Western Journal of Emergency Medicine

Original Research

Out-of-Hospital Surgical Airway Management: Does Scope of Practice Equal Actual Practice? Molly Furin, MD, MS* Melissa Kohn, MD, MS* Ryan Overberger, DO, MS* David Jaslow, MD, MPH†

* Albert Einstein Healthcare Network, Department of Emergency Medicine, Philadelphia, Pennsylvania † Philadelphia University, Department of Emergency Medicine, Philadelphia, Pennsylvania

Section Editor: Andrew W. Phillips, MD, MEd Submission history: Submitted September 21, 2015; Revision received March 5, 2016; Accepted March 21, 2016 Electronically published May 5, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.3.28729

Introduction: Pennsylvania, among other states, includes surgical airway management, or cricothyrotomy, within the paramedic scope of practice. However, there is scant literature that evaluates paramedic perception of clinical competency in cricothyrotomy. The goal of this project is to assess clinical exposure, education and self-perceived competency of ground paramedics in cricothyrotomy. Methods: Eighty-six paramedics employed by four ground emergency medical services agencies completed a 22-question written survey that assessed surgical airway attempts, training, skills verification, and perceptions about procedural competency. Descriptive statistics were used to evaluate responses. Results: Only 20% (17/86, 95% CI [11-28%]) of paramedics had attempted cricothyrotomy, most (13/17 or 76%, 95% CI [53-90%]) of whom had greater than 10 years experience. Most subjects (63/86 or 73%, 95% CI [64-82%]) did not reply that they are well-trained to perform cricothyrotomy and less than half (34/86 or 40%, 95% CI [30-50%]) felt they could correctly perform cricothyrotomy on their first attempt. Among subjects with five or more years of experience, 39/70 (56%, 95% CI [44-68%]) reported 0-1 hours per year of practical cricothyrotomy training within the last five years. Half of the subjects who were able to recall (40/80, 50% 95% CI [39-61%]) reported having proficiency verification for cricothyrotomy within the past five years. Conclusion: Paramedics surveyed indicated that cricothyrotomy is rarely performed, even among those with years of experience. Many paramedics felt that their training in this area is inadequate and did not feel confident to perform the procedure. Further study to determine whether to modify paramedic scope of practice and/or to develop improved educational and testing methods is warranted. [West J Emerg Med.

2016;17(3):372–376.]

INTRODUCTION Emergency airway management has changed dramatically over the past few decades both in the emergency department (ED) and the prehospital setting. The widespread use of rapid sequence intubation, video laryngoscopy, and the development of multiple alternative airway devices have increased oral intubation success. The de-emphasis on endotracheal intubation in resuscitation algorithms has altered the setting in which airway management occurs. From the data in the Western Journal of Emergency Medicine

372

National Emergency Medical Services Information System (NEMSIS), endotracheal intubation is typically performed once every 225 patient care incidents.1 Emergency medical services (EMS) protocols for the use of rapid sequence intubation and cricothyrotomy vary considerably across the country. The training, experience, and confidence levels of prehospital providers are perhaps equally as diverse. Cricothyrotomy is the definitive emergency technique to secure a patient’s airway when endotracheal intubation fails Volume XVII, no. 3 : May 2016

Furin et al.

Out-of-Hospital Surgical Airway Management

and the patient is unable to be ventilated by non-invasive methods. The incidence of surgical cricothyrotomy has declined in recent years and emergency providers may have less exposure to the procedure.2-5 This lack of clinical exposure and performance may affect the success rate and confidence level of providers in accomplishing the procedure. While prior studies examine the prevalence and effectiveness of cricothyrotomy by helicopter EMS and in combat or austere environments, there is an inadequate amount of current literature to evaluate clinical competency in out-of-hospital surgical airway management by ground EMS crews.6-9 Since the procedure is so rare, it is difficult to use standard quality assurance methods to confirm competency. The purpose of this study was to assess clinical exposure, education, and self-perceived competency by ground paramedics in out-of-hospital surgical airway management. METHODS This is a self-administered cross-sectional survey consisting of 22 questions regarding years of experience, surgical airway attempts, types of devices used, training in out-of-hospital surgical airway management, skills verification, and self-perceptions regarding procedural competence and comfort. This study was approved by the institutional review board at the authors’ institution. The state of Pennsylvania includes out-of-hospital surgical airway management within the scope of practice of paramedics. This study includes four ground EMS agencies in four different counties within the state of Pennsylvania. The organizations include fire-based services and non-profit, municipal services. The number of annual emergency calls varies, with the largest agency reporting approximately 13,000 calls per year. The agencies serve areas that range from rural to suburban to urban communities. Annual skills testing under the supervision of the medical director is required for paramedics to verify competency in necessary procedures and knowledge of state protocols. All actively practicing paramedics participate in the annual skills review. The out-of-hospital surgical airway management survey was self-administered during the course of the skills review for medical command reauthorization at the different agencies. The study included both full-time and part-time medics, as long as they were actively practicing in one of the participating 911 emergency response agencies. Medics who practice at more than one participating agency completed only one survey. The survey was self-administered during the skills review sessions and returned to the investigators at the end of each session. We used Likert scales to design the survey. The outcome measures include number of cricothyrotomies performed and observed, the amount of training received in cricothyrotomies, and the providers’ comfort levels and confidence in performing the procedure. We compiled and then analyzed Volume XVII, no. 3 : May 2016

