Volume 25 Issue 3

Volume 25, Number 3, May 2024

Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health

Indexed in MEDLINE

Behavioral Health

303 Addressing System and Clinician Barriers to Emergency Department-initiated Buprenorphine: An Evaluation of Post-intervention Physician Outcomes

Jacqueline J. Mahal, Polly Bijur, Audrey Sloma, Joanna Starrels, Tiffany Lu

312 Factors Associated with Acute Telemental Health Consultations in Older Veterans

Erica C. Koch, Michael J. Ward, Alvin D. Jeffery, Thomas J. Reese, Chad Dorn, Shannon Pugh, Melissa Rubenstein, Jo Ellen Wilson, Corey Campbell, Jin H. Han

320 Implementation and Evaluation of a Bystander Naloxone Training Course

Scott G. Weiner, Scott A. Goldberg, Cheryl Lang, Molly Jarman, Cory J. Miller, Sarah Li, Ewelina W. Stanek, Eric Goralnick

Critical Care

325 Emergency Department SpO2/FiO2 Ratios Correlate with Mechanical Ventilation and Intensive Care Unit Requirements in COVID-19 Patients

Gary Zhang, Michael J. Burla, Benjamin B. Caesar, Carolyne R. Falank, Peter Kyros, Victoria C. Zucco, Aneta Strumilowska, Daniel C. Cullinane, Forest R. Sheppard

Education

332 Geographic Location and Corporate Ownership of Hospitals in Relation to Unfilled Positions in the 2023 Emergency Medicine Match

Zachary J. Jarou, Angela G. Cai, Leon Adelman, David J. Carlberg, Sara P. Dimeo, Jonathan Fisher, Todd Guth, Bruce M. Lo, Laura Oh, Rahul Puttagunta, Gillian R Schmitz

Emergency Department Operations

342 Imaging in a Pandemic: How Lack of Intravenous Contrast for Computed Tomography Affects Emergency Department Throughput

Wayne A. Martini, Clinton E. Jokerst, Nicole Hodsgon, Andrej Urumov

Western Journal of Emergency Medicine:

Integrating Emergency Care with Population Health Indexed in MEDLINE, PubMed, and Clarivate Web of Science, Science Citation Index Expanded

Amin A. Kazzi, MD, MAAEM

The of MMM

The American University of Beirut, Beirut, Lebanon

Brent King, MD, MMM University of Texas, Houston

Christopher E. San Miguel, MD Ohio State University Wexner Medical Center

Christopher E. San Miguel, MD Ohio State University Wexner Medical Center

Daniel J. Dire, MD

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

Douglas Ander, MD Emory University

Emory University

Edward Michelson, MD Texas Tech University

Edward Michelson, MD Texas Tech University

Edward Panacek, MD, MPH

Edward Panacek, MD, MPH University of South Alabama

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

“Maggiore della Carità,” Novara, Italy

Gayle

Gayle Galleta, MD

Editorial Board

Editorial Board

Sørlandet Sykehus HF, Akershus Universitetssykehus, Lorenskog, Norway

Sørlandet Sykehus HF, Akershus Universitetssykehus, Lorenskog, Norway

Hjalti Björnsson, MD

Icelandic Medicine

Hjalti Björnsson, MD Icelandic Society of Emergency Medicine

Jaqueline MD

Desert

Jaqueline Le, MD Desert Regional Medical Center

Jeffrey Love, MD

The George Washington University School of Medicine and Health Sciences

Jeffrey Love, MD The George Washington University School of Medicine and Health Sciences

Katsuhiro Kanemaru, MD University of Miyazaki Hospital, Miyazaki, Japan

Kenneth V. Iserson, MD, MBA University of Arizona, Tucson

Katsuhiro University Miyazaki, Kenneth University

Leslie Chicago Medical School

Leslie Zun, MD, MBA Chicago Medical School

Linda University of

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

Niels K. Rathlev, MD Tufts University School of Medicine

Tufts University School of Medicine

Pablo Aguilera Fuenzalida, MD Pontificia Universidad Catolica de Chile, Región Metropolitana, Chile

Pablo Aguilera Fuenzalida, MD Pontificia Universidad Catolica de Chile

Peter A. Bell, DO, MBA Baptist Health Sciences University

Peter Sokolove, MD University of California, San Francisco

Baptist University of California, San Francisco

Rachel A. Lindor, MD, JD Clinic

Rachel A. Lindor, MD, JD Mayo Clinic

Robert Suter, DO, MHA

Robert Suter, DO, MHA UT Southwestern Medical Center

Derlet, University of California, Davis

Robert W. Derlet, MD University of California, Davis

Rosidah Ibrahim, MD

Hospital Serdang, Selangor, Malaysia

Rosidah Ibrahim, MD Hospital Serdang, Selangor, Malaysia

Scott Rudkin, MD, MBA

Scott Rudkin, MD, MBA University of California, Irvine

Elena Lopez-Gusman, JD

Elena Lopez-Gusman, JD

California ACEP

California ACEP

Mark I. Langdorf, MD, MHPE, MAAEM, FACEP

Isabelle Nepomuceno, BS Executive Editorial Director

Steven Changi

American College of Emergency Physicians

American College of Emergency

Jennifer Kanapicki Comer, MD FAAEM

AAEM School

California Chapter Division of AAEM Stanford University School of Medicine

DeAnna McNett, CAE

DeAnna McNett, CAE

Kimberly Ang, MBA

UC Irvine Health School of Medicine

UC Irvine Health School of Medicine

Robert Suter, DO, MHA American College of Osteopathic

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

MPH FAAEM, FACEP

Shahram Lotfipour, MD, MPH FAAEM, FACEP

Visha Bajaria, BS WestJEM Editorial Director

Emily Kane, MA WestJEM Editorial Director

Scott Zeller, MD University of California, Riverside

Scott Zeller, MD University

Singapore

Steven H. Lim, MD Changi General Hospital, Simei, Singapore

Terry Mulligan, DO, MPH, FIFEM ACEP Ambassador to the Netherlands Society

Terry Mulligan, DO, MPH, FIFEM ACEP Ambassador to the Netherlands Society of Emergency Physicians

Wirachin MSBATS

Wirachin Hoonpongsimanont, MD, MSBATS

Siriraj Mahidol Bangkok,

Editorial Staff Advisory Board

American College of Osteopathic Emergency Physicians

UC Irvine Health School of Medicine

American College of Osteopathic Emergency Physicians School

Randall J. Young, MD, MMM, FACEP

California ACEP

UC Irvine Health School of Medicine

Jorge Fernandez, MD, FACEP

Jorge Fernandez, MD, FACEP

UC San Diego Health School of Medicine

UC San Diego Health School of Medicine

Visha Bajaria, BS WestJEM Editorial Director WestJEM

Stephanie Burmeister, MLIS WestJEM Staff Liaison

Siriraj Hospital, Mahidol University, Bangkok, Thailand

American College of Emergency Physicians

Kaiser Permanente

American College of Emergency Physicians Kaiser Permanente

Physicians, :

Cassandra Saucedo, MS Executive Publishing Director

Nicole Valenzi, BA WestJEM Publishing Director

Cassandra Saucedo, MS BA WestJEM Publishing Director Casey,

June Casey, BA Copy Editor

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

PubMed, Central, CINAHL, Medscape, MDLinx Editorial and Publishing Office: JEM/Depatment of Emergency Medicine, UC Irvine Health, 3800 W. Chapman Ave. Suite 3200, Orange, CA 92868, USA

Available in MEDLINE, PubMed, PubMed Central, Europe PubMed Central, PubMed Central Canada, 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, 3800 W. Chapman Ave. Suite 3200, Orange, CA 92868, USA Office: 1-714-456-6389; Email: Editor@westjem.org

of Emergency

Integrating Emergency Care with Population Health

Indexed in MEDLINE, PubMed, and Clarivate Web of Science, Science Citation Index Expanded

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

345 The Utility of Dot Phrases and SmartPhrases in Improving Physician Documentation of Interpreter Use

Katrin Jaradeh, Elaine Hsiang, Malini K. Singh, Christopher R. Peabody, Steven Straube

350 Best Practices for Treating Blind and Visually Impaired Patients in the Emergency Department: A Scoping Review

Kareem Hamadah, Mary Velagapudi, Juliana J. Navarro, Andrew Pirotte, Christopher Obersteadt

Endemic Infections

358 Association Between Sexually Transmitted Infections and the Urine Culture

Johnathan M. Sheele, Carolyn Mead-Harvey, Nicole R. Hodgson

368 Changing Incidence and Characteristics of Photokeratoconjunctivitis During the COVID-19 Pandemic

Yu-Shiuan Lin, Chih-Cheng Lai, Yu-Chang Liu, Shu-Chun Kuo, Shih-Bin Su

374 Operation CoVER Saint Louis (COVID-19 Vaccine in the Emergency Room): Impact of a Vaccination Program in the Emergency Department

Brian T. Wessman, Julianne Yeary, Helen Newland, Randy Jotte

382 Sexually Transmitted Infection Co-testing in a Large Urban Emergency Department

James S. Ford, Joseph C. Morrison, Jenny L. Wagner, Disha Nangia, Stephanie Voong, Cynthia G. Matsumoto, Tasleem Chechi, Nam Tran, Larissa May

International Emergency Medicine

389 Public Beliefs About Accessibility and Quality of Emergency Departments in Germany

Jens Klein, Sarah Koens, Martin Scherer, Annette Strauß, Martin Härter, Olaf von dem Knesebeck

Neurology

399 Support for Thrombolytic Therapy for Acute Stroke Patients on Direct Oral Anticoagulants: Mortality and Bleeding Complications

Paul Koscumb, Luke Murphy, Matthew Talbott, Shiva Nuti, George Golovko, Hashem Shaltoni, Dietrich Jehle

Pediatrics

407 Patient-related Factors Associated with Potentially Unnecessary Transfers for Pediatric Patients with Asthma: A Retrospective Cohort Study

Gregory A. Peters, Rebecca E. Cash, Scott A. Goldberg, Jingya Gao, Lily M. Kolb, Carlos A.Camargo

Public Health

415 Public Health Interventions in the Emergency Department: A Framework for Evaluation

Elisabeth Fassas, Kyle Fischer, Steven Schenkel, John David Gatz, Daniel B.Gingold

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:

Integrating Emergency Care with Population Health

Indexed in MEDLINE, PubMed, and Clarivate Web of Science, Science Citation Index Expanded

Table of Contents continued

Trauma

423 Trauma-informed Care Training in Trauma and Emergency Medicine: A Review of the Existing Curricula

Cecelia Morra, Kevin Nguyen, Rita Sieracki, Ashley Pavlic, Courtney Barry

Women’s Health

431 Relationship of Beta-Human Chorionic Gonadotropin to Ectopic Pregnancy Detection and Size

Duane M Eisaman, Nicole E. Brown, Sarah Geyer

436 Prevalence and Characteristics of Emergency Department Visits by Pregnant People: An Analysis of a National Emergency Department Sample (2010–2020)

Carl Preiksaitis, Monica Saxena, Jiaqi Zhang, Andrea Henkel

Western Journal of Emergency Medicine:

Integrating Emergency Care with Population Health

Indexed in MEDLINE, PubMed, and Clarivate Web of Science, Science Citation Index Expanded

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 American College of Emergency Physicians

Academic Department of Emergency Medicine Subscriber Albany Medical College Albany, NY

Allegheny Health Network Pittsburgh, PA

American University of Beirut Beirut, Lebanon

AMITA Health Resurrection Medical Center Chicago, IL

Arrowhead Regional Medical Center Colton, CA

Baylor College of Medicine Houston, TX

Baystate Medical Center Springfield, MA

Bellevue Hospital Center New York, NY

Beth Israel Deaconess Medical Center Boston, MA

Boston Medical Center Boston, MA

Brigham and Women’s Hospital Boston, MA

Brown University Providence, RI

Carl R. Darnall Army Medical Center Fort Hood, TX

Cleveland Clinic Cleveland, OH

Columbia University Vagelos New York, NY

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

International Society Partners

Conemaugh Memorial Medical Center Johnstown, PA

Crozer-Chester Medical Center Upland, PA

Desert Regional Medical Center Palm Springs, CA

Detroit Medical Center/ Wayne State University Detroit, MI

Eastern Virginia Medical School Norfolk, VA

Einstein Healthcare Network Philadelphia, PA

Eisenhower Medical Center Rancho Mirage, CA

Emory University Atlanta, GA

Franciscan Health Carmel, IN

Geisinger Medical Center Danville, PA

Grand State Medical Center Allendale, MI

Healthpartners Institute/ Regions Hospital Minneapolis, MN

Hennepin County Medical Center Minneapolis, MN

Henry Ford Medical Center Detroit, MI

Henry Ford Wyandotte Hospital Wyandotte, MI

Emergency Medicine Association of Turkey Lebanese Academy of Emergency Medicine Mediterranean Academy of Emergency Medicine

California Chapter Division of American Academy of Emergency Medicine

INTEGRIS Health Oklahoma City, OK

Kaiser Permenante Medical Center San Diego, CA

Kaweah Delta Health Care District Visalia, CA

Kennedy University Hospitals Turnersville, NJ

Kent Hospital Warwick, RI

Kern Medical Bakersfield, CA

Lakeland HealthCare St. Joseph, MI

Lehigh Valley Hospital and Health Network Allentown, PA

Loma Linda University Medical Center Loma Linda, CA

Louisiana State University Health Sciences Center New Orleans, LA

Louisiana State University Shreveport Shereveport, LA

Madigan Army Medical Center Tacoma, WA

Maimonides Medical Center Brooklyn, NY

Maine Medical Center Portland, ME

Massachusetts General Hospital/Brigham and Women’s Hospital/ Harvard Medical Boston, MA

Great Lakes Chapter Division of the American Academy of Emergency Medicine

Tennessee Chapter Division of the American Academy of Emergency Medicine

Norwegian Society for Emergency Medicine Sociedad Argentina de Emergencias

Mayo Clinic Jacksonville, FL

Mayo Clinic College of Medicine Rochester, MN

Mercy Health - Hackley Campus Muskegon, MI

Merit Health Wesley Hattiesburg, MS

Midwestern University Glendale, AZ

Mount Sinai School of Medicine New York, NY

New York University Langone Health New York, NY

North Shore University Hospital Manhasset, NY

Northwestern Medical Group Chicago, IL

NYC Health and Hospitals/ Jacobi New York, NY

Ohio State University Medical Center Columbus, OH

Ohio Valley Medical Center Wheeling, WV

Oregon Health and Science University Portland, OR

Penn State Milton S. Hershey Medical Center Hershey, PA

Uniformed Services Chapter Division of the American Academy of Emergency Medicine

Virginia Chapter Division of the American Academy of Emergency Medicine

for Emergency Medicine

To become a WestJEM departmental sponsor, waive article processing fee, receive electronic copies for all faculty and residents, and free CME and faculty/fellow position advertisement space, please go to http://westjem.com/subscribe or contact:

Stephanie Burmeister

WestJEM Staff Liaison

Phone: 1-800-884-2236

Email: sales@westjem.org

Sociedad Chileno Medicina Urgencia Thai Association

Western Journal of Emergency Medicine:

Integrating Emergency Care with Population Health

Indexed in MEDLINE, PubMed, and Clarivate Web of Science, Science Citation Index Expanded

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 American College of Emergency Physicians

Academic Department of Emergency Medicine Subscriber Prisma Health/ University of South Carolina SOM Greenville Greenville, SC

Regions Hospital Emergency Medicine Residency Program St. Paul, MN

Rhode Island Hospital Providence, RI

Robert Wood Johnson University Hospital New Brunswick, NJ

Rush University Medical Center Chicago, IL

St. Luke’s University Health Network Bethlehem, PA

Spectrum Health Lakeland St. Joseph, MI

Stanford Stanford, CA

SUNY Upstate Medical University Syracuse, NY

Temple University Philadelphia, PA

Texas Tech University Health Sciences Center El Paso, TX

The MetroHealth System/ Case Western Reserve University Cleveland, OH

UMass Chan Medical School Worcester, MA

University at Buffalo Program Buffalo, NY

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

International Society Partners

University of Alabama Medical Center Northport, AL

University of Alabama, Birmingham Birmingham, AL

University of Arizona College of Medicine-Tucson Tucson, AZ

University of California, Davis Medical Center Sacramento, 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 California, San Francisco San Francisco, CA

UCSF Fresno Center Fresno, CA

University of Chicago Chicago, IL

University of Cincinnati Medical Center/ College of Medicine Cincinnati, OH

University of Colorado Denver Denver, CO

University of Florida Gainesville, FL

University of Florida, Jacksonville Jacksonville, FL

Emergency Medicine Association of Turkey Lebanese Academy of Emergency Medicine Mediterranean Academy of Emergency Medicine

California Chapter Division of American Academy of Emergency Medicine

University of Illinois at Chicago Chicago, IL

University of Iowa Iowa City, IA

University of Louisville Louisville, KY

University of Maryland Baltimore, MD

University of Massachusetts Amherst, MA

University of Michigan Ann Arbor, MI

University of Missouri, Columbia Columbia, MO

University of North Dakota School of Medicine and Health Sciences Grand Forks, ND

University of Nebraska Medical Center Omaha, NE

University of Nevada, Las Vegas Las Vegas, NV

University of Southern Alabama Mobile, AL

University of Southern California Los Angeles, CA

University of Tennessee, Memphis Memphis, TN

University of Texas, Houston Houston, TX

University of Washington Seattle, WA

Great Lakes Chapter Division of the American Academy of Emergency Medicine

Tennessee Chapter Division of the American Academy of Emergency Medicine

Norwegian Society for Emergency Medicine Sociedad Argentina de Emergencias

University of WashingtonHarborview Medical Center Seattle, WA

University of Wisconsin Hospitals and Clinics Madison, WI

UT Southwestern Dallas, TX

Valleywise Health Medical Center Phoenix, AZ

Virginia Commonwealth University Medical Center Richmond, VA

Wake Forest University Winston-Salem, NC

Wake Technical Community College Raleigh, NC

Wayne State Detroit, MI

Wright State University Dayton, OH

Yale School of Medicine New Haven, CT

Uniformed Services Chapter Division of the American Academy of Emergency Medicine

Virginia Chapter Division of the American Academy of Emergency Medicine

for Emergency Medicine

To become a WestJEM departmental sponsor, waive article processing fee, receive electronic copies for all faculty and residents, and free CME and faculty/fellow position advertisement space, please go to http://westjem.com/subscribe or contact:

Stephanie Burmeister

WestJEM Staff Liaison

Phone: 1-800-884-2236

Email: sales@westjem.org

Sociedad Chileno Medicina Urgencia Thai Association

AddressingSystemandClinicianBarrierstoEmergency Department-initiatedBuprenorphine:AnEvaluation ofPost-interventionPhysicianOutcomes

JacquelineJ.Mahal,MD,MBA*†

PollyBijur,PhD†

AudreySloma,MPH*

JoannaStarrels,MD,MS‡

TiffanyLu,MD,MS§

*JacobiMedicalCenter,DepartmentofEmergencyMedicine,Bronx,NewYork

† AlbertEinsteinCollegeofMedicine,DepartmentofEmergencyMedicine, Bronx,NewYork

‡ AlbertEinsteinCollegeofMedicine/Monte fi oreMedicalCenter,Departmentof Medicine,Bronx,NewYork

§ JacobiMedicalCenter,DepartmentofPsychiatry&BehavioralSciences, Bronx,NewYork

SectionEditor:MarcMartel,MD

Submissionhistory:SubmittedMay22,2023;RevisionreceivedJanuary3,2023;AcceptedJanuary4,2023

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18320

Introduction: Emergencydepartments(ED)areintheuniquepositiontoinitiatebuprenorphine,an evidence-basedtreatmentforopioidusedisorder(OUD).However,barriersatthesystemandclinician levellimititsuse.WedescribeaseriesofinterventionsthataddressthesebarrierstoED-initiated buprenorphineinoneurbanED.Wecomparepost-interventionphysicianoutcomesbetweenthestudy siteandtwoaffiliatedsiteswithouttheinterventions.

Methods: Thiswasacross-sectionalstudyconductedatthreeaffiliatedurbanEDswherethe interventionsiteimplementedOUD-relatedelectronicnotetemplates,clinicalprotocols,apeer navigationprogram,education,andreminders.Post-intervention,weadministeredananonymous, onlinesurveytophysiciansatallthreesites.Surveydomainsincludeddemographics,buprenorphine experienceandknowledge,comfortwithaddressingOUD,andattitudestowardOUDtreatment. Physicianoutcomeswerecomparedbetweentheinterventionsiteandthecontrolsiteswithbivariate tests.Weusedlogisticregressioncontrollingforsignificantdemographicdifferencestocompare physicians’ buprenorphineexperience.

Results: Of113(51%)eligiblephysicians,58completedthesurvey:27fromtheinterventionsite,and31 fromthecontrolsites.Physiciansattheinterventionsiteweremorelikelytospend <75%oftheirwork weekinclinicalpracticeandtobeinmedicalpracticefor <7years.Buprenorphineknowledge(including statusofbuprenorphineprescribingwaiver),comfortwithaddressingOUD,andattitudestowardOUD treatmentdidnotdiffersignificantlybetweenthesites.Physicianswere4.5timesmorelikelytohave administeredbuprenorphineattheinterventionsite(oddsratio[OR]4.5,95%confidenceinterval 1.4–14.4, P = 0.01),whichremainedsignificantafteradjustingforclinicaltimeandyearsinpractice, (OR3.5and4.6,respectively).

Conclusion: Physiciansexposedtointerventionsaddressingsystem-andclinician-level implementationbarrierswereatleastthreetimesaslikelytohaveadministeredbuprenorphineintheED. Physicians’ buprenorphineknowledge,comfortwithaddressingandattitudestowardOUDtreatmentdid notdiffersignificantlybetweensites.Our findingssuggestthatED-initiatedbuprenorphinecanbe facilitatedbyaddressingimplementationbarriers,whilephysicianknowledge,comfort,andattitudesmay behardertoimprove.[WestJEmergMed.2024;25(3)303–311.]

INTRODUCTION

Opioid-relatedoverdosedeathsintheUShaveincreased sincethe1990s,andinthe12monthsendingJune2023 provisionaloverdosedeathsexceeded81,000.1 The emergencydepartment(ED)hasbeeninvolvedinaddressing theopioidcrisisbyimplementingopioid-sparingpain managementprotocolsandtreatingopioidoverdoses.Yet patientswithnon-fatalunintentionalopioidoverdosevisits totheEDarestill100timesmorelikelytodieofanoverdose withinayearoftheirindexvisitthanthosefroma demographicallymatchedpopulation.2 Emergency departmentsareintheuniquepositiontoinitiateandlinkto evidence-basedtreatmentforopioidusedisorder(OUD) whenapatientpresentsacutelywithopioidwithdrawalor non-fataloverdose.

Buprenorphine,apartialopioidagonist,isaneffective medicationtotreatOUDthathashistoricallynotbeen offeredinEDsettings.In2015,D’Onofrioetalpublisheda seminal,randomizedcontrolledstudydemonstratingthe efficacyofED-initiatedbuprenorphineandongoing engagementinOUDtreatmentat30-dayspostdischarge.3 Follow-upstudiesalsodemonstratedthatED-initiated buprenorphineisaneffectiveintervention,withongoing OUDtreatmentat30daysin50–86%ofthepatients.4,5 On theheelsofthese findings,theSubstanceAbuseandMental HealthServicesAdministrationpublishedaresourceguidein 2021acknowledgingtheEDasanimportantsitefor provisionofOUDtreatment.6 Inthesameyear,the AmericanCollegeofEmergencyPhysicianspublished consensusrecommendationsforOUDtreatmentincluding useofbuprenorphineintheED.7

WhilebuprenorphineuseintheEDhasincreasedinrecent years, 8 multiplebarriersatthesystemandclinicianlevellimit theimplementationofED-initiatedbuprenorphine.9–13 System-levelbarriersincludelackofstreamlinedordersets forOUDtreatment,difficultyreferringtoongoingtreatment servicesafterdischarge,limitedavailabilityofexpert physiciansandpharmacistsforconsultation,andlackof accesstodedicatedcarecoordinators,socialworkers,orpeer counselors.Clinician-levelbarriersincludelackof knowledge,comfortandexperiencewithbuprenorphineand OUDtreatment,ahistoricalneedforabuprenorphine prescribingwaiver,14 aswellasstigmatowardpatients withOUD.

Fewstudieshaveexaminedspecificinterventionsthat addressclinician-levelbarriersandpost-intervention clinicianoutcomes.Fosteretaldescribeda financial incentiveprogramforemergencyphysicianstocompletethe then-requiredbuprenorphinewaivertrainingandreporteda positivebutvariableincreaseinbuprenorphineprescribingin the fivemonthsaftertheincentive.14 Butleretalreportedona setofbehavioral-scienceinformedinterventionsthat increasedphysicianinitiationofOUD-relatedtreatments15

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

WhileED-initiatedbuprenorphineforthe treatmentofopioidusedisorderhas increased,system-andclinician-levelbarriers continuetolimititsuse.

Whatwastheresearchquestion?

Wesoughttocomparewhetherinterventions addressingbarrierstoED-initiated buprenorphinewouldimproveadministration ofbuprenorphine.

Whatwasthemajor findingofthestudy?

Physiciansattheinterventionsitewere 4.5timesmorelikelytohaveadministered buprenorphine(95%CI1.4 – 14.4,P = 0.01).

Howdoesthisimprovepopulationhealth?

ED-initiatedbuprenorphinecanbefacilitatedby addressingbothsystem-andclinician-level barriers,althoughphysicianknowledge, comfort,andattitudesmaybehardertoimprove.

atasingleacademicEDsitewitharobustaddictionclinic program.Khatrietalrandomizedphysicianstoaclinicianlevelinterventionofeitheradidactic-onlygrouporadidactic plusweeklymessaginganda financialincentivegroup.16 While33%ofallparticipantsprescribedbuprenorphinefor the firsttimeinthe90dayspost-intervention,buprenorphine administrationfrequencyorknowledgedidnotdiffer significantlybetweenthegroups.Inanongoing,multicenter effectivenessstudyofbuprenorphineinitiationintheED, D’Onofrioetaldescribedmultiplesystem-level implementationfacilitatorsthatincludeclinicalprotocols, learningcollaboratives,andreferralprograms.17,18 The implementationfacilitationperiodwasassociatedwitha highernumberofemergencyclinicianswhocompletedthe buprenorphineprescribingwaiver,aswellasEDvisitswhere cliniciansprescribedbuprenorphineandnaloxone.19

Wecontributetothegrowingbodyofliteratureby describingasetofinterventionsthataddressedmultiple system-andclinician-levelimplementationbarrierstoEDinitiatedbuprenorphineinasafety-netED.Weevaluated post-interventionphysicianoutcomesandcomparedthese betweentheEDsitewithtargetedinterventionsandtwo relatedsiteswithouttargetedinterventions.Ouraimwasto determinewhetheraddressingmultipleimplementation

barrierstoED-initiatedbuprenorphineisassociatedwith improvedbuprenorphineknowledge,comfortwith addressingOUD,andattitudestowardOUDtreatment amongphysiciansattheinterventionsite.Wehypothesize thatphysiciansattheinterventionsitehadimproved experiencewithadministeringbuprenorphineintheED.

METHODS

StudyDesign

Weconductedacross-sectionalstudyofattending physiciansatthreeEDsaffiliatedwithalarge,urban emergencymedicine(EM)residencyprogram.Physician knowledge,comfortwith,andattitudestowardOUD treatment,aswellasexperiencewithadministering buprenorphineintheED,werecomparedbetweenone interventionsite(whereamultifacetedsetofinterventions aimedataddressingclinician-andsystem-levelbarriersto ED-initiatedbuprenorphinewasimplemented)vstwo controlsites(whereinterventionsfocusedonED-initiated buprenorphinewerenotimplemented).Thestudywas approvedbytheaffiliatedinstitutionalreviewboards (IRB#2019-10920).

Setting

ThisstudytookplaceatthreeEDsaffiliatedwithalarge academicEMtrainingprogramwith84residentsperyear and100full-timeattendingphysiciansonfaculty.OneED siteispartoftheNewYorkCitymunicipalhospitalsystem, whiletheothertwoEDsitesarepartofalarge,private, academichealthsystem.AllthreeEDsseeahighvisitvolume around70,000perannumpersiteandprovidesafety-netcare toapayormixthatispredominantlypubliclyinsured.All threeEDsareintheboroughofTheBronx,NewYork, wheretheopioid-relatedoverdoseratewas73.6per100,000 in2022,representingthehighestofall fiveboroughsinNew YorkCity.20 ConsistentwithmostEMpracticesacrossthe country,thethreeEDtrainingsiteshavenothistorically offeredbuprenorphineforopioidwithdrawaland OUDtreatment.

InterventionSite

BetweenNovember2018–June2020,themunicipal hospital-basedEDsite(hereinreferredtoas “intervention site”)implementedamultifacetedsetofinterventionsto addresssystem-andclinician-levelbarrierstoED-initiated buprenorphine.System-levelinterventionscustomizedforthe EDincludedthefollowing:1)anelectronichealthrecord (EHR)notetemplateforopioidwithdrawalandOUD assessment;2)aclinicalprotocolforadministering buprenorphineintheED;3)aclinicalworkflowtoprovide naloxonetrainingandtake-homekitsforoverdose prevention;and4)apeernavigationprogramtofacilitate referralandlinkagetooutpatientbuprenorphinetreatment, includinganin-housesubstanceusedisordertreatment

program.System-levelinterventionswerefundedand developedbyacentralizedleadershipteamfromthemunicipal publichospitalsystem.LocalEDimplementationwas facilitatedbyaclinicianchampion(JM)whoworkedclosely withaninterdisciplinaryteamofemergencymedicine, behavioralhealth,pharmacy,andsocialworkleadership. Initialsalarysupportforthisworkwasgrant-funded.

Clinician-levelinterventionsincludedthefollowing:1)a modest financialincentiveforvoluntarycompletionof buprenorphinewaivertrainingandobtainingtheprescribing waiver;2)regularupdatesandremindersaboutsystem-level interventionsatEMfacultymeetingseverytwoweeks;and 3)two,one-hourgrandroundslecturesthatreviewedthe evidenceforED-initiatedbuprenorphineandtheavailability ofclinicalprotocolstosupportbuprenorphinetreatment. Grandroundslecturesatthetimeofinterventionwere conductedinpersonandvoluntarilyattendedbyfacultyand residentsacrosstheEMresidencyprogram.Manyofthe interventionswereintroducedinanoverlappingmannerand refinediterativelyduringthetwo-yearimplementationperiod.

ControlSite

Duringthesameperiod,aclinicalprotocolandanorder settosupporthospital-initiatedbuprenorphinewerealso beingimplementedatthetwootherEDsitesbasedatthe private,academichealthsystem(referredtoas “thecontrol site”);however,theseinterventionsdidnotfocusontheED. PeernavigatorsbasedintheEDwereavailablebutwerenot dedicatedtosupportreferralandlinkagetooutpatient buprenorphinetreatment.Neitherwere financialincentives forcompletionofbuprenorphinewaivertrainingor physicianmeetingsdedicatedtoED-initiated buprenorphineoffered.

Participants

Werecruitedstudyparticipantsbasedonthefollowing criteria:1)licensedphysicianeligibletoobtainawaiverto prescribebuprenorphine;and2)attendingphysicians practicingateithertheinterventionorcontrolsite.Wedid notincluderesidentphysiciansinoursamplebecausethey rotateatboththeinterventionandcontrolsitesandwould haveexperiencedvariableexposuretotheinterventions aimedatED-initiatedbuprenorphine.Neitherdidweinclude physicianassistantswhoareanimportantpartoftheEM workforcebecausetheydidnotreceivethe financialincentive anddidnotattendfacultymeetingsorgrandroundswhere mostoftheclinician-facinginterventionsoccurred.

DataCollection

BetweenSeptember–December2020,weemailed113 eligibleemergencyphysiciansatthethreeEDsitesto introducetheopt-instudyandcontinuedtosendmonthly emailreminders.Wealsoannouncedthestudyinpersonat attendingphysicianmeetingsattwoofthethreesitesthat

couldallocatemeetingtimeduringtheCOVID-19public healthemergency.Individualizedemailremindersweresent toattendingphysiciansatallsitesinthelastmonthofstudy recruitment.Thesurveywasadministeredanonymouslyin EnglishusingtheonlineplatformQualtrics(Qualtrics, Provo,UT).Uponcompletionofthequestionnaire, participantswereeligibletoenteraraffletowinoneof five $50giftcards.

A22-itemsurveywasadaptedfrompreviouslypublished researchonclinicianbarrierstobuprenorphine prescribing.9,11 Thesurveyinstrumentwedeveloped consistedof fivedomains:demographics;buprenorphine experience;buprenorphineknowledge;comfortwith addressingOUD;andattitudestowardOUDtreatment.

Self-reporteddemographicsincludedage,gender,race, ethnicity,yearsinpractice,andamountoftimespent workingclinically(clinicaltime).Thenumberofyearsin practicewasmeasuredbythenumberofyearssince AmericanBoardofEmergencyMedicinecertificationdate, andrespondentswereconsideredjuniorattendingphysicians iftheyhadsevenorfeweryearsinpractice.Clinicaltimewas adichotomousmeasureoflessthanvs ≥ 75%,basedonthe rationalethatattendingphysicianswhospend <75%clinical timerepresentclinician-educators,researchers, oradministrators.

Forbuprenorphineexperience,participantswereaskedto answeryes/notoeveradministeringbuprenorphineinthe ED,completingthebuprenorphinewaivertraining,and receivingtheirbuprenorphineprescribingwaiver. Buprenorphineknowledgewasevaluatedwithseven questionsspecifictotheclinicaluseofbuprenorphineusinga three-pointLikertscale(“agree-neutral-disagree”),where agreeingordisagreeingcorrectlytotheknowledgequestions wasakeyoutcome.ComfortwithOUDtreatmentwasalso evaluatedwithathree-pointLikertscale(“comfortablesomewhatcomfortable-notcomfortable”)regarding managementofopioidwithdrawal,responsetoopioid overdose,counselingonandadministeringmedicationsfor OUD,andreferraltooutpatienttreatmentforsubstanceuse disorder.AttitudestowardOUDtreatmentweremeasured withlevelofagreement(“agree-neutral-disagree”)to stigmatizingstatementsdescribingpatientswithOUDas difficulttotreat,buprenorphineassubstitutingonedrugfor another,andprescribingbuprenorphineforOUDas increasingmedicolegalrisk.

DataAnalysis

Wecalculateddescriptivestatisticsfordemographic characteristics,buprenorphineexperience,and buprenorphineknowledgeforphysiciansattheintervention andcontrolsites.Fisherexacttestswereusedtoassess whetherphysicians’ demographiccharacteristicsand buprenorphineexperiencedifferedbysite.Weexamined buprenorphineknowledgebycalculatingacomposite

knowledgescorebasedonthenumberofcorrectanswersto thesevenknowledgequestionsandcomparedthembysite usingtheMann-WhitneyU-test.Physicians’ comfortwith addressingOUDandattitudestowardOUDtreatmentare describedwithproportionofresponseswith “comfortable” and “ agree, ” andcomparedbysitewithFisherexacttestsand Fisher-Freeman-Haldentests,forvariableswithmorethan twocategories.

Weconductedapost-hocmultivariableanalysisbecause ofastatisticallysignificantdifferencebetweenphysicians’ buprenorphineexperienceof “everadministered buprenorphine” bysite.Weexaminedpossibleconfounding ofthisassociationbythedemographiccharacteristicsthat aresignificantlyassociatedwiththesite.Weusedlogistic regressiontoassesstheassociationbetweenbuprenorphine administrationandsitewhilecontrollingforthesecovariates. Followingtherecommendationthatonevariableshouldbe usedforevery10participantswiththeoutcome,we ascertainedthatonlytwovariablescouldbeincludedina singleanalysisastherewere20participantswhohad “ ever administeredbuprenorphine.” Thus,werananalyseswith siteandeachofthepossibleconfoundersseparately.Alltests weretwo-sidedwithastatisticalsignificancecriterionof0.05. WeusedSPSSversion27(IBMCorp,Armonk,NY)forall statisticalanalyses.

RESULTS

Amongthe113eligibleattendingphysicians,58(51.3%) physiciansfullycompletedthesurvey,with27responses fromtheinterventionsiteand31responsesfromthecontrol site.Asshownin Table1,nosignificantdifferencesinthe demographiccharacteristicsofgender,race,andethnicity werefoundamongemergencyphysiciansbysite.Physicians weremorelikelytospend <75%oftimeinclinicalpracticeat theinterventionvscontrolsites,44.4%vs19.4%,respectively (P = 0.05).Nearlytwiceasmanyphysiciansatthe interventionsitewereinclinicalpracticeforsevenyearsor lesscomparedtothoseatthecontrolsite,70.4%vs38.7%, respectively(P = 0.02).Inotherwords,physiciansatthe interventionsiteweremorelikelytobeclinician-educators, researchersandadministrators,andjunior attendingphysicians.

Forbuprenorphineexperience,morephysiciansatthe interventionsitereported “everadministered buprenorphine” intheirclinicalpracticethanphysiciansat thecontrolsite,51.9%vs19.4%,respectively(P = 0.01).Over halfofthephysicianrespondentscompletedthewaiver trainingatboththeinterventionandcontrolsites,55.6%and 51.6%,respectively.Ofthosewhocompletedthewaiver training,mostphysiciansobtainedtheprescribingwaiver. Therewasnostatisticaldifferenceinwaivertraining completionandstatusbysite.Forbuprenorphine knowledge,themedianscoreofcorrectanswers(oftheseven knowledgequestions)wasthreeforphysiciansatthe

Table1. Demographic,experience,andknowledgeparticipantcharacteristics.

Demographiccharacteristics

Gender

Interventionsite N = 27n(%) Controlsites N = 31n(%) P-value

White 17(63.0%)22(75.9%)

Black 3(11.1%)3(10.3%)

Asian 2(3.7%)3(10.3%)

Other 2(7.4%)1(3.4%) Declinetoanswer4(14.8%)1(3.4%)

Obtainedprescribingwaiveramongthosewhocompletedwaivertraining0.74

Yes 12(80.0)12(75.0)

No 3(20.0)4(25.0)

Buprenorphineknowledge

Mean(SD)numberofcorrectresponses(7items)3.4(2.0)3.3(2.1)0.82 Median(range)3(0–7)4(06)0.88

*Statisticalsignificancewithp-valueforcomparison(p < .05).

interventionsite,whichwassimilartothemedianscoreof fouratthecontrolsite(Mann-WhitneyU = 428, P = 0.88). Asseenin Figures1 and 2,physicians’ comfortwith addressingOUDandtheirattitudestowardOUDtreatment didnotdiffersignificantlybetweentheinterventionand controlsites.

Thepost-hocanalysis(see Table2)oftheassociation betweenbuprenorphineadministrationandsiteindicates thatphysiciansattheinterventionsitewere4.5timesmore

likelytohaveadministeredbuprenorphinethanthoseatthe controlsite(OR4.5,95%CI1.4 – 14.4, P = 0.01).After adjustingforthetwodemographiccharacteristicsthat differedbysite(clinicaltimeandyearsinpractice),the likelihoodofbuprenorphineadministrationremainedhigh andstatisticallysignificantamongphysiciansatthe interventionsitecomparedtothecontrolsite(OR3.5 withclinicaltimecontrolled,4.6withyearsinpractice controlled,respectively).

Trea ng opioid withdrawal with buprenorphine will extend pa ent length of stay in the Emergency Department

Providing pa ents with buprenorphine for OUD will increase my medicolegal risk

Buprenorphine is subs tu ng one drug for another

Pa ents who have a history of OUD are difficult to treat

The Clinical Opiate Withdrawal Scale (COWS) is difficult to administer

PercentAgree(%)

Figure1. Physicianattitudestowardspatientslivingwithopioidusedisorder(OUD)1 anduseofbuprenorphinebysite(percentagree.)

Physicianagreementwiththestatementsalongtheverticalaxisbysite.Nostatisticaldifferencefound.

DISCUSSION

Inthisstudy,wefoundthatemergencyphysicianswho wereexposedtoamultifacetedsetofinterventionsthat addressedsystem-andclinician-levelbarrierstoED-initiated

Refertooutpa entsubstanceuse treatmentprograms

Counselpa entsaboutnaloxone take-homekits

Administermethadone

Administerbuprenorphine

Discussmedica onsforOUD

Respondtoopioidoverdose

TreatOpioidwithdrawal

buprenorphineattheirclinicalsitewereatleastthreetimesas likelytohaveadministeredbuprenorphineafteradjustingfor clinicaltimeandyearsinpractice.Yetphysicians’ buprenorphineknowledge,comfortwithaddressingOUD,

ControlSite Interven onSite

Percent(%)comfort

Figure2. Physicianpercentcomfortwithaddressingopioidusedisorderbysite. Physiciancomfortwiththeactivitieslistedalongtheverticalaxisbysite.*Statisticalsignificancewith P-valueforcomparison(P < .05).

*Statisticallysignificant P-value < 0.05. Ref,referencegroup; OR,oddsratio.

andattitudestowardOUDtreatmentdidnotdiffer significantlybetweentheinterventionandcontrolsites.Our findingssuggestthatED-initiatedbuprenorphinecanbe facilitatedbyaddressingsystem-levelimplementation barriers,whileclinicianknowledge,comfort,andattitudes maybehardertoimproveandmayrequirelong-termand/or differentinterventions.

Thesystem-levelinterventionsdescribedwereaseriesof toolsandservicesintroducedtotheEDbyinterdisciplinary stakeholderstoencouragetheuseofevidence-based,EDinitiatedbuprenorphinethathadnotpreviouslybeen consideredstandardtreatmentforpatientslivingwithOUD. IntegratedEHRtemplatesandclinicalprotocolsand workflowsweretoolstosupportclinicaldecision-making, whilethepeernavigationprogramprovidedharmreduction interventionsandsupportedpost-dischargeplanningand linkagetocare.Theimplementationofthesesystem-level interventionswasintendedtominimizetheburdenon cliniciansandtoreducevariationincare.21 Theimpactofeach interventionwasnotmeasuredindividuallybecausemany componentswereintroducedandrefinedinanoverlapping, iterativemannerduringtheimplementationperiod.(For example,announcementsandeducationregardingtheEHR ordersetsandclinicalprotocolsoccurredatasimilartimeand acrosssubsequentmeetings.)Thecross-sectionalstudy capturedonlyclinicianoutcomesafterreceivingthewholeset ofsystem-levelinterventions,whichisalimitationof measuringreal-worldimplementationfacilitation. Implementationofthesesystem-leveltoolsandservices requiredinterventionsattheclinicianleveltointroduce, familiarize,andremindcliniciansofavailabletoolsand supportservices.Frequentreminders,educational opportunities, financialincentivesforthethen-required buprenorphineprescribingwaivercourseworkwerean attempttoencourageknowledgeofandcomfortwithEDinitiatedbuprenorphinewiththegoaltosupportachangein clinicalpracticeto treat OUD,notjustrespondtoacute

overdoses,intheED.Ourclinician-levelinterventionseased theimplementationofsystem-levelinterventionsinasimilar manner,asthebehavioralscience-based “nudges” wereused toincreasethenumberofphysicianswhoobtainedawaiver atanotherurban,academicED.22 Thesamegroupalsoused clinician-levelnudgesintheformofbestpracticeadvisories intheEHRandmonthlyemailstoincreasetheuseof ambulatoryreferralstoaBridgeClinicandbuprenorphine administration.15 Animportantpartoftheprocessappears toincludeaclinicalchampionwhocanworkwith stakeholderstoovercomeinstitutionalbarriers18,19,23 to refineworkflowsandprotocols,andwhocanalsobea contentexpertresourcetocolleaguestointroduceevidencebasedpracticeupdatesandreminders.

Inourstudy,physicians’ clinicaltimeandyearsinpractice hadanimpactonthelikelihoodofpracticingED-initiated buprenorphine.Clinicaltimeinpracticeisavariableusedto differentiatebetweenphysicianswithorwithoutdedicated timeforclinicaleducation,research,andadministration, whichwashypothesizedtohaveanindependenteffecton adoptionofemergingclinicalpractices.Yearsinindependent clinicalpracticeisusedasameasuretoaccountforsecular trendsinEMtraining;attendingphysicianswithfewerthan sevenyearsinclinicalpracticemayhavebeenexposedto frequentpressontheopioidepidemicandchanging guidelinesforOUDtreatmentintheED.Imetalreportthat junioremergencyphysiciansaremorelikelytoviewOUDas achronicdiseaseandapproveofbuprenorphineinitiationin theED,24 evenifjunioremergencyphysiciansexpresseda similarsenseoffrustrationtreatingpatientswithOUDas seniorphysicians.Ourstudydidnotincluderesident physicianstominimizecross-contaminationofexposureto interventions.Otherstudieshavefoundthatemergency physiciansintheirresidencytrainingareeagertoimplement ED-initiatedbuprenorphine.15,22 Attitudesamong emergencyphysiciansaregenerallychangingtowardOUD, anditisincreasinglybeingviewedasachronicdiseasewith

Table2. Predictorsofbuprenorphineadministrationbyphysiciancharacteristics.

acutemanifestationsthatshouldbetreatedinthe EDsetting.24,25

Theremovalofthebuprenorphine-prescribingwaiver requirementisanacknowledgmentthatthisclinician-level barrierimpededaccesstotreatmentforOUD.26,27 Whilethis studywascompletedatatimewhenthebuprenorphineprescribingwaiverrequirementwasstillineffect(and justified financialincentivesforemergencyclinicianswho voluntarilyobtainedaprescribingwaiver),weexpectthat futureinterventionstoaddressclinician-levelbarriersto buprenorphineinitiationintheEDwillstillrequireaclinical championwhocanregularlyprovideupdatesabout implementationandleadeducationefforts.

LIMITATIONS

Limitationstoourstudyincludearelativelysmallsample sizewitha58%responserate,whichmayhavecontributedto asamplingbias.Ourclinicalsitesareinanurbanareawitha highprevalenceofopioidoverdoseandOUD,whichmay influencephysicianinterestinandknowledgeofOUDand, thus,participationinthesurvey.ImplementationofEDinitiatedbuprenorphineattheinterventionsitereceived financialsupportanddepartmentalresourcesinatertiarycaremunicipalhospitalaswellasinitialgrantfundingfor salarysupportoftheclinicalchampion,whichmaylimit generalizabilitytoEDsettingsinsmaller,ruraland/orunderresourcedhospitals.Withoutpre-/post-evaluationsforeach intervention,wewereunabletoassesswhetheraparticular interventioninfluencedthedifferenceinbuprenorphine administrationattheinterventionsite.Buprenorphine experienceisself-reported;responsesregarding buprenorphineadministrationintheEDarenotlinkedto pharmacydatafromtheinterventionalorcontrolsites. Lastly,cross-contaminationofattendingphysicians’ exposurestointerventionsmayhaveoccurredviaresidents whorotateamongtheinterventionandcontrolsites.Itmay havealsooccurredatthegrandroundslectureswhereall facultyfromtheresidencysitesareinvited;however,total facultyattendancetypicallyhoveredbelow10%forthethen in-personlectures.

CONCLUSION

OurstudycomparestheadministrationofED-initiated buprenorphineattwosimilarandrelatedEDsettingswhere physiciansatonesitewereexposedtoamultifacetedsetof interventionstoED-initiatedbuprenorphine.Physicians exposedtointerventionsdesignedtoaddresssystem-and clinician-levelbarriersweremorelikelytoinitiate buprenorphineforOUDtreatmentintheirclinicalpractice. Futureimplementationeffortsshouldexamineinterventions thataretailoredtoimplementationbarriersevenafterthe buprenorphine-prescribingwaiverrequirementhasbeen eliminated,includingresidencyeducationtoimprovethe

understandinganduptakeofED-initiatedbuprenorphine. Couplingpharmacy-levelbuprenorphineadministrationand prescribingdatawithphysician-reportedoutcomeswillalso helpparseoutimpactoffutureinterventions.

AddressforCorrespondence:JacquelineMahal,MD,MBA,Jacobi MedicalCenter,DepartmentofEmergencyMedicine,1400Pelham ParkwaySouth,Bronx,NY10461.Email: mahalj@nychhc.org

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Mahaletal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.AhmadFB,CisewskiJA,RossenLM,etal.Provisionaldrugoverdose deathcounts.NationalCenterforHealthStatistics.2023.Availableat: https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm AccessedNovember20,2023.

2.Goldman-MellorS,OlfsonM,Lidon-MoyanoC,etal.Mortalityfollowing nonfatalopioidandsedative/hypnoticdrugoverdose. AmJPrevMed. 2020;59(1):59–67.

3.D’OnofrioG,O’ConnorPG,PantalonMV,etal.Emergencydepartmentinitiatedbuprenorphine/naloxonetreatmentforopioiddependence:A randomizedclinicaltrial. JAMA. 2015;313(16):1636–44.

4.KaucherKA,CarusoEH,SungarG,etal.Evaluationofan emergencydepartmentbuprenorphineinductionandmedicationassistedtreatmentreferralprogram. AmJEmergMed. 2020;38(2):300–4.

5.EdwardsFJ,WicelinskiR,GallagherN,etal.Treatingopioidwithdrawal withbuprenorphineinacommunityhospitalemergencydepartment:an outreachprogram. AnnEmergMed. 2020;75(1):49–56.

6.SubstanceAbuseandMentalHealthServicesAdministration (SAMHSA).UseofMedication-AssistedTreatmentinEmergency Departments.2021.Availableat: https://store.samhsa.gov/sites/ default/files/pep21-pl-guide-5.pdf.AccessedAugust16,2023.

7.HawkK,HoppeJ,KetchamE,etal.Consensusrecommendationson thetreatmentofopioidusedisorderintheemergencydepartment. Ann EmergMed. 2021;78(3):434–42.

8.RheeTG,D’OnofrioG,FiellinDA.Trendsintheuseofbuprenorphinein USemergencydepartments,2002–2017. JAMANetwOpen. 2020;3(10):e2021209.

9.HawkKF,D’OnofrioG,ChawarskiMC,etal.Barriersandfacilitatorsto clinicianreadinesstoprovideemergencydepartment-initiated buprenorphine. JAMANetwOpen. 2020;3(5):e204561.

10.MendiolaCK,GalettoG,FingerhoodM.Anexplorationofemergency physicians’ attitudestowardpatientswithsubstanceusedisorder, J AddictMed. March/April2018;12(2):132–5.

11.LowensteinM,KilaruA,PerroneJ,etal.Barriersandfacilitatorsfor emergencydepartmentinitiationofbuprenorphine:aphysiciansurvey. AmJEmergMed. 2019;37(9):1787–90.

12.SavageTandRossM.BarriersandattitudesreportedbyCanadian emergencyphysiciansregardingtheinitiationofbuprenorphine/ naloxoneintheemergencydepartmentforpatientswithopioiduse disorder. CJEM. 2022;24(1):44–9.

13.LoganG,MirajkarA,HouckJ,etal.Physician-perceivedbarriersto treatingopioidusedisorderintheemergencydepartment. Cureus. 2021;13(11):e19923.

14.FosterSD,LeeK,EdwardsC,etal.Providingincentiveforemergency physicianX-waivertraining:anevaluationofprogramsuccessand postinterventionbuprenorphineprescribing. AnnEmergMed. 2020;76(2):206–214.

15.ButlerK,ChavezT,WakemanS,etal.Nudgingemergencydepartmentinitiatedaddictiontreatment. JAddictMed. 2022;16(4):e234–9.