data using descriptive statistics, which are reported here (Microsoft Excel 2010). RESULTS One hundred percent of ground paramedics actively practicing at one of the four agencies (86/86) responded to the survey. The cohort was heavily career staff with slightly more than half having more than 10 years of experience (Table 1) and most working more than 36 hours per week (Table 2). Only 20% of subjects had attempted out-ofhospital surgical airway management (17/86; 95% CI [1128%]). Of these paramedics, 76% have greater than 10 years’ experience in the field (13/17; 95% CI [53-90%]). Of the 17 attempts at out-of-hospital surgical airway management five were by needle cricothyrotomy, 10 with commercial devices, and two by open cricothyrotomy. Seventy-three percent of subjects indicated that they do not believe that they are trained well enough to perform outof-hospital surgical airway management (63/86; 95% CI [6482%]). Similarly, only 40% (34/86, 95% CI [30-50%]) were confident in their ability to perform cricothyrotomy correctly on the first attempt. Figure 1 displays provider responses when asked “which best describes how confident you would be that you could correctly perform a prehospital cricothyrotomy on the first attempt tomorrow?” Half of subjects with five or more years of experience had less than one hour of experience (39/70, 56%, 95% CI [44-68%]) per year of practical cricothyrotomy training within the last five years. Half of the subjects who were able to recall (40/80, 50% 95% CI [39-61%]) reported having proficiency verification for cricothyrotomy within the past five years. Figure 2 shows provider recall of the number of lecture and practical training hours in cricothyrotomy in the past one year and past five years. Table 1. Respondent experience of paramedics in out-of-hospital surgical airway management study (years). Experience (years)

Paramedics (number)

0 to 5

6 to 10

11 to 15

16 to 20

21 to 25

>26

Table 2. Respondent experience (hours). On-duty time (hrs/week) Paramedics (number)

373

0 to 20

21 to 35

36 to 50

>51

Western Journal of Emergency Medicine

Furin et al.

Out-of-Hospital Surgical Airway Management

Figure 1. How confident do you feel that you could correctly perform a prehospital cricothyrotomy on the first attempt tomorrow?

Figure 2. How much and what type of training have you had in prehospital cricothyrotomy in the past 1 year and the past 5 years?

Western Journal of Emergency Medicine

374

Volume XVII, no. 3 : May 2016

Furin et al.

Out-of-Hospital Surgical Airway Management

DISCUSSION Principles of emergency airway management have evolved over the past several decades. Rapid sequence intubation has become commonplace in the ED, and the use of these techniques has been adopted in many prehospital systems. Numerous airway devices and alternative types of video and direct laryngoscopes have been developed and adopted by emergency physicians and prehospital providers. Most recently, the emphasis on obtaining what is considered a “definitive airway,” in most cases endotracheal intubation, has shifted to other aspects of resuscitative care. These changes have affected the setting in which paramedics practice and have likely altered the type of airway management they use. Difficult airway algorithms often include needle or surgical cricothyrotomy when endotracheal intubation cannot be achieved by other means. Surgical cricothyrotomy has decreased in the nation’s EDs, which parallels the use of rapid sequence intubation, video laryngoscopy, and the use of supraglottic airways outside of the operating room.5 Similarly, as these techniques become more common in the prehospital setting, in particular neuromuscular blocking agents, this has also contributed to the diminishing frequency of cricothyrotomy.3 The results from this study demonstrate few attempts at cricothyrotomy even after years of practice. As the use of this procedure becomes less frequent, providers may become less competent to perform the procedure adequately. Establishment and maintenance of procedural competence require in-depth training and skills review. Naturally as prehospital providers become more skilled in difficult airway management, surgical airways become less frequent.10 In the increasingly rare cases where other methods of airway management fail, training of cricothyrotomy on mannequins may improve success rates and performance times.11-13 Competence, however, declines over time from initial training.14 In fact, it is recommended that refresher training be conducted at least every six months to maintain competence.15,16 Initial paramedic training may be brief yet includes multiple life-saving interventions and skills. Continuing education in many states, such as Pennsylvania, is variable and does not have required elements. While medical directors for EMS agencies must confirm paramedic skills annually, there is no uniform process by which to do so and much is left to the discretion of the particular medical director. The amount of training and continuing education that focuses on airway skills can be highly variable. Due to the brief amount of time paramedics may have to focus on continuing education and the maintenance of skills, it is crucial that the most valuable interventions be emphasized and these procedural skills be assessed regularly. If paramedics are not performing this skill in the field, and are not receiving recurrent education and training, their ability to successfully complete the procedure is probably reduced. While this study did not directly assess paramedics’ ability to successfully complete a surgical cricothyrotomy, it did evaluate their comfort level and selfVolume XVII, no. 3 : May 2016

confidence to perform the procedure. The majority of the paramedics did not feel trained well enough to perform an outof-hospital cricothyrotomy, and most did not believe that they could perform it correctly on the first attempt. A meta-analysis of 56 studies by Hubble et al reported needle cricothyrotomy success rates of 65.8% and surgical cricothyrotomy success rates of 90.5%, with the physician success rate slightly higher at 97.1%.17 Paramedic training and skills education were variable in many of these studies though, and results may not be able to be generalized to the broader group of ground paramedics. Some studies report six hours of initial didactic sessions with added airway labs, and others require refresher procedure labs annually or every two years.15,18,19 The results from this survey show that this population, including four services from diverse areas of Pennsylvania, did not receive that amount of education or skills re-verification. This broad sampling may be more similar to other groups of ground paramedics. In this setting, success rates of cricothyrotomies may be significantly less than those reported in prior studies. While variability exists in the performance of surgical airway management, it is clear from existing literature that the procedure is typically reserved for the most critical patients. Several studies show extremely poor survival rates, and even fewer patients who survive with good neurologic outcomes.15,18,20,21 The amount of time and training to establish and maintain procedural competency among ground paramedics must be weighed against the perceived potential for positive patient outcomes. LIMITATIONS Certain limitations exist that limit the utility of this dataset. Cricothyrotomy is rarely performed, and therefore it is difficult to study prospectively or to analyze outcomes. Survey methodology was deemed the most appropriate study design in order to assess paramedics’ previous training, confidence, and experience in performing the skill. Currently, there is no validated survey available on this specific topic. Given the staff turnover rate at many EMS agencies, reproducing the study with the same subjects would be nearly impossible to perform. It was understood that in performing this type of study that reliability would be a limitation. This study relied on paramedics’ recollections of previous patient encounters and is therefore subject to recall bias. Success of the procedure was subjectively determined by the paramedics and is understood that clear parameters for success could not be established. Paramedics are likely to recall attempting to perform a cricothyrotomy due to the unique and stressful circumstances. They may have incorrect recollections about the amount of training on cricothyrotomies they received. However, it was deemed more important to assess paramedics’ subjective recollections about their education and training rather than a medical director’s perception as to whether or not the skill had been taught. This survey attempted to characterize a paramedic’s own