16.KhatriU,LeeK,LinT,etal.Abriefeducationalinterventiontoincrease EDinitiationofbuprenorphineforopioidusedisorder(OUD). JMed Toxicol. 2022;18(3):205–13.

17.D’OnofrioG,EdelmanEJ,HawkKF,etal.Implementationfacilitationto promoteemergencydepartment-initiatedbuprenorphineforopioiduse disorder:protocolforahybridtypeIIIeffectiveness-implementation study(ProjectEDHEALTH). ImplementSci. 2019;14(1):48.

18.WhitesideLK,D’OnofrioG,FiellinDA,etal.Modelsforimplementing emergencydepartment-initiatedbuprenorphinewithreferralforongoing medicationtreatmentatemergencydepartmentdischargeindiverse academiccenters. AnnEmergMed. 2022;80(5):410–9.

19.D’OnofrioG,EdelmanEJ,HawkKF,etal.Implementationfacilitationto promoteemergencydepartment–initiatedbuprenorphineforopioiduse disorder. JAMANetwOpen.2023;6(4):e235439.

20.TuazonE,BaumanM,SunT,etal.UnintentionalDrugPoisoning (Overdose)DeathsinNewYorkCityin2022.NewYorkCityDepartment

ofHealthandMentalHygiene:EpiDataBrief(137);September2023. Availableat: https://www.nyc.gov/assets/doh/downloads/pdf/epi/ databrief137.pdf.AccessedFebruary13,2024.

21.CentersforMedicareandMedicaidServices.QualityMeasurementand QualityImprovement.2023.Availableat: https://www.cms.gov/ Medicare/Quality-Initiatives-Patient-Assessment-Instruments/MMS/ Quality-Measure-and-Quality-ImprovementAccessedNovember23,2023.

22.MartinA,KunzlerN,NakagawaJ,etal.Getwaivered:aresident-driven campaigntoaddresstheopioidoverdosecrisis. AnnEmergMed. 2019;74(5):691–6.

23.WoodK,GiannopoulosV,LouieE,etal.Theroleofclinicalchampionsin facilitatingtheuseofevidence-basedpracticeindrugandalcoholand mentalhealthsettings:asystematicreview. ImplementResandPract. 2020;1:2633489520959072.

24.ImDD,CharyA,CondellaAL,etal.Emergencydepartmentclinicians’ attitudestowardopioidusedisorderandemergencydepartmentinitiatedbuprenorphinetreatment:amixed-methodsstudy. WestJ EmergMed. 2020;21(2):261–71.

25.SokolR,TammaroE,KimJY,etal.LinkingMATTERS:barriersand facilitatorstoimplementingemergencydepartment-initiated buprenorphine-naloxoneinpatientswithopioidusedisorder andlinkagetolong-termcare. SubstUseMisuse. 2021;56(7):1045–53.

26.OfficeofNationalDrugControlPolicy(ONDCP).Dr.Guptaapplauds removalofX-waiverinomnibus,urgeshealthcareproviderstotreat addiction.2022.Availableat: https://www.whitehouse.gov/ondcp/ briefing-room/2022/12/30/dr-gupta-applauds-removal-of-xwaiver-in-omnibus-urges-healthcare-providers-to-treat-addiction AccessedFebruary23,2023.

27.USDepartmentofJustice,DrugEnforcementAdministration,Diversion ControlDivision.EliminationoftheDATAwaiverprogram.Availableat: https://www.deadiversion.usdoj.gov/pubs/docs/dear_registrant_ AM_01122023.html.AccessedNovember20,2023.

ORIGINAL RESEARCH

FactorsAssociatedwithAcuteTelementalHealth ConsultationsinOlderVeterans

EricaC.Koch,MD*†**

MichaelJ.Ward,MD,PhD,MBA*†§

AlvinD.Jeffery,PhD‡§∥

ThomasJ.Reese,PharmD§

ChadDorn,PSM§ ShannonPugh,RN* MelissaRubenstein,MPH*

JoEllenWilson,MD,MPH†¶ CoreyCampbell,DO#

JinH.Han,MD,MSc*†

*VanderbiltUniversityMedicalCenter,DepartmentofEmergencyMedicine, Nashville,Tennessee

† TennesseeValleyHealthcareSystem,GeriatricResearch,Education,andClinical Center,Nashville,Tennessee

‡ VanderbiltUniversitySchoolofNursing,Nashville,Tennessee

§ VanderbiltUniversityMedicalCenter,DepartmentofBiomedicalInformatics, Nashville,Tennessee

∥ TennesseeValleyHealthcareSystem,NursingServices,Nashville,Tennessee

¶ VanderbiltUniversityMedicalCenter,DepartmentofPsychiatry, Nashville,Tennessee

# TennesseeValleyHealthcareSystem,PsychiatricServices,Nashville,Tennessee

**VeteransAffairsQualityScholarsProgram,Nashville,Tennessee

SectionEditor:PhilipsPerera,MD

Submissionhistory:SubmittedMarch23,2023;RevisionreceivedNovember17,2023;AcceptedJanuary10,2024

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.17996

Introduction: TheUnitedStatesVeteransHealthAdministrationisaleaderintheuseoftelementalhealth (TMH)toenhanceaccesstomentalhealthcareamidstanationwideshortageofmentalhealth professionals.TheTennesseeValleyVeteransAffairs(VA)HealthSystempilotedTMHinitsemergency department(ED)andurgentcareclinic(UCC)in2019, withfull24/7availabilitybeginningMarch1,2020. Followingimplementation,preliminarydatademonstratedthatveterans ≥65yearsoldwerelesslikelyto receiveTMHthanyoungerpatients.Wesoughttoexaminefactorsassociatedwitholderveteransreceiving TMHconsultationsinacute,unscheduled,outpatientsettingstoidentifylimitationsinthecurrentprocess.

Methods: ThiswasaretrospectivecohortstudyconductedwithintheTennesseeValleyVAHealth System.Weincludedveterans ≥55yearswhoreceivedamentalhealthconsultationintheEDorUCC fromApril1,2020–September30,2022.Telementalhealthwasadministeredbyamentalhealthclinician (attendingphysician,residentphysician,nursepractitioner,orphysicianassistant)viaiPad,whereasinpersonevaluationswereperformedintheED.Weexaminedtheinfluenceofpatientdemographics,visit timing,chiefcomplaint,andpsychiatrichistoryonTMH,usingmultivariablelogisticregression.

Results: Ofthe254patientsincludedinthisanalysis,177(69.7%)receivedTMH.Veteranswithhighriskchiefcomplaints(suicidalideation,homicidalideation,oragitation)werelesslikelytoreceiveTMH consultation(adjustedoddsratio[AOR]:0.47,95%confidenceinterval[CI]0.24–0.95).Comparedto attendingphysicians,nursepractitionersandphysicianassistantswereassociatedwithincreasedTMH use(AOR4.81,95%CI2.04–11.36),whereasconsultationbyresidentphysicianswasassociatedwith decreasedTMHuse(AOR0.04,95%CI0.00–0.59).TheUCCusedTMHforallbutoneencounter. Patientcharacteristicsincludingtheirvisittiming,gender,additionalmedicalcomplaints,comorbidity burden,andnumberofpsychoactivemedicationsdidnotinfluenceuseofTMH.

Conclusion: High-riskchiefcomplaints,location,andtypeofmentalhealthclinicianmaybekey determinantsoftelementalhealthuseinolderadults.Thismayhelpexpandmentalhealthcareaccessto areaswithashortageofmentalhealthprofessionalsandpreventpotentiallyavoidabletransfersinlowacuitysituations.FurtherstudiesandinterventionsmayoptimizeTMHforolderpatientstoensuresafe, equitablementalhealthcare.[WestJEmergMed.2024;25(3)312–319.]

INTRODUCTION

In2020,52.9millionpeopleintheUnitedStates(US) sufferedfromamentalhealthorsubstanceusedisorder.1,2 Emergencydepartment(ED)visitsandadmissionsfor psychiatricconcernscontinuetoincrease.3–7 Despitethe increaseddemand,thereisawidespreadmentalhealth professionalsshortageintheUS,whichnegativelyaffects accesstotimely,efficientmentalhealthcareforsociety’smost vulnerablepopulations.Anestimated7,632cliniciansare neededtobridgethegapinlow-resourcedareas.8 Approximately66%ofruralorpartiallyruralcountiesare designatedbythefederalgovernmentasmentalhealth professionalshortageareas.8 Patientsintheseareashave beenfoundtohaveworsehealthoutcomes,includingshorter lifeexpectancyandhigherrateofsuicide.9–11 Innovative solutionsareneededtoaddressthesekeygapstoexpand accesstoequitablementalhealthservices,particularlyinthe settingofacutecrises.

Telehealthwas firstdescribedinclinicalpracticeinthelate 1950s.12 Overthepasttwodecades,usehasexpandedina varietyofclinicalsettings.13 TheVeteransHealth Administrationhasadoptedtelehealthacrossavarietyof settings,includingmentalhealthcomplaints.14 By2016, nearlyhalfofEDsintheUSreportedtheuseoftelehealth, with20%usingitformentalhealthpurposes(telemental health[TMH]).15,16 TheuseofTMHinroutineEDclinical practicegrewdramaticallyduringtheCOVID-19 pandemic.5 FormanyEDs,itistheonlyavenueto emergencypsychiatriccare.15

OnMarch1,2020,theTennesseeValleyVeteransAffairs HealthSystemimplementedfull-timeTMHforveteranswho presentedtotheEDformentalhealthcomplaints.Both TMHandin-personconsultationsperformedbyamental healthclinicianwereavailable7daysaweek,24hoursaday, includingholidays,attheEDandduringalloperatinghours attheUCC(daily8 AM – 8 PM).Consultationmodalitywas lefttothechoiceofthementalhealthclinician.In-person cliniciancoveragewasalwaysavailablebyanattending physician,residentphysician,nursepractitioner,or physicianassistantduringfacilityoperatinghours. Capabilitiesdidnotchangedependingontheroleofthe clinician.Amoredetaileddescriptionoftheprogramis providedelsewhere.17 Despitetheimplementationofthis TMHprogram,preliminarydatashowed20%ofmental healthconsultationsstilloccurredinperson.17 Veteranswho receivedin-personmentalhealthevaluationswerenotably oldercomparedtothosereceivingTMH,with31%in-person consultsoccurringinveteransages ≥65vs18%of TMHconsults.17

Olderpatientswithmentalhealthcomplaintsfaceunique challengesintheemergencysetting.Attentiontothese patientsduringtheimplementationofnewprocessesofcare isvitaltoensuretheyreceivehigh-qualitymentalhealth evaluation.Withtheexponentialgrowthprojectedforthe

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue? Thereisawidespreadshortageofmental healthprofessionalsintheUS,which decreasesaccesstotimelyemergency mentalhealthcare.

Whatwastheresearchquestion?

Whatfactorsareassociatedwitholder veteransreceivingacute,unscheduled telementalhealth(TMH)vsinpersonconsults?

Whatwasthemajor findingofthestudy?

High-riskchiefcomplaints(suicidalor homicidalideation,oragitation)were associatedwithdecreasedTMHuse (OR0.39,95%CI0.18 – 0.81).Typeof clinicianandlocationofcarewerealso associatedwithTMHuse.

Howdoesthisimprovepopulationhealth?

TMHrepresentsanopportunitytoexpand accesstomentalhealthcare,therebyreducing potentiallyunnecessarypatienttransfersand shorteningboardingtimes .

olderpopulationintheUS,understandingfactorsassociated withvariabilityofTMHusewillinformfuture implementationandsustainabilityinacutecaresettings.18 In thisstudywesoughttoexaminefactorsassociatedwitholder veteransreceivingTMHconsultationsinacute,unscheduled, outpatientsettingstoidentifypotentialbarriersto widespreaduseofTMHintheED.Encountersinvolving patientsolderthan75,urbanlocation,residentphysicians, andhigheracuitywerehypothesizedtobemorelikelyto occurinperson.

METHODS

StudyDesign,Setting,andPatientPopulation

Thiswasanexploratory,retrospective,cohortstudy conductedattheTennesseeValleyVAHealthSystemED andurgentcareclinic(UCC).20 Describedinmoredetail elsewhere,thisTMHprogramwasinitiallypilotedduring limitedhoursin2019andthenwentlivewith24/7coveragein March2020withtheonsetoftheCOVID-19pandemic.17 PatientswereinitiallyevaluatedbyanEDorUCCclinician (attendingphysician,residentphysician,nursepractitioner, orphysicianassistant)anddeterminedtoneedmentalhealth consultation.Aconsultorderwasthenrequestedthroughthe

electronichealthrecord(EHR)withdirectcommunication betweentheemergencyphysicianandon-callmentalhealth clinician(nursepractitioner,physicianassistant,orattending psychiatrist).Consultmodalitywaslefttothedecisionofthe on-callmentalhealthclinician.TheTMHvisitwasprovided viaAppleiPad(AppleInc,Cupertino,CA)withaudioand visualcapabilities,whereasin-personevaluationswere performedbythesamementalhealthclinicianintheEDor UCC.Bothin-personandTMHconsultationswereavailable 24/7intheEDandduringoperatinghours oftheUCC.

Weincludedveteranswhowere ≥55yearsandreceiveda mentalhealthconsultationintheEDbetweenApril1, 2020–September30,2022.Sincethereisnouniversally acceptedagethatdefines “olderage,” wechose55yearsold asthecut-offtomaximizeoursamplesizewhilemaintaining amedianageof65yearsold,atraditionalcut-point.Nonveteranswithoutservicebenefits,directadmissionswhodid notpresentthroughtheED,andpatientswithamissing modalityofconsultationwereexcluded.Forveteranswith multipleEDmentalhealthencounters,onlythe first consultationencounterwasincluded.Of1,478initialvisits withinthestudyperiod,weselected510chartstoreview;497 hadcompletementalhealthconsultationsinthechart.A substantialproportionofpatientsreceivedTMHduringthe studyperiod.Therefore,2–3TMHconsultationswere includedforeachin-personconsultation.Webalancedthe numberofchartsselectedforeachmonthofthestudyto reducetemporalbias.Wethenexcludedallpatients <55yearsfromthisanalysis.Thisstudywasapprovedbythe localinstitutionalreviewboardasexempt.

DataCollection

Wedesignedthechartreviewmethodologytofollow acceptedguidelines.21 Datawasmanuallyextractedfromthe VAEHRandClinicalDataWarehouse.Thefollowing patientfactorswereincludedinthisanalysis:age;race; gender;maritalstatus;rurality;EDtriagechiefcomplaint; mentalhealthhistory;totalactivenumberofpsychoactive medications;andpresenceofadditionalnon-psychiatric medicalcomplaint(eg,chestpain).Ruralitywasdetermined bytheRural-UrbanCommutingAreaCodesbasedonthe patient’sZIPcode.22 Weconsideredthefollowingsystemlevelfactors:location(EDvsUCC);timingofpresentation (9 AM – 5 PM ornights/weekends)andmentalhealthclinician type(nursepractitioner,physicianassistant,resident physician,orattendingphysician).

Patientdemographics,visitdate,homelessness, psychiatrichistory,andmedicationsweremanually abstractedbyaphysician(ECK)andnurse(SP).Senior authorstrainedabstractorspriortodatacollection.Each reviewerunderwentmentoredtrainingonhowtorevieweach chartwithatrialperiodofmanualdouble-checkingbythe seniorauthortoensurecompetency.Eachchartwas

reviewedbyeitherthephysicianornursereviewerandthen wascarefullydouble-checkedbythesamereviewerfor inaccuracies.Eachchartwasreviewedbyoneperson.Data abstractionformswereused,andthedatawascompiled usingREDCapelectronicdatacapturetoolshostedatUS DepartmentofVeteransAffairs.

Weusedthetotalnumberofpsychiatricconditions documentedintheEHRpriortotheindexEDvisitto determinepsychiatriccomorbidityburden.Anymentionof suicidalideation,homicidalideation,andagitationqualified ashigh-riskmentalhealthchiefcomplaints,regardlessof whetherthiswasthepatient’sprimaryreasonforED evaluation.Additionalmedicalreasonsforthe EDvisitwerecollectedbyreviewingtriageand physiciandocumentation.

OutcomeMeasures

Theprimarydependentvariableofinterestwasreceiptof TMHvsin-personmentalhealthconsultationbyamental healthclinicianwhowasanattendingphysician,resident physician,ornursepractitioner.

DataAnalysis

Wereportedcentraltendencyanddispersionasmedians andinterquartilerangesforcontinuousvariables. Categoricalvariableswerereportedasfrequenciesand percentages.Amultivariablelogisticregressionanalysiswas performedtodeterminefactorsassociatedwithuseofTMH. Wecreatedamoderatelysaturatedmodelwith7–8 covariatestominimizeoverfitting.23 Giventhesmallsample size,independentvariableswererankedaprioribasedon expertopinionfrompsychiatrists(EJW,CC)andemergency physicians(MJW,JHH)whoroutinelycareformental healthpatients.ThetopsevenrankedfactorsforTMHvsinpersonmentalhealthevaluationincludedage,race,high-risk chiefcomplaint,presenceofdementia,urbanlocation, timingofpresentation,andhistoryofsubstanceabuse.To exploreadditionalfactorsassociatedwithTMHvsin-person mentalhealthconsultation,weperformedahighlysaturated modelincorporatingallfactorsintothemultivariablelogistic regressionmodel.Becausesite(EDvsUCC)ofpatient presentationmayhavestronglyinfluencedTMHvsinpersonmentalhealth,thisfactorwasincorporatedintothe models.Adjustedoddsratios(aOR)with95%confidence intervals(CI)arereported.Weconductedallstatistical analyseswithRstatisticalsoftware,v3.6.2(TheRProjectfor StatisticalComputing,Vienna,Austria).

RESULTS

Ofthe510healthrecordsreviewed,254patientsmetage inclusioncriteria(≥55yearsofage)andwereincludedinthe study.Characteristicsofthisoldercohortvstheentirecohort ofchartsreviewedisincludedasasupplementaltable. Ofthoseeligible,177(69.7%)veteransreceivedTMH

Table1. Baselinedemographicdataofpatientspresenting totheemergencydepartmentorurgentcarecenterreceiving psychiatricconsultation.

Variable In-person (n = 77) Telemental health(n = 177)

Age,(years)65[61,71]65[61,70]

Gender,n(%)

Female3(3.9)14(7.9)

Male74(96.1)163(92.1)

Race,n(%)

Black30(39.0)72(40.7)

Non-Black47(61.0)105(59.3)

Maritalstatus,n(%)

Married26(33.8)44(24.9)

Unmarried/unknown51(66.2)133(75.1)

Chiefcomplaintrisk,n(%)

Low49(63.6)130(73.4)

High28(36.4)47(26.6)

Historyofdementia,n(%)

Yes10(13.0)18(10.2)

No67(87.0)159(89.8)

Location,n(%)

ED76(98.7)151(85.3)

UCC1(1.3)26(14.7)

Rural,n(%)

Rural24(31.2)45(25.4)

Urban53(68.8)132(74.6)

ESIscore ≥ 2,n(%)77(100.0)177(100.0)

ESIscore,n(%)

<322(28.6)61(34.5)

≥355(71.4)116(65.5)

Timingofpresentation, n(%)

Offhours28(36.4)64(36.2)

Businesshours49(63.6)113(63.8)

Historyofsubstance abuse,n(%)

No36(46.8)74(41.8)

Yes41(53.2)103(58.2)

Mentalhealthclinician type,n(%)

Attendingphysician62(80.5)123(69.5)

Residentphysician7(9.1)1(0.6)

Nursepractitioneror physicianassistant 8(10.4)53(29.9)

Totalpsychoactive medications, median[IQR]

(Continuedonnextcolumn)

Table1. Continued. Variable In-person (n = 77)

Telemental health(n = 177)

Totalpsychiatric comorbidities,median [IQR] 1.00[1.00,2.00]2.00[1.00,2.00]

Additionaltriagemedical complaint,n(%)

No48(62.3)118(66.7)

Yes29(37.7)59(33.3)

Homelessness,n(%)

No64(83.1)144(81.4)

Yes13(16.9)33(18.6)

CCIscore,median[IQR]2.00[1.00,4.00]2.00[1.00,5.00]

ESI,EmergencySeverityIndex; IQR,interquartilerange; CCI, CharlsonComorbidityIndex; UCC,urgentcareclinic; ED, emergencydepartment.

consultations,and77(30.3%)veteransreceivedanin-person evaluation.Therewerenomissingdatapointsonchart review.Intheunadjustedresults,UCClocationand consultationperformedbynursepractitionersandphysician assistantswasassociatedwithastatisticallysignificanttrend towardsTMHuse(Table1).Consultationsperformedby residentmentalhealthphysiciansweremorelikelytooccurin personbutrepresentedfewconsultsoverall(Table1). Age,race,presenceofdementiaorsubstanceusedisorder inmedicalhistory,totalpsychoactivemedications, psychiatriccomorbidityburden,homelessness,andmarital statuswerenotassociatedwithsignificantdifferencesin consultmodality.

Wethenperformedmultivariablelogisticregression analysis.Modelswereadjustedforlocationtoaccountfor sitepracticedifferencesattheEDandUCC,astheUCC performednearlyallconsultsviaTMH. Table2 demonstratesamoderatelysaturatedriskmodel.Nofactors weresignificantlyassociatedwithTMHusebeyondurgent carelocation(AOR15.15,95%CI1.98–116.04).Inahighly saturatedmodel,patientsevaluatedbyresidentphysicians werelesslikelytoreceiveTMH(AOR0.04,95%CI: 0.00–0.58),whilethoseevaluatedbynursepractitionersand physicianassistantsreceiveditmorefrequently(A5.07,95% CI:2.13–12.03),comparedtoattendingphysicians(Table3). Patientswithhigh-riskchiefcomplaints(suicidalideation, homicidalideation,oragitation)werelesslikelytoreceive TMH(AOR:0.39,95%CI:0.18–0.81)inthehighlysaturated riskmodel(Table3).Gender,age,race,comorbidityburden, timingofpresentation,historyofsubstanceusedisorder, historyofdementia,andhomelessnesswerenotassociated significantdifferencesinconsultmodality.

Table3. Multivariableregressionanalysis – highlysaturatedmodel.

modality.Specifically,weobservedthatpatientswithhighriskpsychiatricchiefcomplaints(suicidalideation, homicidalideation,andagitation)weremorelikelytoreceive in-personconsultations.Residentphysiciansperforming consultswerelesslikelytouseTMH,whilenurse practitionersandphysicianassistantsweremorelikelyto chooseTMH.TheUCCusedTMHnearuniversally.

Themoderatelysaturatedriskmodelofmosthighly rankedapriorifactorsshowedAORsgreaterthan1inurban location,timingofpresentationduringoffhours,andhistory ofsubstanceusedisorder.However,the95%CIweretoo widetobesignificant.These findingsweresimilarinthe highlysaturatedmodel.Whilenotstatisticallysignificant, thesefactorsmayholdclinicalrelevance.Further studieswithahighersamplesizeareneededtoclarify anysignificance.

–2.26

–12.03

–0.58

–1.32

–1.54

–1.40 Homelessness1.130.47–2.71

–1.23

UCC,urgentcareclinic; CCI,CharlsonComorbidityIndex.

DISCUSSION

InanoldercohortofveteranspresentingtotheEDor UCCwithacutepsychiatriccomplaints,wefoundthathighriskpsychiatricchiefcomplaints,cliniciantype,andlocation ofthementalhealthconsultwerekeydriversofconsultation

Onepotentialexplanationforreduceduseamonghigher severitycomplaintsisthatmentalhealthcliniciansmayfeel morecompelledtoconductin-personconsultationinhigher acuitysituationsbecausethisiswhattheyaremostfamiliar with.PracticechangessuchastheuseofTMHmaycreatea disruptionasphysiciansstruggleto “unlearn” whattheyare mostfamiliarwithpriortoestablishinganewpractice pattern.24 Alternatively,asrecognizedbytheSocietyfor AcademicEmergencyMedicineConsensusConferenceon EmergencyTelehealth,littleresearchhasbeendoneonthe qualityandsafetyoftelehealth.25 Recentworkhassoughtto addressthis.Evidencesuggestspatientspresentingwithacute psychosismaytoleratetelehealthwell.26,27 Telementalhealth hasbeenfoundtohavenodifferenceinlong-termoutcomes ofrehospitalizationanddeathinpatientswithsuicidal ideationandsuicideattemptscomparedtoin-person consultation.26,28 Additionally,recentworkhassuggested thatTMHisnotassociatedwithincreased30-dayreturn visits,readmissions,ordeathcomparedtoin-person evaluationsinacutecaresettings.26 Therefore,EDand mentalhealthcliniciansshouldbeeducatedonthesafetyof TMHinolderEDpatientswithhigh-riskmentalhealth chiefcomplaints.

Priorresearchdemonstratedthatclinicianscontribute substantialvariabilitytothedecisiontousetelehealthand maypartiallyexplainwhytherearesuchdifferencesinthe useofTMHbycliniciantype(ie,residentphysiciansvsnurse practitionersandphysicianassistants).29 Therewereno differencesinclinicianschedulingthatcouldaccountforthe findingsinourstudy.Allmentalhealthclinicians,including residents,wereavailabletoperformin-personorTMH evaluations.Therefore,locationdidnotmakeresidentsmore orlesslikelytoevaluatepatientsinperson.Thepandemic demonstratedvariabilityintelehealthusewithclinician factorshavingagreaterinfluenceontheuseofvideo telehealthwhencomparedwithpatientfactors.29 Moreover, priorstudiesindicatetherearevariabilitiesinpatientswho areofferedtelehealthdespitebeingvideo-capable.30 Prior

Table2. Multivariableregressionanalysis – moderatelysaturated model.

qualitativestudiessuggestthatincreasedexposureto telehealthimprovedclinicianattitudes,whileperceptionsof complexitywithintheprocessledtoreducedutilization.31 Furtherresearchisneededtobetterunderstandwhether inequitiesandanycontributingfactorsexist.

Systemswithunanimousleadershipbuy-inandpolicies usetelehealthmorefrequently.31 Despitetheavailabilityof anin-personmentalhealthclinician,theUCCusedTMHfor nearlyeveryencounter.Itisplausiblethatsimilarsystemic factorsmaybecontributingtothisphenomenon. Investigatingthepoliciesanddecision-makingprocesses throughqualitativestudiescouldshedlightontheunderlying reasonsforthenear-universaluseofTMHattheUCC,as factorsnotcapturedinthisstudyarelikelyinvolved.

ReluctancetoadoptTMHmaycontributetopotentially avoidabletransfersinEDswithlimitedmentalhealth resources.Priorresearchfoundthatmentalhealthpatients werethemostlikelytobetransferredfromVAEDsand representthelargestgroupofpotentiallyavoidabletransfers, definedasthosetransfersrapidlydischargedfromtheEDor within24hoursfromhospitaladmission(withouta procedure).32 Our findingssuggestthatmentalhealth cliniciansfeltcomfortableevaluatingpatientsviaTMHin low-acuitysituations.Inplaceswithoutaccesstoin-person mentalhealthconsultation,patientswithloweracuity complaintsmaybeevaluatedandsafelydischargedvia TMH,reducingtheriskofunnecessarytransfer.33

WeidentifiedonlyoneresidentTMHencounter throughouttheentirestudyperiod.Asresidentsgenerally rotatebetweenmultipleVAandnon-VAclinicalservices, this findingmaybeduetolackoffamiliaritywiththeprocess inthissystem.Duetothelowoverallnumberof consultationsperformedbyresidents,itisdifficulttodraw conclusionsregardingthisdata.Educationalinitiatives targetingtelehealthuseamongresidentphysiciansmay increasefamiliaritywithTMH.34,35 Astelehealth expandedacrossmultiplespecialtiesduringthepandemic, medicaltrainingcurriculacouldbeadaptedtoinclude telehealthinitiatives.34,35

LIMITATIONS

Limitationsofthisstudyincludedasmallsamplesize.Our samplesizemayhavebeentoosmalltoidentifyriskfactors forTMHuse.Additionally,becausethisstudywas conductedinasinglecenteritmaynotbegeneralizableto othersettings.Riskfactorsidentifiedinourexploratory analysisandthesignificantassociationsobservedmayhave beensecondarytooverfittingasstatisticalsignificancewas onlynotedinthehighlysaturatedmodel.Asaresult,these findingsshouldbeconfirmedinalargersamplesize. Additionally,theVAhasalowproportionofwomen veterans(estimated11.5%),potentiallylimitingthe generalizabilityofourstudyoutsidetheVApopulation.36 FurtherstudiesoutsidetheVApopulationareneededto

assessforanygender-specificdifferencesthatmayimpact consultmodalitychoice.

TheED/UCCclinicianandmentalhealthclinician generallyhadaverbalconversationoncallpriortomental healthconsultation.Theseconversationsmayhave influencedmodalitychoicebythementalhealthclinician. Ourquantitativedatawouldnothavebeenabletocapture theseconversations.Furtherqualitativeworkmaybridge thisgaptounderstandaclinician’smodalitychoice.

Therearepotentialconfounderstothisstudythatwerenot accountedfor.Severityofillnesslikelyaffectsboththe likelihoodofacutecarepresentationandtheconsult modalitychoice.Whileweadjustedforhigh-riskpsychiatric complaintstoaccountforseverityofillness,residual confoundinglikelystillexists.Encountersthatoccurred duringtheCOVID-19pandemicalsolikelyinfluencedboth thelikelihoodofacutecarepresentationandtheconsult modalitychoice.Morementalhealthcliniciansmayhave optedforTMHtoreducetheriskofvirustransmission, especiallyduringperiodsofwidespreadCOVID-19 transmission.Patientsmayhavealsobeenmorefearfulof presentingtotheEDforcareduringthesetimes.Ongoing post-pandemicdataanalysisbothatthisfacilityand externallyshouldbeperformedtoevaluatetheeffectsofthe COVID-19pandemiconTMHuse.

CONCLUSION

Inthisexploratoryretrospectiveanalysis,illnessseverity, location,andcliniciancharacteristicsappearedtoinfluence useoftelementalhealthinpatientsoverage55.Loweracuity, olderpatientsrepresentapatientpopulationwithwhom moreclinicianswouldbecomfortableusingTMH.For resource-poorsettings,TMHmayrepresentanopportunity toexpandaccesstomentalhealthcareinshortageareasand reducepotentiallyunnecessarypatienttransfersthatcould otherwisebepreventedviaremoteconsultation.Further researchisneededtoexaminehesitancytoadoptTMHin moreacutelyillpopulationsandthegeneralizabilityofthe findingspresentedinthiswork.

AddressforCorrespondence:JinH.Han,MD,MSc,Vanderbilt UniversityMedicalCenter,DepartmentofEmergencyMedicine, 2215GarlandAvenue,LightHallSuite203,Nashville,TN37232. Email: jin.h.han@vumc.org

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Thismaterialisbaseduponwork supported(orsupportedinpart)bytheDepartmentofVeterans Affairs,VeteransHealthAdministration,OfficeofRuralHealth, VeteransRuralHealthResourceCenter–IowaCity(Award#ORH10808).AlvinD.Jefferyreceivedsupportforthisworkfromthe AgencyforHealthcareResearchandQuality(AHRQ)andthe

Patient-CenteredOutcomesResearchInstitute(PCORI)under AwardNumberK12HS026395.Thecontentissolelythe responsibilityoftheauthorsanddoesnotnecessarilyrepresentthe officialviewsofAHRQ,PCORI,ortheUnitedStatesgovernment. Therearenootherconflictsofinterestorsourcesoffunding todeclare.

Copyright:©2024Kochetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.SubstanceAbuseandMentalHealthServicesAdministration.Key substanceuseandmentalhealthindicatorsintheUnitedStates:Results fromthe2020NationalSurveyonDrugUseandHealth.2021.Available at: https://www.samhsa.gov/data/sites/default/files/reports/rpt35325/ NSDUHFFRPDFWHTMLFiles2020/2020NSDUHFFR1PDFW102121. pdf.AccessedMarch22,2023.

2.OwensP,FingarK,McDermottK,etal.InpatientStaysInvolvingMental andSubstanceUseDisorders,2016.HealthcareCostandUtilization Project(HCUP)StatisticalBriefs.2019.Availableat: https://hcup-us. ahrq.gov/reports/statbriefs/sb249-Mental-Substance-Use-DisorderHospital-Stays-2016.jsp.AccessedMarch22,2023.

3.BallouS,MitsuhashiS,SankinLS,etal.Emergencydepartmentvisits fordepressionintheUnitedStatesfrom2006to2014. GenHosp Psychiatry. 2019;59:14–9.

4.TheriaultKM,RosenheckRA,RheeTG.Increasingemergency departmentvisitsformentalhealthconditionsintheUnitedStates. JClin Psychiatry. 2020;81(5):20m13241.

5.ZhuD,PaigeSR,SloneH,etal.Exploringtelementalhealthpractice before,during,andaftertheCOVID-19pandemic. JTelemedTelecare. 2024:30(1):72–8.

6.AgencyforHealthcareResearchandQuality(AHRQ).Trendsin emergencydepartmentvisits,2006–2014.HCUPStatisticalBrief#227. 2017.Availableat:~https://hcup-us.ahrq.gov/reports/statbriefs/ sb227-Emergency-Department-Visit-Trends.jsp#:~:text=Overall%2C% 20the%20number%20of%20ED,U.S.%20population%20grew%206. 9%20percent.AccessedMarch22,2023.

7.FairchildR,Ferng-KuoS-F,RahmouniH,etal.Anobservationalstudy oftelementalcaredeliveryandthecontextforinvoluntarycommitment formentalhealthpatientsinagroupofruralemergencydepartments. TelemedRep. 2020;1(1):22–35.

8.DesignatedHealthProfessionalShortageAreasStatistics:Third QuarterofFiscalYear2022DesignatedHPSAQuarterlySummary (BureauofHealthWorkforce:HealthResourcesandServices Administration(HRSA):U.S.DepartmentofHealth&HumanServices). 2022.Availableat: https://data.hrsa.gov/Default/ GenerateHPSAQuarterlyReport.AccessedMarch22,2023.

9.KuBS,LiJ,CathyLally,ComptonMT,etal.Associationsbetween mentalhealthshortageareasandcounty-levelsuicideratesamong adultsaged25andolderintheUSA,2010to2018. GenHosp Psychiatry. 2021;70:44–50.

10.Ivey-StephensonAZ,CrosbyAE,JackSPD,etal.Suicidetrendsamong andwithinurbanizationlevelsbysex,race/ethnicity,agegroup,and mechanismofdeath-UnitedStates,2001–2015. MMWRSurveill Summ. 2017;66(18):1–16.

11.SinghGKandSiahpushM.Wideningrural-urbandisparitiesinlife expectancy,U.S.,1969–2009. AmJPrevMed. 2014;46(2):e19–29.

12.BrownFW.Ruraltelepsychiatry. PsychiatrServ. 1998;49(7):963–4.

13.ShoreJH.Telepsychiatry:videoconferencinginthedeliveryof psychiatriccare. AMJPsychiatry. 2013;170(3):256–62.

14.BeechEH,YoungS,AndersonJK,etal.Evidencebrief:safetyand effectivenessoftelehealth-deliveredmentalhealthcare.2022. Availableat: https://www.hsrd.research.va.gov/publications/esp/ telehealth-mh-brief.cfm.AccessedMarch22,2023.

15.ZhongC,FreemanRE,BoggsKM,etal.Receiptoftelepsychiatryand emergencydepartmentvisitoutcomesinNewYorkState. PsychiatrQ. 2021;92(3):1109–27.

16.ZachrisonKS,BoggsKM,HaydenME,etal.Anationalsurveyof telemedicineusebyUSemergencydepartments. JTelemedTelecare. 2020;26(5):278–84.

17.WardMJ,ShusterJlJr,MohrNM,etal.Implementationoftelehealthfor psychiatricareinVAemergencydepartmentsandurgentcareclinics. TelemedJEHealth. 2022;28(7):985–93.

18.VincentGK,VelkoffVA,SensusBereauUS.(2010).Thenextfour decades:theolderpopulationintheUnitedStates:2010to2050.City, State/Country:U.S.DepartmentofCommerce,Economicsand StatisticsAdministration,U.S.CensusBureau.

19.Populationestimatesandprojections.U.S.Dept.ofCommerce, EconomicsandStatisticsAdministration,U.S.CensusBureau; 2010:14.

20.vonElmE,AltmanDG,EggerM,etal.TheStrengtheningtheReporting ofObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies. Lancet. 2007;370(9596):1453–7.

21.WorsterA,BledsoeRD,CleveP,etal.Reassessingthemethodsof medicalrecordreviewstudiesinemergencymedicineresearch. AnnEmergMed. 2005;45(4):448–51.

22.OnegaT,WeissJE,Alford-TeasterJ,etal.Concordanceofrural-urban self-identityandZIPcode-derivedRural-UrbanCommutingArea (RUCA)designation. JRuralHealth. 2020;36(2):274–80.

23.HarrellFE.RegressionModelingStrategiesWithApplicationstoLinear Models,LogisticandOrdinalRegression,andSurvival Analysis. 2nded. Cham,Switzerland:Springer;2015.

24.GuptaDM,BolandRJ,Jr,AronDC.Thephysician’sexperienceof changingclinicalpractice:astruggletounlearn. ImplementSci. 2017;12(1):28.

25.HaydenEM,DavisC,ClarkS,etal.Telehealthinemergencymedicine: aconsensusconferencetomaptheintersectionoftelehealthand emergencymedicine. AcadEmergMed. 2021;28(12):1452–74.

26.HanJH,KochE,JefferyAD,etal.Theeffectoftelementalversusinpersonmentalhealthconsultsintheemergencydepartmenton30-day utilizationandprocessesofcare. AcadEmergMed. 2023;30(4):262–9.

27.SharpIR,KobakKA,OsmanDA.Theuseofvideoconferencingwith patientswithpsychosis:areviewoftheliterature. AnnGenPsychiatry. 2011;10(1):14.

28.VakkalankaJP,HarlandKK,WittrockA,etal.Telemedicineisassociated withrapidtransferandfewerinvoluntaryholdsamongpatientspresenting withsuicidalideationinruralhospitals:apropensitymatchedcohort study. JEpidemiolCommunityHealth. 2019;73(11):1033–9.

29.RodriguezJA,BetancourtJR,SequistTD,etal.Differencesintheuseof telephoneandvideotelemedicinevisitsduringtheCOVID-19pandemic. AmJManagCare. 2021;27(1):21–6.

30.BenjenkI,FranziniL,RobyD,etal.Disparitiesinaudio-only telemedicineuseamongMedicarebeneficiariesduringthecoronavirus disease2019pandemic. MedCare. 2021;59(11):1014–22.

31.ConnollySL,SullivanJL,LindsayJA,etal.Factorsinfluencinguptakeof telementalhealthviavideoconferencingathighandlowadoptionsites withintheDepartmentofVeteransAffairsduringCOVID-19:aqualitative study. ImplementSciCommun. 2022;3(1):66.

32.MohrNM,WuC,WardMJ,etal.Potentiallyavoidableinter-facility transferfromVeteransHealthAdministrationemergencydepartments: acohortstudy. BMCHealthServRes. 2020;20(1):110.

33.JumreornvongO,YangE,RaceJ,etal.Telemedicineand medicaleducationintheageofCOVID-19. AcadMed. 2020;95(12):1838–43.

34.KirklandEB,DuBose-MorrisR,DuckettA.Telehealthfortheinternal medicineresident:a3-yearlongitudinalcurriculum. JTelemedTelecare. 2021;27(9):599–605.

35.ZhangJ,BodenM,TraftonJ.Mentalhealthtreatmentandthe roleoftele-mentalhealthattheVeteransHealthaAdministration duringtheCOVID-19pandemic. PsycholServ. 2022;19(2):375–85.

36.AffairsUSDoV.VeteranPopulation:Age/Gender.2020.Availableat: https://www.va.gov/vetData/veteran_population.asp AccessedMarch22,2023.

SUBSTANCE USE DISORDER:ORIGINAL RESEARCH

ImplementationandEvaluationofaBystander NaloxoneTrainingCourse

ScottG.Weiner,MD,MPH*

ScottA.Goldberg,MD,MPH*

CherylLang,MPH†

MollyJarman,PhD,MPH†

CoryJ.Miller,BS*

SarahLi,BA*

EwelinaW.Stanek,PA-C*

EricGoralnick,MD,MS*

SectionEditor:MarcMartel,MD

*BrighamandWomen’sHospital,DepartmentofEmergencyMedicine, Boston,Massachusetts † BrighamandWomen’sHospital,DepartmentofSurgery,Boston,Massachusetts

Submissionhistory:SubmittedMarch13,2023;RevisionreceivedSeptember26,2023;AcceptedJanuary12,2024

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.60409

Introduction: Bystanderprovisionofnaloxoneisakeymodalitytoreduceopioidoverdose-related death.Naloxonetrainingcoursesareavailable,butnostandardizedprogramexists.Aspartofa bystanderempowermentcourse,wecreatedandevaluatedabriefnaloxonetrainingmodule.

Methods: Thiswasaretrospectiveevaluationofanaloxonetrainingcourse,whichwaspairedwithStop theBleedtrainingforhemorrhagecontrolandwasofferedtoadministrativestaffinanofficebuilding. Participantsworkedinanorganizationrelatedtohealthcare,butnonewereclinicians.Thecurriculum includedthefollowingtopics:1)backgroundabouttheopioidepidemic;2)howtorecognizethesignsof anopioidoverdose;3)actionsnottotakewhenencounteringanoverdosevictim;4)thecorrectstepsto takewhenencounteringanoverdosevictim;5)anoverviewofnaloxoneproducts;and6)Good Samaritanprotectionlaws.The20-minutedidacticsectionwasfollowedbyahands-onsessionwith nasalnaloxonekitsandasimulationmannequin.ThecoursewasevaluatedwiththeOpioidOverdose Knowledge(OOKS)andOpioidOverdoseAttitudes(OOAS)scalesfortake-homenaloxonetraining evaluation.WeusedthepairedWilcoxonsigned-ranktesttocomparescorespre-andpost-course.

Results: Twenty-eightparticipantscompletedthecourse.TheOOKS,measuringobjectiveknowledge aboutopioidoverdoseandnaloxone,hadimprovedscoresfromamedianof73.2%(interquartilerange [IQR]68.3%–79.9%)to91.5%(IQR85.4%–95.1%), P < 0.001.ThethreedomainsontheOOASscore alsoshowedstatisticallysignificantresults.Competencytomanageanoverdoseimprovedona fivepointscalefromamedianof2.5(IQR2.4–2.9)toamedianof3.7(IQR3.5–4.1), P < 0.001.Concerns aboutmanaginganoverdosedecreased(improved)fromamedianof2.3(IQR1.9–2.6)tomedian1.8 (IQR1.5–2.1), P < 0.001.Readinesstointerveneinanopioidoverdoseimprovedfromamedianof4 (IQR3.8–4.2)toamedianof4.2(IQR4–4.2), P < 0.001.

Conclusion: Abriefcoursedesignedtoteachbystandersaboutopioidoverdoseandnaloxonewas feasibleandeffective.Weencouragehospitalsandotherorganizationstouseandpromulgatethis model.Furthermore,wesuggesttheconveningofanationalconsortiumtoachieveconsensuson programcontentanddelivery.[WestJEmergMed.2024;25(3)320–324.]

INTRODUCTION

Timeisacriticalcontributingfactorinpatientoutcomes inmanyemergencies.IntheUnitedStates,theaverage responsetimebyemergencymedicalservicestoa9-1-1callis sevenminutes.1 Tobridgethisgap,manyeffortshavebeen launchedtoempowerlaypersons,whoaretypically firston thescene,tointerveneandemployskillsrangingfrom cardiopulmonaryresuscitation(CPR)andautomated externaldefibrillator(AED)usetobleedingcontrol interventions.2 BasicLifeSupport(BLS)coursecontentis baseduponrigorousandfrequentlyupdatedconsensus(ie, AmericanHeartAssociation[AHA]GuidelinesUpdatefor CPRandEmergencyCardiovascularCare).3,4 Thesecourses aretaughtinastandardizedfashionbytheAHAandthe AmericanRedCross.Likewise,theStoptheBleed(STB) program,anationalinitiativelaunchedin2015focusedon empoweringthepublicandpublicsafetyprofessionalsto recognizeandcontrollife-threateningbleeding,hasseveral typesofcourses,themostprominentbeingtheAmerican CollegeofSurgeons’ (ACS)BasicHemorrhageControl Course(BCon).5,6

WhileCPR,AEDandSTBtrainingfocusonpreventable deaths,anothersignificantsourceofpreventabledeathsisthe opioidoverdoseepidemic,whichremainsoneofthemost pressingpublichealthissuesofourtime,havingclaimed about1,000,000livesintheUSsince1999.7 Thenumberof overdosedeathshasincreasedgreatlyinrecentyears,with yetanotherrecordnumberin2021,predominantlydueto fentanyl.8 Bystandernaloxoneadministration,whichcanbe usedtoreverseanopioidoverdose,hasbeenintroducedas onepotentialmitigatingfactor.In2018,theUSSurgeon Generalissuedanadvisoryonnaloxoneandopioidoverdose thatencouragescommunitymemberswhocomeintocontact withpeopleatriskforopioidoverdosetoknowhowtouse naloxoneandkeepitwithinreach.9 Likewise,theUS DepartmentofHealthandHumanServices’ overdose preventionstrategyincludesharmreduction,withagoalto widenaccesstoopioidoverdosereversaltreatments.10

UnlikeCPR,thereisnoonestandardizedcoursefor bystandernaloxonetraining.Onlinecoursesareofferedby agenciessuchastheCentersforDiseaseControland Prevention(CDC),11 theAmericanRedCross,12 individual states(eg,Massachusetts13 andNewYork14),andothernonprofits(eg,GetNaloxoneNow15).Thecourseslacka standardizedcorecontent,measuresofeffectiveness,oragreedupondeliverymethods(inperson,hybrid,remote,simulation, didactic,etc).Althoughanecdotesexistoflaypersonuse,we havealimitedunderstandingofaneffective,layperson naloxone-empowermentcurriculum,andgapsremainin knowledgeabouttrainingparametersandstrategies.16

Inthisstudy,weevaluatedanoverdose-responsenaloxone trainingprogramadministeredtolaypersons.We emphasizedthestructureandcurriculumofthecourseand evaluatedefficacywithavalidatedscreeningtool.

METHODS

Thenaloxonecoursewasdesignedtobeabrief interventionwith20minutesofdidacticsand20minutesof practicalexperiencewithamannequin.Thecoursewas bundledwiththeACSBConcourseaspartofabystander empowermentprogram.Courseinstructorswerethree board-certifiedemergencyphysicians.Thesessiontookplace ataprofessionalofficebuilding.Althoughtheparticipants workedinanorganizationrelatedtohealthcare,allworked asofficestaffandnonewereclinicians.Twoidenticalsessions wereoffered,andbothtookplaceinJune2018during normalbusinesshours.Participantswerenotcompensated specificallyforparticipatingbutattendedinlieuoftheir normalduties.Weadministeredanonymouspre-andpostcourseevaluations.Theprojectwasdeterminedtonotmeet thecriteriaforhumansubjectresearchbytheMassGeneral BrighamHumanResearchOffice.

Curriculum

Createdbythecourseinstructors,thecurriculumincluded thefollowingtopics:1)backgroundabouttheopioid epidemic;2)howtorecognizethesignsofanopioid overdose;3)actionsnottotakewhenencounteringan overdosevictim;4)thecorrectstepstotakewhen encounteringanoverdosevictim;5)anoverviewofnaloxone products;and6)GoodSamaritanprotectionlaws.Content wascreatedby firstsearchingforexistingtrainingresources online,includingtrainingmanualsfromthestatesofNew York(https://www.dhses.ny.gov/naloxone-informationfirst-responders)andTexas(https://txoti.org),andCanadian provinceManitoba(https://www.gov.mb.ca/health/ publichealth/docs/training_manual_overdose.pdf).This informationwasintegratedwithadditionalcontentfrom courseinstructorexpertiseintoadidacticmodulecontaining 30slides(Appendix1),andparticipantswereprovidedwitha hardcopyoftheslides.Thepracticalmoduleentailedsmall groupsaroundasimulationmannequinwithacourse instructor.Participantswereabletopracticewithtwotypes ofnaloxonekits(pre-packagednasalnaloxonesprayandan autoinjector)onthemannequin.Discussionwasencouraged untilallparticipants’ questionsandconcernswereaddressed.

CourseEvaluation

Toevaluatetheefficacyofthecourse,weusedtheOpioid OverdoseKnowledge(OOKS)andOpioidOverdose Attitudes(OOAS)scalesfortake-homenaloxonetraining evaluation.17 The firsthalfofthisvalidatedtool(OOKS)asks objectivequestionsaboutopioidoverdosetoevaluatetrainee knowledge,includingindicatorsofopioidoverdose,howto manageanoverdose,themechanismofactionofnaloxone, anditsdurationofaction.Thesecondpart(OOAS)asks questionspertainingtoperceptionsofcompetenciesto manageanopioidoverdose,concernsaboutmanagingan overdose,andreadinesstointerveneinanopioidoverdose.

StatisticalAnalysis

Allparticipantscompletedpre-andpost-evaluationson paperforms.Subjectswereaskedtowritethesamerandom four-digitnumberoneachofthetwoevaluationsforpaired analysispurposes.Responsesweretransferredtoa spreadsheet,andasecondinvestigatorconfirmedthe accuracyofthetranscription.TheOOKSscaleisaseriesof true/falsestatements,andthecorrectanswersweresummed, withatotalpossible41points.Wemodifiedtheoriginal 45-pointversionslightly,asmultiplepointswerepossiblefor severalindividualquestions(eg, “Whatisnaloxoneused for?” and “Howcannaloxonebeadministered?”)andwe countedthemonlyasonepointeach.Therewasalsoachoice of “don’tknow” forseveralquestions,andthatwas consideredanincorrectanswerasindicatedinthescoring instructions.TheOOASscaleis28questionsdividedinto threedomainsandmeasuredona five-pointLikertscale (5 = completelyagreeand1 = completelydisagree). Althoughthepost-testOOKSresultsandoneofthedomains ontheOOASwerenormallydistributedasdeterminedbythe Shapiro-Wilktest,theremainderofresultswerenon-normal. Thus,allresults,includingthescalesoneachdomainofthe OOASandtheoverallscoreontheOOKS,aredescribed withmediansandinterquartilerange(IQR)andcompared withthepairedWilcoxonsigned-ranktest.Weanalyzeddata withJMPv16(JMPStatisticalDiscoveryLLC,Cary.NC).

RESULTS

Twenty-eightparticipantstookthecourse.Allcompleted thepre-testandthepost-test,althoughthreeparticipantsdid notanswerallquestionsonthepre-testOOASscale. Therefore,thecorrespondinganswersinthedomainsfor thesethreeindividualsonthepost-testwerenotincludedin theanalysis.TheOOKS,measuringobjectiveknowledge aboutopioidoverdoseandnaloxone,hadimprovedscores fromamedianof73.2%(IQR68.3%–79.9%)to91.5%(IQR 85.4%–95.1%), P < 0.001.ThethreedomainsontheOOAS scorealsoshowedstatisticallysignificantresults. Competencytomanageanoverdoseimprovedfroma medianof2.5(IQR2.4–2.9)toamedianof3.7(IQR 3.5–4.1), P < 0.001.Concernsaboutmanaginganoverdose decreased(improved)fromamedianof2.3(IQR1.9–2.6)to median1.8(IQR1.5–2.1), P < 0.001.Readinesstointervene inanopioidoverdoseimprovedfromamedianof4(IQR 3.8–4.2)toamedianof4.2(IQR4–4.2), P < 0.001.

DISCUSSION

Increatingandevaluatinganaloxonetrainingprogram forbystanders,wefoundimprovementinbothsubjective attitudesandobjectiveknowledgeaboutopioidoverdose andnaloxone.Thetrainingisrelativelybrief(lastingunder anhour)andeffective.Wehavesubsequentlytaughtthis curriculumseveraltimestolocalcommunityorganizations, includingthosewhoworkwithpeoplewhousedrugs.

Althoughwedidnotmeasureobjectiveoutcomes subsequently,theconceptofbystanderempowerment, teachingbothnaloxoneandSTBskills,hasbeenwell receivedandrepresentsimportantoutreachfromour hospitaltothelocalcommunity.

Onekeyquestionthatremainsiswhetherthistrainingis necessaryforbystanders.Inourpreviousresearch,wefound that49of50bystanderswereabletocorrectlyadminister naloxoneinasimulatedexperienceonapublicsidewalkwith guidancebyasimulated911dispatcher.18 However,not everyonewillhavetheguidanceofadispatcherwhenusing naloxone,andtheremaybeconfusionabouthowtousethe kitandthetimingofaseconddose(ifneeded)withoutthat assistance.Bystandertrainingmayalsobevaluableasaway tofosterself-efficacy,increasingthelikelihoodthata laypersonwillrecognizeandrespondtoanoverdose.Inour course,wealsocoverwhenbystandersshouldadminister naloxoneanddispelmythsaboutanyharmthatcanbe causedbygivingit,aswellashowtoaccessnaloxone.

Naloxoneforbystandersiscurrentlyavailablevia standingorderinseveralstates,meaningthatindividualscan obtainitfrompharmacieswithoutaprescription.19–22 Standingordersareassociatedwithreductionsinfatal overdosesinthecommunity.23 Thecurrentpackagingof prescriptionnasalnaloxonehasa flapthatopensgivingjustin-time(JIT)instructionstothebystander,butthatmaynot besufficient.TheUSFoodandDrugAdministration(FDA) recentlyapprovedmakingnasalnaloxoneanover-thecountermedication,eventhoughitsbriefingdocument describedseveralcasesofincorrectlyadministerednaloxone, includinganindividualwhodidnotplacethetipofthe dispenserfullyinthenostril,someonewhosqueezedthe devicebutdidnotpushtheplunger,anotherwhoplacedthe deviceupsidedownsothattheplungerwasinthenostril,and severalindividualswhodidnotwait2-3minutesbefore administeringaseconddose.24 WhiletheFDAadvisors votedunanimouslytomakenaloxoneavailablewithouta prescription,25 theseerrorsinadministrationindicate theneedforabystandercoursethatcouldfurther improveoutcomes.

Anotherreasontoteachsuchacourseistoaddressstigma, whichispervasivewhenconsideringopioidusedisorder (OUD).26 Arecentstudyofindividualswhodidnotuseillicit opioidsthemselvesbutknewotherswhodidreportedstigma aboutOUDandmisinformationaboutopioid-related risks.27 Naloxone-basedinterventionscanintroducethe conceptofharmreduction,empowerbystanders,and encourageindividualstocarrynaloxoneincasethey encounteranoverdosevictim.28

Althoughnotapartofourstudy,despitethepositive resultsonourobjectiveandsubjectivetesting,wedo encouragethecreationofstandardizedtraining.TheSTB BConportionofourcoursewascreatedandendorsedby theACS,usingstandardizedcontentandcertifiedtrainers.

Asimilarprocesscouldbeusedfornaloxone,eitheraspartof aBLStraining,suchasfromtheAHAorAmericanRed Cross,fromaspecialtysociety,suchastheAmerican AcademyofEmergencyMedicine,theAmericanCollegeof EmergencyPhysicians,ortheAmericanSocietyofAddiction Medicine,orfromanationaladvocacygroupsuchas Shatterproof.Suchbrandingandpromotionmayempower morebystanderstobecometrainedandfurtherreduce stigmaandmisconceptionsaboutOUDamongthe generalpopulation.

WhileCPRtrainingforlaypersonsisthegoldstandard, manygapsinimplementingbystandertrainingremain,and aninvestmentinthestudyoftheeffectivenessoftherelatively simplestepsofnaloxoneadministrationmayhelpuslearn andimprovetechniquesofCPRandSTBtrainingaswell. Forexample,despiteeducationalinitiativesthatbeganinthe 20thcentury,onlyone-thirdofout-of-hospitalcardiacarrest patientsreceivebystanderCPR.Time,location,and durationhaveallbeenperceivedbythepublicasbarriersto CPRclasses.29 BlacksandHispanicsarelesslikelythan WhitestoreceiveCPRathomeorinpublic.30 Inthelast decade,therehavebeenmanyinitiativeswithvariable efficacy,inmostcasesnotmeasured,touseJITtoolslike flashcards,videoortalkingkitstoprovideuserswithrealtimeinstructionsfortheuseofautomatedexternal defibrillatorsorSTBequipment.Whiletheagreementof coursecontentandidentifyingefficacyisa firststep,future workshouldalsofocusondeveloping,trialing,andscaling effectiveJITnaloxone-administrationtools.

LIMITATIONS

Therearelimitationstoourstudy.Wetaughtthiscourse toasmallsampleofadministrativeprofessionalsinasuburb ofMassachusetts,astatewithahighburdenofopioidrelatedoverdose.Itispossiblethatbystandersfromdifferent backgroundsandgeographiclocationswouldhaveanswered thequestionsdifferently.Wealsodidnotcollectany demographicdataaboutourstudyparticipantstoprotect confidentiality.However,thisinformationmighthave determinedthecharacteristicsofindividualswhomaybenefit mostfromthetraining.Thecontentofthepracticalsessionof thecoursewasnotstandardized.Finally,wedidnot measureknowledgeretentionoruseofnaloxonefollowing thecourse.

CONCLUSION

Abriefcoursedesignedtoteachbystandersaboutopioid overdoseandnaloxonewasfeasibleandeffective.We encouragehospitalsandotherorganizationstouseand promulgatethismodel.Furthermore,wesuggestconvening ofanationalconsortiumtoachieveconsensusonprogram content,delivery,andopportunitiesfordevelopmentofjustin-timetoolstoadministernaloxone.

AddressforCorrespondence:ScottG.Weiner,MD,MPH,Brigham andWomen’sHospital,DepartmentofEmergencyMedicine,75 FrancisStreet,NH-226,Boston,MA02115.Email: sweiner@bwh. harvard.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Outsideofthisresearch,ScottG. WeinerissupportedbyNationalInstitutesofHealthgrant5-R01DA044167,theFoundationforOpioidResponseEfforts,andthe ElevanceFoundation.ScottG.Weinerisanadvisorycommittee memberofVertexPharmaceuticals,Inc.andCessation Therapeutics,Inc.Therearenootherconflictsofinterestorsources offundingtodeclare.

Copyright:©2024Weineretal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.MellHK,MummaSN,HiestandB,etal.Emergencymedicalservices responsetimesinrural,suburban,andurbanareas. JAMASurg. 2017;152(10):983–4.

2.FederalEmergencyManagementAgency.Youarethehelpuntil helparrives.Availableat: https://community.fema.gov/ PreparednessCommunity/s/until-help-arrives?language=en_US AccessedAugust20,2023.

3.AmericanHeartAssociation.2020AmericanHeartAssociation GuidelinesforCPRandECC.Availableat: https://cpr.heart.org/en/ resuscitation-science/cpr-and-ecc-guidelines AccessedAugust20,2023.

4.SoarJ,DonninoMW,MaconochieI,etal.2018InternationalConsensus onCardiopulmonaryResuscitationandEmergencyCardiovascular CareSciencewithTreatmentRecommendationssummary. Circulation. 2018;138(23):e714–30.

5.GoralnickE,ChaudharyMA,McCartyJC,etal.Effectivenessof InstructionalInterventionsforHemorrhageControlReadinessfor LaypersonsinthePublicAccessandTourniquetTrainingstudy (PATTS):arandomizedclinicaltrial. JAMASurg. 2018;153(9):791–9.

6.GoolsbyC,JacobsL,HuntRC,etal.StoptheBleedEducation Consortium:educationprogramcontentanddeliveryrecommendations. JTraumaAcuteCareSurg. 2018;84(1):205–10.

7.CentersforDiseaseControlandPrevention.DrugOverdoseDeaths. Availableat: https://www.cdc.gov/drugoverdose/deaths/index.html AccessedAugust20,2023.

8.NationalCenterforHealthStatistics.ProvisionalDrugOverdoseDeath Counts.Availableat: https://www.cdc.gov/nchs/nvss/vsrr/ drug-overdose-data.htm.AccessedAugust20,2023.

9.UnitedStatesHealthandHumanServices.U.S.SurgeonGeneral’s AdvisoryonNaloxoneandOpioidOverdose.Availableat: https://www. hhs.gov/surgeongeneral/reports-and-publications/addiction-

and-substance-misuse/advisory-on-naloxone/index.html AccessedAugust20,2023.

10.UnitedStatesHealthandHumanServices.HarmReduction.Available at: https://www.hhs.gov/overdose-prevention/harm-reduction AccessedAugust20,2023.

11.CentersforDiseaseControlandPrevention.NaloxoneTraining. Availableat: https://www.cdc.gov/opioids/naloxone/training/index.html AccessedAugust20,2023.

12.AmericanRedCross.FirstAidforOpioidOverdosesOnlineCourse. Availableat: https://www.redcross.org/take-a-class/opioidoverdose AccessedAugust20,2023.

13.CommonwealthofMassachusetts.Overdose&NaloxoneTraining Resources.Availableat: https://www.mass.gov/service-details/ training-resources.AccessedAugust20,2023.

14.NewYorkState.CommunityCalendarofOpioidOverdoseTrainings. Availableat: https://www.health.ny.gov/diseases/aids/general/ opioid_overdose_prevention/training_calendar.htm AccessedAugust20,2023.

15.GetNaloxoneNow.Availableat: https://www.getnaloxonenow.org/ #home.AccessedAugust20,2023.

16.Moustaqim-BarretteA,DhillonD,NgJ,etal.Take-homenaloxone programsforsuspectedopioidoverdoseincommunitysettings:a scopingumbrellareview. BMCPublicHealth. 2021;21(1):597.

17.WilliamsAV,StrangJ,MarsdenJ.DevelopmentofOpioidOverdose Knowledge(OOKS)andAttitudes(OOAS)scalesfortake-home naloxonetrainingevaluation. DrugAlcoholDepend. 2013;132(1-2):383–6.

18.GoldbergSA,DworkisDA,LiaoVT,etal.Feasibilityofbystander administrationofpublic-accessnaloxoneforopioidoverdose. Prehosp EmergCare. 2018;22(6):788–94.

19.IjiomaSC,FernándezEV,GatewoodSB,etal.Communitypharmacists inVirginiadispensingnaloxoneunderastandingorder:aqualitative study. JAmPharmAssoc(2003). 2021;61(6):753–60.e1.

20.SugarmanOK,BreithauptJ,WangX,etal.Characteristicsandhealth serviceuseofMedicaid-insuredindividuals fillingnaloxoneundera

standingorderinLouisiana. JAmPharmAssoc(2003). 2023;63(3):904–8.e1.

21.AbbasB,MarottaPL,Goddard-EckrichD,etal.Socio-ecologicaland pharmacy-levelfactorsassociatedwithnaloxonestockingatstandingordernaloxonepharmaciesinNewYorkCity. DrugAlcoholDepend. 2021;218:108388.

22.ChatterjeeA,YanS,XuanZ,etal.Broadeningaccesstonaloxone: communitypredictorsofstandingordernaloxonedistributionin Massachusetts. DrugAlcoholDepend. 2022;230:109190.

23.TaylorM,PradhanA,OgandoYM,etal.Impactofthenaloxonestanding orderontrendsinopioidfataloverdose:anecologicalanalysis. AmJ DrugAlcoholAbuse. 2022;48(3):338–46.

24.UnitedStatesFederalDrugAdministration.JointNonprescriptionDrug AdvisoryCommitteeandAnestheticandAnalgesicDrugProducts AdvisoryCommitteeMeeting.2023.Availableat: https://www.fda.gov/ media/165338/download.AccessedAugust20,2023.

25.AntrimA.FDAAdvisoryCommitteesRecommendNaloxoneBe AvailableOTC.2023.Availableat: https://www.pharmacytimes.com/ view/fda-advisory-committees-recommend-naloxone-be-available-otc AccessedAugust20,2023.

26.FomiattiR,FarrugiaA,FraserS,etal.Addictionstigmaandthe productionofimpedimentstotake-homenaloxoneuptake. Health (London). 2022;26(2):139–61.

27.SlocumS,OzgaJE,JoyceR,etal.Ifwebuildit,willtheycome? Perspectivesonpharmacy-basednaloxoneamongfamilyandfriendsof peoplewhouseopioids:amixedmethodsstudy. BMCPublicHealth. 2022;22(1):735.

28.MillerNM,Waterhouse-BradleyB,CampbellC,etal.Howdonaloxonebasedinterventionsworktoreduceoverdosedeaths:arealistreview. HarmReductJ. 2022;19(1):18.

29.McGovernSK,BlewerAL,MurrayA,etal.Characterizingbarriersto CPRtrainingattainmentusingTwitter. Resuscitation. 2018;127:164–7.

30.GarciaRA,SpertusJA,GirotraS,etal.Racialandethnicdifferencesin bystanderCPRforwitnessedcardiacarrest. NEnglJMed. 2022;387(17):1569–78.

GaryZhang,MD*

MichaelJ.Burla,DO*†§

BenjaminB.Caesar,MD*

CarolyneR.Falank,PhD* PeterKyros,MD*

VictoriaC.Zucco,BS* AnetaStrumilowska,BS‡

DanielC.Cullinane,MD* ForestR.Sheppard,MD*

SectionEditor:ChristopherTainter,MD

*MaineMedicalCenter,DepartmentofSurgery,Portland,Maine † SouthernMaineHealthcare,DepartmentofEmergencyMedicine,Biddeford,Maine ‡ UniversityofNewEnglandCollegeofOsteopathicMedicine,Biddeford,Maine § TuftsSchoolofMedicine,Boston,Massachusetts

Submissionhistory:SubmittedMarch31,2023;RevisionreceivedJanuary8,2024;AcceptedJanuary10,2024

ElectronicallypublishedApril18,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.17975

Background: Patientswithcoronavirus2019(COVID-19)areathighriskforrespiratorydysfunction. Thepulseoximetry/fractionofinspiredoxygen(SpO2/FiO2)ratioisanon-invasiveassessmentof respiratorydysfunctionsubstitutedforthePaO2:FiO2 ratioinSequentialOrganFailureAssessment scoring.Wehypothesizedthatemergencydepartment(ED)SpO2/FiO2 ratioscorrelatewithrequirement formechanicalventilationinCOVID-19patients.OurobjectivewastoidentifyCOVID-19patientsat greatestriskofrequiringmechanicalventilation,usingSpO2/FiO2 ratios.

Methods: WeperformedaretrospectivereviewofpatientsadmittedwithCOVID-19attwohospitals. HighestandlowestSpO2/FiO2 ratios(percentsaturation/fractionofinspiredO2)werecalculatedon admission.Weperformedchi-square,univariate,andmultipleregressionanalysistoevaluatethe relationshipofadmissionSpO2/FiO2 ratioswithrequirementformechanicalventilationandintensive careunit(ICU)care.

Results: Atotalof539patients(46%female;84%White),withameanage67.6 ± 18.6years,met inclusioncriteria.Patientswhorequiredmechanicalventilationduringtheirhospitalstaywerestatistically youngerinage(P = 0.001),hadahigherbodymassindex(P < .001),andtherewasahigherpercentage ofpatientswhowereobese(P = 0.03)andmorbidlyobese(P < .001).Shortnessofbreath,cough,and feverwerethemostcommonpresentingsymptomswithamediantemperatureof99°F.Averagewhite bloodcountwashigherinpatientswhorequiredventilation(P =<0.001).AhighestobtainedEDSpO2/ FiO2 ratioof ≤300wasassociatedwitharequirementformechanicalventilation.AlowestobtainedED SpO2/FiO2 ratioof ≤300wasassociatedwitharequirementforintensivecareunitcare.Therewasno statisticallysignificantcorrelationbetweenEDSpO2/FiO2 ratios >300andmechanicalventilationor intensivecareunit(ICU)requirement.

Conclusion: TheEDSpO2/FiO2 ratioscorrelatedwithmechanicalventilationandICUrequirements duringhospitalizationforCOVID-19.TheseresultssupportEDSpO2/FiO2 asapossibletriagetooland predictorofhospitalresourcerequirementsforpatientsadmittedwithCOVID-19.Furtherinvestigationis warranted.[WestJEmergMed.2024;25(3)325–331.]

INTRODUCTION

Thecoronavirus2019(COVID-19)pandemicprofoundly impactedhospitalsystemsworldwide.Identifyingpatients presentingwithCOVID-19intheemergencydepartment (ED)atgreatestriskforrequiringmechanicalventilationor intensivecareunit(ICU)careisofparamountimportance sincethiswouldfacilitatemoreefficientallocationoflimited medicalresources.SevereCOVID-19infectioncanbelifethreateningandisassociatedwithsignificanthypoxemiaand thedevelopmentofacuterespiratorydistresssyndrome (ARDS).1,2 Identifyingearlypredictorsofrespiratoryfailure andICUneedisvitalbothforpatientcareandlogisticsinthe settingofaglobalpandemicwithlimitedICUresources.

Thepulseoximetry/fractionofinspiredoxygen(SpO2/ FiO2 ratio)haspreviouslybeenusedasapredictorofhighflownasalcannulafailure,needforintubation,and mechanicalventilation.3 TheSpO2 valuehasbeen demonstratedtobeareliablesurrogateforpartialpressureof oxygeninthearterialblood(PaO2),4,5 andtheSpO2/FiO2 ratiodoesnotrequireanybloodtests.TheSpO2/FiO2 ratiois anon-invasiveassessmentofrespiratorydysfunctionthat canbequicklyobtainedatthebedside.Measuredatthetime ofpresentation,theSpO2/FiO2 ratiohasbeendemonstrated tobeanindependentindicationofARDSdevelopment.6 The abilitytoquicklydeterminerequiredlevelofcarefor vulnerablepatientsisessentialtopreventpooroutcomes, particularlyinresource-limitedenvironments.TheCOVID19pandemicledtoEDcrowdingandadecreaseinventilator andICUavailability.7 Avalidatedprognosticindicatortool akintothesystematicinflammatoryresponsesyndromeor SequentialOrganFailureAssessmentcriteriaforsepsis8 is vitalforEDusetoidentifyCOVID-19patientsathighestrisk ofventilatorandICUneed.TheSpO2/FIO2 ratiopredictive valuehaspreviouslybeenvalidatedinARDS,6 andearly measurementmayserveasanindicatorandtriagetoolin COVID-19withregardtorespiratoryfailure/ventilationrisk andICUneed.

OurobjectiveinthisstudywastoevaluateEDSpO2/FIO2 ratiosinCOVID-19patientsandcorrelatethemwith subsequentrespiratoryfailure,necessitatingtheneedfor ICUlevelofcareand/ormechanicalventilationduring hospitalization.Useofthisratiomayhelphospitalsystems moreefficientlyuseresourcesandeffectivelypreparefora patient’sneedforICUcareormechanicalventilation.

MATERIALSANDMETHODS

StudyDesignandParticipantSelection

Thiswasaretrospectivestudythatevaluatedadmission encountersfrombothMaineMedicalCenter(MMC)and SouthernMaineHealthCare(SMHC).Theseinstitutions workcloselytogether,withMMCbeingtheregion’stertiary carecenterwithover70,000annualEDvisitsandatotalof 45multipurposeICUbeds.TheSMHCisacommunity hospitalwithincloseproximitytoMMC,averaging ≈55,000

totalEDvisitsandnineICUbeds.COVID-19patientswho were ≥18yearsoldandrequiredadmissiontoeitherhospital metinclusioncriteria.Encounterswerecollectedbetween March–December28,2020;thus,nopatientshadbeen vaccinatedagainstCOVID-19.Patientswereexcludedifthey didnotrequireadmission.Thisstudywasperformedunder approvaloftheinstitutions’ reviewboards.

DataVariables

Weperformedretrospectivechartreviewtoidentify patientdemographics,diagnoses,levelofhospitalcare,and hospitaloutcomesdatafromelectronichealthrecords.The FiO2 valueswerecalculatedusingnasalcannula flowrate.9 Werecordedthepatient’slowestandhighestSpO2 andFiO2 valuesintheEDandcalculatedSpO2/FiO2 ratios.

Outcomes

Theprimaryoutcomewastheneedformechanical ventilation.SecondaryoutcomesincludedICUlevelofcare, ventilatordays,in-hospitalcomplications,escalationofcare followinginitialtriage,ICUlengthofstay(LOS),hospital LOS,andin-hospitalmortality.

Analysis

WeanalyzeddatausingRStudio2020(RStudioInc, Boston,MA).Descriptivestatisticswerepresentedas frequencyandpercentage.Normallydistributedcontinuous datawerereportedasmeanwithSDs,andordinalnonnormallydistributedcontinuousdataweredescribedwith medianswithinterquartileranges.Weusedmultivariable logisticregressiontoassesstheassociationbetweeneither loworhighSpO2/FiO2 ratioswithintheED,anticoagulation use,asthma,coronaryarterydisease(CAD),congestive healthfailure(CHF),chronicobstructivepulmonarydisease (COPD),diabetes,hyperlipidemia,hypertension,and gastroesophagealrefluxdisease(GERD),ortheneedfor mechanicalventilation,adjustedforageandbodymass index(BMI).Bivariableanalysisofcategoricalvariableswas doneusingthe χ2test,andnonparametricvariablesbythe Kruskal-Wallistest.Regressionmodelscontrolledforboth ageandBMI.

RESULTS

Atotalof539patients,withameanage67.6 ± 18.6years, metinclusioncriteria.Patientswerestratifiedintotwo cohortsbasedontheneedformechanicalventilation (Table1).Asshowninthetable,patientswhorequired mechanicalventilationduringtheirhospitalstaywere statisticallyyoungerinage(P = 0.001),hadahigherBMI (P < .001),andtherewasahigherpercentageofpatientswho wereobese(P = 0.03)andmorbidlyobese(P < .001). Shortnessofbreath,cough,andfeverwerethemostcommon presentingsymptoms,withamediantemperatureof99°F. Theaveragewhitebloodcountwashigherinpatientswho

Table1. Baselinecharacteristicsofpatientswithcoronavirus2019.

Demographicdata

Mean ± SD,median,rangeorn(%)

Notmechanically ventilatedn = 451 Mechanically ventilatedn = 88 P-value

Age(median,IQR)72,2666,19.750.001

BMI(median,IQR)28.9,9.432.3,10.9 <.001

Gender

Female217(48%)31(35%)0.03

Male234(52%)57(65%)0.03

Race

Asian11(2%)5(6%)0.03

Black31(7%)5(6%)0.73

NativeHawaiianorotherPacificIslander1(0.2%)0(0%)0

Unknown/notreported2(0.4%)2(2%)0.003

Morethanonerace3(0.6%)0(0%)0.47

White397(88%)73(83%)0.20

Other6(1%)3(3%)0.13

Ethnicity

HispanicorLatino9(2%)3(3%)0.56 NotHispanicorLatino440(98%)84(95%)0.10

Unknown/notreported2(0.4%)1(1%)0.46

Origin

Home282(63%)54(61%)0.72

Nursinghome61(14%)9(10%)0.31

Skillednursinghome31(7%)0(0%)0.01

Rehab1(0.2%)2(2%)0.03

Other*76(17%)23(26%)0.05

Comorbidconditions

Alcoholuse23(5%)8(9%)0.14

Anticoagulationtherapy52(12%)13(15%)0.44

Asthma66(15%)14(16%)0.81

Cerebrovascularaccident41(9%)4(5%)0.22

COPD71(16%)16(18%)0.64

Chronicheartfailure67(15%)13(15%)1

Chronickidneydisease73(16%)12(14%)0.64

Cancer57(13%)9(10%)0.44

Coronaryheartdisease/heartfailure105(23%)19(22%)0.84

Currentsmoker30(7%)2(2%)0.08

Dementia75(17%)5(6%)0.01

Diabetesmellitus156(35%)38(43%)0.15

GERD132(29%)26(30%)0.85

Myocardialinfraction39(9%)5(6%)0.36

Hypertension282(63%)57(65%)0.72

Hyperlipidemia222(49%)49(56%)0.23

Morbidlyobese14(3%)11(13%) <.001

Obese81(18%)25(28%)0.03

(Continuedonnextpage)

Table1. Continued.

Demographicdata

Presentingsymptoms

Mean ± SD,median,rangeorn(%)

Notmechanically ventilatedn = 451

Mechanically ventilatedn = 88 P-value

Fever176(39%)34(39%)1

Myalgia72(16%)15(17%)0.82

Arthralgias21(5%)2(2%)0.22

Headache50(11%)4(5%)0.09

GIsymptoms140(31%)17(19%)0.02

Cough229(51%)54(61%)0.09

Shortnessofbreath253(56%)57(65%)0.12

Other233(52%)43(49%)0.61

AveragetemperatureintheED ± SD(Fahrenheit)97.1 ± 12.4,99.1,7.3–104.598.1 ± 10,99,37–1030.48

WBCcountintheED(median,IQR)6.2,4.78,7.4 <.001

Diagnoses

Neurologicaldiagnoses128(28%)40(45%)0.002

Renaldiagnoses129(29%)55(63%) <.001

Liverdiagnoses44(10%)23(26%)

Heartdiagnoses168(37%)56(64%) <.001

Pulmonarydiagnoses280(62%)69(78%)0.004

Shock10(2%)46(52%) <.001

Respiratoryfailure153(34%)75(85%) <.001

Renalfailure29(6%)22(25%) <.001

ICU

PatientswhorequiredICUcareatanypoint75(17%)84(95%) <.001

RequiredmorethanoneICUadmissions2(0.4%)6(7%) <.001

ICULOS(median,IQR)2,313,16 <.001

Intubated

Patientswhowereintubated0(0%)84(95%) <.001

Daysintubatedn/a2,4

Non-procedurebasedintubation0(0%)51(58%) <.001

Mechanicalventilators

Ventilatordays(median,IQR)n/a9,13

Requiredreintubationn/a7(8%)

Escalationofcarefrominitialtriage60(13%)56(64%) <.001

HospitalLOS(median,IQR)6,617.5,19 <.001

Dischargedisposition

Homeorself-care193(43%)8(9%) <.001

Homewithservices97(22%)17(19%)0.53

Hospice/palliativecareunit11(2%)1(1%)0.52

Mentalhealth/psychiatrichospital8(2%)0(0%)0.18

Nursinghome17(4%)1(1%)0.16 (Continuedonnextpage)

Table1. Continued.

Demographicdata

Mean ± SD,median,rangeorn(%)

Notmechanically ventilatedn = 451 Mechanically ventilatedn = 88 P-value

Other62(14%)34(39%) <.001

Rehab15(3%)24(27%) <.001

Skillednursingfacility48(11%)3(3%)0.02

In-hospitalmortality36(8%)28(32%) <.001

*Otherincludeshomeless,transfersin,grouphome,Primarycarephysicianfollowup,mentalhealthfacility.

BMI,bodymassindex; IQR,interquartilerange; COPD,chronicobstructivepulmonarydisease; GERD,gastroesophagealrefluxdisease; WBC,whitebloodcount; ED,emergencydepartment; ARDS,acuterespiratorydiseasesyndrome; GI,gastrointestinal; ICU,intensivecare unit; LOS,lengthofstay.

requiredventilation(P =<0.001)(Table1).Patients requiringmechanicalventilationhadhigherdiagnoses ofARDS(P < .001),pneumonia(P < .001),shock (P < .001),respiratoryandrenalfailure(P < .001), andworsehospitaloutcomeswithanin-hospitalmortality of32%vs8%(P < .001)andamedianhospitalLOSof 17.5vs6days(P < .001).

TheSpO2/FiO2 ratiosintheEDandtheirassociations withmechanicalventilationorneedforICUcareare presentedin Table2.AhighestobtainedEDSpO2/FiO2 ratioof300orbelowwasstatisticallyassociatedwitha requirementformechanicalventilationduring hospitalization.AlowestobtainedEDSpO2/FiO2 ratioof 300orbelowwasstatisticallyassociatedwitharequirement forICUcareduringhospitalization.Therewasno statisticallysignificantrelationshipbetweenEDSpO2/FiO2 ratiosabove >300andmechanicalventilationorICUlevel ofcare.

Chronicobstructivepulmonarydiseasewasa confoundingfactorforCOVID-19patientswhorequired mechanicalventilation(adjustedR2 value = 0.1132; P < .001).Nostatisticallysignificantassociationswere identifiedbetweenthefollowingco-morbidities: anticoagulationuse;asthma(adjustedR2 = 0.096, P = 0.75); CAD(adjustedR2 = 0.102; P = 0.07);CHF(adjustedR2 = 0.096; P = 0.95);diabetes(adjustedR2 = 0.10; P = 0.07); hyperlipidemia(adjustedR2 = 0.11; P = 0.08);hypertension (adjustedR2 = 0.096; P = 0.58);andGERD(adjustedR2 = 010; P = 0.28)fortherequirementofmechanicalventilation.

DISCUSSION

ThisstudydemonstratedthatthehighestobtainedED SpO2/FiO2 ratioof300orbelowcorrelatedwiththeneedfor mechanicalventilationduringhospitalization.Additionally, alowestobtainedEDSpO2/FiO2 ratioof300orbelowwas associatedwitharequirementforICU-levelcare.Although COPDwasaconfoundingfactorforpatientsrequiring mechanicalventilation,otherco-morbiditieswerenot independentlyassociatedwithhigherratesofmechanical

ventilationandtheEDSpO2/FiO2.Thissuggeststhatthe SpO2/FiO2 ratiocanbeusedasaprognosticindicatorto stratifyseverityofillnessinpatientswithCOVID-19during theirinitialevaluationintheED.SincetheSpO2/FiO2 ratiois non-invasiveandcanbequicklyobtainedandtrendedduring apatient’sevaluation,thisratiocouldbeanimportantfactor inpatienttriageanddisposition.

Multipleprognosticindicatorshavebeenproposedinthe previousliteraturetohelpstratifyARDSseverityandpredict outcomes.10–13 ThePaO2:FiO2 (P:F)ratioisawidelyused measureofARDSseverity;however,multiplestudieshave shownthattheP:Fratioisnotanindependentpredictorof mortality.10–13 Anotherprognostictool,theoxygenation index,(OI[FIO2/PaO2 × meanairwaypressure × 100])has beendemonstratedtobeanindependentriskfactorfor mortalityinadultswithARDS,11,12 butitrequires mechanicalventilationandarterialbloodgasanalysisfor calculation.Oxygensaturationindex(OSI[FIO2 × mean airwaypressure × 100)/SaO2])isameasurethatcorrelatesto OIandisanindependentpredictorofclinicaloutcomes.12 AlthoughOSIcalculationdoesnotrequirebloodanalysis,it stillrequiresmechanicalventilation.Anotherprognostic tool,theLungInjuryPredictionScore(LIPS),has applicabilityintheED.13 However,theLIPStoolrequiresa detailedpastmedicalhistory(e.g,alcoholusedisorder)and thepatient’spH,requiringabloodgas.Althoughallthese toolsprovidesomeprognosticvalue,eachhaslimitations, resultinginbarrierstodeploymentfortriagingpatientsin theED.

Incontrast,theSpO2/FiO2 ratiorequiresnobloodtests andisquicklyandeasilyobtainedatthebedside.Measured atthetimeofpresentation,ithasbeenshowntobean independentindicationofARDSdevelopment.6 Thisstudy suggeststhattheSpO2/FIO2 ratiomayofferanestimateof diseaseseverityinpatientswithCOVID-19before progressiontoovertrespiratoryfailure,servingasatriage tooltoidentifythoseatgreatestriskforneedingmechanical ventilationandcriticalcare.TheSpO2/FiO2 ratiocanbeused asatoolorpartofaprotocoltoassesswhetherapatient

Table2. SpO2/FiO2 ratiosandtheirassociationwithintensivecareunitormechanicalventilationneeds. VariableSpO2/FiO2 ratios*

LowestEDSpO2/FiO2

*ForpatientswhohadEDSpO2/FiO2 values. CI,confidenceinterval; OR,oddsratio; ED,emergencydepartment; ICU,intensivecareunit.

meetstransfercriteriawithinahospitalsystem.Many regionalhealthsystemsoperateundera “hubandspoke” modelwherealargecentralinstitutionsupportsanetworkof smallerhospitals.Rapididentificationofpatientsatriskfor decompensationandwithneedforhigherlevelcarewould facilitateaccesstolimitedcriticalcareresourceswhilealso decreasingtheincidenceofover-triagetothehubhospital.

LIMITATIONS

Thestudyisretrospectivewithinherentlimitationsin controllingconfoundingvariables.Thecohortwaslimitedto onehospitalsystem,andthuscannotaccountforpractice variationsinotherhealthcaresystems.Thehospitals evaluatedinthisstudymayhavehaddifferentcriteriafor ICUadmission.Additionally,FiO2 valueswerebasedlargely

onnasalcannula flowrates;limitingtohigh flownasal cannulawouldpermitmoreaccurateFiO2 butwould alsolimitapplicability.Atthetimeofdatacollection, nopatientswerevaccinated,thuslimitingtheapplicability of findingstopopulationswithsomeformof COVID-19vaccination.

CONCLUSION

Insummary,EDSpO2/FiO2 ratioscorrelatewith mechanicalventilationandICUrequirementsduring hospitalizationforCOVID-19infection.Theseresults supportEDSpO2/FiO2 asatriagetoolandpredictorof hospitalresourcerequirementsforpatientsadmittedwith COVID-19.Furtherstudyisrequiredwithaprospective analysisassessingaccuracyoftheSpO2/FiO2 ratioin

predictingmechanicalventilationandneedforICUlevelcare.

AddressforCorrespondence:MichaelJ.Burla,DO,SouthernMaine HealthCare,DepartmentofEmergencyMedicine,SouthernMaine HealthCare,1MedicalCenterDr,Biddeford,ME04005. Email: Michael.burla@Mainehealth.org

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Thisworkwassupportedinpartbythe NorthernNewEnglandClinicalandTranslationalResearchgrant U54GM115516.Therearenoconflictsofinterestorothersourcesof fundingtodeclare.

Copyright:©2024Zhangetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.ArentzM,YimE,KlaffL,etal.Characteristicsandoutcomesof21 criticallyillpatientswithCOVID-19inWashingtonState. JAMA. 2020;323(16):1612–4.

2.BellaniG,LaffeyJG,PhamT,etal.Epidemiology,patternsofcare,and mortalityforpatientswithacuterespiratorydistresssyndromein intensivecareunitsin50countries. JAMA. 2016;315(8):788–800.

3.RocaO,CaraltB,MessikaJ,etal.Anindexcombiningrespiratoryrate andoxygenationtopredictoutcomeofnasalhigh-flowtherapy. AmJRespirCritCareMed. 2019;199(11):1368–76.

4.BilanN,DastranjiA,GhalehgolabBehbahaniA.Comparisonofthe SpO2/FiO2 ratioandthePaO2/FiO2 ratioinpatientswithacutelunginjury

oracuterespiratorydistresssyndrome. JCardiovascThoracRes. 2015;7(1):28–31.

5.RiceTW,WheelerAP,BernardGR,etal.ComparisonoftheSpO2/FIO2 ratioandthePaO2/FIO2 ratioinpatientswithacutelunginjuryorARDS. Chest. 2007;132(2):410–7.

6.FesticE,BansalV,KorDJ,etal.SpO2/FiO2 ratioonhospital admissionisanindicatorofearlyacuterespiratorydistress syndromedevelopmentamongpatientsatrisk. JIntensiveCareMed. 2015;30(4):209–16.

7.SandhuP,ShahAB,AhmadFB,etal.Emergencydepartmentand intensivecareunitovercrowdingandventilatorshortagesinushospitals duringtheCOVID-19pandemic,2020–2021. PublicHealthRep. 2022;137(4):796–802.

8.ZhaoC,XinMY,LiJ,etal.ComparingtheprecisionofthepSOFAand SIRSscoresinpredictingsepsis-relateddeathsamonghospitalized children:amulti-centerretrospectivecohortstudy. WorldJEmergMed. 2022;13(4):259–65.

9.MarkovitzGH,ColthurstJ,StorerTW,etal.Effectiveinspiredoxygen concentrationmeasuredviatranstrachealandoralgasanalysis. Respir Care. 2010;55(4):453–9.

10.WareLB.PrognosticDeterminantsofacuterespiratorydistress syndromeinadults:impactonclinicaltrialdesign. CritCareMed. 2005;33(3Suppl):S217–22.

11.GajicO,AfessaB,ThompsonBT,etal.Predictionofdeathand prolongedmechanicalventilationinacutelunginjury. CritCare. 2007;11(3):R53.

12.DesPrezK,McNeilB,WangC,etal.Oxygenationsaturationindex predictsclinicaloutcomesinARDS. Chest. 2017;152(6):1151–8.

13.Elie-TurenneMC,HouPC,MitaniA,etal.Lunginjurypredictionscore fortheemergencydepartment: firststeptowardspreventioninpatients atrisk. IntJEmergMed. 2012;5(1):33.

ORIGINAL RESEARCH

GeographicLocationandCorporateOwnershipofHospitalsin RelationtoUnfilledPositionsinthe2023Emergency MedicineMatch

ZacharyJ.Jarou,MD,MBA*†

AngelaG.Cai,MD,MBA‡

LeonAdelman,MD,MBA§

DavidJ.Carlberg,MD∥

SaraP.Dimeo,MD,MEHP¶

JonathanFisher,MD,MPH# ToddGuth,MD,MHPE**

BruceM.Lo,MD,MBA††‡‡ LauraOh,MD§§

RahulPuttagunta,MD*†

GillianR.Schmitz,MD∥∥

*AscensionProvidenceHospital,DepartmentofEmergencyMedicine, South field,Michigan

† MichiganStateUniversity,EastLansing,Michigan

‡ UniversityofPennsylvania,DepartmentofEmergencyMedicine, Philadelphia,Pennsylvania

§ IvyClinicians,Raleigh,NorthCarolina

∥ GeorgetownUniversity,DepartmentofEmergencyMedicine,WashingtonDC

¶ DignityHealthEastValleyRehabilitationHospital,ChandlerRegionalMedical Center,Chandler,Arizona

# TexasChristianUniversity,BurnettSchoolofMedicine,FortWorth,Texas

**UniversityofColoradoSchoolofMedicine,DepartmentofEmergencyMedicine, Denver,Colorado

†† SentaraNorfolkGeneralHospital,Virginia

‡‡ EasternVirginiaMedicalSchool,Virginia

§§ EmoryUniversity,DepartmentofEmergencyMedicine,Atlanta,Georgia

∥∥ UniformedServicesUniversityofHealthSciences,DepartmentofMilitaryand EmergencyMedicine,Bethesda,Maryland

SectionEditor:GaryJohnson,MD

Submissionhistory:SubmittedAugust18,2023;RevisionreceivedDecember8,2023;AcceptedDecember14,2023

ElectronicallypublishedMarch29,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.18436

Introduction: Inthe2023NationalResidentMatchingProgram(NRMP)match,therewere554unfilled emergencymedicine(EM)positionsbeforetheSupplementalOfferandAcceptanceProgram(SOAP). WesoughttodescribefeaturesofEMprogramsthatparticipatedinthematchandtheassociation betweenselectprogramcharacteristicsandunfilledpositions.

Methods: Theprimaryoutcomemeasuresincludedtheproportionofpositions filledinrelationtostate andpopulationdensity,hospitalownershiptype,andphysicianemploymentmodel.Secondaryoutcome measuresincludedcomparingprogram-specificattributesbetween filledandunfilledprograms, includingoriginalaccreditationtype,yearoforiginalaccreditation,thetotalnumberofapprovedtraining positions,lengthoftraining,urban-ruraldesignation,hospitalsizebynumberofbeds,resident-to-bed ratio,andthepercentageofdisproportionatesharepatientsseen.

Results: TheNRMPMatchhad276uniqueparticipatingEMprogramswith554unfilledpositions.Six statesoffered52%ofthetotalNRMPpositionsavailable.Fivestateswereassociatedwithtwo-thirdsofthe unfilledpositions.Publichospitalshadastatisticallysignificanthighermatchrate(88%)whencomparedto non-profitandfor-profithospitals,whichhadmatchratesof80%and75%,respectively(P < 0.001). Programswithfacultyemployedbyahealthsystemhadthehighestmatchrateof87%,followedbyclinician partnershipsat79%andprivateequitygroupsat68%(P < 0.001overallandbetweenallsubgroups).

Conclusion: The2023matchinEMsawincreasedratesinthenumberofresidencypositionsand programsthatdidnot fillbeforetheSOAP.Publichospitalshadhighermatchratesthanfor-profitornonprofithospitals.Residencyprogramsthatemployedacademicfacultythroughthehospitalorhealth systemwereassociatedwithhighermatchrates.[WestJEmergMed.2024;25(3)332–341.]

INTRODUCTION

Emergencymedicine(EM)hashistoricallybeenahighly competitivespecialty, fillingallornearlyalltheavailable residencypositionsaspartoftheMainResidencyMatch (match)organizedbytheNationalResidencyMatching Program(NRMP).Afterarecordnumberofapplicantsin 2021,thepasttwoyearshaveseenadeclineinthenumberof studentapplicantswhilethenumberofavailableEM residencypositionshascontinuedtoincrease,ultimately resultinginariseinunfilledprogramsandpositions.Inthe 2022NRMPmatch,therewere219unfilledEMpositions among69programsbeforetheSupplementalOfferand AcceptanceProgram(SOAP),andin2023that figure approximatelydoubledto554unfilledpositionsamong 131programs.Manyareconcernedthatthedramatic increaseinpre-SOAPunfilledpositionsrepresentsa declineinthedesirabilityandcompetitivenessof thespecialty.1

ThisisanobservationalstudydescribingfeaturesofEM residencyprogramsthatparticipatedinthe2023NRMP matchandtheassociationbetweenselectprogram characteristicsandunfilledpositions.Itisunclearwhether certaincharacteristicsincludingstate-basedgeographic locationandpopulationdensity,hospital financingmodels, facultyphysicianemploymentmodels,orspecific programcharacteristicssuchasthesizeofprogramorlength oftrainingareassociatedwithhigherratesofunfilled positions.Transparencyoffactorsassociatedwithunfilled positionswillguidethespecialty’sresponsetothematchand programaccreditationrequirementswithobjectivedata. Priorstudieshaveexaminedsimilarfactorsbutprovided limiteddetailandnuanceonthetopicofcorporate ownership,whichweexpounduponin ourstudy.2,3

METHODS

StudyDesignandSetting

Inthisobservationalstudyweusedpubliclyavailable datasetstoanalyzethematchresultsforEMresidency programsparticipatinginthe2023NRMPMatchbasedon STROBEguidelines.4 Theinstitutionalreviewboard determinedthisstudytobeexempt.AllEMresidency programsandthepositionstheyofferedthatparticipatedin the2023NRMPMatchwereincludedin thestudy.

VariablesandMeasurements

WeobtainedalistofEMresidencyprogramsandtheir numberofofferedand filledpositionsfromtheNRMP.Each NRMPIDwaslinkedtotheprogram’sAccreditation CouncilforGraduateMedicalEducation(ACGME) ProgramID,whichprovidedinformationabouttheyearof accreditation,programlength,numberofapproved

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Priorstudiesexaminedprogramfeaturesand ownershippredictorsofun fi lledpositionsbut withoutdeeperanalysisofcorporate ownershiptrendsandassociations.

Whatwastheresearchquestion?

Whatprogramfeaturesandhospitalor facultyownershipareassociatedwiththe un fi lled2023match?

Whatwasthemajor findingofthestudy?

Public,for-pro fi t,andnon-pro fi tmatched 88%,80%,and75%(P < 0.001).Program facultyemployed,clinicianpartnership,and privateequitymatched87%,79%,and 68%(P < 0.001).

Howdoesthisimprovepopulationhealth?

Understandingfactorsformatchsuccesshelp ensurestableinputstotheemergency medicineworkforce.

positions,andtrainingsites.Wealsoobtainedalistof ACGMEprogramsthatwereformallyaccreditedbythe AmericanOsteopathicAssociation(AOA)andtheyearof earliestAOAaccreditationtype.TheACGMESiteIDfor eachprimarysitewaslinkedtothehospital’sCentersfor MedicareandMedicaidServices(CMS)Certification Numberandthe2023CMSInpatientProspectivePayment SystemFinalRuleData,whichincludesinformationabout hospitalownershiptype,urban-rurallocation,numberof hospitalbeds,resident-to-bedratio,andpercentageof disproportionatesharehospital(DSH)patients.Hospitals werelinkedtothehealthsystemsthatoperatethem. Informationaboutthephysiciangroupstaffingeach hospital’semergencydepartmentandtheownershiptypeof thosegroupsasofMarch2023wasobtainedfromIvy Clinicians.5 Wedefinedphysiciangroupsas “privateequity” iftherewasamajority-ownershipinterestbyaprivateequity firm. “Clinicianpartnerships” weredefinedasbeingmajority ownedbyphysicians.Thisincludedindependentfaculty physiciangroupsaffiliatedwithahealthsystem,equalpartnershipdemocraticgroups,groupswherecertain cliniciansmayownalargerpercentageofshares,andgroups withminority-interestownershipbyaprivateequity firm.We definedphysiciangroupsas “healthsystem” iftheywere

employeddirectlybythephysicianorganizationofthe hospital,healthsystem,medicalschool,oracademic medicalcenter.

OutcomesMeasures

Theprimaryoutcomemeasuresincludedtheproportion ofpositions filledbystateandpopulationdensity,hospital ownershiptype,andphysicianemploymentmodel. Secondaryoutcomemeasurescomparedotherprogramspecificattributesbetween filledandunfilledprograms, includingoriginalaccreditationtype,yearoforiginal accreditation,yearofACGMEaccreditation,thetotal numberofACGME-approvedtrainingpositions,lengthof training,urban-ruraldesignation,hospitalsizebynumberof beds,resident-to-bedratio,andthepercentageofDSH patientsseen.Aprogramwasclassifiedasunfilledifthere wereoneormoreunmatchedpositionsacrossanyofits NRMPIDs;programswithzerounfilledpositionsacrossany ofitsNRMPIDswereclassifiedas filled.

StatisticalMethods

Weperformedalldataextraction,transformation,and analysisusingRStudioversion2023.03.0 + 386runningR version4.2.3(RStudio,PBC,Boston,MA).Wedescribed continuousvariablesusingmediansandinterquartileranges. Categoricalvariablesweredescribedusingfrequencyand percentages.Wecomparedcontinuousvariablesusingthe Wilcoxonrank-sumtest.Wecomparedcategoricalvariables usingPearsonchi-squaredtestingwithBonferroni post-hocanalysiswheremorethantwogroupswere compared. P -valueslessthan0.05wereconsideredtobe statisticallysignificant.

RESULTS

CharacteristicsofStudySubjects

AsofMarch2023,therewere283ACGME-accredited EMresidencies;however, fiveoftheseweremilitary programsthatdonothistoricallyparticipateintheNRMP match,andthereweretwoadditionalprogramsthatdidnot participateinthe2023match.Therewere11EMprograms withdualNRMPIDs,whereoneoftheIDsmaybeusedto offerasinglepositiontoaspecialtypeofapplicant,suchas international/private-fundedpositions,researchpositions,or forthree-yearMDpathresidents.6 Atotalof276EM programsparticipatedinthematch,offering3,010positions in43statesplustheDistrictofColumbiaandPuertoRico. Therewere131programs(48%)with554positions(18%) thatwereunfilledbeforetheSOAP.

Geography

Sixstatesoffered52%ofthetotalNRMPEMpositions available:NewYork(338),California(285),Michigan(236), Florida(234),Pennsylvania(234),andTexas(184).There

wassignificantvariationinthenumberofresidencypositions availableperstatepopulation.Amongthesixstatesthat offeredthelargestnumberofresidencypositions,Michigan hadthemostNRMPpositionsperpopulationat23.5 residentspermillioncitizensinthe2020census,while Texashadonly6.1residentspermillioncitizens.Fivestates wereassociatedwithtwo-thirdsoftheunfilledpositions: Michigan(92);NewYork(83);Pennsylvania(78); Ohio(56);andFlorida(49).Therewasalsosignificant variationinthepercentageofunmatchedpositionsby state(Table1).

HospitalOwnership

Themajority(63%)ofresidencyEMpositionswere offeredby177programsatnon-profithospitals(1,880/ 3,010),while68publichospitalprogramsoffered28%of positions(831/3,010),and31for-profithospitalprograms offered10%ofpositions(299/3,010).Therewasastatistically significantdifferenceinthepercentageofunmatched positionsbyhospitalownershiptype(P < 0.001)(Table2). Publichospitalshadastatisticallysignificanthighermatch rate(88%),comparedtonon-profitandfor-profithospitals, whichhadmatchratesof80%and75%,respectively (P < 0.001).Therewasnodifferenceinmatchratesbetween non-profitandfor-profithospitals.Seventeenhealthsystems operatedthreeormoreresidencyprograms,ofwhich11were non-profit,threewerefor-profit,andtwowerepublic.The healthsystemofferingthelargestnumberofresidency programswasHCAHealthcare(19programs,189positions, 70%matchrate).

GroupOwnershipandEmploymentModel

AmongEMfacultygroupownershipandemployment models,halfofEMresidencypositions(52%)hadprogram facultythatwereemployedbyhealthsystems(1,574/3,010, 134programs),with31%havingclinicianpartnershipfaculty (941/3,010,87programs),and16%ofpositionshaving privateequity-employedfaculty(495/3,010,55programs). Fiveemployergroupsmetthedefinitionofmajorityprivate equityownership.ThesegroupsincludedAmerican PhysicianPartners,EnvisionPhysicianServices,SCP Health,SoundPhysicians,andTeamHealth.

Therewasastatisticallysignificantdifferenceinthe percentageofunmatchedpositionsbytheemployment modeloffacultyphysicians(P < 0.001)(Table3).Programs withfacultyemployedbyahealthsystemhadthehighest matchrateof87%,followedbyclinicianpartnershipsat79% andprivateequitygroupsat68%(Table3).Thirteen physiciangroupsoperatedthreeormoreresidencyprograms. Thephysiciangroupsstaffingthelargestnumberofresidency programswereEnvisionPhysiciansServices(24programs, 230positions,71%matchrate)andTeamHealth (21programs,197positions,75%matchrate).

Table1. Residencymatchresultsbystateandemergencymedicinepositionsperstatepopulation.