375

Western Journal of Emergency Medicine

Furin et al.

Out-of-Hospital Surgical Airway Management perceptions regarding an infrequently used skill in airway management. Medics often work with several different squads and may have received additional training at an outside agency, which would have affected their confidence levels; however, this training would also be reported in the survey. To ensure maximal response rate and limit sampling bias, the survey was only distributed during in-person training sessions, which resulted in a smaller number of subjects. This was mitigated by including diverse agencies in four counties, but as a result the study may not be representative of agencies across the state of Pennsylvania or in other states. Additional investigations with a larger study population in other EMS systems may be warranted.

1998;5(3):247-251. 5. Hessert MJ and Bennett BL. Optimizing emergent surgical cricothyrotomy for use in austere environments. Wilderness Environ Med. 2013;24(1):53-66. 6. Mabry RL and Frankfurt A. Advanced airway management in combat casualties by medics at the point of injury: A sub-group analysis of the reach study. J Spec Oper Med. 2011;22(2):16-19. 7. Mabry RL and Frankfurt A. An analysis of battlefield cricothyrotomy in Iraq and Afghanistan. J Spec Oper Med. 2012;12(1):17-23. 8. McIntosh SE, Swanson ER, Barton ED. Cricothyrotomy in air medical transport. J Trauma. 2008;64(6):1543-1547. 9. Robinson KJ, Katz R, Lenworth MJ. A 12-year experience with prehospital cricothyrotomies. Air Med J. 2001;20(6):27-30. 10. Swanson ER and Fosnocht DE. Effect of an airway education

CONCLUSION Cricothyrotomy is an infrequently performed intervention and patient outcomes are typically poor. Prehospital providers have limited time and access to continuing education and training sessions and must be relied upon to deliver medical care that improves patient outcomes. The responses in this study demonstrate that paramedics may benefit from additional training and recurrent education and skills verification.

program on prehospital intubation. Air Med J. 2002;21(4):28-31. 11. Cho J, Kang GH, Kim EC, et al. Comparison of manikin versus porcine models in cricothyrotomy procedure training. Emerg Med J. 2008;25(11):732-734. 12. Greif R, Egger L, Basciani RM, et al. Emergency skill training- a randomized controlled study of effectiveness of the 4-stage approach compared to traditional clinical teaching. Resuscitation. 2010;81(12):1692-1697. 13. Wong DT, Prabhu AJ, Coloma M, et al. What is the minimum training required for successful cricothyrotomy?: A study in mannequins. Anesthesiology. 2003;98(2):349-353.

Address for Correspondence: Molly Furin, MD, MS, Albert Einstein Healthcare Network, 5501 Old York Road Korman Building, Department of Emergency Medicine, Philadelphia, PA 19141. Email: furinmol@einstein.edu.

14. Siu LW, Boet S, Borges BC, et al. High-fidelity simulation

15. Fortune JB, Judkins DG, Scanzaroli D, et al. Efficacy of

demonstrates the influence of anesthesiologistsâ&#x20AC;&#x2122; age and years from residency on emergency cricothyroidotomy skills. Anesth Analg. 2010;111(4):955-960. prehospital surgical cricothyrotomy in trauma patients. J Trauma. 1997;42(5):832-838. 16. Kudavalli PM, Jervis A, Tighe SQM, et al. Unanticipated difficult airway management in anesthetized patients: a prospective study of

Copyright: ÂŠ 2016 Furin et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

the effect of mannequin training on management strategies and skill retention. Anaesthesia. 2008;63(4):364-369. 17. Hubble MW, Wilfong DA, Brown LH, et al. A meta-analysis of prehospital airway control techniques part II: Alternative airway devices and cricothyrotomy success rates. Prehosp Emerg Care. 2010;14(4):515-530. 18. Marcolini EG, Burton JH, Bradshaw JR, et al. A standing-order

REFERENCES

protocol for cricothyrotomy in prehospital emergency patients.

1. Wang HE, Mann NC, Mears G, et al. Out-of-hospital airway

Prehosp Emerg Care. 2004;8(1):23-28.

management in the United States. Resuscitation. 2011;82(4):378-386.

19. Warner KJ, Sharar SR, Copass MK, et al. Prehospital management

2. Berkow LC, Greenberg RS, Kan KH et al. Need for emergency

of the difficult airway: A prospective cohort study. J Emerg Med.

surgical airway reduced by a comprehensive difficult airway program. Anesth Analg. 2009;109(6):1860-1869.

2009;36(3):257-265. 20. Gerich TG, Schmidt Ul, Hubrich V, et al. Prehospital airway

3. Bulger EM, Copass MK, Maier RV, et al. An analysis of advanced

management in the acutely injured patient: The role of surgical

prehospital airway management. J Emerg Med. 2002;23(2):183-189. 4. Change RS, Hamilton RJ, Carter WA. Declining rate of

cricothyrotomy revisted. J Trauma. 1998;45(2):312-314. 21. Spaite DW and Joseph M. Prehospital cricothyrotomy: An

cricothyrotomy in trauma patients with an emergency medicine

investigation of indications, technique, complications, and patient

residency: implications for skills training. Acad Emerg Med.

outcome. Ann Emerg Med. 1990;19(3):279-285.