Alabama218194%6%5.13.5

Arizona551590%10%7.46.9

Arkansas216944%56%35.3

California242852292%8%397.3

Colorado1170100%0%5.82.9

Connecticut2370100%0%3.610.2

Delaware218667%33%117.7

DistrictofColumbia2220100%0%0.732.7

Florida222344979%21%22.210.5

Georgia558690%10%10.95.3

Illinois12144994%6%12.611.4

Indiana1210100%0%6.83.1

Iowa1100100%0%3.23.1

Kansas110460%40%2.93.4

Kentucky2250100%0%4.55.5

Louisiana4420100%0%4.69.1

Maine1100100%0%1.47.2

Maryland2230100%0%6.23.7

Massachusetts572297%3%710.3

Michigan252369261%39%1023.5

Minnesota3320100%0%5.75.6

Mississippi328968%32%2.99.5

Missouri5511178%22%6.28.3

Nebraska1120100%0%26.1

Nevada3251156%44%3.27.9

NewHampshire160100%0%1.44.3

NewJersey121222778%22%9.313.2

NewMexico1120100%0%2.15.7

NewYork313888379%21%19.719.7

NorthCarolina7852274%26%10.77.9

Ohio171585665%35%11.813.4

Oklahoma533876%24%48.2

Oregon1110100%0%4.22.6

Pennsylvania232347867%33%1318

PuertoRico216194%6%3.25

RhodeIsland222386%14%1.120.1

SouthCarolina555493%7%5.310.4

Tennessee548590%10%7.16.8

Texas151841592%8%306.1

Utah1120100%0%3.43.5

Vermont160100%0%0.69.3

(Continuedonnextpage)

Table1. Continued.

NRMP,NationalResidentMatchingProgram.

ProgramandHospital-specificAttributes

Whencomparing filledandunfilledprogramsby accreditationhistoryandhospital-levelcharacteristics, unfilledprogramsweremorelikelytobesmallerinsizebased

onthenumberofpositionsoffered(P < 0.001),previously accreditedbytheAOA(P < 0.001),andstartedinmore recentyears(P < 0.001).Therewasnodifferencein filledvs unfilledprogramsbyprogramlength(P = 0.78).Unfilled

Table2. Associationofhospitalownershiptypeonunfilledemergencymedicinepositions. Healthsystem

Byhospitalownershiptype(P < 0.001,Pearsonchi-squaredtest)

Overall,theproportionsof filled/unfilledpositionsdidvarybyhospitalownershiptype(X2 = 34.126,df = 2, P < 0.001).Post-hocBonferroni comparisonsbetweenhospitaltypesshowedthatpublichospitalshadalowerproportionofunfilledpositionscomparedtobothfor-profitand non-profithospitals(rawandadjusted P-values <0.001),whiletherewasnodifferenceintheproportionofpositions filledbetweenfor-pro fit andnon-profithospitals(raw P = 0.05,adjusted P = 0.16).

NRMP = NationalResidentMatchingProgram. UPMC,UniversityofPittsburghMedicalCenter.

Table3. 2023emergencymedicinematchratesbyfacultyphysiciangroup/type.

Physiciangroup Group type

Numberofresidency programs NRMPpositions available NRMPpositions matched Unmatched positions(%)

Byresidencyfacultyphysiciangrouptype(P < 0.001,Pearsonchi-squaredtest)

Healthsystem(HS)1341574137513%

Clinicianpartnership(CP)8794174421%

Privateequity(PE)5549533732% Total2763010245618.4%

Byresidencyfacultygroup(operating3+ EMresidencies) AmericanPhysicianPartnersPE426677% ApolloMDCP4362336% EnvisionPhysicianServicesPE2423016329%

IntegrativeEmergency Services

NorthwellHealthHS3393413%

PhysicianAffi

Overall,theproportionsof filled/unfilledpositionsdidvarybyresidencyfacultyphysiciangrouptype(X2 = 99.007,df = 2, P < 0.001).PosthocBonferronicomparisonsbetweengrouptypesshowedthatprogramswithhealthsystememployedfacultyhadthelowestproportionof unfilledpositions,followedbyclinicianpartnershipfaculty,whileresidencieswithprivateequityemployedfacultyhadthehighestproportionof unfilledpositions(rawandadjustedp-valuesforallpairwisecomparisons <0.001).

NRMP,NationalResidentMatchingProgram; UPMC,UniversityofPittsburghMedicalCenter.

programstendedtobeinlessurbanareas(P = 0.03),at hospitalswithasmallernumberofbeds(P < 0.001),lower resident-to-bedratios(P < 0.001),andfewer disproportionatesharepatients(P < 0.001)(Table4).

DISCUSSION

Weexaminethefactorsandprogramcharacteristics associatedwithunfilledpositionsintheEMmatch.Five stateswereassociatedwithtwo-thirdsoftheunfilled positions.Publichospitalshadastatisticallysignificant highermatchrate(88%)whencomparedtonon-profitand for-profithospitals,whichhadmatchratesof80%and75%, respectively(P < 0.001).Publichospitalsincludethose ownedbygovernmententities(local,state,federal government)ortheVeteransHealthAdministration.Nonprofitandfor-profithospitalsareprivatelyownedand differentiatedbytheirtaxstatus(discussedfurtherbelow). Programswithfacultyemployedbyahealthsystemhadthe

highestmatchrateof87%,followedbyclinicianpartnerships at79%andprivateequitygroupsat68%(P < 0.001overall andbetweenallsubgroups).Ouranalysisconfirmsand expands findingsfromrecentstudies.Onestudyidentifiedsix characteristicsofunfilledprograms(indescendingorderof predictivestrength):unfilledpositionsinthe2022match; smallerprogramsize;Mid-Atlanticlocation;priorAOA accreditation;EastNorthCentrallocation;andprivate equitymajorityownershipofphysicianfacultygroup.3 Anotherstudyofcombined2022and2023matchdatafound programsatriskofnot fillinghadaccreditationwithinthe prior fiveyears,hadafor-profitprimaryclinicalsite, andwereingeographicareaswithhighnumbersof positionsoffered.2

ResidencyGrowthTrends

ThenumberofunmatchedpositionsintheEMmatchwas drivenbyadramaticincreaseinthenumberofEMprograms

Table4. Comparingattributesof filled/unfilledprogramsin2023emergencymedicinematch. Filled(n = 145)Unfilled(n = 131)Total

Originalaccreditationtype

ACGME141(97%)84(64%)225(82%)

AOA4(3%)47(36%)51(19%)

Yearoforiginalaccreditation

Median199520102003

Q1,Q31982,20091993,20181988,2016

YearofACGMEaccreditation

Median199520172008

Q1,Q31982,20112006,20191990,2017

TotalapprovedACGMEpositions

Median393036

Q1,Q330,5422,3624,44

Lengthoftraining

3years116(80%)103(79%)219(80%) 4years29(20%)28(21%)57(21%)

Largeurbanarea89(61%)64(49%)153(55%)

Otherurbanarea55(38%)61(47%)116(42%)

Ruralarea1(1%)6(5%)7(3%)

Numberofhospitalbeds <0.001b

Median571359450

Q1,Q3382,730260,534318,680

Resident-to-bedratio(per100beds)

Median472938

Q1,Q330,7016,4521,63

Disproportionatesharehospitalpatients[%]

Median393336

Q1,Q331,5228,4330,47

aPearsonchi-squaredtest. bWilcoxonrank-sumtest.

ACGME,AccreditationCouncilforGraduateMedicalEducation; AOA,AmericanOsteopathicAssociation; Q,quartertile.

andpositionsofferedoverthepastdecade,aswellasamore recentdecreaseinapplicantsoverthepriortwoyears.

Between2014–2023,therewasa29%increaseinthenumber ofEMprogramsanda46%increaseinthenumberof postgraduateyear(PGY)-1positionsofferedinthematch, suggestingthatthegrowthofpositionsisnotonlyrelatedto thecreationofnewprogramsbutalsotheexpansionof existingprograms.Inrecentyears,EMhasexperiencedthe largestgrowthrateofPGY-1positionsacrossallmedical specialties.7 Thematchrateisalsoimpactedbyadecreasein thenumberofapplicantsovertime.ApplicantsinEM peakedin2021at4,391applicants.Itisunclearwhetherthis recordhigh,representinga16%increaseovertheyearbefore, wasanoutlier.Theoverall five-yeartrendisan8%decrease

inapplicantscontrastedwiththe23%increaseinpositions.8 Thisunprecedentedgrowthhasoutstrippedthenumber studentsapplyingtotraininEMandplayedalargerolein thenumberofunfilledspotsin2023.

Between2013–2020,therewassignificantgrowthofEM residenciesinstatesthatalreadyhadmultipleEMtraining programs.Anumberofstatesnearlydoubledthenumberof trainingprogramsinthattimeframe:NewYork(21to31), Pennsylvania(12to21),andCalifornia(14to22),while othersgrewevenmoreOhio(9to18),Michigan(11to25), andFlorida(5to19).9,10 Newprogramsare disproportionatelygrowinginurbanareas,whereassome ruralstatesdonothaveanyEMtrainingprograms.10 Only sevenEMresidencyprogramsarelocatedinruralareas,six

ofwhichdidnot fill.11 Ourdatademonstratesthatmanyof theunfilledspotsin2023occurredinstatesthathadthe highestabsolutenumberofresidentpositionsaswellas numberofresidentspercapitapopulation.Nostate-level regulationsexisttolimitthenumberofresidencytraining programs.WhilesomehavecalledontheACGMEtorestrict thenumberofEMtrainingpositions,itiscurrentlyagainst ACGMEpolicyandaviolationofstateandfederalantitrust lawfortheACGMEtoimplementanationalworkforce policytoestablishthenumberofpracticingphysicians.12 The ACGMEcancreateandadjuststandardsforaccreditationto optimizethelearningenvironment.Somehaveexpressed concernregardingtheacademicqualityofsomeofthenewer programs.Onestudyfoundthatnearly25%ofprograms weregiven “withwarning” accreditationoninitial accreditationcomparedtolessthan3%ofprogramson continuedaccreditation.7,13

Debateexistsoverwhoisresponsiblefortheincreased growthofresidencyprograms.Anewresidencyprogram requiresasponsoringinstitution,whichtheACGMEdefines asan “organizationorentitythatassumesultimate financial andacademicresponsibilityforaprogram.” Sponsoring institutionsmayincludeuniversities,medicalschools, hospitals,healthcaredeliverysystems,orphysiciangroup practices.14 Currently,areviewoftheACGMElistings revealsthatallEMresidencyprogramsaresponsoredby hospitalsandhealthsystems,withnonebeingsponsoredby physicianstaffinggroups. 13 Theroleandmotivationofthe physiciangroupswhoserveasfacultyfornewresidency programsthataresponsoredbyhospitalsandhealthsystems mayvary.Graduatemedicaltrainingprogramsoffer financialbenefitstohospitalsandrecruitmentbenefitsto hostinginstitutionsandstaffinggroups. 15 Newprogram growthcouldbedrivenatthephysiciangroup,hospitalor healthsystemlevel,orboth.Forexample,HCAHealthcare hasatransparentobjectivetoexpandGMEpositionsstating, “With270+ residencyandfellowshipprograms,HCA HealthcareplanstocontinuetogrowthelargestGME communityintheUnitedStates.”16 Itisreasonableto surmisethatfacultygroupsfeelpressuretostartandstaffnew programstoalignwiththehealthsystem’sintenttomaintain contracts.HospitalsthatcreatedGMEprogramsafter2015, knownas “GME-naive,” haveastrongincentivetoincrease thenumberofresidentsattheirsitewithin fiveyearsof startingbecauseCMScalculatestheirtrainingcapafterthe fifthyear.17

Unfilledspotsmayrepresentmarketforcesrightsizingthe numberandgeographicdistributionofresidencyslots, althoughthecomplexitiesofGMEfundingandtrainingcaps createregulatorybarrierstomarketcorrections.9 Unfilled positionsdonotreceiveGMEfunding,whichcouldleadto residencyclosureswithoutalternatesourcesoffunding.18 Whenanesthesiaexperiencedasimilarplightofdecreasing fillratesinthe1990s,acumulativedropof77%ofapplicants

overasix-yearperiodresultedin16%ofallanesthesia residenciesinthecountryclosingtheirdoors.11 However, marketcorrectionswillnotoccurifunfilledspotsintheinitial matcharesubsequently filledintheSOAP,whichoccursa fewdayslater.MostoftheunfilledEMpositionsinthe2022 Matchsubsequently filledintheSOAP.19 Discussion continuesonhowbesttomaintainthequalityandstabilityof theEMworkforce.1

CorporationsandGraduateMedicalEducation

Weobservedsignificantdifferencesinmatchratesby hospitalownershiptypewithpublichospitalshavingthe fewestunmatchedpositions.Non-profithospitalscontinueto makeupthemajorityofEMtrainingsites,andtherewasno statisticaldifferenceinmatchratesbetweennon-profitand for-profithospitals.Overthepast20years,therehasbeen increasedconsolidationandcorporatizationinhealthcare includingEMpracticeandtraining.20–22 Manyfearthat increasedfor-profitandinvestorsponsorshipofresidency programsmayresultinlowerqualitytrainingorthe commoditizationofGME.23,24 Whiletherehasbeen increasedscrutinyoncorporateinvestmentinhealthcareand medicaleducation,andsomestudiesonhealthorworkforce outcomesinotherspecialties,nosuchstudiesexistin EM.20,25,26 TheproportionofEMresidenciescreatedatforprofithospitalshasincreasedconsiderably.7 Priorto2016, only5%ofEMresidencyprogramshadprimarysitesatforprofithospitals(10total),comparedto30%(21/71)ofnew programsbeingbasedatfor-profithospitals.Whilehospitals arefrequentlydifferentiatedbynon-profitorfor-profit status,thisdifferentiationbasedontaxstatushaslimitations incapturingthebusinessincentivesoftheinstitution.27

Ourdatashowsthatpublichospitalswereassociatedwith thehighestmatchrates.Therewasnodifferencebetweenforprofitandnon-profithospitalswithregardtomatchrates.A priorstudysimilarlydidnot findastatisticallysignificant differentgreaterriskofnot fillingatfor-profitsites (comparedtonon-profitorgovernmentsites)butdid finda 50%greaterriskofnot fillingwhenexamining2022and2023 matchdata.2 Wedid findsignificantvariationbetween groupswithinthesametaxdesignation.Forexample,ofthe 17healthsystemsthatoperatethreeormoreprograms, TrinityHealth,anon-profithealthsystem,hadthehighest percentageofunmatchedpositionsat56%(sixprograms total,23/41unmatched)andtheUniversityofCalifornia,a publichealthsystem,had0unfilledpositions(fiveprograms total,67positions).Thehealthsystemoperatingthelargest numberofEMresidenciesisHCAHealth,afor-profithealth system,whichoffered189positionsat19programs,ofwhich 30%wereunmatched.TenetHealthcare,anotherfor-profit healthsystem,whichoffered44positionsatfourprograms, hadamatchrateabovethenationalaverage, filling91%ofits positions.Hence,althoughpublichospitalshadahigher matchrateoverall,thereissignificantvariability.

Muchscrutinyhasfocusedoncorporate,specifically privateequity(PE),ownershipandinvestmentinEM. Amongthedifferenttypesofnon-physiciancorporate investors,PEhasundergoneparticularcriticismdueto significantexpansionwithinEM,evidenceofpooroutcomes inotherareasofhealthcare,andshort-termprofit incentives.28 Privateequityandpubliclytradedcompany controloftheemergencyphysicianstaffingmarketincreased from8.6%to22%from2009to2019.26 Privateequityacquiredhospitalsnowaccountfor8%ofall nongovernmentalhospitals.29 Ourdatashowsthat503/3,010 (17%)ofEMresidencypositionsinthe2023matchwere staffedbyphysiciangroupsthataremajorityownedbyPE. Toourknowledge,therehasneverbeenanoutcomes comparisonstudybetweenemploymentmodelswithin residencytrainingprogramstopredictsuccessinpractice aftergraduation.Employmentmodelsofphysiciansare changingwithincreasedconsolidationinhealthcare. Emergencymedicine-boundstudentshaveexpressedconcern aboutcorporateinfluenceinEM,butitisuncleartherelative contributionofthisonstudentrecruitmentespeciallyinlight ofotherfactors.30 Academicfacultycanbeemployedin multipleemploymentmodelssuchasbyamedicalschool,a healthsystem,alargenationalgroup,aregionalgroup,ora singleownershipgroup.Emergencymedicineprogramswith thehighest fillratesinthematchwereassociatedwith employmentmodelsinwhichfacultyweredirectlyemployed bythehospital,healthsystem,ormedicalschool.Therewas significantvariability,however,betweenemployersand employmenttypes.

LIMITATIONS

Thisanalysishasseveralimportantlimitations.Thereare manyreasonsamedicalstudentmayrankandmatriculateat aresidencyprogram.Uniquecharacteristicsofaprogram thatmayinfluenceaparticularapplicant’sinterestandrank listwerenotcapturedforanalysis.Thenumberofapplicants interviewedandrankedbyprogramsareadditionalfactors thatimpactmatchrates,whichwerenotmeasured.Thepast twoyearsdidnotincludein-personapplicantinterviews, whichmayhavealsoimpactedmatchrates.

Additionally,therelationshipsbetweenhospitals,health systems,physicianfacultygroups,andindividualresidency programsarecomplexandevolving,andthismustbe consideredwheninterpretingresults.Forexample,one healthsystemmayemployphysiciansundermultiplemodels suchasdirectemploymentorathird-partystaffinggroup. Thecurrenthealthsystemorstaffinggroupattheprogramin thisanalysismaynothavebeenthesameonepresentwhen theresidencystartedduetomergersandacquisitions.Since thisanalysistherehavebeenmajorchangesintheemergency physicianstaffinglandscapeincludingtheclosureof AmericanPhysiciansPartnersandChapter11Bankruptcyof

Envision,whichoperatedfourand24residenciesinthe2023 EMmatch,respectively.31

Therearenocurrentlyagreedupondefinitionsfor classifyingphysician-groupownershipstructures.Thevaried spectrumofcorporateinvestor(eg,PE)ownershipstakesin EMgroupsfromminoritytowholecomplicatesthecreation ofdiscretecategories.Ourclassificationofhealthsystems wasnotabletodifferentiatebetweenthevariouscomplex relationshipsthatcomprisehealthsystems,suchasas whetherthehealthsystemphysiciangroupiswhollyowned bythehealthsystemandortheyareownedbyamedical school,academicmedicalcenter,orhospital.Most fundamentally,ownershiponlyservesasaproxyforother importantfeaturessuchasphysicianautonomyand educationalquality.

CONCLUSION

The2023matchinEMsawincreasedratesinthenumber oftrainingslotsandprogramsthatdidnot fillbeforethe SOAP.Publichospitalshadhighermatchratesthanforprofitornon-profithospitalsoverall,buttherewas significantvariabilitywithinhospitalsandhealthsystems. Residencyprogramsthatemployedacademicfacultydirectly throughthehospitalorhealthsystemwereassociatedwith highermatchrates.

AddressforCorrespondence:AngelaG.Cai,Universityof Pennsylvania,DepartmentofEmergencyMedicine,GroundFloor Ravdin,3400SpruceSt.,Philadelphia,PA19104.Email: angela. cai@pennmedicine.upenn.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Jarouetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.SchmitzGRandJarouZJ.Theemergencymedicinematch:Isthesky fallingoristhisjustgrowingpains? AnnEmergMed. 2023;82(5):608–10.

2.GettelCJ,BennettCL,RothenbergC,etal.Unfilledinemergency medicine:ananalysisofthe2022and2023matchbyprogram accreditation,ownership,andgeography. AEMEducationandTraining. 2023;7(4):e10902.

3.PreiksaitisC,KrzyzaniakS,BowersK,etal.Characteristicsof emergencymedicineresidencyprogramswithunfilledpositionsinthe 2023match. AnnEmergMed. 2023;82(5):598–607.

4.vonElmE,AltmanDG,EggerM,etal.TheStrengtheningtheReporting ofObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies. IntJSurg. 2014;12(12):1495–9.

5.AdelmanL.IvyClinicians2023StateoftheEmergencyMedicine EmployerMarket(February2023).;2023.

6.TheMatch-NationalResidentMatchingProgram.MainResidency MatchDataandReports.2023.Availableat: https://www.nrmp.org/ match-data-analytics/residency-data-reports/ AccessedAugust17,2023.

7.CarlbergDJ.Theemergencymedicineeducationalcommunityandthe supplysideoftheworkforce. WestJEmergMed. 2022;24(1):64–7.

8.AAMC.ERAS® statistics.Availableat: https://www.aamc.org/ data-reports/data/eras-statistics-data.AccessedAugust17,2023.

9.CommitteeontheGovernanceandFinancingofGraduateMedical Education,BoardonHealthCareServices,InstituteofMedicine,etal. 2014. Graduatemedicaleducationthatmeetsthenation’shealthneeds WashingtonDC:NationalAcademiesPress.

10.BennettCL,ClayCE,EspinolaJA,etal.UnitedStates2020emergency medicineresidentworkforceanalysis. AnnEmergMed. 2022;80(1):3–11.

11.HaasMRC,HopsonLR,ZinkBJ.Toobigtoofast?PotentialImplications oftherapidincreaseinemergencymedicineresidencypositions. AEM EducTrain. 2020;4(Suppl1):S13–21.

12.NascaTJandCarlsonD.Graduatemedicaleducationspecialtymixand geographicresidencyprogrammaldistribution:IsThereaRoleforthe ACGME? AMAJEthics. 2016;18(3):258–63.

13.ACGME-accreditationdatasystem(ADS).AcademicYear2022–2023. Availableat: https://apps.acgme.org/ads/Public/Reports/Report/1 AccessedJuly22,2022.

14.AccreditationCouncilforGraduateMedicalEducation. Common ProgramRequirements(Residency).2022.Availableat: https://www. acgme.org/programs-and-institutions/programs/common-programrequirements/.AccessedAugust17,2023.

15.KimKandOrnelas-DorianC.Educationasthenewbattleground.In: SchlicherNandHaddockA,eds. EmergencyMedicineAdvocacy Handbook.Irving,Texas:EmergencyMedicineResidents’ Association;2023.

16.HCAHealthcare.GraduateMedicalEducation.Availableat: https:// hcahealthcaregme.com/.AccessedAugust17,2023.

17.CaliforniaHealthCareFoundation.Expandinggraduatemedical educationinCalifornia:theroleofGME-naivehospitals.2021. Availableat: https://www.chcf.org/publication/expandinggraduate-medical-education-gme-naive-hospitals/ AccessedAugust18,2023.

18.U.S.GovernmentAccountabilityOffice.PhysicianWorkforce:Capson Medicare-FundedGraduateMedicalEducationatTeachingHospitals. Availableat: https://www.gao.gov/products/gao-21-391 AccessedAugust17,2023.

19.GutmanA.1in4emergencymedicineresidencyspotsatJefferson Health,EinsteinunfilledaheadofMatchDay,following “unprecedented”

nationaltrend.2023.Availableat: https://www.inquirer.com/health/ match-day-residency-emergency-medicine-spots-jefferson-einsteinnazareth-20230317.html.AccessedAugust17,2023.

20.CouncilonMedicalEducation. GraduateMedicalEducationandthe CorporatePracticeofMedicine.AmericanMedicalAssociation.2020. Availableat: https://councilreports.ama-assn.org/councilreports/ downloadreport?uri=/councilreports/CME_02_I_20_annotated.pdf AccessedAugust17,2023.

21.June2022ReporttotheCongress:MedicareandtheHealthCare DeliverySystem.Availableat: https://www.medpac.gov/document/ june-2022-report-to-the-congress-medicare-and-the-health-caredelivery-system/.AccessedAugust17,2023.

22.AmericanCollegeofEmergencyPhysicians.RequestforInformationon MergerEnforcement.2022.Availableat: https://www.acep.org/ globalassets/acep-response-to-ftc-and-doj-rfi-on-mergerguidelines-04.20.22.pdf.AccessedAugust17,2023.

23.KadakiaKT,ZhengJ,BruchJD,etal.Comparisonofthe financialand operationalcharacteristicsoffor-profitandnonprofithospitalsreceiving federalgraduatemedicaleducationpayments,2011-2020. JAMA. 2023;329(2):173–5.

24.CouncilonMedicalEducation. TheImpactofPrivateEquityonMedical Training.AmericaMedicalAssociation.2022.Availableat: https:// councilreports.ama-assn.org/councilreports/downloadreport?uri=/ councilreports/CME_01_I_22_final_annotated.pdf AccessedAugust17,2023.

25.MedPac.June2022ReporttotheCongress:Medicareand theHealthCareDeliverySystem.Availableat: https://www. medpac.gov/document/june-2022-report-to-the-congressmedicare-and-the-health-care-delivery-system/ AccessedAugust17,2023.

26.AdlerL,MilhauptC,ValdezS.Measuringprivateequitypenetrationand consolidationinemergencymedicineandanesthesiology. Health AffairsScholar. 2023;1(1):qxad008.

27.HerringB,GaskinD,ZareH,etal.Comparingthevalueofnonprofit hospitals’ taxexemptiontotheircommunitybenefits. Inquiry. 2018;55:46958017751970.

28.LevinsH.Theeffectofprivateequityinvestmentinhealthcare.Penn LDI.2023.Availableat: https://ldi.upenn.edu/our-work/researchupdates/the-effect-of-private-equity-investment-in-health-care/ AccessedAugust17,2023.

29.OffodileACII,CerulloM,BindalM,etal.Privateequityinvestmentsin healthcare:anoverviewofhospitalandhealthsystemleveraged buyouts,2003-17. HealthAff. 2021;40(5):719–26.

30.MurphyJA.EMresidencyapplicationsaredown.Whatdoesthismean forourfuture?Availableat: https://www.emra.org/emresident/article/ presidents-message-january-february-march-2023 AccessedAugust17,2023.

31.AdelmanL.APP’sCauseofDeath-ThePEBusinessModel.2023. Availableat: https://emworkforce.substack.com/p/apps-causeof-death-the-pe-business.AccessedNovember5,2023.

ImaginginaPandemic:HowLackofIntravenousContrast forComputedTomographyAffectsEmergency DepartmentThroughput

WayneA.Martini,MD*

ClintonE.Jokerst,MD†

NicoleHodsgon,MD* AndrejUrumov,MD*

*MayoClinicArizona,DepartmentofEmergencyMedicine,Phoenix,Arizona † MayoClinicArizona,DepartmentofRadiology,Phoenix,Arizona

SectionEditor:LeonSanchez,MD,MPH

Submissionhistory:SubmittedOctober20,2023;RevisionreceivedDecember18,2023;AcceptedDecember22,2023

ElectronicallypublishedMarch29,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18515

Introduction: Duringthecoronavirus2019pandemic,hospitalsintheUnitedStatesexperienceda shortageofcontrastagent,muchofwhichismanufacturedinChina.Asaresult,therewasasignificantly decreasedamountofintravenous(IV)contrastavailable.Wesoughttodeterminetheeffectofrestricting theuseofIVcontrastonemergencydepartment(ED)lengthofstay(LOS).

Methods: Weconductedasingle-institution,retrospectivecohortstudyonadultpatientspresentingwith abdominalpaintotheEDfromMarch7–July5,2022.Of26,122patientencountersreviewed,3,028 (11.6%)includedabdominopelvicCTwithacomplaintincluding “abdominalpain.” Weexcludedpatients withoutsideimagingandnon-EDscans.RoutineIVcontrastagentwasadministeredtoapproximately 74.6%ofpatientsbetweenMarch7–May6,2022,whenwealteredusageguidelinesduetoanationwide shortage.BetweenMay6–July5,2022,32.8%ofpatientsreceivedIVcontrastafterinstitutional recommendationsweremadetolimitcontrastuse.Wecomparedpatientdemographicsandclinical characteristicsbetweengroupswithchi-squaretestforfrequencydata.WeanalyzedEDLOSwith nonparametricWilcoxonrank-sumtestforcontinuousmeasureswithfocusbeforeandafternewED protocols.Wealsousedstatisticalprocesscontrolchartsandplottedthe1,2and3sigmacontrollimitsto visualizethevariationinEDLOSovertime.Thechartsincludetheaverage(mean)ofthedataandupper andlowercontrollimits,correspondingtothenumberofstandarddeviationsawayfromthemean.

Results: AfteruseofroutineIVcontrastwasdiscontinued,EDLOS(229.0vs212.5minutes, P =<0.001)declinedby16.5minutes(95%confidenceinterval 10, 22).

Conclusion: IntravenouscontrastaddssignificantlytoEDLOS.DecreaseduseofroutineIVcontrastin theEDacceleratestimetoCTcompletion.ApolicychangetolimitIVcontrastduringanationalshortage significantlydecreasedEDLOS.[WestJEmergMed.2024;25(3)342–344.]

INTRODUCTION

Abdominopelviccomputedtomography(CT)isroutinely orderedfromtheemergencydepartment(ED)toevaluatefor abdominalpain.1 Historically,IVcontrasthasbeenusedto highlightdifferencesbetweensofttissuesthatwould otherwiselookthesame.Intravenous(IV)contrastforCTis oftensourcedfromoverseas,andcurrentestimatesarethat

abouthalfofhospitalsintheUnitedStatesgetmostoftheir IVcontrastagentfromGEHealthcare.Muchofthecontrast dyeismanufacturedatGE’splantinShanghai.Duringthe COVID-19relatedlockdownsinChinatheplantwasclosed oroperatingatreducedcapacityforweeks.Asaresult,many hospitalshadasignificantlydecreasedsupplyofIVcontrast, whichforcedthemtodecreaseutilizationbyupto80%.

AnticipatingcontinueddeficienciesinthesupplyofIV contrast,MayoClinicArizonainMay2022initiatedcritical protocolstolimitcontrastusetopotentiallylife-threatening conditions.ThisdecreasedutilizationwithintheEDcreated auniquecircumstanceinwhichwehadtheopportunityto explorethetheoreticalbenefitofomittingIVcontrast materialfromroutineEDabdominopelvicCTtodetermine whetheritwouldsignificantlydecreaseEDlengthofstay (LOS),whichinourinstitutionwemeasureasthepatient’ s totaltimeinthedepartment.Lengthofstayisabenchmark usedbytheCentersforMedicareandMedicaidServicesasa hospitalqualitymetric.2 Additionally,shorteneddurationof LOShasbeenshowntodecreasetheratethatpatientsleave againstmedicaladvice,whileincreasingpatientsatisfaction, andpotentiallyimprovingtreatmentoutcomes.3

METHODS

Priortothecontrastshortagealert,theIVcontrastagent iohexolwasroutinelyadministeredtoEDpatientsin conjunctionwithCTexaminationsoftheabdomenand pelvis.StartingMay6,2022,ourEDincollaborationwith theradiologydepartmentagreedtodiscontinueIVcontrast materialforroutineCTexceptintwospecificscenarios: patientsrequiringabdominalimagingwhohadabodymass index(BMI) <25;andpatientswithaBMI >25inwhom therewasanacute,time-dependentconcernthatrequiredIV contrasttofurtherdiagnose.

Wedesignatedthe “beforeinterventionperiod” asthe 60dayspriortoMay6,andthe “afterinterventionperiod” as the60daysafterMay6.Sincethestudywasfocusedon processratherthanpatients,thenormalrequirementfor institutionalreviewboardoversightwaswaived.Weincluded inthestudypatientswhopresentedtotheEDwithabdominal painandunderwentabdominopelvicCTatthediscretionof thetreatingattendingemergencyphysician.Theprimary outcomewasEDLOS,whichwasdefinedasthelengthoftime betweenwhenthepatientregisteredforcareintheEDandthe timeofEDdisposition(admitordischargetime).

Medianandinterquartilerange(IQR)valueswere expressedforallcontinuousmeasuresbetweengroups (beforevsafterperiods).Wecomparedpatientdemographics andclinicalcharacteristicsbetweengroupswithchi-square testforfrequencydataandnonparametricWilcoxonranksumtestforcontinuousmeasures.Theprimaryoutcomewas EDLOS.Weanalyzeddatausingstatisticalprocesscontrol charts(with1,2and3sigmacontrollimits),andweadjusted confidencelimitsusinganXmRchart,whichhelpsto determinehowaprocesschangesovertime.TheXmR controlchartisrecommendedforLOSandreal-worldED operationaldata.4 Controlchartswererunforallscansand separatedoutbycontrastadministrationforbothtime periods. P -values <0.05wereconsideredstatistically significant.WeusedRversion4.1.2ggQCpackage (RStudio,Boston,MA)forstatisticalanalysis.

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Intravenouscontrastisusedtohighlight differencesbetweensofttissuesthatwould otherwiselookthesame.

Whatwastheresearchquestion?

Doesdecreaseduseofcontrastforcomputed tomography(CT)improveEDlengthof stay(LOS)?

Whatwasthemajor findingofthestudy?

IfthereisashortageofIVcontrastforCT, usingcontrastonfewerpatientsmayimprove patientthroughputduetoshortenedLOS.

Howdoesthisimprovepopulationhealth?

IfthereisashortageofIVcontrastforCT, usinglesscontrastmayimprovepatient throughputduetoshortenedLOS.

RESULTS

Therewere26,122patientencounterswithinthestudy period,ofwhich3,028(11.6%)metthestudycriteria: complaintattriageofabdominalpain;age >18;and indicationsforCTexclusiveofureterolithiasis.Medianage was60years(IQR40–72).Followingprotocolchange,there wasa41.8%absolutedecreaseinabdominopelvicCTstudies thatusedIVcontrast:74.6%(1,120/1,502)beforevs32.8% (500/11,526)after; P < 0.001).Therewasalsoa16.5-minute decreaseinLOS(95%confidenceinterval 10, 22)from 229.0vs212.5minutes(Table).

DISCUSSION

Webelievethatradiologycansignificantlyimpactpatient throughput.5 Our findingssuggestthatdecreaseduseofIV contrastinnon-essentialimagingoftheabdomenandpelvis isassociatedwithadecreaseinEDLOS,therebyimproving EDthroughput.While16.5minutesmayseemlikeabrief lengthoftime,inthispatientsamplingitreducedLOSby about7.2%(229vs212.5minutes)andreducedaggregate LOSbyacombinedtotalof420hoursoverthecourseofnine weeks.Thistimesavingsmultipliedbythemillionsof patientswhopresenttotheEDannuallyforabdominalpain cantranslateintoalargemagnitudeoftimesaved,further decreasingthestrainontheEDandpotentiallyimproving patientsatisfaction.6 Asourstudywasperformedatan institutionwithhighnursingstafflevels(2–3patientsper nurse)andtechratios(sixpatientspertech),thereby optimizingtimetoIVaccessandkidneyfunctiontestresults,

Table. Beforevsafterrestrictionsonuseofintravenouscontrastforabdominal/pelviccomputedtomography,demonstratingtheimpacton lengthofstayintheemergencydepartment.

Before(n = 1,502)After(n = 1,526)Total(N = 3,028) P-value

TotalLOS(minutes) <0.001

Mean(SD)239.6(89.2)226.3(96.3)232.9(93.1) Median229.0212.5220.0 Q1,Q3176.0,293.0156.3,281.0167.0,287.0

Contrastreceived <0.001

No382(25.4%)1026(67.2%)1408(46.5%)

Yes1120(74.6%)500(32.8%)1620(53.5%)

LOS,lengthofstay; Q,quartile; SD,standarddeviation.

wehypothesizetherewouldbeevenmorepronounced improvementinLOSatfacilitiesthatareshortstaffed. Additionally,discontinuationofcontrastcanhelpto reduceincidenceofneedforIV-lineplacement,andtheriskfor allergy/anaphylaxis.Inconversationswiththeradioloy department,theradiologistsemphasizedthattheyfeltmore confidenceintheaccuracyoftheirdiagnoseswiththeuseof contrastandthatnon-urgent findingssuchascarcinoma wouldmorelikelybemissedwithoutcontrast.Theysuggested thatreduceduseofIVcontrastwouldbeappropriatein settingswhereartificialintelligencehasimprovedpathology recognitionorintheeventofanothershortageofcontrast agent.Moreresearchwillbeneededtoinvestigatetheclinical effectofdiscontinuingIVcontrastinthissetting.

LIMITATIONS

Thereareseverallimitationstoourstudy.Afterreviewing thedataatothersites(MayoClinicRochester,MayoClinic Florida,andMayoClinicHealthSystem)wedecidedtomake thisasingle-centerstudyasothersiteswerenotaffectedinthe samewaybytheshortage,andtheyhadamoregradual rolloutofIVcontrastrestrictions.Whilewenotedareduction inLOS,wewereunabletoclearlyparseoutwhetheritresulted fromdecreasedneedforIVaccessandlabresults,ordecreased timeinradiologydepartment.Additionally,ourstudy encompassedalimitedtimeframeofonlyabout60days,after whichIVcontrastagentbecamemoreavailable.Lastly,more researchisneededtofurtheranalyzethepotentialneedfor repeatimagingorpossiblereturnvisitstotheEDasaresultof notusingIVcontrast.

CONCLUSION

Inthissingle-centerstudy,wefoundthataninstitutional policychangereducingtheuseofcontrastinabdominalpelvicCTduringtheCOVID-19pandemicwassignificantly associatedwithshorterlengthofstayintheED.

AddressforCorrespondence:WayneA.Martini,MD,MayoClinic ArizonaDepartmentofEmergencyMedicine,5777E.MayoBlvd., Phoenix,AZ85054.Email: Martini.Wayne@Mayo.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbe perceivedaspotentialsourcesofbias.Noauthorhasprofessional or financialrelationshipswithanycompaniesthatarerelevantto thisstudy.Therearenoconflictsofinterestorsourcesoffunding todeclare.

Copyright:©2024Martinietal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.FishmanEK.SpiralCT:applicationsintheemergencypatient. Radiographics. 1996;16(4):943–8.

2.ChangAM,LinA,FuR,etal.Associationsofemergencydepartment lengthofstaywithpubliclyreportedquality-of-caremeasures. Acad EmergMed. 2017;24(2):246–50.

3.GarnderRM,FriedmanNA,CarlsonM,etal.Impactofrevisedtriageto improvethroughputinanEDwithlimitedtraditionalfasttrackpopulation. AmJEmergMed. 2018;36(1):124–7.

4.PimentelLandBarruetoF.Statisticalprocesscontrol:separatingsignal fromnoiseinemergencydepartmentoperations. JEmergMed. 2015;48(5):628–38.

5.SweenyA,KeijzersG,O’DwyerJ,etal.Predictorsofalonglengthof stayintheemergencydepartmentforolderpeople. InternMedJ. 2020;50(5):572–81.

6.DeSteenwinkelM,HaagsmaJA,vanBerkelECM,etal.Patient satisfaction,needs,andpreferencesconcerning informationdispensationattheemergencydepartment:a cross-sectionalobservationalstudy. IntJEmergMed. 2022;15(1):5.

TheUtilityofDotPhrasesandSmartPhrasesinImproving PhysicianDocumentationofInterpreterUse

KatrinJaradeh,MD*†

ElaineHsiang,MD*†

MaliniK.Singh,MD,MPH,MBA*†

ChristopherR.Peabody,MD,MPH*†

StevenStraube,MD,MBA*†

SectionEditor:MatthewTews,MD

*DepartmentofEmergencyMedicine,UniversityofCalifornia, SanFrancisco,California

† SanFranciscoGeneralHospital,SanFrancisco,California

Submissionhistory:SubmittedJune6,2023;RevisionreceivedDecember8,2023;AcceptedDecember14,2023

ElectronicallypublishedApril2,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.18352

Background: PatientswithlimitedEnglishproficiency(LEP)experiencesignificanthealthcare disparities.Cliniciansareresponsibleforusinganddocumentingtheiruseofcertifiedinterpretersfor patientencounterswhenappropriate.However,thedataoninterpreterusedocumentationinthe emergencydepartment(ED)islimitedandvariable.Wesoughttoassesstheeffectsofdotphraseand SmartPhraseimplementationinanadultEDontheratesofdocumentationofinterpreteruse.

Methods: Weconductedananonymoussurveyaskingemergencyclinicianstoself-report documentationofinterpreteruse.WealsoretrospectivelyrevieweddocumentationofinterpreterservicesuseinEDchartsatthreetimepoints:1)pre-interventionbaseline;2)post-implementationofa clinician-drivendotphraseshortcut;and3)post-implementationofaSmartPhrase.

Results: Mostemergencycliniciansreportedusinganinterpreter “almostalways” or “often.” Ourmanual auditrevealedthatatbaseline,interpreterusewasdocumentedin35%oftheinitialcliniciannote,4%of reassessments,and0%ofprocedurenotes;52%ofdischargeinstructionswerewritteninthepatients’ preferredlanguages.AfterimplementationofthedotphraseandSmartPhrase,respectively,ratesof interpreter-usedocumentationimprovedto43%and97%ofinitialcliniciannotes,9%and6%of reassessments,and5%and35%ofprocedurenotes,with62%and64%ofdischargeinstructions writteninthepatients’ preferredlanguages.

Conclusion: Therewasadiscrepancybetweenreportedratesofinterpreteruseandinterpreter-use documentationrates.Thelatterincreasedwiththeimplementationofaclinician-drivendotphraseand thenaSmartPhrasebuiltintothenotes.Ensuringaccuratedocumentationofinterpreteruseisan impactfulstepinlanguageequityforLEPpatients.[WestJEmergMed.2024;25(3)345–349.]

INTRODUCTION

Asof2019,over65millionpeopleintheUnitedStates (US)speakalanguageotherthanEnglish,with approximately20%ofhouseholdsreportingspeaking Englishlessthan “verywell,” alsoknownaslimited-English proficiency(LEP).1 IntheUS,presidentialExecutiveOrder 13166,enactedin2000,ensuresthatLEPpatientsareoffered interpretationservicesathealthcarefacilitiesreceiving federalassistance.2,3 Thelackofaccesstolanguage-

concordantcarecontributestohealthcaredisparitiesamong LEPpatients.4

Intheemergencydepartment(ED),LEPpatientswere morelikelytohaveunplannedrevisitswithin72hours5 with limitedevidencesuggestingdifferencesintriageoradmission decisionsdependingoninterpreteruse.6 Recentdata demonstratesincreasedunnecessarytestingandhospital admissionwithlongerlengthsofstayamongLEPpatients whodidnotreceiveprofessionalinterpretingservices.7,8

Documentationofinterpreteruseisoftenusedasa proxyforinterpreteruse.Severalgroupsofresearchers studiedtherateofinterpreter-usedocumentationinthe hospital.Onefoundthat41%(30/74)ofpatientshada consentformintheirnativelanguageorthataninterpreter hadsignedit.8

Interventionshavebeenimplementedtoimprove documentationofinterpreteruse.Benderetalfoundthat whentheyplaced flyersintheEDandmadepre-work shiftannouncements,documentationofinterpreteruse increasedfromabaselinerateof5%to25%.9 In2021,astudy amongpatientsadmittedtoapediatricservicefoundthat usingadotphraseincreasedinterpreterusefrom64%to81%, andinterpreter-usedocumentationincreasedfrom69%to 98%.10 Toourknowledge,therehavebeennostudies investigatingtheuseofadotphrase(textinsertedwith keyboardshortcuts)oraSmartPhrase(abbreviationsor wordsusedtopulllongphrasesintoaphysician’snote)inan adultEDtoimprovedocumentationofinterpreteruse.We assessedtheeffectsofadotphraseandaSmartPhraseinan adultEDontheratesofdocumentationofinterpreteruse. Wehypothesizedtheseinterventionswouldincrease documentationrates.

METHODS

WeconductedthisstudyataLevelIacademictrauma centerinanadultED,whereinterpretersareavailableover thephone24/7andinpersonduringdesignatedhours.First, wegatheredpatients’ medicalrecordnumbers(MRN)from interpreterservicesthatdocumentedaninterpreterhadbeen used.Apre-interventionretrospectivechartreviewwas conductedtoassessthebaselinerateofinterpreter-use documentationintheelectronichealthrecord(EHR). Second,wesurveyedemergencyclinicianstoassesstheir perspectiveoninterpreteruseanddocumentation.Third,we implementedadotphraseandthenaSmartPhraseand retrospectivelyreviewedchartsfordocumentationof interpreteruse.Bothinstrumentswerebeingdevelopedatthe sametime,butthedotphrasewascompletedmorequickly andimplemented first.Documentationofinterpreterusewas capturedwithinthehistoryandphysical(H&P), reassessments,procedurenotes,anddischargeinstructions (DCI),whichincludesaverbaldiscussion,written instructions,andattachments.Weexcludedchartsfromthe studyifthepatientonlyspokeorpreferredtospeakin English,leftwithoutbeingseen,MRNswerenotfound,orif ititwasaduplicaterecord.Therewasnopriortrainingon documentationofinterpreteruse.Weanalyzeddatausing descriptivestatistics.Thisstudywasdeemedexemptbyour institutionalreviewboard.

Pre-Intervention

WeverifiedMRNsfromtheinterpreterservicedata intheEHR.Anumbergeneratorwasusedtorandomize

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

PatientswithlimitedEnglishpro fi ciency experiencehealthcaredisparities.Using interpretersreducesunnecessarytestingand hospitalizationsforthispopulation.

Whatwastheresearchquestion?

Doesimplementingadotphraseand SmartPhraseincreasedocumentationof interpreteruse?

Whatwasthemajor findingofthestudy?

Documentationofinterpreteruseinthe historyandphysicalrosefrom35%to 43%(dotphrase)andthento 97%(SmartPhrase).

Howdoesthisimprovepopulationhealth?

Aninterventiontoimprovedocumentationof interpreterusehelpsensurelanguageequity forlimitedEnglishpro fi ciencypatients.

andidentifypatientsforchartreview.Tominimize clinician-specificpracticepatterns,weauditedone chartperdayfromJuly–September2021fromvarious shifttimestoestimatethepre-interventionrateofinterpreter usedocumentation.

ClinicianSurvey

Weemailedananonymoussurveyto128EDattendings, fellows,residents,andnursepractitionersregarding interpreter-usedocumentationafterthepre-intervention datawascollected.Onefollow-upemailwassent.Wecreated asurveyof14multiple-choicequestionshostedonQualtrics (Qualtrics,Provo,UT).Thesurveyincludeddemographics, questionsaboutinterpreteruse,documentation,andwaysto improvedocumentation.

DotPhrase

Adotphraseisablockoftextinsertedusingakeyboard shortcutproceededbyadotthatfacilitatesclinician’ s documentation.Clinicianscaninputthephrase “.EDinterpreter” forthestatement “A[phone,in-person] [languageoptions]interpreterwasusedon[dateandtime], [INTERPRETERID#]” tobeaddedintheEHR.Thedot phrasewasavailableonJuly1,2022.Allchartsfrom interpreterservicesdatawereauditedbetween July1–October14,2022.

SmartPhrase

WeembeddedaSmartPhraseintotheH&Pandprocedure notescreatinga “hardstop” whereclinicianscouldnotsign theirnotesuntiltheSmartPhrasewascompleted.Thiscould bebypassedbydeletingtheSmartPhrase.Ifanon-English languagewasselected,theSmartPhrasewouldpromptto choosethepatient’spreferredlanguage.TheSmartPhrase wasavailableonNovember1,2022.Allchartsfrom interpreterservicesdatawereauditedbetween November1–February1,2023.

RESULTS

Pre-Intervention

Of91auditedcharts,Spanish(61%)wasthemost preferredlanguage,followedbyCantonese/Mandarin/ Taishanese(37%),andRussian(2%).Useofaninterpreter wasdocumentedin35%ofH&Ps,4%ofreassessments,and in0%ofprocedurenotes.Withinthedischargeinstructions, 6%ofchartsindicateddiscussinginstructionsusingan interpreter;52%ofwrittenDCIsand89%ofattachments wereprovidedinthepatient’snativelanguage(Figure1).

ClinicianSurvey

Of128emergencyclinicianswhoreceivedthesurvey,67 (52%)initiatedand65(51%)completedit.Ofthe respondents,46%wereresidents,37%attendings,9%NPs, and8%fellows.Cliniciansreporteduseofaninterpreter “almostalways” or “often” 66%and25%ofthetimewhen interactingwithLEPpatients.Additionally,23%and8%of cliniciansreported “almostalways” or “often” documenting useofaninterpreterintheH&P(Figure2a).Clinicians reported “almostalways” documentinguseofaninterpreter

inthereassessment(3%),procedure(15%),andDCI(8%) portionsofthenote(Figure2b, 2c, 2d).Whenaskedwhatcan makedocumentationeasier,41%suggestedadditionstothe EDnotetemplatewith29%recommendingadotphrase.

DotPhrase

Of866auditedcharts,weanalyzed809(93%).Spanish (67%)wasthemostpreferredlanguage,followedby Cantonese/Mandarin/Taishanese(32%),andRussian(1%). Forty-threepercentofH&Ps,9%ofreassessments,and5%of procedurenoteshaddocumentationofinterpreteruse. Documentationofinterpreteruseduringdischargeremained at6%.ThewrittenportionandattachmentsoftheDCIwere inthepatient’snativelanguagein62%and94%ofcharts.

SmartPhrase

Of779auditedcharts,weanalyzed646(83%).Spanish (64%)wasthemostpreferredlanguage,followedby Cantonese/Mandarin/Taishanese(35%),andRussian (0.62%).Ninety-sevenpercentofH&P,6%ofthe reassessments,and35%ofprocedurenoteshad documentationofinterpreteruse.RegardingtheverbalDCI, 4%documentedinterpreteruse.Thewrittenportionand attachmentswereinthepatient’snativelanguagein64%and 94%ofcharts(Figure1).

DISCUSSION

Documentationratesofinterpreteruseincreasedafter implementationofadotphraseandaSmartPhrase.After implementingtheSmartPhrase,almost100%oftheH&Ps and35%ofprocedurenotesdocumentedinterpreteruse. BecausetheSmartPhrasewasembeddedonlyinH&Psand

Figure1. Percentageofpatientchartswithdocumentationofinterpreteruseatbaseline(blue),afterthecreationofthedotphrase(orange), andafterthecreationoftheSmartPhrase(gray).

H&P,historyandphysical; DCI,dischargeinstructions; DP,dotphrase; SP,SmartPhraseandprocedurenoteimplementation.

Figure2. (a)Emergencyclinicians’ perspectiveondocumentationofinterpreteruseinthehistoryandphysical.(b)Emergencyclinicians’ perspectiveondocumentationofinterpreteruseinthereassessment.(c)Emergencyclinicians’ perspectiveondocumentationofinterpreter useintheprocedurenote.(d)Emergencyclinicians’ perspectiveondocumentationofinterpreteruseintheDCI. H&P,historyandphysical; DCI,dischargeinstructions.

procedurenotes,wedidnotexpectincreasesintheDCIand reassessments.Thegeneralratesofinterpreter-use documentationinthisstudyandpreviousstudiesvary. Behairyetalfoundthatattheirchildren’shospital documentationofinterpreterusewas0%,11 whereasTaira etalfounddocumentationofinterpreteruseintheirpublic EDtobe4.6%.12 Toourknowledge,thisisoneofthe firststudiesontheimpactofadotphraseandaSmartPhrase ondocumentationofinterpreteruseinan adultED.

Therewasadiscrepancybetweenreportedratesof interpreteruseanddocumentationofinterpreteruse.Despite 66%ofcliniciansreporting “almostalways” usingan interpreter,only23%reported “almostalways” documentingtheiruseintheH&P.Thesamediscrepancy wasseenamongreassessments(3%),procedurenotes(35%), andDCIs(8%)wherecliniciansreportedthey “almost always” documentedtheiruse.Whilewedidnotspecifically askclinicianswhentheyuseaninterpreter(whilegathering theH&P,etc,settingdocumentationasaproxyfor interpreteruse,manycliniciansspeakingtotheirpatients withaninterpreterwouldnothavethedocumentationto supporttheirclaim.Lastly,cliniciansmayhaveusedanad hocinterpreter(familymemberoramemberofthe healthcareteam),asthesurveydidnotspecifyuseof professionalinterpretation.Thismayaccountforsome

ofthediscrepancybetweenthereportedandactualratesof interpreteruseperinterpreterservicesdata.

Next,wehopetoassesstheimpactofimproved documentationonpatientcare.