Western Journal of Emergency Medicine

376

Volume XVII, no. 3 : May 2016

Original Research

Pilot Study to Determine Accuracy of Posterior Approach Ultrasound for Shoulder Dislocation by Novice Sonographers Shadi Lahham, MD, MS* Brent Becker, MD† Alan Chiem, MD, MPH‡ Linda M. Joseph, MD* Craig L. Anderson, PhD, MPH* Sean P. Wilson, MD* Mohammad Subeh, MD, MS* Alex Trinh, BS* Eric Viquez, BS* John C. Fox, MD*

*University of California Irvine, Department of Emergency Medicine, Orange, California † Wellspan York Hospital, Department of Emergency Medicine, York, Pennsylvania ‡ Olive View-UCLA Medical Center, Department of Emergency Medicine, Los Angeles, California

Section Editor: Mark I. Langdorf, MD, MHPE Submission history: Submitted November 16, 2015; Revision received January 10, 2016; Accepted February 12, 2016 Electronically published April 26, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29290

Introduction: The goal of this study was to investigate the efficacy of diagnosing shoulder dislocation using a single-view, posterior approach point-of-care ultrasound (POCUS) performed by undergraduate research students, and to establish the range of measured distance that discriminates dislocated shoulder from normal. Methods: We enrolled a prospective, convenience sample of adult patients presenting to the emergency department with acute shoulder pain following injury. Patients underwent ultrasonographic evaluation of possible shoulder dislocation comprising a single transverse view of the posterior shoulder and assessment of the relative positioning of the glenoid fossa and the humeral head. The sonographic measurement of the distance between these two anatomic structures was termed the Glenohumeral Separation Distance (GhSD). A positive GhSD represented a posterior position of the glenoid rim relative to the humeral head and a negative GhSD value represented an anterior position of the glenoid rim relative to the humeral head. We compared ultrasound (US) findings to conventional radiography to determine the optimum GhSD cutoff for the diagnosis of shoulder dislocation. Sensitivity, specificity, positive predictive value, and negative predictive value of the derived US method were calculated. Results: A total of 84 patients were enrolled and 19 (22.6%) demonstrated shoulder dislocation on conventional radiography, all of which were anterior. All confirmed dislocations had a negative measurement of the GhSD, while all patients with normal anatomic position had GhSD>0. This value represents an optimum GhSD cutoff of 0 for the diagnosis of (anterior) shoulder dislocation. This method demonstrated a sensitivity of 100% (95% CI [82.4-100]), specificity of 100% (95% CI [94.5100]), positive predictive value of 100% (95% CI [82.4-100]), and negative predictive value of 100% (95% CI [94.5-100]). Conclusion: Our study suggests that a single, posterior-approach POCUS can diagnose anterior shoulder dislocation, and that this method can be employed by novice ultrasonographers, such as non-medical trainees, after a brief educational session. Further validation studies are necessary to confirm these findings. [West J Emerg Med. 2016;17(3):377–382.] Volume XVII, no. 3 : May 2016

377

Western Journal of Emergency Medicine

Lahham et al.

Posterior Approach Ultrasound for Shoulder Dislocation INTRODUCTION Acute shoulder dislocation is common in the emergency department (ED) with an estimated incidence rate of 23.9 per 100,000 person-years.1 This affects 1.7% of the population and results in nearly 200,000 ED visits annually.2 Overall, 95-98% of shoulder dislocations are anterior. 3-5 Prompt recognition of this condition is essential for effective treatment, as reduction becomes increasingly difficult the longer the shoulder remains dislocated.6 The current standard approach for patients with suspected shoulder dislocation involves conventional radiography, performed both before and after reduction of the joint.1,2,7 Such imaging is often delayed in the ED due to a myriad of factors, prolonging patient discomfort and potentially complicating subsequent attempts at reduction. Ultrasound (US) provides a rapid, point-of-care (POC) imaging modality that may facilitate the physician’s diagnosis of shoulder dislocation at the bedside and expedite definitive treatment.3,4,8 The orthopedic literature has demonstrated applicability of US in the evaluation of chronic shoulder conditions, such as instability, rotator cuff injuries and labral tears in the outpatient clinic setting. However, there is sparse literature documenting the use of US to diagnose acute shoulder injury or suspected dislocation.9-13 The emergency medicine literature contains several case reports detailing successful emergency physician (EP) use of US to evaluate shoulder dislocations. However, only one prospective study, by Abbasi et al, examines POCUS for the diagnosis of acute shoulder injuries.8,14,15,16 In the Abbasi study, all USs were performed by two sonographers, one of whom was particularly experienced in shoulder US, and entailed obtaining multiple sonographic views of the shoulder.16 A streamlined, single-view US technique to evaluate for shoulder dislocation that is accessible to even the most inexperienced sonographers would be of clinical benefit to EPs. The primary objective of this study was to derive a standardized method for diagnosing shoulder dislocation using a single-view, posterior approach POCUS performed by undergraduate research students. A secondary objective was to determine the accuracy of this sonographic method in the diagnosis of shoulder dislocation by inexperienced sonographers. MATERIALS AND METHODS Study Design This was a prospective, observational study of adult patients presenting with acute, traumatic shoulder pain aimed at deriving a single-view, sonographic method of diagnosing shoulder dislocation. This was based on the relative positioning of the glenoid fossa and the humeral head, as well as the distance between these two structures. Undergraduate research students enrolled all patients, performed the USs and calculated sonographic measurements. The study was approved by the study site institutional review board. Western Journal of Emergency Medicine