LIMITATIONS

Thiswasasingle-institutionstudyandresultsmaynotbe generalizable.Variabilityindocumentationamong emergencyclinicians,andintimeanddayofshiftwerenot captured.Sinceonlyonechartperdaywasreviewedforpreinterventiondata,documentationratesmayhavebeenmore influencedbytimeofdaythanpost-interventionrates, affectingthedifferencesinpre-/post-interventionchanges. WedidnottrackthedataofdotphraseandSmartPhraseuse. Further,despitetheSmartPhraseleadingtoa “hardstop,” clinicianscoulddeletetheSmartPhrase.However,we includedboththedotphraseandSmartPhraseas interventionssinceclinicianscouldaddthedotphrase intootherelementsoftheEHRwhentheyusedan interpreter(eg,reassessments).

CONCLUSION

Documentationofinterpreteruseisvaried.Therewasa discrepancybetweenreportedratesofinterpreteruseand interpreter-usedocumentation.Implementationofadot phraseandaSmartPhraseimproveddocumentationof

interpreteruse,suggestingitsfeasibilitytoimprove cliniciandocumentation.

AddressforCorrespondence:KatrinJaradeh,MD,Universityof CaliforniaSanFrancisco,DepartmentofEmergencyMedicine,533 ParnassusAve.,SanFrancisco,CA94143.Email: katrin.jaradeh@ ucsf.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.ChristopherPeabodyworksasa consultantinunrelatedcapacitiesforFujiFilmandSonoSite.There arenootherconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Jaradehetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.UnitedStatesCensusBureau.2022.WhatLanguagesDoestheUnited StatesSpeak?AGeographicAnalysisoftheLanguagesSpokenat HomeintheUnitedStates,2015–2019.Availableat: https://www. census.gov/library/visualizations/2022/demo/lang-geogr-distrob.html AccessedOctober27,2022.

2.U.S.DepartmentofJusticeCivilRightsDivision.2015.ExecutiveOrder 13166.Availableat: https://www.justice.gov/crt/executive-order-13166 AccessedApril21,2023.

3.U.S.DepartmentofJusticeCivilRightsDivision.2015.TitleVIofthe CivilRightsActof196442U.S.C.§2000dEtSeq.Availableat: https:// www.justice.gov/crt/fcs/titlevi-overview.AccessedApril21,2023.

4.Eneriz-WiemerM,SandersLM,BarrDA,etal.ParentallimitedEnglish proficiencyandhealthoutcomesforchildrenwithspecial healthcareneeds:asystematicreview. AcadPediatr 2014;14(2):128–36.

5.NgaiKM,GrudzenCR,LeeR,etal.Theassociationbetweenlimited Englishproficiencyandunplannedemergencydepartmentrevisitwithin 72hours. AnnEmergMed.2016;68(2):213–21.

6.MarteD,BairdJ,BristerJ,etal.240Evaluatingtheimpactoflimited Englishproficiencyoncaredeliveryinapediatricemergency department. AnnEmergMed.2019;74(4):S94.

7.LindholmM,HargravesJL,FergusonWJ,etal.Professionallanguage interpretationandinpatientlengthofstayandreadmissionrates. JGenInternMed.2012;27(10):1294–9.

8.SchenkerY,WangF,SeligSJ,etal.Theimpactoflanguage barriersondocumentationofinformedconsentatahospital withon-siteinterpreterservices. JGenInternMed 2007;22(2):294–9.

9.BenderD.(2021) ImprovingInterpreterDocumentationinthe EmergencyDepartment.RutgersUniversity-Schoolof Nursing – RBHS.

10.RajbhandariP,KeithMF,BraidyR,etal.Interpreteruseforlimited Englishproficiencypatients/families:aQIstudy. HospPediatr 2021;11(7):718–26.

11.BehairyM,AlencheryA,Cuesta-FerrinoC,etal.Increasing languageinterpreterservicesuseanddocumentation: aqualityimprovementproject. JHealthcQualJHQ 2023;45(1):19.

12.TairaBRandOrueA.LanguageassistanceforlimitedEnglish proficiencypatientsinapublicED:determiningtheunmetneed. BMCHealthServRes.2019;19(1):56.

BestPracticesforTreatingBlindandVisuallyImpairedPatientsin theEmergencyDepartment:AScopingReview

KareemHamadah,BS*‡

MaryVelagapudi,DO,BSN†‡

JulianaJ.Navarro,MD,MPH†‡

AndrewPirotte,MD†‡§¶

ChrisObersteadt,EMT†‡#

*UniversityofKansasSchoolofMedicine,KansasCity,Kansas

† UniversityofKansasHealthSystem,KansasCity,Kansas

‡ UniversityofKansasMedicalCenter,KansasCity,Kansas

§ UniversityofKansasMedicalCenter,DepartmentofEmergencyMedicine, KansasCity,Kansas

¶ UniversityofKansasMedicalCenter,OfficeofStudentAffairs, KansasCity,Kansas

# RockhurstUniversity,KansasCity,Kansas

SectionEditor:MuhammadWaseem,MD

Submissionhistory:SubmittedAugust8,2023;RevisionreceivedNovember21,2023;AcceptedJanuary19,2024

ElectronicallypublishedMay3,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.61686

Introduction: Blindandvisuallyimpairedindividuals,anunder-representedpopulationofthe emergencydepartment(ED),possesscomorbiditiesandhaveahigherchanceofin-hospitalsequelae, includingfalls.Thispotentiallyvulnerablepopulation,ifnottreatedmindfully,canbesubjecttodecreased qualityofcare,recurrentand/orlongerhospitalizations,persistenceofhealthissues,increased incidenceoffalls,andhigherhealthcarecosts.Forthesereasons,itiscrucialtoimplementholistic practicesandtrainclinicianstotreatblindandvisuallyimpairedpatientsintheEDsetting.

Methods: Weidentifiedandusedacomprehensivearticledescribingbestpracticesforthecareofblind andvisuallyimpairedpatientstoestablishtheED-specificrecommendationspresentedinthispaper. AscopingreviewoftheliteraturewasthenperformedusingPubMedtoidentifyadditionalarticlesto supporteachrecommendation.Toensurethatrecommendationscouldbeimplementedina representative,scalable,andsustainablemanner,weconsultedanadvocatefortheblindtohelprefine andprovideadditionalsuggestions.

Results: Weidentified14recommendationsthatfocusoncommunicationstrategies,EDresource access,andcontinuityofcare.Themainrecommendationisforthecliniciantosupporttheunique healthcareneedsofthevisuallyimpairedindividualandmaintainthepatient’sautonomy.Another recommendationistheconsistentuseofassistivedevices(eg,canes,guidedogs)toaidpatientsto safelyambulateintheED.Alsoidentifiedasbestpracticesweredischargeeducationwiththeuseofa screenreaderandtimelyfollow-upwithaprimarycarephysician.

Conclusion: Whilewesummarizeavarietyofrecommendationsinthisarticle,itisimportantto implementonlythestrategiesthatworkbestforthepatients,personnel,andenvironmentspecifictoyour ED.Afterimplementation,itisvitaltorefine(asfrequentlyasneeded)theinterventionstooptimizethe strategies.Thiswillenabletheprovisionofexceptionalandequalcaretoblindandvisuallyimpaired patientsintheED.[WestJEmergMed.2024;25(3)350–357.]

INTRODUCTION

Theblindandvisuallyimpaired(VI)areasmallbuthighly marginalizedpopulationintheUnitedStatesandaroundthe world.1,2 ThereareapproximatelyninemillionVIpeoplein theUS,withblindpeoplemakingupslightlylessthan1%of thepopulation.Globally,about2%ofchildrenare consideredVI.3 EstimatesintheUSareexpectedtodouble by2050,withtheVIpopulationprojectedtobemore prevalentinracialminoritiesandinthesouthernUS.The reasonforthisincreaseismultifactorialbutmaybeduetoan ever-increasingagingpopulationanddifferentialaccessto preventativeservicesamongminoritygroups.4,5 Datais limitedontheexactnumberofVIpatientswhoareseeninthe emergencydepartment(ED).However,approximately0.2% ofpatientsadmittedtoUShospitalsareconsideredVI.6 StatesthathavenotexpandedMedicaidcoverageseehigher ratesofVIpatientsintheirEDs.7 Patientswithother disabilities,suchasthoseinthedeafandhard-of-hearing (DHH)community,arealsomorelikelytoseekEDcare comparedtothosenotintheDHHcommunity.8–12

AlthoughVIpatientsrepresentarelativelysmall proportionofpatientsseenintheEDandadmittedtothe hospital,theyhavesignificantlyworseoutcomes:Theyare admittedandreadmittedmoreoften,incurhigherhealthcare costs,andmayhaveahigherin-hospitalmortalityrate.6,13,14 Visuallyimpairedpatientsaremorelikelytoexperience multiplemorbidities,thusfurtherincreasingtheirriskof needingEDcare.15 Fallsandtheirsequelae,suchaship fractures,areamongthemostcommonreasonsforblind patientstobeseenintheED.16–18 PediatricVIpatientsare particularlylikelytoincurorthopedicinjuriesandarealso morelikelytohavefracturesuponpresentation.19 HospitalizedpatientswhoareVIarealsomorelikelyto experiencedelirium,20 awell-knownriskfactorformorbidity andmortality.21 TheseinjuriesandconditionsamongtheVI occurinothercountries22 andtootherdisabledgroups.12,23 IntheUS,theseissuesarefurthercompoundedbythe intersectionsofraceandage.2 Blackpatientsandpatients insuredbyMedicare(ie,those ≥65yearsold)arethemost likelytohaveextendedhospitalstays.13

OptimalcareforallpatientsintheEDremainsanongoing challenge;careofVIpatientspresentsuniquechallengesthat offeranumberofopportunities.Amindfulapproachtocare ofVIpatientsrequiresthatEDsandclinicianspursuebest practices,supportstaff,impactfuleducation,andspecialized considerations.Aswithmanypopulations,theneedsofVI patientsimpacttheirexperienceduringEDcare.Inthis article,wepresentbestpracticeconsiderations.Thisscoping reviewisintendedtopromptimprovedpracticeandto furtherdiscussiontooptimizeEDcareforVIpatients.

METHODS

WeperformedascopingreviewtoidentifyPubMed articlesrelatedtoblindnessandVIintheED,withparticular

emphasisonassessingtheexperiencesofVIpatientsinthe ED.Articleswereincludediftheymetoneormoreofthe followingcriteria:centeredontheexperiencesofdisabled people,particularlyVIpeople,inhealthcare;discussedthe experiencesorepidemiologyofdisabledpatientsinEDsor hospitalsregardlessofgeographicregion;providedbest practicerecommendationsforthecareofVIpatients regardlessofspecialty;anddiscussedoutcomesofdisabled patientsintheEDorhospital.Weexcludedarticles discussingthecareofacuteblindnessorVI,asthefocusof thisreviewwasonpatientswithpre-existingvisual impairment.Duetotheoveralllackofdataonthistopic, guidelinesfromotherspecialties(eg,ophthalmology)were includedandadaptedtotheEDsetting.

WeusedtheAmericanswithDisabilitiesAct(ADA) ChecklistbyMarshallandJoffee(2006)asthebasisofour recommendations,asitprovidesacomprehensivelistofbest practicesforallhealthcareclinicians.Fromthispaper,we selected15recommendationsmostrelevantandapplicableto theEDsetting(Table).Recommendationswere supplementedusingfocusgroupandsurveydatafoundon PubMed.Thesearchphrase “(visuallyimpaired)AND (accessibility))AND(emergencydepartment)” resultedin28 results.WefoundonerelevantstudybyCarmichaeletal (2023),inwhich12disabledindividualswereinterviewed(six ofwhomwereVI).DuetoalackofdataspecifictoVI patients,thesearchwasexpandedtoincludetheexperiences ofpatientswithotherphysicalandcognitivedisabilities, whichyieldedanadditionalstudybyMorrisetal(2021).

WeevaluatedtrendsinEDuseamongdisabledpatientsto contextualizetherecommendationsprovided.Finally,we usedarticlesbytheNationalFederationoftheBlind(NFB) toensurethatthevoicesofVIauthorsandacademicswere wellrepresentedandtoinformseveralrecommendations (eg,language).Mostofthedatawereobservationaland retrospective.Wealsoconsultedasubjectmatterexpertwho wasbornblindanddedicatedhercareertoadvocatingfor otherVIpeopletoensurethatwewerebestrepresentingthe needsofVIpatients.Usingthisdata,weidentifiedactionable recommendationsandbestpractices.

RESULTS

WeperformedPubMedsearchestoidentifysupporting articlesforall14recommendations(see Table).Articleswere selectedusingthepreviouslydescribedinclusionand exclusioncriteria.ExcludingtheADAChecklistby MarshallandJoffee,whichwasusedtodevelop eachrecommendation,wefoundfourarticlessupporting recommendationone.Threearticleswerefoundsupporting recommendationtwo.Fivearticleswerefound supportingrecommendationthree,andonearticlewas foundsupportingrecommendationfour.Wefoundthree articlessupportingrecommendation five,sixarticles supportingrecommendationsix,andfourarticlessupporting

Table. Summaryofrecommendationsforinteractingwithvisuallyimpairedpatientsintheemergencydepartment. RecommendationsRationaleReferences

Useoptimallanguage:disability- firstoftenpreferred.Betterrepresentsthepatient’slivedexperiences24

Introduceyourselfeverytimeyouentertheroom(consider placingsignagetoalertstaff).

Ensurespatientisawareofwhoisintheroomatall timesandmayhelppreventdelirium

Tellthepatientwhatyou’regoingtodobeforedoingit, includingbeforeleavingtheroom.

Listentothepatient’scaregiver(s),ifapplicable,butonly aftergatheringasmuchinformationfromthepatientas youcan.

Ifavailableatyourfacility,askwhetherthepatientwould likeanadvocate.

Accommodatetheneedsofthepatient,butdonot over-focusonvisualimpairmentduringtheHPI.

PlacethepatientinquietestpartoftheED.Mayhelppreventdelirium32,35,41

Ensurethepatienthasaccesstomobilityequipment (eg,cane,guidedog)atalltimes.

Ensurethepatienthasaccesstopersonaltechnology (eg,phone,smartwatch,etc).

Ensurethepatientknowswherethecalllightisand howtouseit.

Ensuresmaximalpatientautonomyandmayhelp preventdelirium

preventdelirium

Usethecorrectstrategieswhenguidingapatient.Helpsensurepatientsafety31,32 Clearlynotethepatient’svisualimpairmentinthemedical record(ICD-9:369;ICD-10:H54).

Helpsensureallhealthcareworkersareawareofthe patient’sVIandcanproviderelevant accommodations

AdvocateforMedicaidexpansionatthestateandmedical society(eg,AAEM)level,andencouragepatientstoapply. MayhelpdecreasefrequencyofEDvisits7,32

HelpthepatientestablishcarewithaPCP.HelpstopreventrecurrentEDvisits32,44,45

HPI,historyofpresentillness; ED,emergencydepartment; VI,visuallyimpaired; ICD,InternationalClassificationofDiseases,Rev9or10; AAEM,AmericanAcademyofEmergencyMedicine;PCP,primarycarephysician.

recommendationseven.Threearticleswerefound supportingrecommendationeight.Onearticlewasfound supportingrecommendation9–13.Finally,wefoundtwo articlessupportingrecommendation14.Allsupporting articlesandwhichrecommendationstheyinformedcanbe foundin Table

DISCUSSION

CommunicationStrategies

OptimalLanguage

Theuseofperson-first(eg,personwhoisblind)and disability-first(eg,blindperson)languageisacontentious issue.Academicsconsiderperson-firstlanguagetobemore dignifyingasitplacesfocuslessonthedisabilityandmoreon theindividual.24,25 However,manyblindpeopleandblind advocatesstronglydisagreewithperson-firstlanguageasit mayinadvertentlystigmatizedisability.Blindadvocatesalso arguethatdisability-firstlanguagemoreaccurately representsdisabledexperiences.25–27 Thiscontentionfurther emphasizestheimportanceoflisteningtodisabledpatients andusingtheterminologytheyprefer.Ifablindpatient

preferstobecalleda “blindpatient” ora “patientwhois blind,” thatpreferenceshouldbeaccommodatedlikeany other.Disability-firstlanguagewillbeusedinthispaperfor brevityand,moreimportantly,becauseitisgenerally preferredbytheVIcommunity.

EnteringandExiting

Consentisanintegralcomponentofpatientcare,andall effortsshouldbemadebyemergencycliniciansandpatient carestafftoobtaininformedconsentatalltimes.28 However, thewaythatconsentisobtainedcannotbeuniformlyapplied toallpatients.Forexample,blindpatientscannotseewhois enteringtheirroom,sotheymaynotimmediatelybeableto tellwhetherthepersonwhojustwalkedinisadoctor,nurse, familymember,etc.Thus,itisimperativeforeachperson enteringablindpatient’sroomtoverballyinformthepatient oftheirnameandroleeverytimetheyentertheroom.29 This isespeciallyimportantintheED,anoftenhecticand disorientingplaceforallpatients,andparticularlyforthose withdisabilities.30 Justasimportantasannouncingwhenyou walkintoapatient’sroomisannouncingwhenyouor

othersinvolvedinpatientcareleavetheroom.29 Ifthis isnotdone,thepatientmayattempttospeaktosomeone whotheylogicallyassumeisstillintheroomonlytobe metwithsilence.Thisisnotonlypotentiallyembarrassing butdisorienting.31,32

InformedConsent

InformedconsentdiscussionsalsomustbetailoredforVI patients.Inadditiontothetypicaldiscussionstogain consent,VIpatientsbenefitfromtheclinicianmaintainingan ongoingdialogueduringaprocedure,explainingwhatwillbe donenextandprovidingclear,actionableinstructionswhen necessary.29,33 Addingthisextralayerofcommunicationcan beinstrumentalinensuringpatientsafetyandadherence,and theoverallefficacyofthemedicalinterventionforblind patients.Furthermore,itservestomaintainrespectfortheir autonomy,helpsfosteracooperativeenvironment,and minimizessurpriseordiscomfortduringtheprocedure, aparticularlyimportantconsiderationinanEDsetting wherethepaceofcareisoftenrapidandpotentially anxiety-inducing.34,35

MindfulnessofUniqueNeeds

NavigatingCaregivers

Ifacaregiverisnotpresent,youmayaskthepatientor checkthepatient’schartforapotentialcaregiver’scontact information.However,donotassumeapatienthasor requiresacaregiverbecausetheyareVI.Duringthecourseof treatmentofaVIpatient,thecaregiver(ifapplicable)maybe abletoprovidehelpfulinformationorcontextregardingthe patient.36 Forexample,thecaregivermayprovide informationaboutthepatient’sbaselineindependenceand ActivitiesofDailyLiving theskillsneededto independentlycareforoneself.Thisinformationcanbe helpfulduringthecourseoftreatmentintheED,aswellas upondischargetocustomizeinstructionstothepatient. However,itisimportanttorememberthatcaregivers areanadjuncttopatientcareandnotthepatientsthemselves. Thus,besuretogatherasmuchinformationfromthe patientaspossibleaswellasfromtheircaregiver.29,32 This helpsmaintainarespectfulandautonomouspatientclinicianrelationship.

UsingaPatientAdvocate

Patientadvocatescanplayasignificantroleintheholistic careofapatient.37,38 DuringthecourseoftreatmentforaVI patient,itisimportanttoaskthepatientwhethertheyhave anadvocate,whichcanbedoneasearlyasthetriageprocess. Ifthepatientdoesnotalreadyhaveanadvocateorcannot thinkofsomeone,itisimportanttoworkcollaboratively withthepatienttoidentifyanadvocate,iftheywouldlike one.Thereareseveralpotentialpeoplewhocanbeadvocates includingfamilyandfriendsofthepatient,workcolleagues,

caregivers,socialworkers,andhospitalvolunteers (eg,premedicalstudentsandnavigators).32,33

Theroleofanadvocatemayvary;therefore,itiscriticalto establishclearrolesandresponsibilitiesfortheadvocate.One oftheirkeyresponsibilitiescanbetoaccompanythepatientin thewaitingroom.Iftheadvocateisanemployeeofthe hospitalorfamiliarwiththeED,itcanbehelpfulforthe advocatetodiscusstheoverallEDprocess.Thiswillprovide predictabilityofwhattoexpectandclarifytheEDprocessfor thepatient.37 Afterthewaitingroom,theadvocatecanalso providesupportduringtransporttotheroomandinmeeting healthcarepersonnelandexplainingtheworkupand proceduresforlabsorimaging.Finally,duringdisposition,the advocatecanappropriatelyadvocateonbehalfofthepatient forresourcesrequiredfollowingdischargeorduringthe admissionprocess.Theoverallrolesandresponsibilitiescan varybypatientandEDsetting,butitisimportantforthe patientandtheadvocatetoestablishamutualunderstanding.

HistoryofPresentIllnessConsiderations

Whengatheringthehistoryofpresentillness(HPI)onaVI patient,emergencycliniciansshouldstrivetotreatthepatient assimilarlytootherpatientsaspossible.Forexample, lookingatthepatientdirectlywhenyouarespeaking,asyou wouldforotherpatients,isconsiderateand thoughtful.29,31–33 ItisalsoimportanttorecognizethatVI existsonaspectrumfromslightlydecreasedvisualacuitytoa completelackofvision,andmostpeopletypicallyconsidered blindhavesomelevelofvisualfunction.39 Acknowledging thisspectrum,cliniciansshouldattempttodiscernthe patient’suniqueneedstoprovideoptimalcare.Itisalso important not topresumelowercognitiveabilityorother disabilitiesduetovisualimpairment.36 Ininteractionswith thepatient,beconsiderateoftheirvisualimpairment,butdo notoverlyfocusonit.Remember,EDvisitsforblindnessand lowvisionareexceedinglyrare40;thus,ablindpatientis unlikelytobeseekingemergencycarefortheirblindness. Treattheblindpatientasyouwouldyourotherpatientsas muchaspossible,anddonotoverlyplacatethepatient.For example,iftheblindpatientneedstosignaconsentform,you canmakethenecessaryaccommodationssuchasreadingthe formoutloud.29,32,33

PlacementStrategiesandAccessibility

OptimalLocationforPatientsintheED

ItiscommonforpeoplewhoareVItohaveheightened sensorysensitivity,particularlytosound.41,42 Thisis especiallytrueforpeoplewithearlyvisionloss.43 Therefore, makingconsiderationsforadaptingthecareenvironment cancontributetoamorecomfortablepatientexperience.For example,placingthepatientinthequietestpartoftheED canhelp.32,35 Thismayalsohelppreventdelirium, particularlyifapatientneedstostayintheEDfora prolongedperiodoftime.

EnsureAccesstoAssistiveDevices

Accessibilitytopersonalassistivedevices,suchasmobility equipment,shouldbeconsidered.32 Thesedevices,likecanes orguidedogs,areconsideredanextensionofthepersonand arelegallyrecognizedasmedicalequipmentundertheADA. Forpatientswithaguidedog,cliniciansandotherhealthcare staffshouldunderstandthatthedoghasaspecificjoband, thus,shouldnotbebotheredorinhibited.Healthcarestaff arenotrequiredtodirectlycareforaguidedogbutmayassist withcaretasksifthepatientrequestsandtimepermits.By ensuringthatVIpatientshavecontinualaccesstotheseaids, wecanhelpfacilitateindependentnavigationandmobility, whichservestopreservetheirautonomyandreduces potentialdistressduringtheirstay.29,33,35

Phonesorsmartwatchescanalsohelpbridgegapsin healthcareequitybyservingseveralfunctions.Forexample, VIpatientsoftenusespeech-to-textsoftwareornavigational aids,whichtheymayaccessthroughtheirpersonaldevices.30 Manyhospitalsofferappsoronlinetoolstotrack appointments,viewlabresults,orcommunicatewith clinicians.Ensuringaccesscan,therefore,facilitate communicationwithmedicalstaffandcontributetoamore comprehensiveunderstandingoftheircare.Finally,personal devicesenablepatientstomaintaincontactwiththeirsocial networks,friends,orfamily,whichcanhelppromote emotionalwell-beingduringapotentiallystressfulhospital stay.30 Somepatientsmayrelyontheirdevicesfor entertainmentordistraction,whichcanmakethestressful EDenvironmenteasiertocopewith.Inall,maintaining accesstopersonaltechnologyisnotmerelyaconveniencefor VIpatients;itplaysacrucialroleinensuringequityand inclusivitybyfosteringamorepatient-centeredapproachto careandempoweringtheminthemanagementoftheir healthcare.30,32 Finally,ensuringthatpatientsareawareof thelocationandoperationofthecalllightcanfurther empowerthemandfacilitateimmediatecommunication, especiallyinemergencysituations.30,32 Thesesimple strategiesmayalsohelppreventdeliriuminVIpatientswho arealreadyathigherrisk.

GuidingPatients

IfaVIpatientneedstomovesomewhere(eg,tousethe bathroom),andisstableenoughtoambulate,itisimportant toknowhowtobestassistthepatient.Allowingambulatory patientstowalkalsoprovidesthemwithautonomy.Guiding canbeadauntingtaskforthosewhohaveneverdoneit,but thistaskisrelativelysimple.First,thehealthcarestaffshould askthepatientwhethertheywouldlikeaguideandwhether theywouldliketobringtheirassistivedevice(ie,caneor guidedog).Iftheysayyes,allowthemtostand;then,the healthcareworkershouldstandnexttothepatientandtap thepatient’sarm.Thepatientwillthentaketheperson’ sarm orelbowandwillbereadytobeguided.Thehealthcare workershouldwalkatanormalpace.Iftheworkerispassing

throughatightarea,theyshouldsimplymovetheirelbow behindtheirbackandholditthere.Thiswillsignalthe patienttowalkbehindthestaffmember.Whenit’ssafefor thepatienttoreturntotheclinicianorhealthcareworker’ s, theworkershouldmovetheirelbowbacktotheirside;this willsignalitissafetoreturntowalkingbytheworker’sside. AlthoughunlikelyintheED,ifthehealthcareworker encountersaledgeorstairs,theyshouldinformthepatient andpausewhentheygettothearea.Thiswillgivethepatient enoughtimetogainstablefooting.After,walkupordown thestairsatanormalpace.IftheEDstaffmember encountersadoor,openthedoorandensurethepatienthasa handonthedoor.Thiswillensuretheyareabletocontrol whenthedoorcloses.Ifthepatientisusingthebathroom, assistthemin findingthetoiletandsink;thenleavethe bathroomandgivethepatientprivacy.When finished,the patientwillletthestaffmemberknow,andtheycanbe guidedbacktotheirroom.31,32

EnsuringQualityContinuanceofCare

OptimalDocumentation

Whentreatingablindpatient,itisimportanttonote visualimpairmentasearlyaspossibleandasclearlyas possibleinthechartand/oronthewristbandthatthe personiswearing,forexample.30,32 Theidealtimetonote visualimpairmentwouldbeduringtheintakeortriage process.TheInternationalClassificationofDiseases,Rev9 and10codesforBlindnessandLowVisionare369andH54, respectively.Thiswouldenablethedownstreamhealthcare workerstoappropriatelyadjusttheircaretoapatientwith visualimpairment.

Uponrecognizingthatthepatientisblind,thepatient’ s chartshouldbeupdatedtoclearlyreflectthevisual impairment,asperhospitalorEDprotocol.Ifyour healthcaresettingdoesnothaveaprotocol,youcanseekto establishastandardizedprotocol.Beforeimplementing, considerthattheprotocolshouldbeimplementableacross bothelectronicandpaperhealthrecords.Oneexamplecould bean “ eye ” iconinanelectronichealthrecord(EHR)ora coloredstickerforpapercharts.Additionally,thesamecoloredstickercanalsobeappliedasapatientwristband. Finally,ensurethattheprotocoldoesnotoverlaporconflict withanotherexistingdepartment/hospitalprotocol.For example,ifyourhospitalusesayellowwristbandtosignifya fall-riskpatient,itisbesttouseanalternatecolortosignifya patientwithvisualimpairment.Similarsignageusedfor “fall-risk” orinfectionprecautionscanbeusedonthe patient’sdoor,ifadmitted.32

DischargeConsiderations

Duringdischarge,patientsareoftengivenpapercopiesof theirdischargeinstructions.However,thisisnotaccessible forVIpatients.Thus,itisimportantto findalternativemeans ofprovidingthisinformation.29,32,33 ManyEHRsystems

havewebsitesorappspatientscanusetoaccesstheirhealth information.Forexample,Epic(EpicSystemsCorporation, Verona,WI)usestheMyChartsystem,whichisscreenreaderaccessible.Screenreadersaresoftwarenatively installedordownloadedontodevicesthatusethedevice’ s microphonetoreadoutloudwhatisonscreen.TheMyChart appcanbeusedwithIOSandAndroidscreenreaders, VoiceoverandTalkback,respectively,andthewebsitecanbe accessedwithJAWSandNVDA,thetwomostcommonly usedWindowsscreenreaders.Althoughit’simpossibleto testeveryEHR,youcanreachouttoyourinformation technologydepartmenttodeterminewhetheryoursystemis screen-readeraccessible,andifnot,toadvocateforupdates tobemadesoallpatientscanaccesstheirhealthrecordsand dischargeinstructions.

SupportMedicaidExpansion

StatesthathaveexpandedMedicaidcoverageseea decreasedrateofEDvisitsamongdisabledpatients.Thisis likelybecauseitdecreasesthe financialburdenfordisabled patientstoseekpreventativecare.7 Importantly,thismay alsodecreaseclinicianburden.Werecommendadvocating forMedicaidexpansioninyourstate.Thiscanbedonein manyways,suchascontactingyourmemberofcongressor representativesatyourmedicalsociety(eg,American AcademyofEmergencyMedicine).Additionally,hospital financialservicesorsocialworkersmaybeabletoassist patientsinapplyingtoMedicaid.32

ConnectPatientstoaPrimaryCarePhysician

Itisknownthataccesstoaprimarycarephysician(PCP) isassociatedwithsignificantlyreducedEDvisits.44 ForVI patientswhohaveamyriadofuniqueneeds,itisespecially importanttoconnectthemwithaPCPbeforetheyare discharged.32 Thishasalsobeenfoundtodecreaserecurrent EDvisitsamongdisabledpatients.45

LIMITATIONS

ThisreviewislimitedbythelackofdataonVIpatientsin theED.Itisalsoimportanttonotethatdisabledindividuals’ experiencesarevariedandhighlypersonal,sothe recommendationsprovidedinthispaperaregeneral.Alldata usedinthisreviewareretrospectiveandobservationaland, thus,subjecttothelimitationsinherenttothosestudytypes. Moreresearchisneededtodeterminetheshortcomingsof EDcareofVIpatients.

CONCLUSION

Thereareavarietyofimpactfulinterventionsthatcan improveEDcareforvisuallyimpairedpatients.These interventionsarereproducible,notresource-intensive,and profoundlyhelpfulforVIpatientsintheED.LikemanyED interventions,theserecommendationsarenotstaticor comprehensivebutratherservethepurposeoffurtheringa

much-neededconversation.Theserecommendationsshould alsobefurtherstudiedtodeterminetheirpatient-centered impact,ideallyinpartnershipwithnationalandstate organizationsrepresentingVIpeople.Optimalcareinthe EDforvisuallyimpairedpatientsisoptimalcareforall patients.Pleaseconsiderimplementingsomeorallofthese interventionsandapproachingthecareofVIEDpatients mindfullyandintentionally.

ACKNOWLEDGMENTS

TheauthorswouldliketothankSheilaStyronforher feedbackonthispaperfromtheperspectiveofaVIperson andadvocateatTheWholePerson.

AddressforCorrespondence:KareemHamadah,Universityof KansasSchoolofMedicine,4000CambridgeSt,KansasCity, KS66103.Email: khamadah@kumc.edu

ConflictsofInterest: Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Hamadahetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.UlldemolinsAR,LansinghVC,ValenciaLG,etal.Socialinequalitiesin blindnessandvisualimpairment:areviewofsocialdeterminants. Indian JOphthalmol. 2012;60(5):368–75.

2.BesagarS,YonekawaY,SridharJ,etal.Associationofsocioeconomic, demographic,andhealthcareaccessdisparitieswithseverevisual impairmentintheUS. JAMAOphthalmol. 2022;140(12):1219–26.

3.YektaA,HooshmandE,SaatchiM,etal.Globalprevalenceandcauses ofvisualimpairmentandblindnessinchildren:asystematicreviewand meta-analysis. JCurrOphthalmol. 2022;34(1):1–15.

4.ChanT,FriedmanDS,BradleyC,etal.Estimatesofincidenceand prevalenceofvisualimpairment,lowvision,andblindnessintheUnited States. JAMAOphthalmol. 2018;136(1):12–9.

5.VarmaR,VajaranantTS,BurkemperB,etal.Visualimpairment andblindnessinadultsintheUnitedStates:demographicand geographicvariationsfrom2015to2050. JAMAOphthalmol. 2016;134(7):802–9.

6.HarrisCMandWrightSM.Severevisionimpairmentandblindnessin hospitalizedpatients:aretrospectivenationwidestudy. BMC Ophthalmol. 2021;21(1):263.

7.Horner-JohnsonW,LindnerS,LevyA,etal.Timetrendsinemergency departmentuseamongadultswithintellectualanddevelopmental disabilities. DisabilHealthJ. 2022;15(2):101225.

8.RaschEK,GulleySP,ChanL.Useofemergencydepartmentsamong workingageadultswithdisabilities:aproblemofaccessandservice needs. HealthServRes. 2013;48(4):1334–58.

9.KuehnBM.Disabilitiesandemergencycare. JAMA. 2013;309(8):759.

10.JamesTG,CoadyKA,StacciariniJMR,etal. “They’renotwillingto accommodatedeafpatients”:communicationexperiencesofdeaf AmericanSignLanguageusersintheemergencydepartment. Qual HealthRes. 2022;32(1):48–63.

11.JamesTG,McKeeMM,MillerMD,etal.Emergencydepartment utilizationamongdeafandhard-of-hearingpatients:aretrospective chartreview. DisabilHealthJ. 2022;15(3):101327.

12.McKeeMM,WintersPC,SenA,etal.Emergencydepartmentutilization amongdeafAmericanSignLanguageusers. DisabilHealthJ. 2015;8(4):573–8.

13.MorseAR,YatzkanE,BerberichB,etal.Acutecarehospital utilizationbypatientswithvisualimpairment. ArchOphthalmol. 1999;117(7):943–9.

14.CreweJM,SpilsburyK,MorletN,etal.Healthserviceuseandmortality oftheelderlyblind. Ophthalmology. 2015;122(11):2344–50.

15.ZhengDD,ChristSL,LamBL,etal.Patternsofchronicconditions andtheirassociationwithvisualimpairmentandhealthcareuse. JAMAOphthalmol. 2020;138(4):387–94.

16.SinghRRandMauryaP.Visualimpairmentandfallsamongolder adultsandelderly:evidencefromlongitudinalstudyofageinginIndia. BMCPublicHealth. 2022;22(1):2324.

17.LoriautP,LoriautP,BoyerP,etal.Visualimpairmentandhipfractures: acase-controlstudyinelderlypatients. OphthalmicRes. 2014;52(4):212–6.

18.PatinoCM,McKean-CowdinR,AzenSP,etal.Centralandperipheral visualimpairmentandtheriskoffallsandfallswithinjury. Ophthalmology. 2010;117(2):199–206.e1.

19.S¸ahinAA,Keles¸S¸ahinA,UzunA,etal.Therelationship betweenextremityfracturesandvisualimpairmentinchildhood: acase-controlstudy. UlusTravmaVeAcilCerrahiDerg. 2022;28(5):662–667.

20.MorandiA,InzitariM,UdinaC,etal.Visualandhearingimpairment areassociatedwithdeliriuminhospitalizedpatients:resultsof amultisiteprevalencestudy. JAmMedDirAssoc. 2021;22(6):1162–7.e3.

21.AungTheinMZ,PereiraJV,NitchinghamA,etal.Acalltoactionfor deliriumresearch:meta-analysisandregressionofdeliriumassociated mortality. BMCGeriatr. 2020;20(1):325.

22.KimAM,LeeJY,KimJ.Emergencydepartmentutilizationamong peoplewithdisabilitiesinKorea. DisabilHealthJ. 2018;11(4):598–605.

23.MerrifieldJ.Meetingtheneedsofpeoplewithalearningdisabilityinthe emergencydepartment. IntEmergNurs. 2011;19(3):146–51.

24.ForgeJ.Preferredlanguagepracticeinprofessionalrehabilitation journals. JRehabil. 1991;57:49.

25.GernsbacherMA.EditorialPerspective:Theuseofperson-first languageinscholarlywritingmayaccentuatestigma. JChildPsychol Psychiatry. 2017;58(7):859–61.

26.VaughanCE.People-firstlanguage:anunholycrusade. Braille Monitor.2009.Availableat: https://nfb.org/sites/default/files/images/ nfb/publications/bm/bm09/bm0903/bm090309.htm AccessedApril9,2023.

27.StreeterL.Thecontinuingsagaofpeople-firstlanguage. Braille Monitor.2010.Availableat: https://nfb.org/sites/default/files/images/ nfb/publications/bm/bm10/bm1005/bm100509.htm AccessedApril9,2023.

28.HallDE,ProchazkaAV,FinkAS.Informedconsentforclinicaltreatment. CMAJ 2012;184(5):533–40.

29.SchneiderK.Caringbetterforpatientswhoareblindorvisually impaired. AmFamPhysician. 2013;88(11):774.

30.MorrisMA,WongAA,DorseyHollimanB,etal.Perspectivesofpatients withdiversedisabilitiesregardinghealthcareaccommodationsto promotehealthcareequity:aqualitativestudy. JGenInternMed. 2021;36(8):2370–7.

31.StevensS.Assistingtheblindandvisuallyimpaired:guidelines foreyehealthworkersandotherhelpers. CommunityEyeHealth. 2003;16(45):7–9.

32.MarshallSandJoffeeE.ADAChecklist:HealthCareFacilitiesand ServiceProviders–EnsuringAccesstoServicesandFacilitiesby PatientsWhoAreBlind,Deaf-Blind,orVisuallyImpaired.TheAmerican FoundationfortheBlind.2006.Availableat: https://www.afb.org/ blindness-and-low-vision/your-rights/advocacy-resources/ ada-checklist-health-care-facilities-and.AccessedMay2,2023.

33.CarmichaelJH,KalagherKM,ReznekMA,etal.Improvingaccessibility intheemergencydepartmentforpatientswithdisabilities:aqualitative study. WestJEmergMed. 2023;24(3):377–83.

34.RotoliJ,BacksterA,SappRW,etal.Emergencymedicineresident educationoncaringforpatientswithdisabilities:acalltoaction. AEMEducTrain. 2020;4(4):450–62.

35.PoffenbergerCM,CoatesWC,BacksterA,etal.Makingemergency medicineaccessibleforall:thewhat,why,andhowofproviding accommodationsforlearnersandphysicianswithdisabilities. AEMEducTrain. 2022;6(S1):S71–6.

36.HeydarianNM,HughesAS,MoreraOF,etal.Perspectivesof interactionswithhealthcareprovidersamongpatientswhoareblind. 2021.Availableat: https://nfb.org/images/nfb/publications/jbir/jbir21/ jbir110203.html.AccessedApril3,2023.

37.AmericanCollegeofEmergencyPhysicians(ACEP).Emergency departmentpatientadvocateroleandtraining.Policystatement. AnnEmergMed. 2015;65(1):130–1.

38.ChoiPP.Patientadvocacy:theroleofthenurse. NursStand. 2015;29(41):52–8.

39.LeeSYandMesfinFB.(2023)Blindness.In: StatPearls Availableat: http://www.ncbi.nlm.nih.gov/books/NBK448182/ AccessedJuly30,2023.

40.VaziriK,SchwartzSG,FlynnHW,etal.Eye-relatedemergency departmentvisitsIntheUnitedStates,2010. Ophthalmology. 2016;123(4):917–9.

41.NilssonMEandSchenkmanBN.Blindpeoplearemoresensitivethan sightedpeopletobinauralsound-locationcues,particularlyinter-aural leveldifferences. HearRes. 2016;332:223–32.

42.KolarikAJ,RamanR,MooreBCJ,etal.Theaccuracyofauditoryspatial judgmentsinthevisuallyimpairedisdependentonsoundsource distance. SciRep. 2020;10(1):7169.

43.BauerCM,HirschGV,ZajacL,etal.MultimodalMR-imagingreveals large-scalestructuralandfunctionalconnectivitychangesinprofound earlyblindness. PLoSONE. 2017;12(3):e0173064.

44.TsaiMH,XirasagarS,CarrollS,etal.Reducinghigh-users’ visitstothe emergencydepartmentbyaprimarycareinterventionfortheuninsured: aretrospectivestudy. InqJMedCareOrganProvisFinanc. 2018;55:0046958018763917.

45.ChuL,SoodN,TuM,etal.Reductionofemergencydepartment useinpeoplewithdisabilities. AmJManagCare. 2017;23(12):e409–15.

ORIGINAL RESEARCH

AssociationBetweenSexuallyTransmittedInfections andtheUrineCulture

JohnathanM.Sheele,MD,MHS,MPH*

CarolynMead-Harvey,MS†

NicoleR.Hodgson,MD

‡

SectionEditor:IoannisKoutroulis,MD,MBA,PhD

*MayoClinic,DepartmentofEmergencyMedicine,Jacksonville,Florida † MayoClinic,DivisionofClinicalTrialsandBiostatistics, Scottsdale,Arizona ‡ MayoClinicHospital,DepartmentofEmergencyMedicine, Phoenix,Arizona

Submissionhistory:SubmittedFebruary1,2023;RevisionreceivedDecember13,2023;AcceptedJanuary16,2014

ElectronicallypublishedMay3,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.60033

Introduction: Bacterialurinarytractinfections(UTI)andsomesexuallytransmittedinfections(STI)can haveoverlappingsignsandsymptomsornonspecific findings,suchaspyuriaonurinalysis. Furthermore,resultsfromtheurinecultureandthenucleicacidamplificationtestforanSTImaynotbe availableduringtheclinicalencounter.Wesoughttodeterminewhethergonorrhea,chlamydia,and trichomoniasisareassociatedwithbacteriuria,informationthatmightaidinthedifferentiationofSTIs andUTIs.

Methods: Weusedmultinomiallogisticregressiontoanalyze9,650encountersoffemalepatientswho wereaged ≥18yearsandwhounderwenttestingforSTIs.TheEDencounterstookplacefromApril18, 2014–March7,2017.Weusedamultivariableregressionanalysistoaccountforpatientdemographics, urinalysis findings,vaginalwet-mountresults,andpositiveornegative(orno) findingsfromtheurine cultureandtestingfor Neisseriagonorrhoeae,Chlamydiatrachomatis, or Trichomonasvaginalis.

Results: Inmultivariableanalysis,infectionwith Tvaginalis, Ngonorrhoeae, or Ctrachomatis wasnot associatedwithhavingaurinecultureyielding10,000ormorecolony-formingunitspermililiter(CFU/mL) ofbacteriacomparedwithaurinecultureyieldinglessthan10,000CFU/mLornourinecultureobtained. ThediagnosisofaUTIintheEDwasnotassociatedwithhavingaurinecultureyielding10,000ormore CFU/mLcomparedwithaurinecultureyieldinglessthan10,000CFU/mL.

Conclusion: Afteradjustingforcovariates,noassociationwasobservedbetweenurinecultureresults andtestingpositivefortrichomoniasis,gonorrhea,orchlamydia.Ourresultssuggestthathavinga concurrentSTIandbacterialUTIisunlikely.[WestJEmergMed.2024;25(3)358–367.]

Keywords: chlamydia;emergencydepartment;emergencymedicine;gonorrhea;Trichomonas; urinarytractinfection.

INTRODUCTION

Urinarytractinfection(UTI)isoneofthemostcommon bacterialinfectionsdiagnosedintheemergency department(ED). 1 , 2 SymptomsofUTIarethereasonfor approximately1%ofallambulatoryvisitsandresultin2 – 3 millionEDvisitsintheUSeachyear. 2 Urinecultureresults cantakemorethanaday,andtheurinalysis fi ndingscan

causediagnosticuncertaintyabouttheexistenceofa bacterialUTI.Addingtotheproblemisthattheincidence ofsomesexuallytransmittedinfections(STI)suchas gonorrhea,chlamydia,andtrichomoniasisisincreasingin theUS, 3 , 4 andclinicalmanifestationsofUTIsandSTIs mayoverlap.Theseoverlappingsignsandsymptomsmay leadtounderdiagnosingSTIsinpatientswithurinary

concernsandovertreatingforSTIsinpatientswithgenital concerns. 5 – 9 Previousstudy fi ndingshaveshownthatSTIs areassociatedwithsterilepyuriaandothernon-speci fi c fi ndingsonurinalysis. 5 , 9 , 10 Diagnosticconfusionmaybe mostcommonwhentrichomoniasisisidenti fi edintheED byurinalysisorwetmountandtheclinicianmustconsider whethertheurinein fl ammatorychangesarecausedby Trichomonasvaginalis onlyorbyaconcurrent bacterialUTI.

Inthisanalysis,wesoughttodeterminewhether infectionwithgonorrhea,chlamydia,andtrichomoniasis wasassociatedwithspeci fi curinecultureresults. Speci fi cally,weattemptedtodeterminethefrequencyof STIsandhavingaurinecultureyield10,000ormore colony-formingunitspermilliliter(CFU)/mL)ofbacteria. Theresearchquestionwesoughttoanswerwasasfollows: Forawomansuspectedofhavingorfoundtohave gonorrhea,chlamydia,ortrichomoniasisduringtheED encounterwhohasgenitourinaryconcerns,arethe in fl ammatorychangesobservedonurinalysismostlikely causedonlybytheSTI,orisconcurrentbacteriuria (eg,UTI)contributing?

METHODS

Dataset

Weusedanexistingdatasetof75,000EDencountersof patients ≥18yearsinagefromasinglehealthsystem.11–22 All patientsinthedatasetreceivedtestingforgonorrhea, chlamydia,ortrichomoniasisorunderwentbothurinalysis andurineculture.Patientsundergoingonlyurinalysis, regardlessofSTItesting,werenotincludedinthedataset.All EDencounterstookplaceApril18,2014–March7,2017. Thedatasetwascreatedbytheinstitution’sinformation technologyteamwhoextractedretrospectivedatafromthe electronichealthrecords(EHR).Forouranalysis,we includedwomenwhowerenotadmittedtothehospitaland whohadanucleicacidamplificationtest(NAAT)for gonorrhea,chlamydia,ortrichomoniasisorhadavaginal wetmount.DataontheNAATswabsitewerenotavailable. IndividualpatientscouldhavemorethanoneEDencounter. Ourprojectwasapprovedbytheinstitutionalreviewboard withanexemptionfromfullreview,andinformedconsent waswaived.Articleshavebeenpublishedusingthe originaldataset.11–22

Patientsinthedatasetwereidentifiedashaving trichomoniasisiftheparasitewasseenwithurinemicroscopy (amethodhavingverylowsensitivitybuthighspecificity), vaginalwetmount(moderatesensitivityandhigh specificity),orNAAT(highsensitivityandspecificity).22–24 Toavoidmulticollinearityinthemultivariableanalysis,we consolidated findingsfromvaginalwetmountandurine microscopyfor Tvaginalis intoasinglevariablelabeled T vaginalis infectionstatusknownduringtheEDencounter. The Tvaginalis NAAT(Aptima,Hologic,Inc,

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Thereisanoverlapinthesigns,symptoms, and fi ndingsonurinalysisforwomenwith urinarytractinfections(UTIs)andsexually transmittedinfections(STIs).

Whatwastheresearchquestion?

Forawomansuspectedofhavingorfoundto haveanSTI,arethein fl ammatorychanges observedonurinalysismostlikelycausedonly bytheSTI,orcouldshehave concurrentbacteriuria?

Whatwasthemajor findingofthestudy?

Afteradjustingforcovariates,noassociation wasobservedbetweenurinecultureresults andtestingpositiveforanSTI,suggesting concurrentSTIandbacterialUTI areunlikely.

Howdoesthisimprovepopulationhealth?

ConcurrentSTIsandbacterialUTIs areunlikely.

Marlborough,MA)result,orthe Neisseriagonorrhoeae or Chlamydiatrachomatis NAAT(Aptima),wasconsidered separatelybecausetheresultwasnotobtaineduntilafterthe EDvisit.Womenmayhavetestedpositivefor Tvaginalis by morethanonetestduringtheirencounter,andanypatient withapositive Tvaginalis testwasconsideredtobeinfected with Tvaginalis. AllSTItestingwasperformedatthe discretionofthetreatingclinician.

Wereportedthevaginalwetmountasnotperformedifthe patienthadnoresultsfromthevaginalwetmountforwhite bloodcells(WBC),yeast, Tvaginalis ,orcluecells.The vaginalwetmountWBCswereanalyzedas0–10/11ormore cellsperhigh-power field(HPF).16 Forthevaginalwet mount,yeast,cluecells,and Tvaginalis werereportedbythe clinicallaboratoryaspresentorabsent.

Weconsideredaurinalysistohavebeenperformedifany componenttestfromtheurinalysiswasreported.Theurine samplewasreportedtohavebeencollectedbythefollowing methods:cleancatch/voided;missingornotdocumentedby nursing;or “other” (eg,bladdercatheter,straightcatheter, ileostomy,nephrostomy,suprapubic,orurostomy).From theurinalysis,weconsideredthefollowingvariables:bacteria (0–4+);blood(0–3+);glucose(positiveornegative);ketones

(positiveornegative);leukocyteesteraselevel(0–3+);mucus (0–4+);nitrites(positiveornegative);protein(positiveor negative);redbloodcells(RBC)(0–101cells/HPF); Trichomonas (positiveornegative);WBCclumps(positive vsnegative);WBCs(0–101cells/HPF);andyeast(presentor absent).IfarangeofurineRBCsandWBCswasreported, weusedthemedianoftherangeintheanalysis,andifmore than100cells/HPFwerereported,weusedtheresult “101 cells/HPF” foranalysis.Allurinetestswereorderedatthe discretionofthetreatingclinician.

Weincludedthefollowingdemographicandtriage informationifitwasavailableduringtheEDencounter: methodofEDarrival;maritalstatus;age;race;andthetriage EmergencySeverityIndex.Ageinyearswasconvertedtoa categoricalvariabletoaccountforthenonlinearrelationship withSTIs.25

WomenwereconsideredtohaveaUTIdiagnosisifthey hadaspecificEDcodeonthe InternationalClassi fi cationof Diseases,9 th or10 th Rev(ICD-9/ICD-10) (Supplement1). Womenwereconsideredpregnantiftheyhadadocumented positivepregnancytestoraspecific ICD-9 or ICD-10 code(Supplement1).

StatisticalAnalysis

Wesummarizedcontinuousvariablesasmedianand interquartilerange,withanalysisofvariance F testsusedto testassociations.Wereportedcategoricalvariablesascounts andpercentages,witha χ 2 testusedtotestassociations.We perfomedmultinomiallogisticregressionanalysis accountingformultipledemographic,clinical,and diagnostictestingvariables,withtheWaldtestusedto determine P values.Multivariableanalyseswereperformed forpatientswhohadcompletedataforallmodelcovariates. Oddsratiosand95%confidenceintervalswerecalculated fromthemultivariablemodel.A P valuelessthan.05was consideredsignificant.Weconductedstatisticalanalyses withstatisticalsoftwareJMPPro14(JMPStatistical Discovery,LLC,Cary,NC)andSASversion9.4(SAS Institute,Inc,Cary,NC).

RESULTS

Amongthe75,000EDencountersintheoriginaldataset, 16,755womenmetourinclusioncriteria.Asummaryofthe clinicalencountersisshownin Table1.Amongthe1,631 patientencounterswithapositivetestresultforgonorrhea, chlamydia,orboth,1,443(88.5%)hadurinalysis,443 (27.2%)hadurineculture,and438(26.9%)hadboth urinalysisandurineculture.Amongthe1,354womenwith Tvaginalis identifiedonvaginalwetmountand418women withapositiveNAATresultfor Tvaginalis ,1,203(88.8%) and374(89.5%)patients,respectively,hadurinalysis.

Table2 showsencounterswithapositiveSTItestresultand theresultsoftheurineculture.Amongthe443patientswith gonorrhea,chlamydia,orbothwhohadaurineculture

result,341(77.0%)hadlessthan10,000CFU/mLofbacteria, and102(23.0%)had10,000ormoreCFU/mLofbacteriain theurineculture.

Intotal,1,804patientencountershadapositivetestresult for Trichomonas byurinemicroscopy,vaginalwetmount,or NAAT.Ofthese,1,612(89.4%)hadaurinalysistestresult, 548(30.4%)hadaurinecultureperformed,and538(29.8%) hadbothaurinalysisandurinecultureresult.Atotalof9,650 clinicalencountershadcompleteobservationsforallmodel covariatesandwereincludedinthemultivariableanalysis (Table3).Thisnumberincluded2,414patientencounters withlessthan10,000CFU/mLofbacteria,722patientswith 10,000ormoreCFU/mLofbacteria,and6,514patientswith nourinecultureperformed.

Thefollowingvariablesweresignificantlymorelikelyto beassociatedwithaurineculturewith10,000ormore CFU/mLcomparedwithlessthan10,000CFU/mL:higher bacteriuriaonurinalysis;higheramountofbloodintheurine nitrite-positiveurinepresenceofurinaryWBCclumps; higherurinaryWBCcount;andfewerWBCsonthevaginal wetmount(all P ≤ 0.01; Table3).Thesevariableshada significantlylowerlikelihoodofbeingassociatedwithaurine culturewith ≥10,000CFU/mL:no Tvaginalis NAATresult (comparedwithanegative Tvaginalis NAAT)andprotein intheurine(both P ≤ 0.01; Table3).Thefollowingvariables weresignificantlymorelikelytobeassociatedwith ≥10,000 CFU/mLofbacteriaintheurineculture(comparedwithno urinecultureperformed):married/lifepartner(vssingle); higherbacteriuriaonurinalysis;higherurineleukocyte esteraselevel;nitrite-positiveurine,proteinintheurine, presenceofurinaryWBCclumps,UTIdiagnosedintheED, andhigherurinaryWBCcount(all P ≤ 0.02; Table3).These variableshadasignificantlylowerlikelihoodofbeing associatedwithaurineculturewith ≥10,000CFU/mL (comparedwithnourinecultureperformed):no Tvaginalis NAATresult(comparedwithanegative Tvaginalis NAAT),proteinintheurine,andnovaginalwetmountclue cells(comparedwithpresent)(all P ≤ 0.01; Table3). Neisseriagonorrhoeae or CtrachomatisdetectedbyNAAT, orknown Tvaginalis infectionintheED,wasnot associatedwithaurinecultureyielding10,000ormore CFU/mL.Additionally,UTIdiagnosedintheED wasnotassociatedwithaurinecultureyielding 10,000ormoreCFU/mLcomparedwithlessthan 10,000CFU/mL.

DISCUSSION

BothUTIsandSTIscanhaveoverlappingsignsand symptomsandcancauseinflammatorychangesintheurine. DistinguishingbetweenUTIandSTIcanbechallengingin theED.5,6,26 Wesoughttoassesstherelationshipbetween bacteriuriaandSTIs.Ourresearchquestionwasasfollows: Forawomansuspectedofhavingorfoundtohave gonorrhea,chlamydia,ortrichomoniasisduringtheED

Table1. Demographicsandclinicalcharacteristicsbyurinecultureresult.

Characteristic

(N = 16,755)

(n = 12,372)

Urineculture, <10,000 CFU/mL(n = 3,534)

Urineculture, ≥10,000 CFU/mL(n = 849) P

Age,y,no.(%) .002a

18–2810,524(62.8)7,769(62.8)2,201(62.3)554(65.3)

29–394,328(25.8)3,252(26.3)894(25.3)182(21.4)

≥401,903(11.4)1,351(10.9)439(12.4)113(13.3)

Race,no.(%)(n = 16,683)(n = 12,311)(n = 3,523) <.001a

Black14,855(89.0)11,090(90.1)3,017(85.6)748(88.1)

NotBlack1,828(11.0)1,221(9.9)506(14.4)101(11.9)

Maritalstatus,no.(%)(n = 16,708)(n = 12,336)(n = 3,526)(n = 846) <.001a

Marriedorlifepartner1,488(8.9)1,050(8.5)359(10.2)79(9.3)

Separated,divorced, orwidowed

670(4.0)460(3.7)168(4.8)42(5.0)

Single14,550(87.1)10,826(87.8)2,999(85.1)725(85.7)

Pregnant,no.(%) <.001a

No13,105(78.2)9,725(78.6)2,681(75.9)699(82.3)

Yes3,650(21.8)2,647(21.4)853(24.1)150(17.7)

ESI,no.(%)(n = 15,793)(n = 11,810)(n = 3,365)(n = 798)0.06a 1and2353(2.2)255(2.2)77(2.3)21(2.6)

311,937(74.7)8,777(74.3)2,574(76.5)586(73.4) 4and53,683(23.1)2,778(23.5)714(21.2)191(23.9)

MechanismofEDarrival, no.(%) (n = 16,663)(n = 12,309)(n = 3,507)(n = 847)0.46a

EMSorpolice1,122(6.7)815(6.6)244(7.0)63(7.4)

Publictransportationor onfoot 852(5.1)630(5.1)170(4.8)52(6.1)

Privatevehicle14,689(88.2)10,864(88.3)3,093(88.2)732(86.4)

Urinespecimensource,no.(%)

Cleancatheter/voidedurine3,309(19.7)0(0.0)2,703(76.5)606(71.4)

Other71(0.4)0(0.0)51(1.4)20(2.4)

Notdocumentedormissing13,375(79.8)12,372(100.0)780(22.1)223(26.3)

NAATfor Chlamydia trachomatis, no.(%)

Negative14,985(89.4)11,123(89.9)3,127(88.5)735(86.6)

Positive1,303(7.8)958(7.7)266(7.5)79(9.3)

Notestresult467(2.8)291(2.4)141(4.0)35(4.1)

NAATfor Neisseria gonorrhoeae,no.(%) <.001a

Negative15,819(94.4)11,745(94.9)3,292(93.2)782(92.1)

Positive477(2.8)342(2.8)104(2.9)31(3.7)

Notestresult459(2.7)285(2.3)138(3.9)36(4.2)

NAATfor Trichomonas vaginalis, no.(%)

Negative4,505(26.9)3,409(27.6)854(24.2)242(28.5)

Positive418(2.5)293(2.4)94(2.7)31(3.7)

Notestresult11,832(70.6)8,670(70.1)2,586(73.2)576(67.8)

(Continuedonnextpage)

Characteristic

DiagnosedwithUTIintheED, no.(%)

<10,000 CFU/mL(n = 3,534)

≥10,000 CFU/mL(n = 849)

No14,849(88.6)11,456(92.6)2,900(82.1)493(58.1)

Yes1,906(11.4)916(7.4)634(17.9)356(41.9)

Treatmentofgonorrheaand chlamydia,no.(%)

No13,593(81.1)10,051(81.2)2,855(80.8)687(80.9)

Yes3,162(18.9)2,321(18.8)679(19.2)162(19.1)

Vaginalwetmount, WBCs/HPF,no.(%)

11–1005,296(31.6)3,716(30.0)1,287(36.4)293(34.5) ≤1010,868(64.9)8,233(66.5)2,119(60.0)516(60.8)

Notperformed591(3.5)423(3.4)128(3.6)40(4.7)

Vaginalwetmount, yeast,no.(%)

Present1,027(6.1)762(6.2)217(6.1)48(5.7)

None14,538(86.8)10,765(87.0)3,036(85.9)737(86.8)

Notperformed1,190(7.1)845(6.8)281(8.0)64(7.5)

Vaginalwetmount,cluecells, no.(%)

None8,826(52.7)6,449(52.1)1,908(54.0)469(55.2)

Present6,941(41.4)5,232(42.3)1,386(39.2)323(38.0)

Notperformed988(5.9)691(5.6)240(6.8)57(6.7)

.001a

Leukocyteesterase(urine) <.001b

No.(missing)14,616(2,139)10,381(1,991)3,403(131)832(17)

Median(IQR)0.0(0.0–1.0)0.0(0.0–1.0)1.0(0.0–2.0)2.0(1.0–3.0)

Range0.0–3.00.0–3.00.0–3.00.0–3.0

Nitrite(urine),no.(%)(n = 14,818)(n = 10,505)(n = 3,480)(n = 843) <.001a

Negative14,257(96.2)10,236(97.4)3,417(98.5)604(71.6)

Positive561(3.8)269(2.6)53(1.5)239(28.4)

WBCs(urine) <.001b

No.(missing)10,692(6,063)7,199(5,173)2,699(835)794(55)

Median(IQR)5.0(2.5–13.0)3.0(2.5–12.5)8.0(2.5–16.0)31.5(10.0–101.0)

Range0.0–101.00.0–101.00.0–101.00.0–101.0

Bacteria(urine) <.001b

No.(missing)10,688(6,067)7,194(5,178)2,700(834)794(55)

Median(IQR)1.0(0.0–1.0)1.0(0.0–1.0)1.0(0.0–1.0)1.0(1.0–2.0)

Range0.0–4.00.0–4.00.0–4.00.0–4.0

Blood(urine) <.001b

No.(missing)14,604(2,151)10,361(2,011)3,411(123)832(17)

Median(IQR)0.0(0.0–1.0)0.0(0.0–1.0)0.0(0.0–1.0)1.0(0.0–2.0)

Range0.0–3.00.0–3.00.0–3.00.0–3.0 (Continuedonnextpage)

Table1. Continued.

Characteristic

Total (N = 16,755)

Nourineculture (n = 12,372)

Urineculture, <10,000 CFU/mL(n = 3,534)

Urineculture, ≥10,000 CFU/mL(n = 849) P value

Glucose(urine),no.(%)(n = 14,809)(n = 10,500)(n = 3,467)(n = 842)0.39a

Negative14,216(96.0)10,092(96.1)3,322(95.8)802(95.2)

Positive593(4.0)408(3.9)145(4.2)40(4.8)

Ketones(urine),no.(%)(n = 14,786)(n = 10,477)(n = 3,467)(n = 842) <.001a

Negative12,220(82.6)8,740(83.4)2,808(81.0)672(79.8)

Positive2,566(17.4)1,737(16.6)659(19.0)170(20.2)

Mucus(urine) 0.88b

No.(missing)10,692(6,063)7,202(5,170)2,696(838)794(55)

Median(IQR)1.0(0.0–2.0)1.0(0.0–2.0)1.0(0.0–2.0)1.0(0.0–2.0)

Range0.0–4.00.0–4.00.0–4.00.0–4.0

Protein(urine),no.(%)(n = 14,800)(n = 10,494)(n = 3,464)(n = 842) <.001a

Negative10,553(71.3)7,716(73.5)2,366(68.3)471(55.9)

Positive4,247(28.7)2,778(26.5)1,098(31.7)371(44.1)

RBCs(urine) <.001b

No.(missing)10,693(6,062)7,196(5,176)2,701(833)796(53)

Median(IQR)2.5(2.0–12.5)2.5(1.0–12.5)2.5(2.0–12.5)5.0(2.3–22.8)

Range0.0–101.00.0–101.00.0–101.00.0–101.0

WBCclumps(urine),no.(%)(n = 10,578)(n = 7,116)(n = 2,672)(n = 790) <.001a

None10,118(95.7)6,915(97.2)2,549(95.4)654(82.8)

Present460(4.3)201(2.8)123(4.6)136(17.2)

Yeast(urine),no.(%)(n = 10,628)(n = 7,154)(n = 2,684)(n = 790)0.04a

Present280(2.6)171(2.4)80(3.0)29(3.7)

None10,348(97.4)6,983(97.6)2,604(97.0)761(96.3)

Tvaginalis statusduringED encounter,no.(%) (n = 16,308)(n = 11,987)(n = 3,478)(n = 843) <.001a

Nowetmountperformed720(4.4)451(3.8)216(6.2)53(6.3)

Negativec 14,176(86.9)10,554(88.0)2,922(84.0)700(83.0)

Positive1,412(8.7)982(8.2)340(9.8)90(10.7)

a χ 2 test.

bAnalysisofvariance F test.

cNegativevaginalwetmountandurinemicroscopy(ifperformed).

CFU,colony-formingunits; ED,emergencydepartment; EMS,emergencymedicalservices; ESI,EmergencySeverityIndex; HPF,highpower field; NAAT,nucleicacidamplificationtest; RBCs,redbloodcells; UTI,urinarytractinfection; WBCs,whitebloodcells.

encounterwhohasgenitourinaryconcerns,arethe inflammatorychangesobservedonurinalysismostlikely causedonlybytheSTI,orisconcurrentbacteriuria(eg,UTI) contributing?Ourresultsshowthatinfectionwith gonorrhea,chlamydia,ortrichomoniasiswasnotassociated withalsohavingaurinecultureyielding ≥10,000CFU/mLof bacteriacomparedwith <10,000CFU/mLornourine cultureperformed.Animportant findingwasthatwhen Tvaginalis wasidentifiedduringtheEDencounteronurine microscopyorvaginalwetmount,therewasnosignificant associationwithbacteriaintheurineculture.Whenan emergencyclinicianisevaluatingawomanwith

genitourinaryconcernsandanSTIissuspectedoractually identified,asisthecaseonurinemicroscopyorvaginalwet mountfor Tvaginalis, bacteriuriaisnotmorelikelyto coexist.Our findingssupportrecommendationsforscreening forbothUTIsandSTIsinappropriatepatients.7,8 For instance,womenundergoingpelvicexamination whowerealsodiagnosedwithaUTIintheEDwere subsequentlyfoundtohavehighratesofSTIs.7 However, emergencycliniciansfrequentlydonotscreenforSTIsin womenwithdysuriawhoarediagnosedwithaUTI.8 Furthermore,our findingssupportthoseof smallerstudies.6,27

Table2. PositiveSTItestresultsbyurineculture.a

Urinecultureresult, CFU/mL

Positiveforgonorrhea,chlamydia,orboth onNAAT(n = 1,631)

Positivefor Trichomonasvaginalis bytestmethod

Urinemicroscopy (n = 275)

Vaginalwetmount (n = 1,354) NAAT (n = 418)

Nourineculture1,188(72.8)186(67.6)943(69.6)293(70.1)

0 <10,000341(20.9)71(25.8)326(24.1)94(22.5)

10,000 <100,00015(0.9)5(1.8)16(1.2)6(1.4) >100,00087(5.3)13(4.7)69(5.1)25(6.0)

aDataispresentedasNo.(%).Womenmayhavetestedpositivefor Tvaginalis bymorethan1test. CFU,colony-formingunits; NAAT,nucleicacidamplificationtest; STI,sexuallytransmittedinfection.

Table3. Multinomiallogisticregressionusingurinecultureresultastheoutcomevariable(N = 9,650).

VariableComparisongroupReference

NAATfor Chlamydia trachomatis

NAATfor Neisseria gonorrhoeae

Separated,divorced, orwidowed Single1.04(0.66–

4and51and21.42(0.72–

NotestresultNegative0.87(0.17–

NotestresultNegative1.20(0.23–

NAATfor Trichomonas vaginalis PositiveNegative0.90(0.53–1.52)0.690.95(0.58

MechanismofEDarrivalEMS/policePrivate vehicle

Publictransportation/ onfoot Private vehicle

DiagnosedwithUTIintheEDYesNo1.17(0.94–

Treatmentofgonorrheaand chlamydia YesNo0.79(0.62–

Tvaginalis statusduringED encounter Nowetmount performed Negativec

Bacteria(urine)1-UnitincreaseNone1.13(1.05–1.23) <.0011.19(1.10–1.28) <.001 (Continuedonnextpage)

Glucose(urine)PositiveNegative0.91(0.59–1.42)0.680.98(0.65–1.49)0.93

Ketones(urine)PositiveNegative1.03(0.80–1.31)0.831.13(0.90–1.42)0.31

Leukocyteesterase(urine)1-UnitincreaseNone0.98(0.88–1.09)0.661.16(1.05–1.29) <.001

Mucus(urine)1-UnitincreaseNone0.99(0.92–1.06)0.810.98(0.92–1.05)0.62

Nitrite(urine)PositiveNegative15.7(10.8–22.76) <.0015.72(4.45–7.34) <.001

Protein(urine)PositiveNegative0.71(0.57–0.88).0020.76(0.62–0.93) <.001

WBCclumps(urine)PresentNone1.54(1.10–2.15)0.011.89(1.39–2.56) <.001

Yeast(urine)PresentNone1.18(0.70–1.98)0.531.33(0.82–2.18)0.25

WBCs(urine)1-UnitincreaseNone1.02(1.02–1.02) <.0011.02(1.02–1.02) <.001

Vaginalwetmount,cluecellsPresentNone0.91(0.75–1.11)0.350.78(0.65–0.94) <.001

NotperformedNone0.33(0.06–1.70)0.190.53(0.12–2.43)0.42

Vaginalwetmount,WBCs/HPF11–100 ≤100.68(0.55–0.84) <.0010.84(0.69–1.02)0.09 Notperformed ≤101.01(0.36–2.88)0.980.96(0.35–2.64)0.94

Vaginalwetmount,yeastPresentNone0.96(0.64–1.42)0.820.77(0.53–1.12)0.17 NotperformedNone2.68(0.50–14.45)0.253.50(0.74–16.59)0.11

aCovariateWaldtestfromthemultinomiallogisticregressionmodel.

bAgewasgroupedas18–28,29–39,and ≥40years.

cNegativetestresultbyvaginalwetmountandurinemicroscopy(ifperformed).

CFU,colony-formingunits; ED,emergencydepartment; EMS,emergencymedicalservices; ESI,EmergencySeverityIndex; HPF,highpower field; NAAT,nucleicacidamplificationtest; OR,oddsratio; RBCs,redbloodcells; UTI,urinarytractinfection; WBC,whitebloodcell.

AstudybyPrentissetalshowedthatamongadolescent girlswithurinarytractsymptomsintheED,9%hadanSTI, 57%hadaUTI,and6%hadbothanSTIandaUTI.6

Clinicianaccuracywas83%forSTIsand71%forUTIs, whereasonly23%correctlydiagnosedpatientswithboth UTIandSTI.6 Shapiroetal27 foundthatamong92women withurinarytractsymptoms,STIrateswerenotdifferent betweenwomenwithapositivevsanegativeurineculture (102 CFU/mL).Additionally,aretrospectivestudyofED patientsfoundthatpatientswhoweretreatedforaUTI, testedpositiveforgonorrhea,chlamydia,ortrichomoniasis, andhadpyuriaweresignificantlymorelikelytohavea negativeurineculturethanapositiveurineculture.9

Relianceonpositiveurinenitriteandpyuriatotreatfor UTIinpatientswithconfirmedorsuspectedSTImay resultinovertreatmentwithantibiotics.However,patients withanSTIandapositiveurineculturehadsignificantly higherurineleukocytesthanthosewithnegative cultureresults.9

Wealsofoundthatclinicalencountersinwhichpatients werediagnosedwithaUTIintheEDwerenotmorelikelyto haveaurinecultureof ≥10,000CFU/mLofbacteria comparedwith <10,000CFU/mLPossibly,patientswho werediagnosedwithaUTIbutwhohad <10,000CFU/mL

weremorelikelytohaveanSTI,butthisassociationwasnot examinedinthecurrentstudy.Becausethediagnosisofa UTIwasnotpartofourinclusioncriteria,notallwomen withaUTIdiagnosisarerepresentedinourcohort.Wewere abletostudyonlywomenwhohadbothaurinalysisand urineculture,notjustaurinalysis.Therefore,theassociation betweenaUTIdiagnosisandbacteriuriadeserves furtherinvestigation.

LIMITATIONS

Althoughourstudyusedalargedataset,ithassome limitations.First,notallwomenfromourdatasetunderwent urinalysis,urineculture,vaginalwetmount,andNAATfor STIs.Furthermore,notallwomendiagnosedwithaUTI underwentSTItestingoravaginalwetmount.Second, modeling Tvaginalis intheEDhasinherentlimitations becausetheurinalysisandvaginalwet-mountresultsare availabletotheclinicianduringtheencounter,buttheylack highsensitivity,whereasNAATishighlysensitiveand specificbuttypicallydoesnotyieldresultsduringthepatient encounter.Third,womenundergoingSTItestingwhoalso hadaurineculturemayhavebeenmorelikelytobe concernedabouturinarysymptoms,whichcouldhavebiased ouranalysistothosewomenwithgenitourinaryconcerns. Table3. Continued.

Becausewewereunabletoincludehistoryandphysical examination findingsinouranalysis,wecouldnot differentiatebetweenpatientswithmoregenital concernsandthosehavingmoreurinarysymptoms. Alternatively,somewomenincludedintheanalysismay havehadasymptomaticbacteriuriaoranasymptomaticSTI, althoughthispossibilityisthoughttobelesslikely. Wedidnotattempttodifferentiatebetween contaminatedurineculturesandthoseyielding classicuropathogens.5

Fourth,ourdatasetrepresentedalimitedgeographical areaintheUS,andthepatientswerepredominantlyBlack; therefore,ourresultsmaynotbegeneralizabletoother locationsandraces.Fifth,thedatadidnotincludepediatric patientsormen;soourresultscannotbeextrapolatedto thosegroups.Sixth,patientswhoweretreatedpresumptively forSTIswithoutspecifictestingwereexcludedfrom analysis,andthiscouldhaveresultedinselection bias.Finally,inherentlimitationsexisttousinga datasetextractedfromtheinstitution’sEHRand ICD codes.

CONCLUSION

Inourregressionanalysis,positivegonorrhea,chlamydia, andtrichomoniasistestresultswerenotassociatedwith bacteriuriayielding ≥10,000CFU/mLcomparedwith <10,000CFU/mLornourinecultureobtained.

AddressforCorrespondence:JohnathanM.Sheele,MD,MHS, MPH,MayoClinic,DepartmentofEmergencyMedicine, 4500SanPabloRd.,Jacksonville,FL32224. Email: sheele.johnathan@mayo.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Sheeleetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.LongBandKoyfmanA.Theemergencydepartmentdiagnosisand managementofurinarytractinfection. EmergMedClinNorthAm. 2018;36(4):685–710.

2.Flores-MirelesAL,WalkerJN,CaparonM,etal.Urinarytractinfections: epidemiology,mechanismsofinfectionandtreatmentoptions. NatRev Microbiol. 2015;13(5):269–84.

3.BarrowRY,AhmedF,BolanGA,etal.Recommendationsforproviding qualitysexuallytransmitteddiseasesclinicalservices,2020. MMWR RecommRep. 2020;68(5):1–20.

4.CrowleyJS,GellerAB,VermundSH. SexuallyTransmittedInfections: AdoptingaSexualHealthParadigm.Washington,DC:National AcademiesPress;2021.

5.TomasME,GetmanD,DonskeyCJ,etal.Overdiagnosisofurinarytract infectionandunderdiagnosisofsexuallytransmittedinfectioninadult womenpresentingtoanemergencydepartment. JClinMicrobiol. 2015;53(8):2686–92.

6.PrentissKA,NewbyPK,VinciRJ.Adolescentfemalewithurinary symptoms:adiagnosticchallengeforthepediatrician. PediatrEmerg Care. 2011;27(9):789–94.

7.BergE,BensonDM,HaraszkiewiczP,etal.Highprevalenceofsexually transmitteddiseasesinwomenwithurinaryinfections. AcadEmerg Med. 1996;3(11):1030–4.

8.WilbanksMD,GalbraithJW,GeislerWM.Dysuriaintheemergency department:misseddiagnosisof Chlamydiatrachomatis WestJEmerg Med. 2014;15(2):227–30.

9.ShipmanSB,RisingerCR,EvansCM,etal.Highprevalenceofsterile pyuriainthesettingofsexuallytransmittedinfectioninwomen presentingtoanemergencydepartment. WestJEmergMed. 2018;19(2):282–6.

10.HuppertJS,BiroFM,MehrabiJ,etal.Urinarytractinfectionand chlamydiainfectioninadolescentfemales. JPediatrAdolescGynecol. 2003;16(3):133–7.

11.BonnerM,SheeleJM,Cantillo-CamposS,etal.Adescriptiveanalysisof mendiagnosedwithepididymitis,orchitis,orbothintheemergency department. Cureus. 2021;13(6):e15800.

12.CamposSC,ElkinsJM,SheeleJM.Descriptiveanalysisofprostatitisin theemergencydepartment. AmJEmergMed. 2021;44:143–7.

13.ElkinsJM,HamidOS,SimonLV,etal.AssociationofBartholincystsand abscessesandsexuallytransmittedinfections. AmJEmergMed. 2021;44:323–7.

14.ElkinsJM,Cantillo-CamposS,SheeleJM.Frequencyofcoinfectionon thevaginalwetpreparationintheemergencydepartment. Cureus 2020;12(11):e11566.

15.ElkinsJM,Cantillo-CamposS,ThompsonC,etal.Descriptive evaluationofmaleemergencydepartmentpatientsintheUnitedStates withgonorrheaandchlamydia. Cureus. 2020;12(10):e11244.

16.SheeleJM,ElkinsJM,MohseniMM,etal.Vaginalleukocytecountsfor predictingsexuallytransmittedinfectionsintheemergencydepartment. AmJEmergMed. 2021;49:373–7.

17.MohseniM,SimonLV,SheeleJM.Epidemiologicandclinical characteristicsoftubo-ovarianabscess,hydrosalpinx,pyosalpinx, andoophoritisinemergencydepartmentpatients. Cureus. 2020;12(11):e11647.

18.MohseniMM,BenardRB,Mead-HarveyC,etal.Sexuallytransmitted infectionsintheemergencydepartmentarenotassociatedwithholidays orschoolbreaks. AmJEmergMed. 2021;45:642–4.

19.SheeleJM,LibertinCR,FinkI,etal.Alkalineurineintheemergency departmentpredictsnitrofurantoinresistance. JEmergMed. 2022;62(3):368–77.

20.SheeleJM,NiforatosJD,ElkinsJM,etal.Predictionmodelfor gonorrhea,chlamydia,andtrichomoniasisintheemergency department. AmJEmergMed. 2022;51:313–9.

21.FoxHTandSheeleJM.Associationofmaritalstatusinthetestingand treatmentofsexuallytransmittedinfectionsintheemergency department. Cureus. 2021;13(8):e17489.

22.LawingLF,HedgesSR,SchwebkeJR.Detectionoftrichomonosisin vaginalandurinespecimensfromwomenbycultureandPCR. JClin Microbiol. 2000;38(10):3585–8.

23.BlakeDR,DugganA,JoffeA.Useofspunurinetoenhancedetectionof Trichomonasvaginalis inadolescentwomen. ArchPediatrAdolesc Med. 1999;153(12):1222–5.

24.ChapinKandAndreaS.APTIMA® Trichomonasvaginalis,a transcription-mediatedamplificationassayfordetectionof Trichomonas vaginalis inurogenitalspecimens. ExpertRevMolDiagn 2011;11(7):679–88.

25.WorkowskiKA,BachmannLH,ChanPA,etal.Sexuallytransmitted infectionstreatmentguidelines,2021. MMWRRecommRep. 2021;70(4):1–187.

26.SheeleJM,SmithJ,NiforatosJD,etal.History,physicalexamination, andlaboratory findingsassociatedwithinfectionandtheempiric treatmentofgonorrheaandchlamydiaofwomenintheemergency department. Cureus. 2019;11(12):e6482.

27.ShapiroT,DaltonM,HammockJ,etal.Theprevalenceofurinarytract infectionsandsexuallytransmitteddiseaseinwomenwithsymptomsof asimpleurinarytractinfectionstratifiedbylowcolonycountcriteria. AcadEmergMed. 2005;12(1):38–44.

ORIGINAL RESEARCH

ChangingIncidenceandCharacteristicsof

PhotokeratoconjunctivitisDuringtheCOVID-19Pandemic

Yu-ShiuanLin,MD*

Chih-ChengLai,MD†

Yu-ChangLiu,MD‡§

Shu-ChunKuo,MD*∥

Shih-BinSu,MD,PhD¶#

*ChiMeiMedicalCenter,DepartmentofOphthalmology,Tainan,Taiwan

† ChiMeiMedicalCenter,DivisionofHospitalMedicine,DepartmentofInternalMedicine, Tainan,Taiwan

‡ ChiMeiMedicalCenter,DepartmentofEmergencyMedicine,Tainan,Taiwan

§ NationalChengKungUniversity,CollegeofMedicine,DepartmentofEnvironmentaland OccupationalHealth,Tainan,Taiwan

∥ ChungHwaUniversityofMedicalTechnology,DepartmentofOptometry,Jen-Teh, Tainan,Taiwan

¶ ChiMeiMedicalCenter,DepartmentofOccupationalMedicine,Tainan,Taiwan

# ChiMeiMedicalCenter,DepartmentofFamilyMedicine,Tainan,Taiwan

SectionEditor:WirachinHoonpongsimanont,MD,MSBATS

Submissionhistory:SubmittedFebruary12,2023;RevisionreceivedJanuary11,2024;AcceptedJanuary17,2024

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.17882

Introduction: Photokeratoconjunctivitis(PKC)isprimarilycausedbywelding.However,inappropriate useofgermicidallamps,whichhavebeenwidelyusedfollowingtheCOVID-19outbreak,canalsocause PKC.OurgoalinthisstudywastoinvestigatetheincidenceofandchangesinthecausesofPKCduring thecoronavirus2019(COVID-19)pandemic.

Methods: Weconductedasingle-center,retrospectiveobservationalstudy.Thehealthrecordsof patientswhovisitedtheemergencydepartmentinatertiarycarehospitalfromJanuary1, 2018–December31,2021andwerediagnosedwithPKC,werereviewed.Wethenconductedan analysistocomparethecharacteristicsofPKCbeforeandafterCOVID-19beganandthefeaturesof PKCcausedbyweldingandgermicidallamps.

Results: Therewere160PKCcaseswithaclearetiologybeforetheCOVID-19pandemicand147cases duringtheCOVID-19pandemic.Nosignificantdifferencesinageandgenderweredetectedbetweenthe twogroups.TheincidenceofPKCinducedbytheuseofgermicidallampsduringtheCOVID-19 pandemicwassignificantlyhigher(10.2%)thantheincidencebeforethepandemic(3.1%).Theratioof femalestomalesinthegermicidallampsubgroupwassignificantlyhigherthantheratiointhewelding subgroup.Limitationsincludedincompleteinformationduetotheretrospectivenatureofthestudy, underestimationofincidence,andpossiblerecallbias.

Conclusion: IntheeraofCOVID-19,cliniciansshouldbeawareofthehazardsofgermicidallamps. AlthoughtheCOVID-19pandemicseemstoshowsignsofeasing,newinfectiousdiseasesthatrequire protectivemeasurescouldstillemergeinthefuture.Therefore,injuriesrelatedtogermicidallamps deservemorepublichealthattention.[WestJEmergMed.2024;25(3)368–373.]

Keywords: COVID-19;SARS-CoV-2;ultravioletlight;photokeratoconjunctivitis;germicidallamp; welding.

INTRODUCTION

Photokeratoconjunctivitis(PKC),orphotophthalmia,is relatedtoultravioletradiation(UVR)exposure.Exposureto ultravioletB(UV-B)andultravioletC(UV-C)candamage theocularsurface,includingthecornealorconjunctival epithelialcells.1 Theclinicalmanifestationsincludeocular painorforeignbodysensation,tearing,photophobia,and evenblurredvisioninseverecases.Thephotochemical reactiontypicallytakes6–12hourstocausesymptoms.2 Therefore,patientsoftenexperiencesymptomsatnightafter daytimeexposure,leadingtoemergencydepartment(ED) visitsatnight.3 ExposuretoUVRcanbeclassifiedinto naturalandartificialsources.Naturalsourcesincludedirect orreflectedsunlightduringskiingortimespentatthebeach.4 Artificialsourcesincludeworkplacewelding flashes,which arethemostcommoncauseofPKC,andcuringlights, printingmachines,high-techindustrialprocesses,laser engravers,andgermicidallamps.3

Thecoronavirusdisease2019(COVID-19)iscausedby thesevereacuterespiratorysyndromecoronavirus2(SARSCoV-2).TheCOVID-19outbreakbeganinDecember2019 andwasdeclaredapandemicbytheWorldHealth Organizationin2020.5,6 Tocontrolthespreadofthishighly contagiousvirus,severalpreventativeandcontrolmeasures wereimplemented,includingwearingmasks,social distancing,handhygiene,anduseofpersonalprotective equipmentanddisinfectants,suchasdilutedbleachsolutions or70%ethanol.6,7 Ultraviolentradiationwasinvestigated forirradiationofthecoronavirus8;UVRdisinfectsby damagingDNAstructures,includingviralDNA.Different wavelengthshavedifferentdisinfectingeffects.9 The potentialhazardofgermicidallampstotheeyeswas recognizedbeforetheCOVID-19pandemic.10,11 However, germicidallampuseincreasedsignificantlyduringthe COVID-19pandemic,andtheincreasedusagemayleadto morecasesofPKC.

ChangesinophthalmicEDvisitsaftertheCOVID-19 pandemicbeganwererecentlydiscussed.Adecreased numberofoveralleyeinjurieswasnotedinseveral studies.12,13 Additionally,severalstudiesreportedcasesof PKCduetogermicidallampsaftertheCOVID-19pandemic began,andaneight-weekcomparisonstudysuggestedan upwardtrend.14–16 However,long-termdataaboutthe causesofPKCaftertheCOVID-19pandemicislimited.In thisstudyweaimedtoinvestigatetheincidenceofand changesinthecausesofPKCbeforeandafterthe COVID-19pandemic.

METHODS

StudyDesign

Wedividedpatientsintotwogroups:patientswithPKC beforetheCOVID-19pandemic(betweenJanuary1, 2018–December31,2019)andpatientswithPKCafterthe COVID-19pandemicbegan(betweenJanuary1,

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Photokeratoconjunctivitis(PKC)ismainly causedbywelding.Germicidallamps,which alsocausePKCifimproperlyused,were widelyusedaftertheCOVID-19outbreak.

Whatwastheresearchquestion?

Whataretheincidenceofandchangesinthe causesofPKCduringtheCOVID-19 pandemic(overatwo-yearperiod)?

Whatwasthemajor findingofthestudy?

TheincidenceofPKCinducedbyuseof germicidallampsincreasedsigni fi cantlyafter COVID-19began(10.2%vs3.1%).

Howdoesthisimprovepopulationhealth?

Thepotentialinjuriesfromgermicidallamps deservemorepublichealthattentionboth duringtheCOVID-19eraandfornew infectiousdiseasesinthefuture.

2020–December31,2021).Demographicdata,including gender,age,month,time,etiology,occupation,andexposure time,werecollected.

Participants

Patientswhovisitedthetertiary-carecenterEDin TaiwanbetweenJanuary1,2018 –December31,2021and werediagnosedwithPKCwereenrolledinthestudy. Photokeratoconjunctivitiswasdiagnosedaccordingtothe followingcriteria:UVexposurehistorywithinoneday; typicalsymptoms,includingocularforeignbodysensation, pain,photophobia,andtearing;andophthalmicclinical fi ndings,suchasconjunctivalhyperemiaandcorneal super fi cialpunctatelesions.Weexcludedpediatricpatients, giventhepossibilityofuncooperativephysicalandocular examination.BasedonthecausesofPKCmentionedinthe medicalrecords,patientsweredividedintothreegroups basedontheetiology:germicidallamps;welding;andother. Patientswhosemedicalrecordsspeci ficallymentionedthe useofgermicidallampswereassignedtothegermicidal lampgroup,patientswhosemedicalrecordsmentioned exposuretoweldingwereassignedtotheweldinggroup, andthosewhosecauseswereunknownorrelatedtodirector re fl ectivesunlightwereassignedtotheothersgroup. ThisstudywasapprovedbytheInstitutionalReviewBoard

ofChiMeiMedicalCenter,Tainan,Taiwan(Applicant’ s No:11108-010).

OutcomeMeasures

Theprimaryoutcomeofthestudywastheincidenceand causesofPKCbeforeandafterCOVID-19began.The secondaryoutcomewasthecharacteristicsofthegermicidal lampandweldinggroups.

StatisticalAnalysis

WeanalyzeddatausingSPSSStatisticsforWindows, version22(IBMCorp,Armonk,NY).ThePearsonchisquaredtestandFisherexacttestwereusedtocompare categoricalvariables.Continuousvariableswerecompared usingtheStudent t -test.Thethresholdforstatistical significancewasdefinedasa P -valuelessthan0.05.

RESULTS

PatientCharacteristics

Duringthestudyperiod,307PKCpatientswererecruited. FewercasesofPKCoccurredafterCOVID-19began(147vs 160)comparedwiththenumberofPKCcasesbeforethe COVID-19period.Themeanpatientageswere41.85 ± 0.97years(range,20–71years)beforeCOVID-19and 40.07 ± 1.09years(range,20–78years)afterCOVID-19.No significantdifferencesinageorgenderweredetectedbetween thegroupsbeforeandafterCOVID-19began.Amajorityof patientswenttotheEDatnight;90%wenttotheED between8 PM–07:59 AM,andthemostprevalentperiodwas 12 AM–03:59 PM.ThecharacteristicsofthepatientswithPKC aresummarizedin Table1.

IncidenceandDemographicDataofGermicidalLamprelatedPKC

ThetotalnumberofPKCcasesslightlydecreasedafterthe COVID-19pandemicbegan.TheetiologiesofPKCwere differentbeforeandafterCOVID-19began.Thepercentage ofpatientsinthegermicidallampgroupbeforeCOVID-19 (5,3.1%)waslowerthanthepercentageofpatientsinthe germicidallampgroupafterCOVID-19began(15,10.2%); thus,theincidenceofPKCinthegermicidallampgroup increasedsignificantlyafterCOVID-19began(P = 0.03). Withinthegermicidallamp-relatedPKCsubgroup,themean ageswere39.20 ± 3.69years(range,31–51years)and42.73 ± 3.88years(range,21–78years)beforeandafterCOVID-19 began,respectively,andnosignificantdifferencesinageor genderweredetectedbetweenthegroups(Table2).All patientsinthegermicidallampgroupwenttotheED between8 PM–07:59 AM

ComparisonBetweentheWeldingandGermicidal LampGroups

MostpatientswithPKCweremales(morethan90% beforeandafterCOVID-19began),andmostpatientsinthe

Table1. Characteristicsofpatientswithphotokeratoconjunctivitis.

Before COVID-19 (N = 160)

After COVID-19 (N = 147) P value

Age:mean ± SD41.85 ± 0.9740.07 ± 1.090.22

Gender(%)0.88

Male154(96.3)141(95.9)

EtiologyofPKC(%) P = 0.03

Welding144(90.0)118(80.3)

Germicidallamp5(3.1)15(10.2)

Othercauses11(6.9)14(9.5)

TimeofEDvisit(%)0.89

00:00–03:59100(62.5)93(63.3)

04:00–07:5918(11.3)16(10.9)

08:00–11:592(1.3)2(1.4)

12:00–15:591(0.6)2(1.4)

16:00–19:592(1.2)3(2.0)

20:00–23:5937(23.1)31(21.1)

COVID-19,coronavirusdisease2019; ED,emergencydepartment; PKC,photokeratoconjunctivitis.

Table2. Demographicdataofpatientswithgermicidallamp-induced photokeratoconjunctivitisbeforeandafterCOVID-19began. Before COVID-19 (n = 5) After COVID-19 (n = 15)

Male2(40.0)10(66.7)

TimeofEDvisit(%)0.51

00:00–03:592(40)10(66.67)

04:00–07:591(20)1(6.67)

08:00–11:590(0)0(0)

12:00–15:590(0)0(0)

16:00–19:590(0)0(0)

20:00–23:592(40)4(26.67)

*Fisherexacttest.

COVID-19,coronavirusdisease19; ED,emergencydepartment; PKC,photokeratoconjunctivitis.

weldingsubgroupweremales(Tables1 and 3).However,the percentageoffemalesinthegermicidallampsubgroupwas higherthanthepercentageoffemalesintheweldinggroup (P < 0.001).ThetimespatientswenttoEDwerenot significantlydifferentbetweenthegermicidallampgroup andtheweldinggroup.

Table3. Thecomparisonbetweenthephotokeratoconjunctivitis subgroupsofgermicidallampandwelding.

InformationAboutIndividualsintheGermicidal LampGroup

Informationaboutpatientsinthegermicidallampgroup ispresentedin SupplementalTables1and2).Theexposure durationrangedfromafewsecondstotwohours.The domesticcomponentaccountedfor60%(3/5)ofthePKC casesbeforeCOVID-19and40%(6/15)ofthecasesafter COVID-19began.BeforeCOVID-19,theoccupational componentincludedmedicalpersonnel/staff(2/5).After COVID-19began,theoccupationalcomponentexpandedto medicalstaff(3/15),workersintherestaurantandhotel industry(3/15),acleaner(1/15),aschoolemployee(1/15), andaconstructionindustryworker(1/15).

DISCUSSION

Inthisstudywecomparedtheincidenceandcausesof PKCbeforeandafterCOVID-19began.Weselected January1,2020,asthestartingpointfortheobservation periodfortwoprimaryreasons.The firstreasonwasthe

geographicalproximitybetweenTaiwanandMainland Chinaandthefrequentbusinesstripsbetweenthetwo nations.Secondly,basedonthepreviouspainfulexperience oftheSARSoutbreakinTaiwanin2003,ourgovernment andpeopletreatedthisincidentwithgreatcautionatavery earlystage.AtapressconferenceonDecember31,2019,the TaiwanMinistryofHealthandWelfare,announced epidemicinformationandinitiationofaborderquarantine inaccordancewithstandardprocedure.

Theproportionofgermicidallamp-relatedPKCcases significantlyincreasedafterCOVID-19began.Theincrease inPKCcasesisattributedtotheincreasednumberof germicidallamp-relatedPKCcasesandthedecreased numberoftotalPKCcases.ThenumberofgermicidallamprelatedPKCcaseslikelyincreasedafterCOVID-19began duetotheincreaseduseoftheselampsfordisinfection. Previousstudiesconcerninggermicidallamp-relatedPKC casesafterCOVID-19beganareshownin Table4.Leung reportedthreecases(sixeyes)inHongKongandSengillo reportedsevencases(14eyes)intheUnitedStates.Wangetal comparedgermicidallamp-relatedPKCcaseseightweeks beforeandeightweeksafterCOVID-19beganandreported thepercentageofPKCduetodisinfectionincreased significantlyfrom9.1%to56.9%afterCOVID-19began.14 Wangetalalsomentionedthatthecasenumberdecreased substantiallyaftergoodpublichealtheducation.InTaiwan, germicidallampshadproductinstructionsandwarnings aboutimproperuse.Thenewsmediaalsoemphasizedthe hazardsofgermicidallamps.Despitethespreadofpublic educationontheuseofgermicidallamps,somepatientswere

Table4. Summaryofrecentstudiesaboutgermicidallamp-inducedphotokeratitisafterCOVID-19pandemicbegan.

Reference

Wang 202114 109cases inChina

Sengillo 202115 7cases (14eyes) inUSA

Leung 202116 3cases (6eyes)in HongKong

Lin202215cases inTaiwan

Retrospective32.1(range, 21–54) M:F = 55:54NorecordAverage:16.7 minutes

Caseseries40(range, 24–59)

CasereportOneis17;no recordabout othertwo patients

Retrospective42.7(range, 21–78)

M:F = 5:2P3:38W UV-C germicidal lamp(AURA)

± 0.02 logMAR

P6:38W UV-C germicidal lamp(Uvlizer) 10minute-4 hoursin5 cases,2 without documentation 20/30or betterin 13/14eyes (93%) Norecord

M:F = 1:2UV-C Effective illumination area:40m2 15,20,60 minutes

All3were 6/12 bilaterally

All3were6/6 bilaterally

M:F = 10:5Norecordfewseconds to2hours,5 caseswithout documentation NorecordNorecord

UV,ultraviolet; VA,visualacuity; M:F,maletofemale; logMAR;logarithmoftheminimumangleofresolution.

unawareofthedangers.However,publiceducationmay havepreventedupwardspikesintheincidenceofPKC.

OurstudyalsoshowedthatthenumberoftotalPKC(160 before;147after)patientsandthenumberofpatientsinthe weldingsubgroup(144before;118after)slightlydecreased afterCOVID-19began.Thereareseveralexplanationsfor thisdecrease.First,patientsdidnotgototheEDdueto concernsaboutSARS-CoV-2.17,18 Second,theoutbreak forcedmanycompaniesandfactoriestohaltproduction (lockdown),whichpredominantlyinfluencedshort-termor part-timeworkers.AccordingtoastudybyYenetal,3 longtermworkerscompliedwithsafetyregulationsbetterthan short-termworkersandworeprotectiveequipmentmore. Therefore,theweldingcasesdecreased,andgermicidallamprelatedPKCcasesincreased,leadingtothesignificantly increasedproportionofgermicidallamp-relatedPKCcases afterCOVID-19began.

Thedomesticcomponentinthegermicidallampsubgroup declinedfrom60%(3/5)beforeCOVID-19to40%(6/15) afterCOVID-19began.Incontrast,workplacecases increasedfrom40%(2/5)beforeCOVID-19to60%after COVID-19(9/15)began.This findingimpliesthathospitals/ clinics/nursinghomes,restaurants,andhotelsrequiredmore germicidallampuse.InTaiwan,manyhotelsservedas quarantinehotels,wheregermicidallampswerefrequently usedfordisinfection.InLeung’sreport,thethreecases belongedtoclusteringathome.16 InthestudyofWang etal,14 clusteringplayedanimportantrole.Ourcasesof germicidallamp-relatedPKCwereallsporadicratherthan clusteringepisodes,whichmayindicategoodpublicpolicies andstaffsafetyeducationinmostcompanies andworkplaces.

Therearesomedifferencesbetweenwelding-associated andgermicidallamp-relatedPKC.Thewavelength emanatingfromgermicidallampsismostlyUV-C (254nanometers),andthecornea,whichabsorbsmostofthe UV-C,ispredominantlydamaged.9,19,20 Thewavelength emanatingfromweldingequipmentisintheUV-Bspectrum. Theoretically,theenergyofUV-Cisgreaterthantheenergy fromUV-Band,therefore,causesmoredamagetothe corneaatthesamedistanceandexposuretime.However, mostpreviousstudiesreportedvisualacuityasagood prognosisingermicidallamp-relatedPKCcases(Table4) Becauseitisprimarilymenwhoworkinthewelding industry,theproportionoffemalesinthegermicidal-lamp groupwassignificantlyhigherthanthefemalesinthewelding groupinourstudyandpreviousstudies.14

LIMITATIONS

Thereweresomelimitationstoourstudy.Duetoits retrospectivenature,somedata(suchasexposuretime, visualacuity,brand,andwavelengthofthemachine)was incomplete.However,thismissingdatadidnotaffectour results.Second,wemayhaveunderestimatedtheincidenceof

germicidallamp-relatedPKC.Ourstudyfocusedonthe adultpopulation;sopediatricpatientsshouldbetakeninto accountinfutureresearch.Inaddition,wedidnotanalyze patientswithoutaclearPKCcausetomaintaintherigorof ourstudy.Thus,theremightberecallbias.Inourclinical experience,patientsdeniedUVRoranyotherstronglight exposureuntiltheywerespecificallyaskedaboutexposureto germicidallampsorUV-enableddishdryers.Ifthemedical staffdidnotdirectlyaskpatientsabouttheirexposureto certainmachines,theexactcauseofthePKCwasdifficultto determine.Third,theresultsmaynotbegeneralizedtoother nationsbecauseofdifferencesinrace,culture,education status,pandemicseverity,andaccessibilitytoUVR machines.Despitetheselimitations,wehopeourstudy focusesmorepublicattentionontherelatedissuesand potentialhazards.Theeffectsofnewsmediaandpublic safetyeducationonthetrendofgermicidallamp-related PKCafter2022mayrequirefurtherstudiestoevaluate.

CONCLUSION

Germicidallamp-relatedPKCincreasedduringthe COVID-19era.Wefoundthattheincidenceincreased significantlyoveratwo-yearperiodfrom3.1%before COVID-19to10.2%afterCOVID-19began.Whileit appearsthattheCOVID-19pandemicisgraduallysubsiding, itisimportanttorecognizethatnewinfectiousdiseasesmay emergeinthefuture,necessitatingprotectivemeasures. Therefore,cliniciansshouldpayattentiontothispotential causeofPKCandtakemoreaccuratehistories.Thepotential hazardofgermicidallampsisanimportantpublichealth issuethatshouldbeemphasizedtopreventfurtherinjury fromthissourceofultravioletradiation.

ACKNOWLEDGMENTS

WewouldliketothankEnagofortheEnglishrevision.

AddressforCorrespondence:Shih-BinSu,MD,PhD,ChiMei MedicalCenter,DepartmentofOccupationalMedicine,901 ZhonghuaRoad,YongkangDistrict,TainanCity710,Taiwan. Email: shihbin1029@gmail.com

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Linetal.Thisisanopenaccessarticledistributed inaccordancewiththetermsoftheCreativeCommonsAttribution (CCBY4.0)License.See: http://creativecommons.org/licenses/by/ 4.0/

REFERENCES

1.YamJCandKwokAK.Ultravioletlightandoculardiseases. Int Ophthalmol. 2014;34(2):383–400.

2.CullenAP.Photokeratitisandotherphototoxiceffectsonthecorneaand conjunctiva. IntJToxicol. 2002;21(6):455–64.

3.YenYL,LinHL,LinHJ,etal.Photokeratoconjunctivitiscausedby differentlightsources. AmJEmergMed. 2004;22(7):511–5.

4.IzadiM,Jonaidi-JafariN,PouraziziM,etal.Photokeratitisinducedby ultravioletradiationintravelers:amajorhealthproblem. JPostgradMed. 2018;64(1):40–6.

5.MarfakA,AchakD,AziziA,etal.ThehiddenMarkovchainmodellingof theCOVID-19spreadingusingMoroccandataset. DataBrief. 2020;32:106067.

6.SharmaA,AhmadFaroukI,LalSK.COVID-19:areviewonthenovel coronavirusdiseaseevolution,transmission,detection,controland prevention. Viruses. 2021;13(2):202.

7.GunerR,HasanogluI,AktasF.COVID-19:Preventionandcontrol measuresincommunity. TurkJMedSci. 2020;50(SI-1):571–7.

8.KumarA,RajA,GuptaA,etal.PollutionfreeUV-Cradiationtomitigate COVID-19transmission. GondwanaRes. 2023;114:78–86.

9.ChangCP,LiuHH,PengCY,etal.UVRmeasurementofaUV germicidallamp. HealthPhys. 2007;92(3):242–50.

10.ZaffinaS,CamisaV,LemboM,etal.AccidentalexposuretoUV radiationproducedbygermicidallamp:casereportandrisk assessment. PhotochemPhotobiol. 2012;88(4):1001–4.

11.StrippMP,DulaneyAR,BeuhlerMC,etal.Photokeratitisamong restaurantworkers:acaseseries. AmJIndMed. 2018.Inpress.