Setting The study was conducted at a single, urban, universitybased hospital ED with an emergency medicine residency and an annual census of approximately 50,000 patients. Selection of Participants All adult ED patients at least 18 years of age presenting with an acute, traumatic shoulder pain necessitating conventional radiography were eligible for inclusion. We excluded patients if they were under 18 years of age, not expected to undergo conventional radiography, incarcerated or could not otherwise provide consent. A convenience sample of patients was enrolled between April 2011 and January 2015 daily between 8:00am and midnight. Undergraduate research students approached all adult ED patients presenting with a complaint of acute, traumatic shoulder pain for whom the treating physician ordered conventional radiographs of the shoulder. The students were present in the ED daily between 8:00am and midnight, but no patients were enrolled between the hours of midnight and 8:00am due to the lack of research staff availability. As part of the undergraduate research program, all participating students were fully trained in the consenting and enrollment process. After providing written consent and before receiving conventional radiographs, enrolled patients underwent bedside ultrasonography of the affected shoulder. Undergraduate students with no previous sonographic experience collected all data. These students were enrolled in an emergency medicine research course and collected data for various studies. These novice sonographers were blinded to the results of the radiographs. While it was not possible to blind the undergraduate research students to the physical appearance of the shoulder, they did not possess formal medical training in anatomy or physical examination, which may have effectively blinded them from a clinical assessment. Similarly, conventional radiographs were formally read by attending radiologists blinded to the results of the POCUS. US Technique Prior to the start of patient enrollment, each undergraduate research student received a 30-minute lecture on basic shoulder anatomy followed by a 30-minute hands-on US training session. This session was standardized and given by the US director at our institution. This was repeated once per year for a total of three sessions over the entire enrollment period. Each undergraduate research student performed the US study using a Sonosite Edge (FUJIFILM Sonosite Inc). A 10-5 Mhz linear transducer was placed transversely on the posterior aspect of the patient’s shoulder with the probe indicator to the patient’s right with the patient seated upright. Both the glenoid rim and humeral head were visualized and identified (Figures 1, 2). For both the glenoid rim and the humeral head, the student operator placed a horizontal line, tangent to the most

378

Volume XVII, no. 3 : May 2016

Lahham et al.

Posterior Approach Ultrasound for Shoulder Dislocation (GhSD<0cm) represented an anterior position of the glenoid rim relative to the humeral head. A GhSD value of 0cm implied that the most posterior aspect of the glenoid rim and humeral head were in perfect vertical alignment on the screen (Figures 3, 4). The GhSD values were measured in real time by the undergraduate students and recorded using standardized data collection forms. Outcome Measures The primary sonographic outcome measure of interest was the GhSD. This measurement was correlated with the presence or absence of a dislocation seen on conventional radiography to derive an appropriate GhSD cutoff for diagnosing shoulder dislocation with US. Primary Data Analysis Study data were entered in Excel (Microsoft, Redmond WA) and analyzed using Stata (version 12.1, StataCorp, College Station TX). We compared patient characteristics using chi-square tests for independence and t-tests, with calculation of exact binomial confidence intervals. The optimum GhSD cutoff value was determined to be used in defining the derived US method. Finally, we calculated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the POCUS method.

Figure 1. Superior view illustrating a posterior approach to a right shoulder ultrasound.

Figure 2. Schematic illustration demonstrating anatomic probe position of the right shoulder with an overhead view. G, glenoid; HH, humeral head

posterior aspect of each of the two anatomic structures. The distance between these two parallel lines was measured by a third line placed perpendicular to the previous two and defined the Glenohumeral Separation Distance (GhSD) (Figure 3, 4). The GhSD, measured in centimeters (cm), was given a positive or negative value based on the location of the glenoid rim relative to the humeral head. A positive GhSD value (GhSD>0cm) represented a posterior position of the glenoid rim relative to the humeral head and a negative GhSD value Volume XVII, no. 3 : May 2016

RESULTS A total of 103 patients were approached for enrollment in the study, of whom nine declined to participate and one was a prisoner. Nine patients were considered for enrollment but were diagnosed clinically and reduced at the bedside by provider prior to radiography. We included 84 patients in the data analysis, enrolled by 31 of 54 trained undergraduate research students (Figure 5). Patient ages ranged from 19 to 72 years old with a median age of 45 years old. Nineteen (22.6%) patients had radiography-confirmed dislocations, all of which were anterior. Fourteen of 52 male patients (26.9%) and 5 of 32 (19.6%) female patients had shoulder dislocations. The GhSD in patients with confirmed dislocations and in patients without dislocation are shown in Table 1 and Table 2, respectively. All confirmed dislocations demonstrated a GhSD<0cm and all confirmed non-dislocations demonstrated a GhSD>0cm. This is represented graphically in Figure 6. Thus GhSD=0cm was chosen as the cutoff value for the diagnosis of anterior shoulder dislocation. This derived POCUS method demonstrated a sensitivity of 100% (95% CI [82.4-100]), specificity of 100% (95% CI [94.5-100]), PPV of 100% (95% CI [82.4-100]), and NPV of 100% (95% CI [94.5-100]) in the diagnosis of anterior shoulder dislocations. DISCUSSION Shoulder dislocations are commonly diagnosed and managed in the ED. Unfortunately, history and physical examination is often insufficient to definitively make the

379

Western Journal of Emergency Medicine

Lahham et al.

Posterior Approach Ultrasound for Shoulder Dislocation

Figure 3. Ultrasound image depicting dislocated right shoulder. Both the humeral head (HH) and glenoid fossa (GF) are depicted with an illustration of the measured glenohumeral separation distance (GhSD).