12.YanM,XiangyuS,JingM,etal.Analysisofdiseasespectrumand epidemiologicalcharacteristicsofemergencyoculartraumaduring

outbreakofnovelcoronaviruspneumonia. OphthalmologyinChina. 2020;29(3):224.

13.LiangH,ZhangM,ChenM,etal.OculartraumaduringCOVID-19 pandemic:asystematicreviewandmeta-analysis. AsiaPacJ Ophthalmol(Phila). 2022;11(5):481–7.

14.WangY,LouJ,JiY,etal.Increasedphotokeratitisduringthe coronavirusdisease2019pandemic:clinicalandepidemiological featuresandpreventivemeasures. Medicine(Baltimore). 2021;100(24):e26343.

15.SengilloJD,KunklerAL,MedertC,etal.UV-photokeratitisassociated withgermicidallampspurchasedduringtheCOVID-19pandemic. Ocul ImmunolInflamm. 2021;29(1):76–80.

16.LeungKCPandKoTCS.Improperuseofthegermicidalrange ultravioletlampforhouseholddisinfectionleadingtophototoxicityin COVID-19suspects. Cornea. 2021;40(1):121–2.

17.TankelJ,KeinanA,BlichO,etal.Thedecreasingincidenceofacute appendicitisduringCOVID-19:aretrospectivemulti-centrestudy. World JSurg. 2020;44(8):2458–63.

18.AlenaziTH,AlArbashH,El-SaedA,etal.Identifiedtransmission dynamicsofMiddleEastrespiratorysyndromecoronavirusinfection duringanoutbreak:implicationsofanovercrowdedemergency department. ClinInfectDis. 2017;65(4):675–9.

19.WorldHealthOrganization,WorldMeteorogicalOrganization,United NationsEnvironmentProgramme,etal. GlobalSolarUVIndex: APracticalGuide.Geneva,Switzerland:WorldHealth Organization;2002.

20.McCartyCAandTaylorHR.Areviewoftheepidemiologicevidence linkingultravioletradiationandcataracts. DevelopOphthalmol 2002;35:21–31.

BrianT.Wessman,MD*†

JulianneYeary,PharmD‡

HelenNewland,PharmD‡

RandyJotte,MD*

ORIGINAL RESEARCH

*WashingtonUniversityinSaintLouis,DepartmentofEmergencyMedicine, St.Louis,Missouri

† WashingtonUniversityinSaintLouis,DepartmentofAnesthesiology, DivisionofCriticalCareMedicine,St.Louis,Missouri

‡ BJCHealthCare,DepartmentofPharmacyServices,St.Louis,Missouri

SectionEditor:AsitMisra,MD,MSDEd,CHSE

Submissionhistory:SubmittedMay22,2023;RevisionreceivedJanuary4,2024;AcceptedJanuary9,2024

ElectronicallypublishedApril30,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18325

Introduction: Coronavirus2019(COVID-19)inequitablyimpactedminoritypopulationsandregionswith limitedaccesstohealthcareresources.TheBarnes-JewishEmergencyDepartmentinSt.Louis,MO, servessuchapopulation.TheCOVID-19vaccineisanavailabledefensetohelpachievecommunity immunity.Theemergencydepartment(ED)isapotentialsocietalresourcetoprovideaccesstoa vaccinationintervention.OurobjectiveinthisstudywastodescribeandevaluateanovelEDCOVID-19 vaccineprogram,includingitsimpactonthelocalsurroundingunderservedcommunity.

Methods: Thiswasaretrospective,post-protocolimplementationreviewofanEDCOVID-19 vaccinationprogram.Overtheinitialsix-monthperiod,wecompileddataonallvaccinatedpatientsoutof theEDtoevaluatedemographicdataandtheimpactonunderservedregionalareas.

Results: WereportasuccessfulED-basedCOVID-19vaccineprogram(withover1,000vaccines administered).Thisprogramhelpedraiseregionalandstatevaccinationrates.Over50%ofthe populationthatreceivedtheCOVID-19vaccinefromtheEDwerefromdefinedsociallyvulnerable patientpopulations.Noadverseeffectsweredocumented.

Conclusion: OperationCoVER(COVID-19VaccineintheEmergencyRoom)SaintLouiswasableto successfullyvaccinateasociallyvulnerablepatientpopulation.Thisfree,COVID-19ED-basedvaccine programwithdedicatedpharmacysupport,wasnovelinemergencymedicinepractice.SimilarED-based vaccineprogramscouldhelpwithfuturevaccinedistribution.[WestJEmergMed.2024;25(3)374–381.]

INTRODUCTION

Coronavirus2019(COVID-19) firstimpactedtheUnited Statesinearly2020.ByFebruary2021,morethen500,000 individualshaddiedintheUSafterbecominginfected.1–4 Variousstrategieswereemployedtolimitthespreadofthe virusincludingcommunitylockdowns,socialdistancing, contacttracing,andmasking,withvariedsuccessand waningadherenceovertime.5 TheCOVID-19pandemic inequitablyimpactedminoritypopulationsandregionswith limitedaccesstohealthcareresources.6–8 TheBarnes-Jewish

HospitalEmergencyDepartment(BJHED)inSaintLouis, MO,staffedbyWashingtonUniversitySchoolofMedicine inSt.Louisemergencyphysicianswithdedicatedpharmacy supportservessuchapopulationforthebi-stateregionof MissouriandIllinois.

Oneofthestrongestdefensesagainstthisnovelvirusis vaccines.Developmentandmorewidespreaddistributionof vaccinesbeganinSpring2021.BySeptember2021,COVID19vaccinationwasestimatedtoprevent56%ofexpected hospitalizationsand58%ofexpecteddeaths.9 Toachieve

community or “herd” immunityandtherebyreducing theriskofcommunityspread,approximately67–90%ofthe populationmustachieveimmunity,eitherbyvaccinationor infection.However,vaccinehesitancy,miseducation,and lackofaccesstovaccinesaremajorbarrierstoachievingthis herdimmunitygoal.10–16 Asthesafetynetformany communities,theemergencydepartment(ED)providesa multitudeofhealthcare,educational,andsocialservices.17,18 WehypothesizedthattheEDcouldalsoplayapivotalrole withvaccineeducation,distribution,andaccessforan underservedpopulation.Priorstudieshaveevaluatedthe theoreticalbenefitofusingtheEDasapotentialvaccination resourcesite.19–25 Atthetimeofprojectinitiation,thestateof Missourirankednationallyinthebottom10ofstatesfor populationvaccinationrates,presentingopportunityfor improvement.4

WestartedofferingCOVID-19vaccinestopatients presentingtoBJHEDonJuly21,2021,initiallyboththe PfizerandJohnson&Johnson(J&J)vaccines.Ofnote,this initiationdatewaswellintothedeltavariantsurgeofthe pandemic.Additionally,vaccinationwasapprovedand availableforthepublicthroughotherpublichealthsources. Toourknowledge,thisfreeCOVID-19vaccineprogramby emergencyphysicianswithpharmacysupport,basedoutof anED,wasnovelinemergencymedicinepractice.Wenamed ourproject:OperationCoVER(COVID-19Vaccineinthe EmergencyRoom)STL(SaintLouis).

METHODS

CollaborationontheEDvaccinationimplementation projectbetweenBJHEDhospitaladministration,Washington Universityemergencyphysicians,andthepharmacy departmentbeganinSpring2021.Ona24/7basis,theEDteam offersat-the-momenthealthcarewithconfidentiality,patientcenterededucation,andaccesstofollow-upresources (includingcompletionoftheinitialvaccinationseries,if indicated).Suchavailabilitydifferedfromothercommunity resources.Barriersthroughouttheprocesswereidentifiedand resolved.TheseincludedPfizervaccinestoragerequirements (ultra-lowtemperaturefreezer[ 80°Cto 60°C]);avoidanceof vaccinewastage(asonceavialwasdiluted,contentshadtobe usedwithinsixhours);handdeliveryfrominpatientpharmacy totheED;administrationofthevaccinewithintwohoursfrom vialextraction;record-keepingofappropriateviallotnumber; expirationdateandinjectionsite;andclinician/nursetraining. Vaccineeducationwasprovidedtoourphysicians,nurse practitioners,physicianassistants,andnurses.Allwere encouragedtooffereverypatienttheCOVID-19 vaccination.Signage,educationalmaterials,andadvertising weredevelopedanddistributedtoraiseawarenessoffree vaccinationaccessintheED.Schedulingsubsequentdosesto completetheinitialvaccineserieswasfacilitatedbyour dischargenursecoordinators.Weeklyemailreminders trackingvaccinesadministeredwerecirculatedtothe

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

TheCOVID-19vaccineisanavailable defensetohelpachievecommunityimmunity. TheEDisapotentialsocietalresourceto provideaccesstoavaccinationintervention.

Whatwastheresearchquestion?

Ourgoalwastodescribeandevaluateanovel EDCOVID-19vaccineprogram,including itsimpactonthelocalsurrounding underservedcommunity.

Whatwasthemajor findingofthestudy?

ThisED-basedCOVID-19vaccineprogram resultedinover1,000vaccinesadministered.

Howdoesthisimprovepopulationhealth?

Theprogramhelpedraiseregional/state vaccinationrates.Over50%ofthose whoreceivedthevaccinefromtheED werefromde fi nedsociallyvulnerable patientpopulations.

WashingtonUniversityemergencyphysician/nurse practitionerandphysicianassistant/residentgroup.

Basedonregionalvaccinationratesandthehealthcare accessofourpatientpopulation,anassumptionwasmade thatapproximatelyone-thirdofpatientswouldarrive vaccinated.Alsoconsideringcriticalillness/trauma presentation,acuteillness,vaccinehesitancy,andclinician forgetfulness,weanticipatedanotherone-thirdofpatients wouldnotbeavailabletoconsentforvaccination.Ofthe remainingpatients,avaccinationgoalrateof5–10% (approximately5–15patientsaday)wasencouraged.This wasdiscussedatlengthwithpharmacytosupportthe componentofvaccinestorage,preparation,and administrationinatimelyfashionfortheED patientpopulation.

Allpatientswererequiredtoconsenttoreceivingthe vaccine,whichwasdocumentedelectronicallyuponorder entrybytheclinician.Patientsnoteligibletoreceivethe vaccineincludedthosewithacontraindicationtothevaccine, anactiveCOVID-19infection,orthosewithadocumented COVID-19infectionwithintherecentpast(current recommendationofpriorseven-dayperiod).Allvaccines werekeptinacentralizedpharmacylocationtomeetstorage requirementsofboththePfizerandJ&Jvaccines.Pharmacy staffwithdrewdosesfortherequestedvaccineandhand deliveredittotheEDbedsidenursealongwithvaccinevial

information(manufacturer,expirationdate,lotnumber,and timeofdosewithdrawal)andablankstandardizedCOVID19vaccinecard(issuedbytheUSCentersforDisease ControlandPrevention[CDC]).

Thebedsidenursewouldadministerthevaccinedoseas soonaspossibleinviewofthetwo-hourlimitbetweenvial withdrawalandadministration.Nursesalsoprovidedpatient educationregardingpotentialsideeffectsandadverseevents afterreceivingthevaccine.Finally,nursesprovidedeach vaccinatedpatientwitheducationregardingfollow-up requirements.UponBJHEDdischarge,patientswould receivevaccineinformationsheetsandscheduling informationforthesecondvaccinedeadline,ifapplicable. Dischargenursecoordinatorswouldreceiveareportofall patientswhoreceivedtheir firstvaccineintheseriesand wouldcallpatientstoconfirmtheyhadasecondvaccine completedorscheduled,asapplicable.

Thiswasaretrospectivepost-protocolimplementation reviewofallBJHEDpatientsreceivingtheCOVID-19 vaccinethroughOperationCoVERSTLbetweenJuly1, 2021–January20,2022.Wereportimpactonvaccineefforts forvariousdemographicsofourregion.Datawascollected fromtheelectronichealthrecord.Weanalyzedadditional CDCdatatocomparevaccineregionaluptake.Specifically, theSocialVulnerabilityIndex(SVI)wascollectedfromCDC data,specifictoourpatientpopulation’saffectedarea. Sociallyvulnerablepopulationsareespeciallyatriskduringa publichealthemergencyduetofactorssuchas socioeconomicstatus,householdcomposition,minority status,accesstotransportation,housingtype,andlackof resources. 5–7 TheCDCusesthisindextohelpdetermine wheretoleveragehealthcareresourcestohelpalleviate humansufferingandeconomicloss(estimatesupplies,need foremergencyshelters,evacuationplanning,required emergencypersonnel).TheSVIdatabasewasimportant duringtheCOVID-19outbreaktodeterminewhich communitieswouldbeaffectedmoreandrequireadditional support(ie,vaccineimplementation).26 Thedataisfurther dividedintoquartilesinwhichquartile “A” representsthe lowest/leastlevelofvulnerabilityandquartile “D” represents thehighest/most.Weobtainedappropriateinstitutional reviewboardapproval(classifiedas “exempt”)toconduct thisretrospectivestudyatourinstitution.

RESULTS

Atotalof874COVID-19vaccinedoseswereadministered betweenJuly21,2021–January20,2022(averageof4.78 vaccinedosesperday).Thetotalnumberofimpacted patientswas824individuals.(Aminorityofpatientsusedthe EDfortheirsecondvaccinedoseadministration.)Themean patientagewas44.4yearsold(±15.6years).Thedistribution inraceincluded76%(626/824)Black,27.2%(224/824) White,and2.91%(24/824)AmericanIndian,Asian,or “other” ethnicitypatients(Table1, Figure1).

Table1. Demographicdataonpatientswhoreceivedthe COVID-19vaccine.

N = 874vaccinesadministered;N = 824patient

Meanage(years)44.4 ± 15.6

MeanEDduration(hours)7.49 ± 5.05

Admitted(yes%[n])21.6[189] DischargedfromED(Yes%[n])78.4[685] Deceased(yes%[n])0.11[1] MeannumberofEDvisitsinprior5years11 Gender(female%[n])45.3[396] Race(%[n])

Black76.0[626/824] White27.2[224/824] AmericanIndian0.73[6/824] Asian0.24[2/824]

Unabletoanswer1.94[16/824]

Meanweight(kg)82.38 ± 24.1

Meanheight(cm)170.6 ± 10.64

Insurancestatus

Self-pay%[n]29.5[258] Insurance%[n]70.5[616] MOMedicaid%[n]22.3[195] MOmanagedcare%[n]10.1[88]

Primarycareprovider

Yes%[n]51.4[449] No%[n]48.6[425]

COVID-19vaccinegiven Pfizer%[n]81.1[709] Johnson&Johnson%[n]18.9[165]

Historyof +COVID-19priortovaccination (Yes%[n])

8.7[76]

COVID-19+ aftervaccination(Yes%[n])4.9[43]

Timeofvaccinegivenpershift

1st shift(0700–1500)%[n]42.4[371] 2nd shift(1501–2300)%[n]30.8[269]

3rd shift(2301–0659)%[n]26.0[227]

Medicationsgiven

EpiPen%[n]0.23[2]

Diphenhydramine%[n]1.03[9]

Steroids%[n] (Methylprednisolone,prednisolone, prednisone,ordexamethasone)

Patientaddress/homestates

0.80[7]

Missouri89.7[784] Illinois9.61[84] Indiana0.11[1]

(Continuedonnextpage)

Texas0.11[1]

Unknown0.11[1]

CDCData

Missouridata

AveragepercentageoftheMOpopulationthat received1doseofanyCOVID-19vaccine by7/21/21

AveragepercentageoftheMOpopulationthat completedthevaccineseriesby7/21/21

AveragepercentageoftheMOpopulationthat received1doseofanyCOVID-19vaccine by1/20/22

AveragepercentageoftheMOpopulationthat completedtheseriesby1/20/22

Countydata

AveragepercentageoftheSt.LouisCity Countypopulationthatreceived1doseofany COVID-19vaccineby7/21/21

AveragepercentageoftheSt.LouisCity Countypopulationthatcompletedtheseries by7/21/21

AveragepercentageoftheSt.LouisCity Countypopulationthatreceived1doseofany COVID-19vaccineby1/20/22

AveragepercentageoftheSt.LouisCity Countypopulationthatcompletedtheseriesby 1/20/22

AveragepercentageoftheSt.LouisCounty populationthatreceived1doseofany COVID-19vaccineby7/21/21

AveragepercentageoftheSt.LouisCounty populationthatcompletedtheseries by7/21/21

AveragepercentageoftheSt.LouisCounty populationthatreceived1doseofany COVID-19vaccineby1/20/22

AveragepercentageoftheSt.LouisCounty populationthatcompletedtheseries by1/20/22

SocialVulnerabilityIndex (timeframe:7/21/21 – 1/20/22) A(0–

C(0.5001–0.75)36%

D(0.7501–1.0)16%

ED,emergencydepartment; MO,Missouri; COVID-19,coronavirus 2019; CDC,USCentersforDiseaseControlandPrevention.

NumberofBarnes-JewishHospitalEmergency Departmentpatientsvaccinatedbyrace.

Thegeographicdistribution(basedonlistedhomeZIPcode) included89.7%(784/874)ofMissouripatientsand9.61%(84/ 874)ofIllinoispatients(Figure2).Otherrepresentedstates includedIndiana,Kentucky,Mississippi,Tennessee,and Texas.Approximately22%ofvaccinatedpatientswere admitted,and78%weredischargedfromtheED.Themean numberofEDvisitsintheprior fiveyearsperpatientinthe vaccinatedcohortwas11totalEDvisits.

AtthetimeoftheirBJHEDvisitandvaccination,29.5% (258/874)ofpatientslackedhealthinsurance.Ofthe70.5% (616/874)ofpatientswithinsurance,22.3%(195/874)had MissouriMedicaidand10.1%(88/874)hadMissouri ManagedCare,bothofwhichprovidemedicalinsuranceto lowerincomehouseholds.AtthetimeoftheirBJHEDvisit, 51.4%(449/874)ofpatientshadaknownprimary carephysician.

Duringthestudiedtimeframe,16%ofthepatients vaccinatedbytheBJHEDvaccineprogramlivedinareasof highsocialvulnerability(quartileDoftheSVI),withan additional37%residinginareasofmedium-highsocial vulnerability(quartileC).Altogether,greaterthan50% (51.3%)ofthepatientsimpactedbytheBJHEDvaccine administrationprogramwerefromareasofmedium-high andhighsocialvulnerability(Figure3).Seeincludedmaps (Figure2)demonstratinggeographicimpactonourregion (MissouriandIllinois).

DatafromSaintLouisCityandSaintLouisCounty(the twolargestsurroundingregions)showeda21.1%increasefor St.LouisCityandan18.2%increaseforSt.LouisCountyfor patientsreceivingatleastonedoseoftheCOVID-19vaccine overthetemporalperiodofOperationCoVERSTL.Wealso revieweddataonadverseoutcomes,specificallyreviewingall medicationsprovidedduringeachpatientencounter.Useof agentsforanaphylacticreactions(epinephrine, corticosteroids,antihistamines)werelimitedinthepatient cohort.Twopatientsreceivedepinephrine0.3milligrams intramuscularinjectionsduringtheirEDstay;however,both wereunrelatedtothevaccineadministration(onepresented totheEDafteraninsectstingandanotherwithangioedema asthepresentingchiefcomplaint,priortoreceivingtheir COVID-19vaccineatdischarge).Wewereunabletoassess

Figure1.

Figure2. GeographicdistributionofvaccinatedBarnes-JewishemergencydepartmentpatientsbylistedZIPcode(forstatesofMissouri andIllinois).

Figure3. Socialvulnerabilityindexforimpactedpostalcodesfor vaccinatedBarnes-Jewishemergencydepartmentpatients(Ais low/leastvulnerableandDishigh/mostvulnerable).

forotherpotentialadverseeventssuchaspericarditisorlocal siteirritation;however,wedidnotrecordanyrepeatvisitsin thispatientcohortforthesepresentingdiagnoses.

DISCUSSION

OperationCoVERSTLisanovel,ED-basedvaccination programthatmeetstheneedsofanunderservedcommunity withahighsocialvulnerabilityrisk.TheWashington UniversityDepartmentofEmergencyMedicineservesasthe locusofprimarycareformanyofourregionalpatients.The BJHEDcensusaverages185–240patientsdaily,with upwardof80,000adultpatientvisitsperyear.Emergency cliniciansareadeptatordering,administering,and documentingvaccines;themostcommonexampleisthe tetanus,diphtheria,andpertussisvaccine,whichis administeredalmostdailyintheEDforopen-wound prophylaxisintraumapatients.Wehavepreviouslybeen

involvedwithotherpublicvaccinationeffortsincluding offeringtheinfluenzavaccineinprior “fluseasons, ” although withvariablesuccess.

TheaveragenumberofEDvisitsperpatientsinthis vaccinatedcohortwas11(overtheprior fiveyears), demonstratingtheuniqueroletheBJHEDservesfor healthcareinourregionalcommunity.Populationswithinour communityaredependentontheBJHEDtoreceivemuchof theirhealthcare,reflectingwhyOperationCoVERSTLwas impactful.Thispracticeissimilaramongotherlargeurban areas,withanEDfulfillingtheroleofcentralandessential “healthcare” deliveryforanunderservedpatientpopulation.

WeevaluatedCDCdataonvaccinationratesforCOVID19vaccineuptakeinMissouriduringourED-based initiative.Onday0ofOperationCoVERSTL,32.9%ofthe statepopulationhadreceivedonedoseofanyCOVID-19 vaccineand28.6%ofthestatepopulationhadcompletedthe COVID-19vaccineseries(two-doseregimenformRNA vaccines).OnJanuary20,2022(enddatacohortdate),this ratehadincreasedto45.2%ofthestatepopulationhaving receivedonedoseofanyCOVID-19vaccineand39.0%ofthe populationhavingcompletedtheCOVID-19vaccineseries (Figure4).ProgramssuchasOperationCoVERSTLhelped alongwithotherinitiativesandprogramstoachievethis 12.3%increaseininitialvaccinationratesfortheMissouri population(16.6%increaseincompletedvaccinationseries).

WewereabletoaccessdatafromtheCDCtoanalyzeand understandourimpactontheregion.TheSVIdatawas assessedbytheimpactedZIPcodesassociatedtotheED visit.TheZIPcodesofourpatientcohortwereinhigh-risk sociallyvulnerableregions,indicatingapproximately52%of ourvaccinationrecipientswerefromsociallyvulnerable populations.TheSVIrateshelpdemonstratethatmany patientsservedbyBJHEDandWashingtonUniversity emergencyphysiciansarethosewithahigherimpactfrom publichealthemergenciesandresideinareasinneedof additionalsupport.

WealsolookedattheracialdistributionofOperation CoVERSTLefforts.Thedistributionofvaccinesprovidedin ourcohortincluded76%Black(626/824patients)and27.2% White(224/824patients).Thiscorrespondstotheracial distributionofthegeographicareaservedbytheBJHED, whichincludesSt.Louiscityandthesurroundingbistate regionsofMissouriandIllinois.Asof2022,thecensusofSt. Louiscitydemonstratedthat “BlackorAfricanAmerican alone” madeup44.8%ofthecitypopulationand “White alone” madeup46.3%ofthecitypopulation.27 However,our hospitalcensusnumberstypicallyreflectahigherpercentageof “BlackorAfricanAmerican” patientsusingtheBJHEDto accesshealthcare.Duringthesix-monthperiodofthiscohort, theBJHEDprovidedcarefor39,570patients.Theracial distributionoftheEDpopulationincluded61.29% “Black” and33.92% “White” (Table2).Thismayagainreflecttherole theBJHEDservesforspecificpopulationsinourcity(higher SVIZIPcodes)whoaresociallyvulnerableandwhyOperation CoVERSTLdidprovideauniquepublichealthresource.

Wehavealsobeguntolookatclinicianattitudesand supportofthisprogramthroughsurveystobetterunderstand

Table2 RacialdistributionofBarens-JewishhospitalemergencydepartmentpopulationduringOperationCoVERSTL.

Figure4. COVID-19vaccinationratesofuptakeinMissouri duringOperationCoVERSTL.

allparametersofthispilot.WehopeourED-based vaccinationprogramcanserveasamodelforotherEDswith similarsociallyvulnerablepopulations.

WehavecontinuedtoofferOperationCoVERSTL throughourBJHED.WenowofferthePfizervaccineand booster(s),ifeligible.Ofnote,wedidremoveaccesstothe J&JvaccineunderCDCpublicguidance.Wehaveexpanded ourvaccinationeffortstoincludeboosterimmunizationsfor eligiblepatients.Asweapproachedtheone-yearanniversary ofthestartofthisinitiative,wehadvaccinatedover1,236 patientsasofJanuary20,2023).

LIMITATIONS

Thisretrospectiveanalysisisnotwithoutlimitations includingitsobservationalnatureandoursingle-center analysis.Vaccinationsweregivenonaclinician-preference basis,andwereliedheavilyuponcliniciansinitiatingthe conversationofvaccineswithpatients.Ofnote,thepublic visualannouncementsofvaccineaccessintheEDwaiting room,individualpatientcareareas,andrestroomsdidlead somepatienttoinitiatethe vaccineconversationwith WashingtonUniversiyemergphysicians.Vaccinehesitancy wasnotscreenedfororassessedinthisstudy,butanecdotally wasacommonthemelimitingvaccineuptake.Withthe retrospective,blindeddesignofourdatacohort,wewere unabletoinvestigateindividualfactorsimpactingpatient vaccinationdecisions.OnJanuary20,2022(enddatacohort date),only39%oftheMissouripopulationhadcompleteda COVID-19vaccineseries(Figure4),demonstratingthatless thanhalfofourstatepopulationhadgoneforwardwitha decisiontovaccinate.Weare awareofmultipleemergency cliniciansatourinstitutionreportingpatientsrefusingtoreceive thevaccinewhenofferedasanadditionalbenefitoftheirED visit.Ouroriginalvaccinationgoalwassetat5–15vaccinesper day.Weendedupadministering4.78vaccinedoseseachday; thus,vaccinehesitancycouldhaveimpactedourdailyrates.

Duetocrowdingissues,theBJHEDhasaprolongedwait timeandlengthofstay.Itisnotuncommonformiddle-to low-acuitypatientstowait4-6hoursinthetriageareaprior tohavingaccesstoaclinicianinanEDroom.Itispossible thattheseprolongedEDtimescouldhaveimpacted vaccinationrates.Typically,anEDpatientarriveswithan acute “ emergency ” chiefcomplaint.SomeEDsmayhave fasterevaluationanddispositiontimes,duringwhichtime additionalrequirements(vaccinedefrosting,administration) maynegativelyimpactpatient flow.However,withalonger EDlengthofstay,theEDstaffmayhavemoreopportunities toengagewiththepatienttodiscussspecificconcernsabout vaccineadministration.Furthermore,thepatientmaywant togetasmanypotentialavailableservicestomaximizecare duringtheirprolongedwait.Ourlarge,academicEDhas directaccesstopharmacywithadedicatedEDclinical pharmacist.SmallerEDswithoutdirectpharmacistaccess maybelimitedwithasimilarvaccineprotocoldesign

requiringpharmacysupport.Finally,thepatientpopulation inourareaisprimarilyurban,potentiallylimiting applicabilitytoruralareas.

CONCLUSION

Herewereportonthedevelopmentandimplementation ofasuccessfulED-basedCOVID-19vaccinationprogram. Ourprogramwasabletovaccinateanunderservedpatient populationbymeetingthepatientswheretheyreceivedtheir standardhealthcare.Thisprogramcanserveasamodelfor otheremergencydepartmentslookingtoimpacttheirregions throughvaccinationefforts.Futurestudiesshouldevaluate longevityofsuchprograms,aswellaspublicperceptionand clinicianattitudes.

AddressforCorrespondence:BrianT.WessmanMD,Washington UniversityinSaintLouis,DepartmentofEmergencyMedicine, 660S.EuclidAve.,CB8054,St.Louis,MO,63110.Email: brianwessman@wustl.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Wessmanetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.YesudhasD,SrivastavaA,GromihaMM.COVID-19outbreak:history, mechanism,transmission,structuralstudiesandtherapeutics. Infection. 2021;49(2):199–213.

2.ShahriarS,RanaMS,HossainMS,etal.COVID-19:epidemiology, pathology,diagnosis,treatment,andimpact. CurrPharmDes. 2021;27(33):3502–25.

3.QuL,LiJ,RenH.COVID-19:theepidemiologyandtreatment. BrJHospMed(Lond). 2020;81(10):1–9.

4.CentersforDiseaseControlandPrevention(CDC).UnitedStates COVID-19Cases,Deaths,andLaboratoryTesting(NAATs)byState, Territory,andJurisdiction:caseslast7daysrateper100,000.2021. Availableat: https://stacks.cdc.gov/view/cdc/107545 AccessedMay12,2023.

5.CraneMA,ShermockKM,OmerSB,etal.Changeinreported adherencetononpharmaceuticalinterventionsduringtheCOVID-19 pandemic,April-November2020. JAMA. 2021;325(9):883–5.

6.KhanijahaniA,IezadiS,GholipourK,etal.Asystematicreviewofracial/ ethnicandsocioeconomicdisparitiesinCOVID-19. IntJEquityHealth. 2021;20(1):248.

7.MackeyK,AyersCK,KondoKK,etal.Racialandethnicdisparitiesin COVID-19-relatedinfections,hospitalizations,anddeaths:asystematic review. AnnInternMed. 2021;174(3):362–73.

8.WebbHooperM,NápolesAM,Pérez-StableEJ.COVID-19andracial/ ethnicdisparities. JAMA. 2020;323(24):2466–7.

9.SteeleMK,CoutreA,ReedC,etal.EstimatedmumberofCOVID-19 infections,hospitalizations,anddeathspreventedamongvaccinated personsintheUS,December2020toSeptember2021. JAMANetw Open. 2022;5(7):e2220385.

10.JoshiA,KaurM,KaurR,etal.PredictorsofCOVID-19vaccine acceptance,intention,andhesitancy:ascopingreview. FrontPublic Health. 2021;9:698111.

11.GhasemiyehP,Mohammadi-SamaniS,FirouzabadiN,etal.Afocused reviewontechnologies,mechanisms,safety,andefficacyofavailable COVID-19vaccines. IntImmunopharmacol. 2021;100:108162.

12.ShenSCandDubeyV.Addressingvaccinehesitancy:clinicalguidance forprimarycarephysiciansworkingwithparents. CanFamPhysician. 2019;65(3):175–81.

13.SevinAM,RomeoC,GagneB,etal.Factorsinfluencingadults’ immunizationpractices:apilotsurveystudyofadiverse,urban communityincentralOhio.2016.Availableat: https://doi.org/10.1186/ s12889-016-3107-9.AccessedMay12,2023.

14.AttwellK,NavinMC,LopalcoPL,etal.Recentvaccinemandatesinthe UnitedStates,EuropeandAustralia:acomparativestudy. Vaccine. 2018;36(48):7377–84.

15.SmithMJ.Promotingvaccineconfidence. InfectDisClinNorthAm. 2015;29(4):759–69.

16.PaltielAD,SchwartzJL,ZhengA,etal.ClinicaloutcomesofaCOVID19vaccine:implementationoverefficacy. HealthAff(Millwood). 2021;40(1):42–52.

17.LaneBH,MallowPJ,HookerMB,etal.TrendsinUnitedStates emergencydepartmentvisitsandassociatedchargesfrom2010to 2016. AmJEmergMed. 2020;38(8):1576–81.

18.WallsCA,RhodesKV,KennedyJJ.Theemergencydepartmentas usualsourceofmedicalcare:estimatesfromthe1998NationalHealth InterviewSurvey. AcadEmergMed. 2002;9(11):1140–5.

19.RodriguezRM,TorresJR,ChangAM,etal.Therapid evaluationofCOVID-19vaccinationinemergencydepartments forunderservedpatientsstudy. AnnEmergMed. 2021;78(4):502–10.

20.ChenG,KazmiM,ChenD,etal.Identifyingassociationsbetween influenzavaccinationstatusandaccess,beliefs,andsociodemographic factorsamongtheuninsuredpopulationinSuffolkCounty,NY. J CommunityHealth. 2020;45(6):1236–41.

21.GuM,TaylorB,PollackHA,etal.ApilotstudyonCOVID-19vaccine hesitancyamonghealthcareworkersintheUS. PLoSOne. 2022;17(6):e0269320.

22.KhubchandaniJ,SharmaS,PriceJH,etal.COVID-19vaccination hesitancyintheUnitedStates:arapidnationalassessment. JCommunityHealth. 2021;46(2):270–7.

23.ReiterPL,PennellML,KatzML.AcceptabilityofaCOVID-19vaccine amongadultsintheUnitedStates:Howmanypeoplewouldget vaccinated? Vaccine. 2020;38(42):6500–7.

24.KaighC,BlomeA,SchreyerKE,etal.Emergencydepartment-based hepatitisAvaccinationprograminresponsetoanoutbreak. WestJEmergMed. 2020;21(4):906–8.

25.KapurAKandTenenbeinM.Vaccinationofemergency departmentpatientsathighriskforinfluenza. AcadEmergMed. 2000;7(4):354–8.

26.DasguptaS,BowenVB,LeidnerA,etal.Associationbetweensocial vulnerabilityandacounty’sriskforbecomingaCOVID-19hotspot UnitedStates,June1–July25,2020. MorbMortalWklyRep. 2020;69(42):1535–41.

27.UnitedStatesCensusBureau.QuickFacts:St.Louis,Missouri. Availableat: https://www.census.gov/quickfacts/fact/table/ stlouiscitymissouri/PST045222.AccessedSept25,2023.

ORIGINAL RESEARCH

SexuallyTransmittedInfectionCo-testinginaLargeUrban EmergencyDepartment

JamesS.Ford,MD*

JosephC.Morrison†

JennyL.Wagner,PhD,MPH‡ DishaNangia§

StephanieVoong,MPH∥

CynthiaG.Matsumoto,PhD¶ TasleemChechi,MPH∥

NamTran,PhD# LarissaMay,MD,MSPH∥

*UniversityofCaliforniaSanFrancisco,DepartmentofEmergencyMedicine, SanFrancisco,California

† UniversityofCaliforniaDavis,Davis,California

‡ CaliforniaStateUniversity,DepartmentofPublicHealth,Sacramento,California

§ UniversityofCaliforniaDavis,SchoolofMedicine,Sacramento,California

∥ UniversityofCaliforniaDavisHealth,DepartmentofEmergencyMedicine, Sacramento,California

¶ UniversityofCaliforniaDavisHealth,LearningHealthSystem,Departmentof PopulationHealthandAccountableCare,Sacramento,California

# UniversityofCaliforniaDavisHealth,DepartmentofPathologyandLaboratory Medicine,Sacramento,California

SectionEditor:JuanAcosta,DO,MS

Submissionhistory:SubmittedJuly19,2023;RevisionreceivedNovember17,2023;AcceptedDecember22,2023

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18404

Introduction: Theincidenceofsexuallytransmittedinfections(STI)increasedintheUnitedStates between2017–2021.ThereislimiteddatadescribingSTIco-testingpracticesandtheprevalenceofSTI co-infectionsinemergencydepartments(ED).Inthisstudy,weaimedtodescribetheprevalenceof co-testingandco-infectionofHIV,hepatitisCvirus(HCV),syphilis,gonorrhea,andchlamydia,inalarge, academicED.

Methods: Thiswasasingle-center,retrospectivecross-sectionalstudyofEDpatientstestedforHIV, HCV,syphilis,gonorrheaorchlamydiabetweenNovember27,2018–May26,2019.In2018,thestudy institutionimplementedanED-basedinfectiousdiseasesscreeningprograminwhichanypatientbeing testedforgonorrhea/chlamydiawaseligibleforopt-outsyphilisscreening,andanypatient18–64years whowashavingblooddrawnforanyclinicalpurposewaseligibleforopt-outHIVandHCVscreening.We analyzeddatafromallEDpatients ≥13yearswhohadundergoneSTItesting.Theoutcomesofinterest includedprevalenceofSTItesting/co-testingandtheprevalenceofSTIinfection/co-infection.We describedatawithsimpledescriptivestatistics.

Results: Duringthestudyperiodtherewere30,767EDencountersforpatients ≥13years(meanage: 43 ± 14years,52%female),and7,866(26%)weretestedforatleastoneofHIV,HCV,syphilis, gonorrhea,orchlamydia.Weobservedthefollowingtestingfrequencies(andprevalenceofinfection): HCV,7,539(5.0%);HIV,7,359(0.9%);gonorrhea,574(6.1%);chlamydia,574(9.8%);andsyphilis,420 (10.5%).Infectiousetiologieswithuniversaltestingprotocols(HIVandHCV)madeupthemajorityofSTI testing.Inpatientswithsyphilis,co-infectionwithchlamydia(21%,9/44)andHIV(9%,4/44)washigh.In patientswithgonorrhea,co-infectionwithchlamydia(23%,8/35)andsyphilis(9%,3/35)washigh,andin patientswithchlamydia,co-infectionwithsyphilis(16%,9/56)andgonorrhea(14%,8/56)washigh. PatientswithHCVhadlowco-infectionproportions(<2%).

Conclusion: PrevalenceofSTIco-testingwaslowamongpatientswithclinicalsuspicionforSTIs; however,co-infectionprevalencewashighinseveralco-infectionpairings.Futureeffortsareneededto improveSTIco-testingratesamonghigh-riskindividuals.[WestJEmergMed.2024;25(3)382–388.]

Keywords: emergencydepartment;sexuallytransmittedinfection;sexuallytransmitteddisease;public health;humanimmunodeficiencyvirus;HIV;hepatitisCvirus;syphilis;gonorrhea;chlamydia.

INTRODUCTION

Anestimatedonein fiveindividualsintheUnitedStates (US)areinfectedwithasexuallytransmittedinfection (STI).1,2 Between2017–2021,theincidenceofsyphilisand gonorrheaincreasedandtheincidenceofchlamydia infectionsremainedhigh.2 Withwidespreaduseof antiretroviraltreatment,theoverallincidenceofHIVhas declinedoverthesameperiod,butincidencehasplateauedin certainhigh-riskgroups,suchaspeoplewhoinjectdrugs.3 WhilecurativetreatmentforHCVbecameavailableinthe USin2011,theincidenceofHCVdoubledbetween 2013–2020.4,5 Moreover,just33%ofthosewithchronic HCVhavebeencured,andlessthan17%ofyoung (<40years),uninsuredpatientshaveachievedsustainedviral clearance.6 Lowtestingfrequencies,patientunawarenessof infection,pooraccesstotraditionaltreatmentsettings(ie, primarycareclinics)andre-infectionfollowingcureall contributetothesesub-optimaldata.6,7

Theemergencydepartment(ED)isanimportantsafety netforunderserved,high-riskpopulations,makingitavital settingtodeliverhealthcareservicestopatientswithout accesstoprimarycare.8,9 Emergencydepartment-based infectiousdiseasesscreeningprogramshavedemonstrated successinidentifyingSTIsandlinkingpatientsto treatment.10–12 ItiswellknownthatcontractionofoneSTI increasesapatient’sriskofco-infectionwithotherSTIs.13,14 OneED-basedstudyshowedthatamongpatientswho receivedtestingforSTI,co-testingforasecondSTIwasas lowas8%;however,thisstudydidnotreportprevalenceof infection/co-infection.15 OtherED-basedstudiesreport prevalenceofco-infectionbutonlysingleSTI pairings.10,15–17 UnderstandingEDSTIco-testing frequenciesandprevalenceofco-infectionsisimperativefor optimizingpublichealthinfectiousdiseasesurveillanceand treatment,particularlyamongpatientswithoutaccessto traditionalprimarycareservices.Inthisstudy,weaimedto describeco-testingandco-infectionprevalenceofHIV, hepatitisCvirus(HCV),syphilis,gonorrhea,andchlamydia, inalarge,academicED.

METHODS

Overview

In2018,thestudyinstitutionimplementedanopt-out, ED-basedinfectiousdiseasesscreeningprogramthat employedelectronichealthrecord(EHR)bestpracticealerts (BPA).Anypatientbeingtestedforgonorrhea/chlamydia waseligibleforopt-outsyphilisscreening.Additionally,any

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Dataonsexuallytransmittedinfection(STI) testingandprevalencearelimitedinthe emergencydepartment(ED)setting.

Whatwastheresearchquestion?

WhatistheprevalenceofSTItesting,cotestingandco-infectionamongEDpatients .

Whatwasthemajorquantitative finding ofthestudy?

Co-testingforSTIswasinfrequent,butcoinfectionwithchlamydiawashighamong patientswithsyphilis(21%)and gonorrhea(23%).

Howdoesthisimprovepopulationhealth?

ThisstudyhighlightstheneedtoimproveSTI co-testingratesamonghigh-riskindividuals.

EDpatient18–64yearsofagewhowashavingblooddrawn foranyclinicalpurpose,waseligibleforopt-outHIVand HCVscreening.Fundingforlabtestswasobtainedby chargingthepatient’sinsurance,abillingstrategyemployed bysimilarscreeningprogramsandstudies.18 Ifapatient requestedthattheirinsurancenotbecharged,ortheydidnot haveinsurance,testingwaspaidforbytheprogramgrant. Physicians(includingallresidents),nursepractitionersand physicianassistantscouldordertesting.Thefulldetails ofthesescreeningprogramshavebeenpreviously described.10–12 AnexampleBPAisavailablein Figure1.In thisstudyweexaminedSTItesting/co-testingfrequencies andinfectionprevalenceintheED.Asdatawasinitially collectedforqualityassurancepurposes,thestudywas deemednottobehumansubjectsresearchbythe InstitutionalReviewBoardQualityImprovementSelfCertificationTool.

StudyDesignandSetting

Thiswasaretrospective,cross-sectionalstudyofED patientstestedforHIV,HCV,syphilis,gonorrhea,or

Figure1. Exampleofabestpracticealertinsidetheelectronichealthrecord.

chlamydiaduringthesix-monthperiodfollowing implementationoftheED-basedinfectiousdiseases screeningprogram.Thestudyinstitutionwasaquaternary care,academic,LevelItraumacenterinNorthernCalifornia thatseesmorethan80,000patientvisitsannually.

SelectionofParticipants

Weincludeddatafromallpatients ≥13yearswhohad undergonetestingforoneormoreofHIV,HCV,syphilis, gonorrhea,orchlamydiaintheEDbetweenNovember27, 2018–May26,2019.

Measurements

WeabtracteddatafromtheEHR(EpicSystemsCorp, Verona,WI)usingcomputer-generatedreportsbyquerying patientswhohadreceivedEDSTItestsduringthestudy period.Weincludeddemographicfactors(age,gender,race, ethnicity)andresultsofSTItesting.Thedataanalyst responsibleforprocuringthesereportswasblindedtothe hypothesisofthestudy.WedefinedSTIco-testingastesting fortwoormoreofthefollowingSTIs:gonorrhea,chlamydia, syphilis,HIV,andHCV.Topreventduplicatedata,we includedonlyapatient’ s firstEDvisitwheretheyreceived HCVtestingwhencalculatingco-testing/co-infection prevalence.Weexaminedsubsequenttestingthatoccurredin futureEDvisitstoidentifyinstanceswherebroaderSTI testingcouldhaveidentifiedinfectionsearlier.Datawas storedinde-identifieddatasets,andpatientsweregiven uniqueidentifierstomaintainpatientconfidentiality. HIVscreeningwasperformedusingaHIVP24antigen (Ag)andHIV-1/HIV-2antibody(Ab)combinationtestwith theARCHITECTi1000SRimmunoanalyzer(Abbott Laboratories,AbbottPark,IL),anddiagnoseswere confirmedusingBio-RadRapidTestMultispotHIV-1/HIV2Abreflextesting(Bio-RadLaboratories,Inc,Hercules, CA).ScreeningforHCVintheEDwasperformedusinga chemiluminescentanti-HCVARCHITECTi1000SR immunoassay,anddiagnoseswereconfirmedbyHCV ribonucleicacidviralload(VL)withCobasHCV4800assay (RocheMolecularSystems,Pleasanton,CA).Patientswere consideredpositiveforHCVonlyiftheyhadadetectable

VL.Multiplexgonorrheaandchlamydiaurinepolymerase chainreactiontestingwasalsoperformedviatheCobas4800 assay.Patientsweretestedfor Treponemapallidum IgM/IgG antibody(TPA)usingBio-Rad’smultiplex flow immunoassay(MFI),Bioplex2200.19 Specimenswith reactiveTPAMFIresultsunderwentreflexiveconfirmatory quantitativenon-treponemalassaytesting,usingrapid plasmareagin.Iftestresultswerediscordant,thespecimen wastestedreflexivelyusingthe Tpallidum particle agglutinationtestasanadditionalconfirmatory treponemaltest.

Outcomes

Theoutcomesofinterestincludedtheprevalence ofSTItesting/co-testingandprevalenceofSTI infection/co-infection.

Analysis

Wedescribeddatausingdescriptivestatistics.Categorical variableswereexpressedaspercentagesandproportions,and continuousvariableswereexpressedasmeans ± standard deviations.WeperformedallstatisticalanalysesusingStata 15.1(StataCorpLLC,CollegeStation,TX).

RESULTS

PatientCharacteristics

Therewere30,767EDpatientencountersforpatients aged ≥13yearsduringthestudyperiod.Ofthese,7,866(26%) weretestedforatleastoneofHIV,HCV,syphilis, gonorrhea,orchlamydia.Themeanageofpatientswas 43 ± 14years,and4,077(52%)werefemale.Themost commonracewasWhite(39%),andmostpatientswerenonHispanic(76%).MostpatientstestedhadMedicaid insurance(56%).See Table1 forfullpatientcharacteristics.

PrevalenceofSexuallyTransmittedInfection Testing/Co-Testing

ThemostcommonlytestedSTIswerethosewithuniversal screeningindications:HCV(24.5%,7,539/30,767);andHIV (23.9%,7,359/30,767).Gonorrhea/chlamydia(1.9%,74) testingwasmorecommonthansyphilistesting(1.4%,

Table1. Characteristicsof7,866emergencydepartmentpatients whounderwenttestingforatleastonesexuallytransmittedinfection.

CharacteristicValue

Meanage(years)1

Genderr

43 ± 14

Male48%(3,789)

Female52%(4,077)

Race/ethnicity2

White39%(3,034)

Black22%(1,698)

Asian7%(517)

Mixed/other32%(2,517)

Ethnicity3

Hispanic24%(1,845)

Non-Hispanic76%(5,933)

Sexuality(self-identified)4

Heterosexual93%(1,386)

LGBTQ7%(105)

Housingstatus5

Domiciled91%(5,982)

Undomiciled9%(614)

Insurancetype

Private27%(2,162)

Medicare13%(1,025)

Medicaid56%(4,399)

Self/uninsured4%(280)

1Reportedasmean ± standarddeviation.

2Datamissingfor100patients.

3Datamissingfor88patients.

4Datamissingfor6,375patients.

5Datamissingfor1,270patients.

LGBTQ,Lesbian,Gay,Bisexual,Transgender,Queer.

420/30,767).OfthosewhoreceivedtestingforSTIs,6.5% (508/7,866)weretestedforasingleSTI.Patientsweretested fortwoormoreSTIsin95.6%(7,521/7,866)ofcasesand threeormoreSTIsin5.6%(437/7,866)ofcases.Patientswere testedforHIV,HCV,syphilis,andgonorrhea/chlamydiain 3.6%(286/7,866)ofcases.See Table2 foroverallco-testing.

PrevalenceofInfection/Co-Infection

Table2. Overalltesting/co-testingproportionsamongemergency department(ED)patientstestedforsexuallytransmittedinfections duringtheir firstEDvisit.

OneSTItestedTestingproportion

HCVonly24.5%(7,539/30,767)

HIVonly23.9%(7,354/30,767)

Syphilisonly1.4%(420/30,767)

Gonorrhea1.9%(574/30,676)

Chlamydia1.9%(574/30,676)

TwoSTIsco-testedCo-testingproportion

Gonorrhea + Chlamydia1 100%(574/574)

HCV + HIV95%(7,240/7,650)

HCV + syphilis4.4%(333/7,626)

HCV + Gonorrhea4.4%(344/7,769)

HCV + Chlamydia4.4%(344/7,769)

HIV + syphilis4.5%(357/7,417)

HIV + Gonorrhea4.6%(346/7,582)

HIV + Chlamydia4.6%(346/7,582)

Syphilis + Gonorrhea57%(361/633)

Syphilis + Chlamydia57%(361/633)

ThreeSTIsco-testedCo-testingproportion

HCV + HIV + syphilis2.1%(314/14,680)

HCV + HIV + Gonorrhea2.2%(319/14,829)

HCV + HIV + Chlamydia2.2%(319/14,829)

HCV + syphilis + Gonorrhea2.0%(303/14,861)

HCV + syphilis + Chlamydia2.0%(303/14,861)

HIV + syphilis + Gonorrhea4.0%(306/7,736)

HIV + syphilis + Chlamydia4.0%(306/7,736)

FourSTIsco-testedCo-testingproportion

HCV + HIV + Syphilis + Gonorrhea286(3.6%)

HCV + HIV + Syphilis + Chlamydia286(3.6%)

All fiveSTIsco-testedCo-testingproportion

HCV + HIV + syphilis + Gonorrhea + Chlamydia 286(3.6%)

1Gonorrheaandchlamydiawerealwaystestedtogether. STI,sexuallytransmittedinfection; HCV,hepatitisCvirus.

Theseroprevalenceofinfectionwashighestforsyphilis (44/420,10.5%[95%CI7.7–13.8]),followedbychlamydia (56/574,9.8%[95%CI7.4–12.5]),gonorrhea(35/574,6.1% [95%CI4.3–8.4]),HCV(373/7,470,5.0%[95%CI4.5–5.5]), andHIV(67/7,354,0.9%[95%CI0.7–1.2]).Among67 patientswhotestedpositiveforHIV,HCVwasthemost commonco-infection(sevenpatients,10.4%).Among373 patientswhotestedpositiveforHCV,HIVwasthemost commonco-infection(seven,0.9%).Among44patientswho testedpositiveforsyphilis,chlamydiawasthemostcommon co-infection(nine,20.5%).Among35patientswhotested positiveforgonorrhea,chlamydiawasthemostcommoncoinfection(eight,22.9%).Amongpatientswhotestedpositive forchlamydia,syphiliswasthemostcommonco-infection (nine,16.1%).OnepatientwasinfectedwiththreeSTIs (HCV,HIV,andsyphilis).Nopatientswereinfectedwith morethanthreeconcurrentSTIs.Overallco-infectiondatais availablein Table3.

Table3. Infectionandco-infectionproportionsforsexually transmittedinfections.

Infectiontype%(Proportion)

Syphilisinfection10.5%(44/420)

Chlamydiaco-infection20.5%(9/44)

HIVco-infection9.1%(4/44)

Gonorrheaco-infection6.8%(3/44)

HCVco-infection2.3%(1/44)

Gonorrheainfection6.1%(35/574)

Chlamydiaco-infection22.9%(8/35)

Syphilisco-infection8.6%(3/35)

HIVco-infection2.9%(1/35)

HCVco-infection2.9%(1/35)

Chlamydiainfection9.8%(56/574)

Syphilisco-infection16.1%(9/56)

Gonorrheaco-infection14.3%(8/56)

HCVco-infection1.8%(1/56)

HIVco-infection0%

HIVinfection0.9%(67/7,354)

HCVco-infection10.4%(7/67)

Syphilisco-infection5.9%(4/67)

Gonorrheaco-infection1.5(1/67)

Chlamydiaco-infection0%

HCVinfection5.0%(373/7,470)

HIVco-infection1.9%(7/373)

Chlamydiaco-infection0.3%(1/373)

Gonorrheaco-infection0.3%(1/373)

Syphilisco-infection0.3%(1/373)

Gonorrheaandchlamydiawerealwaystestedtogether. HCV,hepatitisCvirus.