Figure 4. Ultrasound image depicting normal right shoulder anatomy. Both the humeral head (HH) and glenoid fossa (GF) are depicted with an illustration of the measured glenohumeral separation distance (GhSD).

diagnosis.1,17 Radiographic imaging is therefore almost always employed prior to attempting reduction, as well as after reduction to confirm successful reduction. Radiographs are also used to determine whether the patient has developed common complications of shoulder dislocations, such as a Bankart lesion or Hill-Sachs lesion. Currently, typical care dictates that these patients obtain pre- and post-reduction imaging to confirm dislocation and that ED reduction is successful.17, 18 Waiting to obtain radiographs to diagnose shoulder dislocations has potential to delay definitive care. A study by Shuster et al has demonstrated that pre-reduction plain film radiographs increases ED length of stay by 29.6 minutes.19 Multiple case reports have found US useful in identifying acute dislocations.8,14,15,20 This is a promising Western Journal of Emergency Medicine

imaging modality as not all healthcare facilities may have access to plain film imaging, especially in developing countries and austere environments. US confers a number of potential benefits in the diagnosis and management of acute shoulder dislocation. It has been shown that sonographic diagnosis of shoulder dislocation can be obtained within five minutes of initial evaluation.16 This efficiency is especially helpful in patients requiring procedural sedation for reduction. An unsuccessful reduction may necessitate repeat sedation if a patient recovers before imaging can be performed.21 Therefore, POCUS may be used to confirm adequate shoulder reduction, which could eliminate the need for repeat sedation and its associated risks. There is also the potential to decrease radiation exposure, especially in patients who require multiple manipulations. A recent study by Abbasi et al, employing a combined anterior and lateral approach, showed a sensitivity and specificity of 100% for US in the diagnosis of shoulder dislocation in 69 dislocated patients among 73 studied.16 The study demonstrated no difference in image acquisition and interpretation between an experienced emergency sonographer and a senior emergency medicine resident who had received a one-hour lecture and performed 10 supervised shoulder USs. This study suggests that physicians with focused training are able to accurately identify shoulder dislocations using POCUS; however, it remains unclear if non-medically trained practitioners can do the same. In our study, we attempted to derive a standardized method for diagnosing shoulder dislocation using a singleview, posterior approach POCUS technique. Given that there is no way to blind clinicians from physical exam findings and to eliminate the possibility of physical exam bias, undergraduate research students without formal anatomy or medical training enrolled patients and performed USs. While there are several physical exam findings indicative of shoulder dislocation, there are no sonographic findings that are universally agreed upon.15 We propose the following method of diagnosing anterior shoulder dislocations using a posterior transverse US approach: A positive humeral head distance relative to the glenoid indicates a normally placed shoulder, while a negative humeral head distance relative to the glenoid indicates an anteriorly dislocated shoulder. Use of this measurement for the diagnosis of anterior shoulder dislocation resulted in a sensitivity of 100%, and specificity of 100%, with no overlap in the GhSD between the dislocated and non-dislocated groups (Figure 5). Our data suggest that despite minimal training, image acquisition and measurements obtained by non-clinical personnel were found to be adequate for clinical use. More importantly, given the large number of true negatives, we hope to use these measurements to help create a standard approach to shoulder US US and standardize a sonographic definition of a dislocated shoulder. While these results may not 380

Volume XVII, no. 3 : May 2016

Lahham et al.

Posterior Approach Ultrasound for Shoulder Dislocation clinicians can employ this technique with minimal training to expedite the diagnosis of anterior shoulder dislocation.

Figure 5. 103 patients were screened for enrollment; 19 were ineligible and 84 patients consented and were enrolled. Of these, 19 had shoulder dislocations and 64 patients did not. POCUS, point of care ultrasound

LIMITATIONS There are several limitations to our study. We used a convenience sample that was small and enrollment was limited to a single center. Nine patients were reduced prior to enrollment without additional imaging. We were also unable to blind sonographers to the visual appearance of the shoulder and potential deformities. US scans were not repeated by more experienced practitioners to confirm findings. Data on injuries other than glenohumeral dislocation, such as fractures, were not specifically collected and the ability of US to detect other potential shoulder injuries was not evaluated. No posterior dislocations were diagnosed during the course of the study. While posterior dislocations are rather uncommon, we cannot comment on the ability of this technique to diagnosis a posterior shoulder dislocation. Given glenohumeral anatomy, posterior shoulder ultrasonography may not be able to correctly identify a posterior dislocation. CONCLUSION Our findings suggest that a single, posterior-approach POCUS technique can diagnose anterior shoulder dislocation and can be employed by novice, non-medical trainees after a brief educational session. We propose a definition that using a posterior transverse US of the shoulder, a positive humeral head distance relative to the glenoid indicates a normally placed shoulder while a negative step off distance of the humeral head relative to the glenoid indicates an anteriorly dislocated shoulder. Further validation studies are necessary to confirm these findings.

Address for Correspondence: Shadi Lahham, MD, MS, University of California Irvine, Department of Emergency Medicine, 333 City Boulevard West Suite 640, Orange, CA 92868. Email: Slahham@ uci.edu.

Figure 6. The measured glenohumeral separation of patients arranged by positive and negative dislocation.

be generalizable to all ED clinicians, they do show promise in the sense that with minimal training, clinicians can potentially obtain images that can be used for clinical decision-making. Furthermore, given their lack of clinical training, these students were effectively blinded from making a clinical diagnosis of shoulder dislocation. Given the strong correlation of the direction of the humeral head relative to the glenoid to the presence of a shoulder dislocation, we believe that Volume XVII, no. 3 : May 2016

Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: ÂŠ 2016 Lahham et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES 1. Zacchilli MA and Owens BD. Epidemiology of shoulder dislocations presenting to emergency departments in the United States. J Bone

381

Western Journal of Emergency Medicine

Lahham et al.

Posterior Approach Ultrasound for Shoulder Dislocation Joint Surg Am. 2010;92(3):542-9.

changes in the rotator cuff interval with arthroscopic correlation.

2. Youm T, Takemoto R, Park BKH. Acute Management of Shoulder Dislocations. J Am Acad Orthop Surg. 2014; 22(12):761.

Skeletal Radiol. 2005;34(9):522–7. 13. Jacobson JA, Lancaster S, Prasad A, et al. Full-thickness and partial-

3. Davids JR and Talbott RD. Luxatio erecta humeri. A case report. Clin

thickness supraspinatus tendon tears: value of US signs in diagnosis.

Orthop Relat Res. 1990;144-9. 4. Yamamoto T, Yoshiya S, Kurosaka M, et al. Luxatio erecta (inferior

Radiology. 2004;230(1):234–42. 14. Blakeley CJ, Spencer O, Newman-Saunders T, et al. A novel use

dislocation of the shoulder): a report of 5 cases and a review of the

of portable ultrasound in the management of shoulder dislocation.

literature. Am J Orthop (Belle Mead NJ). 2003;32(12):601-3. 5. Westin CD, Gill EA, Noyes ME, et al. Anterior shoulder dislocation.

Emerg Med J. 2009;26:662-3. 15. Yuen CK, Mok KL, Kan PG, et al. Bedside ultrasound for verification

A simple and rapid method for reduction. Am J Sports Med.

of shoulder reduction with the lateral and anterior approaches. Am J

1995;23(3):369-71. 6. G Allen. Shoulder ultrasound imaging – integrating anatomy,

Emerg Med. 2009;27:503 16. Abbasi S, Molaie H, Hafezimoghadam P, et al. Diagnostic accuracy

biomechanics and disease process. Eur J Radiol. 2008;68:137-46.

of ultrasonographic examination in the management of shoulder

7. McNamara R. Management of common dislocations. In: Roberts JR,

dislocation in the emergency department. Ann Emerg Med.

Hedges JR, editors. Clinical procedures in emergency medicine. 3rd ed. Philadelphia, PA: Saunders; 1998. p.818-52.

2013:62(2):170-5. 17. Hendey G. Necessity of radiographs in the emergency department

8. Halberg M, Sweeney TW, Owens WB. Bedside ultrasound for

management of shoulder dislocations. Ann Emerg Med. 2000;36:108-

verification of shoulder reduction. Am J Emerg Med. 2009;27:134. e5-134.e6.

13. 18. Riebel GD and McCabe JB. Anterior shoulder dislocation: a review of

9. Schydlowsky P, Strandberg C, Galbo H, et al. The value of ultrasonography in the diagnosis of labral lesions in patients with

reduction techniques. Am J Emerg Med 1991;9(2):180-8. 19. Shuster M, Abu-Laban RB, Boyd J. Prereduction radiographs in

anterior shoulder dislocation. Eur J Ultrasound. 1998;8:107-13.

clinically evident anterior shoulder dislocation. Am J Emerg Med.

10. Kolla S and Motamedi K. Ultrasound evaluation of the shoulder. Semin Musculoskelet Radiol. 2007;11(2):117-25.

1999;17:653-8. 20. Stone M and Sutijono D. Dynamic Emergency Medicine: Intraarticular

11. Morag Y, Jacobson JA, Lucas D, et al. US appearance of the rotator

injection and closed glenohumeral reduction with emergency

cable with histologic correlation: preliminary results. Radiology. 2006;241(2):485–91.

ultrasound. Acad Emerg Med. 2009;16(12):1384-5. 21. Wen DY. Current concepts in the treatment of anterior shoulder

12. Lee JC, Sykes C, Saifuddin A, et al. Adhesive capsulitis: sonographic

Western Journal of Emergency Medicine

dislocations. Am J Emerg Med 1999;17(4):401-7.

382

Volume XVII, no. 3 : May 2016

Images in Emergency Medicine

Point-of-Care Ultrasound Diagnosis of Left-Sided Endocarditis Charles W. Bugg, MD, PhD Kristin Berona, MD

Keck School of Medicine, LAC+USC Medical Center, Department of Emergency Medicine, Los Angeles, California

Section Editor: Rick A. McPheeters, DO Submission history: Submitted January 30, 2016; Accepted February 22, 2016 Electronically published May 2, 2016 Full text available through open access at http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.2016.2.29921

[West J Emerg Med. 2016;17(3):383–383.]

A 56-year-old male presented to the emergency department (ED) with fatigue, generalized weakness, cough without hemoptysis or dyspnea, occasional fever, night sweats, dark stools, and weight loss for months. Physical exam revealed a cachectic gentleman with bilateral distal pitting edema. Cardiac and respiratory exams revealed no adventitious sounds. Stool guaiac was negative for blood. Complete blood count revealed mild anemia without leukocytosis, and chest radiograph showed cardiomegaly with left pleural effusions. Initial providers were most concerned for malignancy, although differential included tuberculosis or endocarditis. Given the reassuring exam and chronicity of symptoms, initial disposition plan was discharge with close follow up for malignancy evaluation. With the consideration of endocarditis, a point-of-care cardiac ultrasound was performed revealing a mobile mass on the aortic valve and an irregular mitral valve (Video). After blood cultures and broad-spectrum antibiotics, the patient was admitted. Transthoracic echocardiogram (TTE), demonstrated vegetations on the mitral and aortic valves. Blood cultures grew enterococcus. The patient underwent mitral and aortic valve replacements without complications and was discharged home. Infective endocarditis (IE) is an uncommon disease and a challenging diagnosis to make in the ED. Symptoms are nonspecific, murmurs can be difficult to auscultate, and skin manifestations are rare. Though transesophageal echocardiogram is the most sensitive modality for evaluation of suspected IE, it is not available in most EDs. TTE only has sensitivity of 50-70%; however, larger lesions (>10mm) have a sensitivity of 84%.1 Vegetations are mobile, irregularly shaped structures usually on the upstream side of valves, with motion irrespective of valve. Differential includes myxomatous process, tumors, thrombi, or imaging artifact.2 Point-of-care ultrasound (POCUS) has been reported to diagnose both right- and left-sided IE.3,4 Though not definitive, POCUS can dramatically change disposition and expedite care in patients for Volume XVII, no. 3 : May 2016

whom emergency physicians are considering IE. Address for Correspondence: Charles W. Bugg, MD, PhD, Keck School of Medicine, LAC+USC Medical Center, Department of Emergency Medicine, 1200 N State St #4250, Los Angeles, CA 90033. Email: wally.bugg@gmail.com. Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. The authors disclosed none. Copyright: © 2016 Bugg et al. This is an open access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) License. See: http://creativecommons.org/ licenses/by/4.0/

Video. Point-of-care ultrasound videos showing aortic and mitral valve vegetations. REFERENCES

383

1. Reynolds HR, Jagen MA, Tunick PA, et al. Sensitivity of transthoracic versus transesophageal echocardiography for the detection of native valve vegetations in the modern era. J Am Soc Echocardiogr. 2003;16(1):67–70.  2. Infective Endocarditis. In: Feigenbaum H, ed. Feigenbaum’s Echocardiography. 6th ed. Philadelphia, PA: Lippincott, Williams and Wilkins; 2005. 3. Cheng AB, Levine DA, Tsung JW, et al. Emergency physician diagnosis of pediatric infective endocarditis by point-of-care echocardiography. Am J Emerg Med. 2012;30(2):386.e1-3. 4. Seif D, Meeks A, Mailhot T, et al. Emergency department diagnosis of infective endocarditis using bedside emergency ultrasound. Crit Ultrasound J. 2013;5(1):1.

Western Journal of Emergency Medicine

CHALLENGE YOURSELF AS A U.S. ARMY EMERGENCY MEDICINE PHYSICIAN.

As part of the U.S. Army health care team, you can make a real difference treating the immediate medical needs of Soldiers and their families while also growing in your career: • You’ll work with the most advanced emergency technology and have the opportunity to experience a variety of settings. • You’ll develop as a leader, learn from the best and have the ability to make quality patient care your main focus. • You’ll receive excellent benefits such as special pay, as well as the potential for continuing education and career specialty options.

To learn more about joining the U.S. Army health care team, visit healthcare.goarmy.com/fa09 or call 415-398-2329 or 714-953-9507.

TAB COPY GOES HERE

SAVE THE DATE!

November 1 - 5, 2016 Hilton Union Square, San Francisco, CA

ACOEP’s flagship conference brings you more than ever before. Highlights include: • Specialized Tracks in Critical Care, Pain Management, and Wellness

• Advanced Airway Course • Active Shooter Training • Advanced EKG Course • FOEM Legacy Gala: Dinner & Awards Ceremony

• Meeting of ACOEP’s Council for Women in Emergency Medicine • 1-on-1 Access to Expert Speakers

ON SALE NOW! Visit www.acoep.org/scientific for more information. Paramount Sponsor

Prime Sponsor

Supporting Sponsor

SEPTEMBER 20TH HYATT REGENCY ORANGE COUNTY, CA

Volume XVII, Number 3, May 2016

Open Access at www.westjem.com

ISSN 1936-900X

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health Indexed in MEDLINE

ENDEMIC INFECTIONS 238 Identify-Isolate-Inform: A Tool for Initial Detection and Management of Zika Virus Patients in the Emergency Department

KL Koenig, A Almadhyan, MJ Burns

VOLUME XVII, NUMBER 3, May 2016

2016

Western Journal of Emergency Medicine

PRESENTS

West

SOCIETAL IMPACT ON EMERGENCY CARE 245 Emergency Medical Treatment and Labor Act (EMTALA) 2002-15: Review of Office of Inspector General Patient Dumping Settlements

N Zuabi, LD Weiss, MI Langdorf

TREATMENT PROTOCOL ASSESSMENT 252 Prospective Validation of Modified NEXUS Cervical Spine Injury Criteria in Low-risk Elderly Fall Patients

J Tran, D Jeanmonod, D Agresti, K Hamden, RK Jeanmonod

PRACTICE VARIABILITY 258 Quality Improvement Initiative to Decrease Variability of Emergency Physician Opioid Analgesic Prescribing

JH Burton, JA Hoppe, JM Echternach, JM Rodgers, M Donato

CRITICAL CARE 264 Anticoagulation Reversal and Treatment Strategies in Major Bleeding: Update 2016

S Christos, R Naples

271

Academic Emergency Medicine Physiciansâ&#x20AC;&#x2122; Knowledge of Mechanical Ventilation

SR Wilcox, TD Strout, JI Schneider, PM Mitchell, J Smith, L Lutfy-Clayton, EG Marcolini, A Aydin, TA Seigel, JB Richards

DIAGNOSTIC ACUMEN 280 Anti-N-Methyl-D-Aspartate Receptor Encephalitis, an Underappreciated Disease in the Emergency Department

DR Lasoff, J Corbett-Detig, R Sell, M Nolan, G Wardi

PAGES 238-383

Contents continued on page iii

CALIFORNIAACEP.ORG > INFO@CALIFORNIAACEP.ORG A Peer-Reviewed, International Professional Journal