PotentiallyMissedDiagnoses

Atotalof633patientsreceivedtargetedSTItestingdueto clinicalconcernduringtheir firstEDvisit(testedforcombo gonorrhea/chlamydiaand/orsyphilis).However,co-testing betweensyphilisandgonorrhea/chlamydiaoccurredinonly 57%(361/633)ofthesetestingencounters.Only63% (397/633)ofthesepatientsreceivedHIVco-testing,andonly 59%receivedHCVco-testing.SomepatientsreceivedSTI testingforoneormoreSTIs,butnotall fiveSTIs(gonorrhea, chlamydia,syphilis,HIV,HCV),duringtheir firstEDvisit. InthisgroupwithincompleteSTItesting,weassessed whetherpatientsreceivedfurtherSTItestinginanyoftheir nextfourdocumentedEDvisitswithinthestudyperiodand foundthefollowingtestingcountsandpositiveresults:HCV, 81(9positive[11.1%]);HIV,61(1positive[1.6%]);syphilis, 49(3positive[6.1%]);andgonorrhea/chlamydia, 55(1gonorrheapositive[1.8%]).

DISCUSSION

InourstudyweexaminedSTItesting/co-testingand infection/co-infectionprevalenceinEDpatientswhowere testedforatleastoneofHIV,HCV,syphilis,gonorrhea,or chlamydia.Toourknowledge,thisisthe firstED-based studytoreportEDSTIco-testingandco-infection frequenciesforeverycombinationofgonorrhea,chlamydia, syphilis,HIV,andHCV.Overall,STIco-testingwas infrequent,butco-infectionprevalencewashighamong severalSTIco-testpairings.

TheHIV/HCVtestingco-testingoccurredfrequently, likelyrelatedtothepresenceoftheuniversalscreeningBPA. AccordingtotheUSCentersforDiseaseControland Prevention,approximately21%ofindividualswithHIVin high-riskpopulations(ie,menwhohavesexwithmen,people whousedrugs)areco-infectedwithHCV.20 Whilepublished dataislimited,previousED-basedstudiesfoundthat8–33% patientswithHIVwereco-infectedwithHCV.21–24 Inour study,10.5%ofpatientswithHIVwereco-infectedwith HCV,butonly1.9%ofpatientswithHCVwereco-infected withHIV.PreviousstudiesinthisEDpopulationfoundthat patientssharedsome(malegender,unhousedstatus,history ofillicitdruguse,andMedicareinsurancestatus)butnotall riskfactorsforinfection.25,26 Itispossiblethatco-infection proportionsmaydifferamongpatientswithHCVandHIV duetosomeotherunmeasuredriskfactor.Alternatively, giventheimmunosuppressivepropertiesofHIV,patients whoareexposedtoHCVmaybemorelikelytoprogresstoa chronicinfection.27

Amongpatientswithgonorrhea,23%wereco-infected withchlamydia.Conversely,only14%ofthoseinfectedwith chlamydiawereco-infectedwithgonorrhea.Thisdifferential co-infectionpatternhasbeenpreviouslyreportedinatleast oneotherED-basedstudy(gonorrhea+,chlamydia+:44%; chlamydia+,gonorrhea+:17%).28 Amongpatientswith syphilis,wealsoobservedahighprevalenceofchlamydiacoinfection(21%).Inourstudy,therewasaBPAthatprompted clinicianstotestpatientsforsyphiliswhowereundergoing gonorrhea/chlamydiatesting,andco-testingoccurredinjust 57%ofpatients.Previousstudiesreporttheprevalenceof syphilisandgonorrhea/chlamydiaco-testing(9–39%); however,wecouldnot findanyED-basedstudiesthat reportedproportionsofco-infection.29 Similarly,wefound thatpatientswhoweretestedforsyphilisand/orgonorrhea/ chlamydia,co-testingforHIVandHCVoccurred infrequently.Giventhatpatientswithsyphilis,gonorrhea, andchlamydiahadthehighestprevalenceofco-infection withotherSTIs.Theseinstancesrepresentpotentialmissed opportunitiesfordiagnosisandlinkagetocare.

Inourstudy,therewereseveralpatientswhotested positiveforspecificSTIsinsubsequentEDvisitsandwere nottestedfortheseSTIsintheirinitialvisit.Itispossiblethat patientscontractedtheSTIexposureaftertheindexEDvisit, andhadtheybeentestedattheindexvisittheymayhavebeen

negative.However,itisalsopossiblethatthesediagnoses werepresentattheindexvisitandweremissed,suggesting thatincreasedco-testingcanleadtoincreaseddiagnosisof clinicallysignificantco-infectionswiththepotentialtoreduce transmissioninthecommunity.

LIMITATIONS

Thiswasaretrospectivestudywithdataobtainedfromthe EHRatasingleinstitution;thus,our findingsmaynotbe generalizabletoallsettings.OurstudyhadmultipleBPAsin placethatlikelyinfluencedclinicianco-testingbehavior.We didnotreportchiefcomplaints,makingitdifficultto differentiatepatientswithtrueclinicalindicationfortesting, andpatientswhowerebeingscreenedaspartofascreening protocol.Sincetestingforsyphilisandgonorrhea/chlamydia wasnotuniversal,thereportedproportionsareunlikelyto representtheprevalenceforthewholeED,butratherthe prevalenceofinfectionamongpatientswithclinical suspicionforSTI.WhileHIVandHCVscreeningwas universallyorderedformostpatientsundergoingbloodwork, patientscouldstilloptout,whichmayhavebiasedthe prevalenceestimatesfortheseinfections.

CONCLUSION

Prevalenceofco-testingforsexuallytransmittedinfection waslowamongpatientswithclinicalsuspicionforSTI; however,co-infectionprevalencewashighinseveralcoinfectionpairings.EncounterswithsingleSTItesting representamissedopportunitytoscreenforco-infections. FutureeffortsareneededtoimproveSTIco-testingrates amonghigh-riskindividuals.Withtheincidenceofmany STIsincreasing,theEDcanserveasanimportantscreening settingforSTIs,especiallyinpatientswithoutaccessto traditionaloutpatientservices.

AddressforCorrespondence:LarissaMay,MD,MSPH,Universityof CaliforniaDavisHealth,DepartmentofEmergencyMedicine, 4150V.St.,PSSB2100,Sacramento,CA95817.Email: lsmay@ ucdavis.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.ThisworkwassupportedbyGileadfundedFrontlinesofCommunitiesintheUnitedStates(FOCUS). TheFOCUSProgramisapublichealthinitiativethatenables partnerstodevelopandsharebestpracticesinroutineblood-borne virus(HIV,HCV,HBV)screening,diagnosis,andlinkagetocarein accordancewithscreeningguidelinespromulgatedbytheUS CentersforDiseaseControlandPrevention,theUSPreventive ServicesTaskForce(USPSTF),andstateandlocalpublichealth departments.FOCUSfundingsupportsHIV,HCV,andHBV screeningandlinkagetothe firstmedicalappointmentafter diagnosis.FOCUSpartnersdonotuseFOCUSawardsforactivities

beyondlinkagetothe firstmedicalappointment.Therearenoother conflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Fordetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.CentersforDiseaseControlandPrevention(CDC).PressRelease: CDCestimates1in5peopleintheU.S.haveasexuallytransmitted infection.2021.Availableat: https://www.cdc.gov/nchhstp/newsroom/ 2021/2018-STI-incidence-prevalence-estimates-press-release.html AccessedMay20,2022.

2.CentersforDiseaseControlandPrevention(CDC),U.S.Departmentof HealthandHumanServices,NationalCenterforHIV,ViralHepatitis, STD,andTBPrevention,etal.SexuallyTransmittedDisease Surveillance2019.2021.Availableat: https://www.cdc.gov/std/ statistics/2019/std-surveillance-2019.pdf.AccessedMay20,2022.

3.CentersforDiseaseControlandPrevention(CDC).HIVIncidence. Availableat: https://www.cdc.gov/hiv/statistics/overview/in-us/ incidence.html.AccessedJuly18,2023.

4.MannsMPandMaasoumyB.BreakthroughsinhepatitisCresearch: fromdiscoverytocure. NatRevGastroenterolHepatol. 2022;19(8):533–50.

5.CentersforDiseaseControlandPrevention(CDC).HepatitisC Surveillance2020.2020.Availableat: https://www.cdc.gov/hepatitis/ statistics/2020surveillance/hepatitis-c.htm.AccessedJuly18,2023.

6.WesterC,OsinubiA,KaufmanHW,etal.HepatitisCvirusclearance cascade-UnitedStates,2013–2022. MorbMortalWklyRep. 2023;72(26):716–20.

7.RyersonAB,SchillieS,BarkerLK,etal.VitalSigns:newlyreported acuteandchronichepatitisCcases-UnitedStates,2009–2018. MMWR MorbMortalWklyRep. 2020;69(14):399–404.

8.StanfordKA,HazraA,SchneiderJ.Routineopt-outsyphilisscreeningin theemergencydepartment:apublichealthimperative. AcadEmerg Med. 2020;27(5):437–8.

9.FordJS,MarianelliLG,FrassoneN,etal.HepatitisBscreeninginan ArgentineED:increasingvaccinationinaresource-limitedsetting. AmJ EmergMed. 2020;38(2):296–9.

10.FordJS,ChechiT,OtmarM,etal.EDsyphilisandgonorrhea/chlamydia cotestingpracticesbeforeandaftertheimplementationofanelectronic healthrecord-basedalert. EmergMedJ. 2021.21:emermed-2020210331.

11.FordJS,ChechiT,ToosiK,etal.UniversalscreeningforhepatitisC virusintheEDusingabestpracticeadvisory. WestJEmergMed. 2021;22(3):719–25.

12.FordJ,RouleauS,VoongS,etal.UniversalHIVscreeninginthe emergencydepartment:aninterruptedtimeseriesanalysis. AIDS. 2022.Inpress.

13.ChoudhryS,RamachandranVG,DasS,etal.Characterizationof patientswithmultiplesexuallytransmittedinfections:Ahospital-based survey. IndianJSexTransmDisAIDS. 2010;31(2):87–91.

14.PakianathanMR,RossJD,McMillanA.Characterizingpatientswith multiplesexuallyacquiredinfections:amultivariateanalysis. IntJSTD AIDS. 1996;7(5):359–61.

15.BarnesA,JetelinaKK,BettsAC,etal.Emergencydepartmenttesting patternsforsexuallytransmitteddiseasesinNorthTexas. SexTransm Dis. 2019;46(7):434–9.

16.SeballosSS,LopezR,HusteyFM,etal.Co-testingforhuman immunodeficiencyvirusandsexuallytransmittedinfectionsinthe emergencydepartment. SexTransmDis. 2022;49(8):549–50.

17.WhiteDA,AlterHJ,IrvinNA,etal.Lowrateofsyphilisscreeningamong high-riskemergencydepartmentpatientstestedforgonorrheaand chlamydiainfections. SexTransmDis. 2012;39(4):286–90.

18.Schechter-PerkinsEM,MillerNS,HallJ,etal.Implementationand preliminaryresultsofanemergencydepartmentnontargeted,opt-out hepatitisCvirusscreeningprogram. AcadEmergMed. 2018;25(11):1216–26.

19.MarangoniA,NardiniP,FoschiC,etal.EvaluationoftheBioPlex2200 syphilissystemasa first-linemethodofreverse-sequencescreeningfor syphilisdiagnosis. ClinVaccineImmunol. 2013;20(7):1084–8.

20.CentersforDiseaseControlandPrevention(CDC).Peoplecoinfected withHIVandviralhepatitis.Availableat: https://www.cdc.gov/hepatitis/ populations/hiv.htm#ref07.AccessedFebruary14,2023.

21.BurrellCN,SharonMJ,DavisS,etal.Usingtheelectronic medicalrecordtoincreasetestingforHIVandhepatitisCvirusinan

Appalachianemergencydepartment. BMCHealthServRes. 2021;21(1):524.

22.Martel-LaferriereV,BarilJG,AlarieI,etal.Opt-outuniversalHCVand HIVscreeninginaCanadianemergencyroom:across-sectionalstudy. BMJOpen. 2022;12(1):e048748.

23.CieplyL,SimmonsR,IjazS,etal.SeroprevalenceofHCV,HBVandHIV intwoinner-cityLondonemergencydepartments. EpidemiolInfect. 2019;147:e145.

24.TorianLV,FelsenUR,XiaQ,etal.UndiagnosedHIVandHCVinfection inaNewYorkCityemergencydepartment,2015. AmJPublicHealth. 2018;108(5):652–8.

25.FordJS,HollywoodE,SteubleB,etal.RiskfactorsforhepatitisCvirus infectionatalargeurbanemergencydepartment. JViralHepat. 2022;29(10):930–7.

26.FordJS,MousaMA,VoongS,etal.RiskfactorsforHIVinfectionata largeurbanemergencydepartment:across-sectionalstudy. Sex TransmInfect. 2023;99(6):404–8.

27.ThomasDL,LeoutsakasD,ZabranskyT,etal.HepatitisCin HIV-infectedindividuals:cureandcontrol,rightnow. JIntAIDSSoc. 2011;14:22.

28.WilsonSP,VohraT,KnychM,etal.Gonorrheaandchlamydiainthe emergencydepartment:continuedneedformorefocusedtreatmentfor men,womenandpregnantwomen. AmJEmergMed. 2017;35(5):701–3.

29.EricksonJL,WuJ,FertelBS,etal.Multidisciplinaryapproachtoimprove humanimmunodeficiencyvirusandsyphilistestingratesinemergency departments. OpenForumInfectDis. 2022;9(12):ofac601.

PublicBeliefsAboutAccessibilityandQualityofEmergency DepartmentsinGermany

JensKlein,PhD* SarahKoens,PhD*

MartinScherer,MD†

AnnetteStrauß,MD†

MartinHärter,PhD,MD‡

OlafvondemKnesebeck,PhD*

SectionEditor:GayleGalletta,MD

*UniversityMedicalCenterHamburg-Eppendorf,InstituteofMedicalSociology, Hamburg,Germany

† UniversityMedicalCenterHamburg-Eppendorf,DepartmentofGeneralPractice andPrimaryCareHamburg,Germany

‡ UniversityMedicalCenterHamburg-Eppendorf,DepartmentofMedical Psychology,Hamburg,Germany

Submissionhistory:SubmittedMay16,2023;RevisionreceivedDecember20,2023;AcceptedJanuary22,2024

ElectronicallypublishedMay3,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18224

Background: Itiswellestablishedthatemergencydepartment(ED)crowdingleadstoworsehealth outcomes.AlthoughvariouspatientsurveysprovideinformationaboutreasonstovisitEDs,lessis knownintermsofbeliefsaboutEDsamongthegeneralpopulation.Thisstudyexaminespublicbeliefs regardingaccessibilityandqualityofEDsandtheirassociationswithsocialcharacteristics(gender,age, education,immigrationbackground)aswellasknowledgeaboutemergencycareservicesand healthliteracy.

Methods: Weconductedacross-sectionalstudybasedonarandomsampleof2,404adultslivingin Hamburg,Germany,inwinter2021/2022.Wedevelopedeightstatementsregardingaccessibilityand qualityofEDsleadingtotwoscales(Cronbach’s α accessibility = 0.76andqualityofcare = 0.75). Descriptivestatisticsoftheeightitemsareshownandlinearregressionwereconductedtodetermine associationsofthetwoscaleswithsocialcharacteristicsaswellasknowledgeaboutemergencycare servicesandhealthliteracy(HLS-EU-Q6).

Results: Nearly44%oftherespondentsagreedthat “youcanalwaysgotoanED,ifyoudonotgeta short-termappointmentwithageneralpractitionerorspecialist.” And38%agreedwiththestatement, “If youdonothavethetimeduringnormalpracticehoursduetoyourwork,youcanalwaysgotoanED.” In termsofquality,38%believedthatdoctorsinEDsaremorecompetentthandoctorsingeneralpractice, and25%believedthatdoctorsinEDsaremorecompetentthandoctorsinspecializedpractices.Inthe fullyadjustedmodel,publicbeliefsaboutemergencycareaccessibilityandqualityofEDswere significantlyassociatedwithallsocialcharacteristicsandknowledgeofemergencycareoptionswiththe strongestassociationsbetweenknowledgeandaccessibility(β = 0.17; P < 0.001)andbetween educationandquality(β = 0.23; P < 0.001).

Conclusion: WefoundendorsementofpublicbeliefsaboutaccessibilityandqualityofEDsthatcan leadtoinappropriateutilization.Ourresultsalsosuggestthatknowledgeofdifferentemergencyservices playsanimportantrole.Therefore,aftersystem-relatedreorganizationsofemergencycare,information campaignsaboutsuchservicestailoredtosociallydeprivedpopulationsmayhelpalleviatetheissueof crowding.[WestJEmergMed.2024;25(3)389–398.]

INTRODUCTION

Crowdingofemergencydepartments(EDs)hasbecome animportantissueinmanycountries.1–3 Twocontributing causesofcrowdingareboardingofadmittedpatients (primary)andinappropriateutilizationoftheEDfornonurgentconditions(secondary).3–5 Intermsofthe firstcause, accessblock(ie,accesstohospitalbedsisblockedandno admissiontoaninpatientwardispossible)andhospital admissionsforambulatorycare-sensitiveconditions(ACSC) havebeenextensivelydiscussed.6,7 Inthisstudyweaimedto addressthesecondcause.AmongOrganizationfor EconomicCooperationandDevelopmentcountries,the increaseofEDuseinGermanyiscomparativelyhigh.8

Emergencydepartmentcrowdinghasbeenshownto negativelyaffectpatientsafety.3–5 Variousstudieshave examinedassociationsbetweencrowdedEDsandworse healthcareoutcomes(eg,delaysincriticaltreatments, medicationerrors,returnvisits,complicationrates,and mortality).9–11 Forinstance,recentresearchintheUnited KingdomfoundthatEDcrowdingisassociatedwith treatmentdelayandanincreaseinall-cause,30-day mortality.12 ToreducepatientnumbersinEDs,researchis focusedonavoidableEDvisitsofpatientswithnon-urgent conditions.StudieshaveshownthatthepercentageofallED visitsjudgedtobenon-urgentisabout30–40%,eventhough studydesignswereveryheterogeneous.13 Moreover,astudy fromGermanyfoundthatmorethanhalfofthepatients visitinganEDdidnotthinkthattheirconditionrequired urgenttreatmentandthusdidnotmeetthedefinitionofa medicalemergency.14

Innumerouspatientsurveys,differentreasonsforvisiting EDswithnon-urgentconditionswerereported.Access barrierstooutpatientcare,assumptionsofhigherqualityof careandmorehealthcareoptionsatEDs(aswellasnegative perceptionsaboutprimarycarephysicians),perceivedneed andanxiety,convenience(eg,24/7availability,no appointments,transport),andreferralfromhealthcare professionalsweremostfrequentlymentionedinvarious internationalsurveys.13,15–17 Patientsurveysconductedin GermanyfoundfourmainmotivationsforpatientswhoselfreferredtotheED:distress/perceivedurgency;access;quality ofcare;andconvenience.14,18,19

Alowersocioeconomicstatus(SES) mostlyassessedby educationallevel,income,occupationonindividualor regionallevel,andimmigrationstatus predictmore frequentEDutilizationandahigheruseforlow-acuity presentations,20–23 eventhoughsomecurrent findingsdidnot completelyconfirmtheseinequalitiesforGermany.19 Inthis context,theconceptofhealthliteracyplaysanimportant role.24 Lowhealthliteracywasshowntobeassociatedwith preventableEDvisitsduetominorornon-urgentproblems andwithmorefrequentutilizationofEDsandemergency services,25–27 althoughsomeotherstudiesdidnotshow thisassociation.28

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

CrowdedEDsareassociatedwithpoorhealth outcomes.Patientsurveyshaveshown problematicassumptionsaboutED accessibilityandquality.

Whatwastheresearchquestion?

WesoughttoexaminebeliefsabouttheED andtheirassociationswithvarious characteristicsinapopulationsurvey.

Whatwasthemajor findingofthestudy?

44%ofrespondentsagreed, “ youcanalways gotoanED,ifyoucan ’ tgetanappointment withanof fi cedoctororspecialist, ” and38% saidyoucouldusetheEDforcareduring non-businesshours.

Howdoesthisimprovepopulationhealth?

ByunderstandinginappropriateEDuse, wecandevelopeducationprogramsfor vulnerablegroupstoinformthemabout alternativevenuestoobtaincare.

Whilecurrentevidenceprovidesinformationabout reasonsandpredictorsoffrequentorinappropriateEDuse, nearlyall findingsarederivedfrompatientsurveysthatwere conductedatEDsorwerebasedonEDrecords.These surveysexaminetherecordedhealthcareutilizationofactual patientsratherthanthebeliefsaboutEDsamongthegeneral population.Thisresearchgapconcerningpublicbeliefs aboutEDsandtheiraccessibilityandqualitywasour rationaleforconductingthisstudy.Publicbeliefsabout emergencycarearehighlyrelevantastheymaycontributeto abetterunderstandingofinappropriateEDuseandtothe developmentofcampaignstoimprovehealthliteracy.29 Againstthisbackground,weexploredtworesearch questions:1)Whatarethepublic’sbeliefsaboutEDsinterms ofaccessibilityorconvenienceandqualityofcare;and2)Are therevariationsinthesebeliefsaboutEDsaccordingtosocial characteristics(age,gender,educationlevel,and immigrationstatus)andhealthliteracy(generalhealth literacyandknowledgeofemergencycareoptions)?

METHODS

StudyDesignandPopulation

Across-sectionaltelephonesurveywasconductedin Hamburg,Germanyinwinter2021/2022viacomputerassistedtelephoneinterviews.Weobtainedarandomsample

ofGerman-speakingpeopleaged ≥18yearsusingallpossible telephonenumbersinHamburg,includingnon-registered numbers,viarandomdigitaldialing.Onlylandlinenumbers couldbeincludedasmobiletelephonenumbersarenot providedonaregionallevel.About83%ofallhouseholdsin Germanyhavealandlinetelephone.30 Thus,alargemajority ofthepopulationcanbereachedvialandlinenumbers. Repeatedcallsweremadebytrainedinterviewersofa professionalsurveyresearchinstitute(USUMA,Berlin, Germany)ondifferentweekdays.WeappliedtheKish selectiongridtorandomlyselectthetargetpersoninthe contactedhouseholds.31 Priortothis,thesamesurvey researchinstituteconductedapilotstudyamong30 individualsinthegeneralpopulation.

Wechoseatelephonesurveyasthemethodfordata collectionduetothevignettedesignofthestudy.Atthe beginningofthesurvey,recordedaudio filesdescribing differentsymptomsweredirectlyplayedtotherespondents. Toguaranteeastandardizedstimulusandimmediate response,telephonesurveysareusuallyfavoredandan establishedmethod.32 Subsequently,astandardized questionnairewasapplied.Samplesizewascalculatedbased onavignettedesign(48vignettesintotal)appliedinthe study.Accordingtopowercalculations,asamplesizeof50 respondentspervignettewascalculatedtoidentifymedium sizedifferencesresultinginabout2,400requiredparticipants (statisticalpower0.8,andtype-Ierror0.05).Thesevignettes werenotusedinthepresentanalyses.Thesampleconsisted of2,404respondents.

Duetodifferentapproachesforthedefinitionofeligibility intelephonesurveys,differentresponserates(RR)canbe calculated.33 Thus,aRRinthissurveyvariedbetween 10.9–46.0%(AmericanAssociationofPublicOpinion ResearchRR334 17.3%).Togainarepresentativesample,we weighteddataforhouseholdsize,gender,age,educational level,andplaceofresidence(districtinHamburg)usingthe officialstatisticsregardingtheadultpopulationlivingin Hamburg.35–37 InaccordancewithHalbeslebenand Whitman,38 weconductedasample/populationcomparison toassessnonresponsebias. Table1 showsthattheweighted sampleadequatelyrepresentsthegeneraladultpopulationof Hamburgregardingthedistributionofgender,age,and educationallevel.35,36 ThesurveywasapprovedbytheLocal PsychologicalEthicsCommitteeattheCenterfor PsychosocialMedicine,UniversityMedicalCenter Hamburg(No.LPEK-0200).Respondentsgavetheir informedconsentfortheparticipationandtheuseof theirdata.Consentsandrefusalsweredocumentedby theinterviewers.

Measures

Toassessthepublic’sbeliefsaboutEDs,wedevelopedeight items(statementsaboutEDs)basedonareviewofthe literature.13–16,18,19,39 Asdescribedabove,themain

Table1. Samplecharacteristicsofsurveyrespondentscompared withofficialstatisticsforthepopulationinHamburgbypercentage.

Samplea (N = 2,404)

Adultpopulationof Hamburg2020b Pc

Gender (0)d

Male48.548.40.95

Female51.551.6

Age(years) (0)

18–249.69.40.83 25–3419.719.6

35–4417.217.5

45–5417.516.6

55–6414.115.1 65–7410.110.0 ≥7511.811.8

Educationlevel (71)

Migration background (46) Nomigration background

2nd generation11.2 –g 1st generation10.9 –g

aWeighted; b34,35; cPearson’schi2; dNumberofmissingcasesinbracketsinitalics; eDataforeducationonlyavailableforpeople ≥15yearsold. fNoexactdataavailable. gAstherewasnodiscreteweightingforimmigrationbackground,test statisticswerenotconducted.

motivationsforpreferringEDsin patientsurveyswererelated tobarrierstoaccessofoutpatientcare,convenience, assumptionsofhigherqualityofcare,anddistressorsubjective need.Wedevelopedfourstatementsregardingaccessbarriers andconvenience,aswellasfourstatementsrelatedtothe qualityofcareprovidedinEDs(Figure1).Asthesurveywas conductedamongthegeneralpopulationandnotacute patientsinEDs,wedidnotincludestatementsregarding distressandsubjectiveneed.Responsecategorieswere “fully agree, ”“ratheragree,”“ratherdisagree,”“fullydisagree” and, additionally, “don’tknow,” withhighervaluesindicating strongeragreement.Validitywastestedinaccordancewith someaspectsofMessick'sunifiedframework.40

Wecollectedcontentvalidityevidencethroughan extensiveliteraturescreeningofpatientsurveysidentifying themainmotivationsforpreferringEDs.Additionally,

Figure1. Publicbeliefsaboutaccessibilityandqualityofemergencydepartments(N = 2,404). ED, emergencydepartment.

expertsinemergencycarewereinvolvedintheitem development.Usingapilotstudy,wepretesteditemsand responseconsistencies.Furthermore,internalstructure validitywastestedthroughCronbach’ s α andthefactorial structureoftheinstrumentviaprincipalcomponentanalysis. Theweightingofdemographicsinourstudywasaimedto meetexternalvalidity.

Gender,age,educationallevel,andimmigration background(noimmigrationbackground, firstgeneration, secondgeneration)wereintroducedassocialcharacteristics oftherespondents.Educationlevelwasbasedonyearsof schooling(9years = low;10years = middle; ≥12 = high).A personwasconsideredtohaveanimmigrationbackgroundif heorsheoratleastoneparentwasbornabroad. Respondentswithanimmigrationbackgroundwhowere borninGermanyareconsideredsecond-generation immigrants,whilethosewithimmigrationexperienceare subsequentlytermed first-generationimmigrants.We assessedgeneralhealthliteracyusingaEuropeanhealth literacysurveyquestionnaire,theHLS-EU-Q6,ashortform oftheestablishedHLS-EU-Q47.41 Onafour-point Likertscale,theanswercategorieswere “verydifficult,” “fairlydifficult,”“fairlyeasy,” and “ veryeasy, ” including “don’tknow.” TheCronbach’ s α = 0.60ofthescaleis acceptableforaninstrumentthatisshortandfeatures discreteelementsofliteracy.42 Wecomputedasumscaleby averagingtheresponsestothesixitemsresultinginarange between1–4,withhigherscoresindicatinganincreased healthliteracy.

Tospecificallyassessknowledgeaboutavailable emergencycareservices,weaskedtherespondentstoname alloptionsofemergencycaretheyknewof(open-ended question).IntheGermanhealthcaresystem,patients basicallyhavefouroptionsofemergencycare.43 Theycan1) calltherescueservice(telephonenumber112);2)gotoan ED;3)gotoanemergencypractice(practicesthatareusually openfrom6 PM tomidnightforurgentconditions); or4)contactthemedicalon-callservice(alsoknownas “116117,” referringtothetelephonenumber)inurgentor emergencycases.Inthesurvey,thisquestionwaslocated beforetheitemaboutarespondent’sbeliefsconcerningEDs. Basedontheresponses(respectiveemergencycareservice mentioned = 1,notmentioned = 0),wecalculatedasum scalewithapossiblerangefrom0–4withhigherscores indicatingmoreknowledge.

Analyses

Wepresentpercentagesofagreementoftheeightsingle itemstoassessbeliefsaboutEDsasdescriptiveresults. Furthermore,weconductedaprincipalcomponentanalysis includingtheeightitemsassessingpublicbeliefs.Theanalysis revealedtwocomponentswitheigenvalues ≥1reflecting accessibility(eigenvalue:3.22,explainedvariance:40.3%) andqualityofcare(eigenvalue:1.41,explainedvariance: 17.6%),whichaccountedfor57.9%ofthetotalvariance (rotatedloadingsbetween0.66–0.78).Eigenvaluesareused todeterminetherelativeimportanceandtheexplained varianceofeachprincipalcomponent.Usually,onlyfactors

witheigenvalues ≥1areconsidered.44 Accordingly,forthe multivariateanalyses,twoscalesrepresentingaccessbarriers andconvenience(subsequentlylabelledas “accessibility,” fouritems)andqualityofcare(“quality,” fouritems)were calculatedrangingfrom1–4.Higherscoresindicatestronger agreementwitheasyaccessibilityandsuperiorcarequality withregardtoEDs.Intermsofreliability,internal consistencyofthetwoscalesrevealedsatisfactoryresults (Cronbach’ s α = 0.76[accessibility]and0.75[quality]).

Wecalculatedlinearregressionmodelstoanalyze associationsbetweensocialcharacteristics,healthliteracy, andpublicbeliefsaboutaccessibilityandqualityofEDs. Dependentvariableswerethetwoscalesregarding accessibilityandqualityofEDs.Aspredictorvariables, gender,age,educationlevel,immigrationbackground, generalhealthliteracy,andknowledgeofemergencyservices wereintroduced.Ina firststep,wecalculatedsimple unadjustedmodelsshowingthesingleestimatesand significancesofeachpredictorvariable.Thereafter,inthefull model,thepredictorvariableswereenteredsimultaneously adjustingallvariablesforeachother.Wedocumented regressionestimates(B),standardizedB(β),significances(P ), andexplainedvariance(R2).Resultswith P < 0.05were consideredstatisticallysignificant.Asmanyparticipants chosetheoption “don’tknow” whencompletingtheHLSEU-Q6,whichhadtobeconsideredasmissingvalue,the multivariateanalyseswereconductedwithasamplesizeof 1,751(quality)or1,826(accessibility),respectively. Moreover,variouskeyassumptionsforlinearregression models(linearrelationship,normaldistributionofresiduals, auto-correlation,homoscedasticity,andmulticollinearity) weresuccessfullytested.Allanalyseswerecalculatedwith

weighteddataandcarriedoutusingtheStatisticalPackage fortheSocialSciencesV27(SPSS,Inc,ChicagoIL).45

RESULTS

Meanageoftherespondentswas48.8years(SD19.0); 51.5%werefemale.Almosthalfofthesample(48.5%)hada higheducationallevel(middlelevel:24.1%;lowlevel: 27.4%).About11%eachbelongedtothegroupof first-or second-generationimmigrants,whileabout78%ofthe samplehadnoimmigrationbackground(Table1).Themean (SD)was2.56(0.49)forhealthliteracy(HLS-EU-Q6)score (range1–4).Regardingknowledgeofavailableemergency services,onaveragetherespondentsknewtwooffour options. Figure1 showsthedistributionoftheeightitems measuringbeliefsaboutaccessibilityandqualityofEDs. Agreement(percentageofrespondentswho “fully” and “rather” agreedsummedup)totheitemsrelatedtoaneasy accessofEDsrangedbetween38.4%(“Ifyoudonothavethe timeduringnormalpracticehoursduetoyourwork,youcan alwaysgotoanED”)and66.9%(“Youcanalwaysgotoan EDwhenpracticesareclosed”).Intermsofqualityofcare providedinEDs,25.3%oftherespondents “fully” or “rather” agreedwiththeitem “DoctorsinEDsaremore competentthandoctorsinspecializedpractices,” while68% “rather” or “fully” disagreed.Regardingtheitem “Youget bettercareinEDsbecauseallspecialistsarepresentthere,” 59.7%expressedagreement.

Table2 showstheresultsoflinearregressionanalyseswith thesumscaleindicatingaccessibilityofEDsasthedependent variable.Ascanbeseenintheunadjustedmodels,all predictorvariablesindicatedsignificantassociationswith beliefsaboutaccessibility.Strongestassociationswereshown

B = regressionestimate, β = standardizedB, P = significance(significantassociations[P < 0.05]arebold). aHighervaluesindicatestrongeragreement(range1to4). bAllanalysesbasedonthesamplesizeofthefullyadjustedmodel.

cGender = reference:male,age = range18–96years,education = range1–3;migrationbackground = reference:nomigrationbackground; healthliteracy = range1–4;knowledgeofemergencycareservices = range0–4. dEmergencydepartment/emergencypractice/rescueservice/medicalon-callservice.

Table2. Beliefsaboutemergencydepartments:sumscaleaccessibilitya (N = 1,826b)(linearregressions).

Table3. Beliefsaboutemergencydepartments:sumscalequalitya (N = 1,751b)(linearregressions) UnadjustedmodelsFullyadjustedmodel

B = regressionestimate, β = standardizedB,p = significance(significantassociations[P < 0.05]arebold). aHighervaluesindicatestrongeragreement(range1to4).

bAllanalysesbasedonthesamplesizeofthefullyadjustedmodel.

cGender = reference:male,age = range18–96years,education = range1–3;migrationbackground = reference:nomigrationbackground; healthliteracy = range1–4;knowledgeofemergencycareservices = range0–4. dEmergencydepartment/emergencypractice/rescueservice/medicalon-callservice.

foreducationlevelandknowledgeofemergencycareservice. Inthefullyadjustedmodel,femalerespondentslessoften agreedthatEDsarecharacterizedbyeasyaccessibility. Moreover,agreementincreasedwithage,whileit decreasedwitheducationlevelandknowledgeof emergencycareserviceoptions.Comparedtorespondents withoutanimmigrationbackground, first-andsecondgenerationimmigrantsmorestronglybelievedinthe easyaccessibilityofEDs.Highest β-valuesinthefully adjustedmodelwereshownforeducation(β = 0.13, P < 0.001),knowledge(β = 0.17, <0.001)andage (β = 0.13, P < 0.001).

IntermsofbeliefsaboutsuperiorcarequalityinEDs, significantassociationswereshownforallpredictorsexcept immigrationbackground(secondgeneration)inthe unadjustedmodels(Table3).Again,educationleveland knowledgeofemergencycareserviceindicatedhighest βvalues.Regardingthefullyadjustedmodel,significant negativeassociationswitheducationlevel,emergencycare knowledge,andhealthliteracyemerged.Furthermore,these beliefsincreasedwithageandweremorepronouncedamong first-generationimmigrantsandmales.Educationlevel showedthestrongestassociation(β = 0.23, P < 0.001)in thefullyadjustedmodel.

DISCUSSION

BasedonapopulationsurveyinaGermanmetropolis,we assessedthepublic’sbeliefsaboutaccessibilityandqualityof careofEDs.Nearly44%oftherespondentsagreedthat “ you canalwaysgotoanED,ifyoudonotgetashort-term appointmentwithageneralpractitioneroraspecialist.” Still, 38%agreedtothestatement “Ifyoudonothavethetime

duringnormalpracticehoursduetoyourwork,youcan alwaysgotoanED.” Intermsofsuperiorquality,38% believedthatdoctorsinEDsaremorecompetentthan doctorsingeneralpracticeand25%regardeddoctorstobe morecompetentinspecializedpractices.Inaddition,nearly 60%agreedthat “yougetbettercareinEDsbecauseall specialistsarepresentthere.” Furthermore,thepublic’ s perceptionsofemergencycarearesignificantlyassociated withsocialcharacteristics(gender,age,educationlevel, immigrationbackground)andknowledgeofemergencycare options.Regardingaccessibility,knowledgeshowedthe strongestassociation:Themoreoptionsofemergencycare respondentsnamed,thelessrespondentsagreedthatEDsare alwaysaccessible.Intermsofbeliefsaboutqualityofcare, educationlevelturnedouttobethestrongestpredictor:The lesseducatedtherespondentswerethemoretheyagreedthat thequalityofcareissuperiorinEDs.

Astherearemanypatientsurveysbutveryfew population-basedstudies,comparabilityofourresultswith previousresearchislimited.Someresearchersalsoaimedto assesspublicperceptionsaboutEDs,buttheirmethodsvary considerably.27,46 Regardingattitudestowardaccessibility andquality,males,olderpeople,ethnicminorities,and peoplewithlowerSESshowedatendencytouseemergency services,evenforminorproblems,morefrequently.An Australianstudyamongthegeneralpopulationshowedthat perceivedurgency,goodaccessibility,andbetterhealthcare provisionwerestatedasreasonstovisitanED.46 However, nofurtheranalysisaboutpredictingfactorswasconducted. InaBritishsurveyusingcasevignettes,thetendencytocall foranambulanceortovisitanEDinlessurgentcaseswas significantlyincreasedformales,olderage,andthosewho

wereofethnicminoritiesandhadalowerpaidoccupation andalowerlevelofhealthliteracy.27

Whendevelopingthestatementsconcerningaccessibility, wedeliberatelychosestrictwording(“always”),sothatthe itemswerenottooleading.Toagreetothefourstatements wasnotcompletelywrong,butitwasinappropriateinterms offavorednavigationwithintheGermanhealthcaresystem. ServicesintheEDaregenerallyprovidedforlife-threatening conditionsorseriousacuteproblemsthatcannotwaitand needtobetreatedbyadoctorimmediately.Inlessurgent cases,otheralternativesshouldbepreferred.Forthesecases, mainlytwoservicesareprovidedwhenpracticesareclosed: emergencypracticesandthemedicalon-callservice(also knownas “116117” referringtothetelephonenumber).In fact,thesetwoserviceswereimplementedtounburdenEDs. Thiswastakenintoaccountwhenwedevelopedthefour statementsconcerningbeliefsaboutaccessibility.Itissimilar inthecaseofthestatementsregardingbetterqualityofcare inEDs.Thereisconcentratedexpertiseinhospitals,butthe ratingofworseexpertiseofoutpatientdoctorsandthe assumptionthatallspecialistsareavailableintheemergency wardaredoubtfulandcouldleadtounrealisticexpectations regardingtheuseofEDs.Inthisregard,thepresentstudy couldhelpustounderstandthepublic’sbeliefsonwhich inappropriateutilizationofEDsarebased.However,it cannotberuledoutthatsomeparticipantsdidnotcorrectly understandtheitems.

The findingsprovidedataaboutthelackofhealth educationamongthegeneralpopulation.Particularly, males,olderandlesseducatedpeople,andthosewithlimited knowledgeofemergencycareoptionsshowedapotentially inappropriateutilizationoftheED.Intermsofimmigration status,especially first-generation,immigrantsshowedalack ofinformationthatcouldbeduetolessexperiencewiththe healthcaresystem,languagebarriers,differentexpectations andpreferences,aswellasformalaccessbarriers(eg,waiting timesortraveldistances).47 Regardinggender-specific differences,previousresearchshowedahigherED attendancefornon-urgentproblemsandahigheruseofoutof-hourshelp-seekingamongmen.27,48 Potentially,this preferencecouldbeduetolongerworkinghoursamongmen andlesswillingnesstobeabsentfromworkbecauseof healthcare.Thus,socialinequalitiesshouldbeconsidered whenimplementinginterventions(eg,information campaigns).Tomodifypublicbeliefsabouthealthcarein generaloremergencycareinparticular, “ emergency literacy” campaignsareawaytoaddresstheproblemofED crowding.Inthisregard,knowledgeabouttheavailabilityof differentemergencycareservices,navigationwithinthe healthcaresystem,andtheassessmentofsymptomscould beaddressed.

PeopleshouldbeeducatedthattheEDisforlifethreateningandseriousconditionssuchasheavybleeding, brokenbones,chestpainorstroke,andthatmanysymptoms

canbetreatedmoreappropriatelyelsewhere.AnAustralian behaviorchangecampaignthatfocusedonattitudes, awareness,andknowledgewassuccessfulinreducingthe numberofinappropriateornon-urgentcallstoambulance servicesormedicalemergencyphonenumbers.29 Currently, aqualitativestudyfromGermanypositivelyevaluatedan educationalinterventiontailoredforEDpatientswithlowacuityconditions.49 Anotherstudyexaminedphysiciandirectedstrategiesforimprovingpatienthealthliteracyin EDs.50 Furthermore,lowerhealthliteracywasfoundamong peoplewhowereofolderageandhadlowereducationlevels, lessaffluence,andwithimmigrationbackgrounds,51,52 which arefactorsthatwerealsoshowntobeassociatedwithhigher andinappropriateEDuseinsomestudies.25–27 Asourdata ofpublicbeliefssupportsthe findingsofsocialinequalitiesin inappropriateEDuse,tailoredhealtheducationhastotake placeinmoredeprivedareaswherevulnerablegroupsare livingandtheavailabilityofhealthcareservicesispotentially limited.Informationindifferentlanguagesandindigitaland non-digitalversionscouldhelptoreachthepopulationina betterway.

Inthisstudy,wefocusedonbeliefsthatmayfosteran inappropriateutilizationoftheEDfornon-urgent conditionsasonecauseofcrowding.Anotherandmore importantreasonisrelatedtoboardingofadmittedpatients.3 Inthiscontext,accessblockandhospitaladmissionsfor ambulatorycare-sensitiveconditions(ACSC)are discussed.6,7 Accessblockisthesituationinwhichaccessto hospitalbedsisblockedandnoadmissiontoaninpatient wardispossible.6 HospitaladmissionsforACSCaredefined asadmissionsinhospitalwardsincludingEDsformedical conditionsthatarepotentiallyavoidableiftheyaremanaged intheoutpatientcare.7 ThroughACSC,theavailability, accessandqualityofoutpatientcarecanbeevaluated,and socialinequalitiescanberevealed.7 Somereviewssummed uppossibleimplicationsandinterventionsintermsof reorganizationofEDwardsandavailabilityofoutpatient care. 53–55 Recently,reformsofemergencycarehavebeen discussedinGermanyintermsofallocatingandtriaging patients(ie,implementationofacoordinationcenterfor first telephonecontactandfurtherallocation,andageneral counterforinitialassessmentandtriagingatEDs).

LIMITATIONS

Thisstudyhassomelimitationsthatneedtobediscussed. Eventhoughthedatawasweightedforgender,age,and educationlevel,andthecomparisonofsampleand populationshowedreasonableresults,apotentialselection biasduetonon-responseandtheexclusiveuseoflandline numberscannotberuledout.Inthisregard,aresponserate ofbetween10.9–46%(dependingondefinitionofeligibility) canbeconsideredacceptablecomparedtoothertelephone surveys. 56 Moreover,83%ofhouseholdscanbereachedvia landlinenumbersinGermany.30 Ourdatareferstothe

situationofhealthcareprovisioninaGermanmetropolis. Theconditionsinothercountriesandinmoreruralregions couldbedifferent.

Astherewasnovalidatedmeasureforpublicbeliefsabout availabilityofEDs,wedevelopedeightitemsbasedona reviewoftheliteratureofpatientsurveys.Although psychometricpropertiesofthetwoscalesseemadequate (Cronbach’ s α = 0.76[accessibility]and0.75[quality]),42 thesescalesneedtobefurtherdevelopedandtested.Interms ofmissingvalues,someitemsoftheaccessibilityandquality scale(n = 185andn = 320),andnotablyitemsoftheHLSEU-Q6scaleyieldedahighnumberofmissingvaluesdueto “don’tknow” answers(434).Althoughthisprocedurewasin accordancewiththeoriginalHLS-EUinstrument,theoption of “don’tknow” inquestionnairesshouldnotbe automaticallytreatedasmissingvalues.Therefore,amissing analysiswasconducted.Theresultsrevealedonlya consistentpatternforage(significantlyincreasedmissing valuesamongpeoplewitholderage).Thus,therelevanceof agecouldbeunderestimatedintheregressionanalyses,and duetosubjectivedataacommonmethodbiascouldnotbe ruledout.Finally,theevaluationofgeneralhealthliteracy wasconductedwithanestablishedinstrumentoftheHLSEUconsortium,butwiththeshortestversionavailable (HLS-EU-Q6).41 So,amorecomprehensiveinstrument wouldpossiblyleadtoanimprovedassessment.

CONCLUSION

Thisstudysuggeststhatthepublic’ sperceptionsabout EDqualityandaccessibilitycontributetoinappropriate EDutilizationandcrowdinginGermany.Particularly,this holdstrueforpeopleofolderage,malegender,lower educationlevel,andthosewhoare fi rst-generation immigrantsandwhohavelessknowledgeaboutavailable emergencycareservices.The fi ndingshelpinunderstanding inappropriateutilizationofemergencycareservicesand developinghealtheducationprogramstailoredtosocially deprivedpopulations.

AddressforCorrespondence:JensKlein,PhD,UniversityMedical CenterHamburg-Eppendorf,InstituteofMedicalSociology, Martinistr.52,Hamburg,Germany20146.Email: j.klein@uke.de

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Kleinetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.MorleyC,UnwinM,PetersonGM,etal.Emergencydepartment crowding:asystematicreviewofcauses,consequencesandsolutions. PLoSOne. 2018;13(8):e0203316.

2.VashiAA,UrechT,CarrB,etal.Identificationofemergencycaresensitiveconditionsandcharacteristicsofemergencydepartment utilization. JAMANetwOpen. 2019;2(8):e198642.

3.HigginsonI.Emergencydepartmentcrowding. EmergMedJ. 2012;29(6):437–43.

4.PinesJM,HiltonJA,WeberEJ,etal.Internationalperspectives onemergencydepartmentcrowding. AcadEmergMed. 2011;18(12):1358–70.

5.TrzeciakSandRiversEP.Emergencydepartmentovercrowdinginthe UnitedStates:anemergingthreattopatientsafetyandpublichealth. EmergMedJ. 2003;20(5):402–05.

6.AffleckA,ParksP,DrummondA,etal.Emergencydepartment overcrowdingandaccessblock. CJEM. 2013;15(6):359–84.

7.JohnsonPJ,GhildayalN,WardAC,etal.Disparitiesinpotentially avoidableemergencydepartment(ED)care:EDvisitsforambulatory caresensitiveconditions. MedCare. 2012;50(12):1020–28.

8.BerchetC.Emergencycareservices:trends,driversandinterventions tomanagethedemand.2015.Availableat: https://www.oecd-ilibrary. org/social-issues-migration-health/emergency-care-services_ 5jrts344crns-en.AccessedDecember11,2023.

9.SunBC,HsiaRY,WeissRE,etal.Effectofemergencydepartment crowdingonoutcomesofadmittedpatients. AnnEmergMed. 2013;61(6):605–11.e6.

10.KimD-U,ParkYS,ParkJM,etal.Influenceofovercrowdinginthe emergencydepartmentonreturnvisitwithin72hours. JClinMed. 2020;9(5):E1406.

11.KimJ-S,BaeH-J,SohnCH,etal.Maximumemergencydepartment overcrowdingiscorrelatedwithoccurrenceofunexpectedcardiac arrest. CritCare. 2020;24(1):305.

12.JonesS,MoultonC,SwiftS,etal.Associationbetweendelaystopatient admissionfromtheemergencydepartmentandall-cause30-day mortality. EmergMedJ. 2022;39(3):168–73.

13.Uscher-PinesL,PinesJ,KellermannA,etal.Emergencydepartment visitsfornonurgentconditions:systematicliteraturereview. AmJManag Care. 2013;19(1):47–59.

14.SchererM,LühmannD,KazekA,etal.Patientsattendingemergency departments. DtschArzteblInt. 2017;114(39):645–52.

15.CosterJE,TurnerJK,BradburyD,etal.Whydopeoplechoose emergencyandurgentcareservices?Arapidreviewutilizinga systematicliteraturesearchandnarrativesynthesis. AcadEmergMed. 2017;24(9):1137–49.

16.PearsonC,KimDS,MikaVH,etal.Emergencydepartmentvisitsin patientswithlowacuityconditions:factorsassociatedwithresource utilization. AmJEmergMed. 2018;36(8):1327–31.

17.AbarB,HolubA,HongS,etal.Latentclassanalysisofbarriersto careamongemergencydepartmentpatients. WestJEmergMed. 2019;20(2):256–61.

18.SchmiedhoferM,MöckelM,SlagmanA,etal.Patientmotivesbehind low-acuityvisitstotheemergencydepartmentinGermany:aqualitative studycomparingurbanandruralsites. BMJOpen. 2016;6(11):e013323.

19.HolzingerF,OslisloS,MöckelM,etal.Self-referredwalk-inpatientsin theemergencydepartment – whoandwhy?Consultationdeterminants inamulticenterstudyofrespiratorypatientsinBerlin,Germany. BMC HealthServicesResearch. 2020;20(1):848.

20.KriegC,HudonC,ChouinardM-C,etal.Individualpredictorsoffrequent emergencydepartmentuse:ascopingreview. BMCHealthServRes. 2016;16(1):594.

21.CredéSH,SuchE,MasonS.Internationalmigrants’ useofemergency departmentsinEuropecomparedwithnon-migrants’ use:asystematic review. EurJPublicHealth. 2018;28(1):61–73.

22.LynchB,FitzgeraldAP,CorcoranP,etal.Driversofpotentially avoidableemergencyadmissionsinIreland:anecologicalanalysis. BMJQualSaf. 2019;28(6):438–48.

23.CarlsonLC,ZachrisonKS,YunBJ,etal.Theassociationof demographic,socioeconomic,andgeographicfactorswithpotentially preventableemergencydepartmentutilization. WestJEmergMed. 2021;22(6):1283–90.

24.QuaglioG,SørensenK,RübigP,etal.Acceleratingthehealthliteracy agendainEurope. HealthPromotInt. 2017;32(6):1074–80.

25.BalakrishnanMP,HerndonJB,ZhangJ,etal.Theassociationofhealth literacywithpreventableemergencydepartmentvisits:acrosssectionalstudy. AcadEmergMed. 2017;24(9):1042–50.

26.SchumacherJR,HallAG,DavisTC,etal.Potentiallypreventableuseof emergencyservices:theroleoflowhealthliteracy. MedCare. 2013;51(8):654–8.

27.O’CathainA,SimpsonR,PhillipsM,etal.Tendencytocallan ambulanceorattendanemergencydepartmentforminorornon-urgent problems:avignette-basedpopulationsurveyinBritain. EmergMedJ. 2022;39(6):436–42.

28.VandenboschJ,VandenBrouckeS,VancorenlandS,etal.Health literacyandtheuseofhealthcareservicesinBelgium. JEpidemiol CommunityHealth. 2016;70(10):1032–38.

29.BorgK,WrightB,SannenL,etal.Ambulancesareforemergencies: shiftingattitudesthrougharesearch-informedbehaviourchange campaign. HealthResPolicySyst. 2019;17(1):31.

30.FederalStatisticalOfficeofGermany.Equipmentofhouseholdswith informationandcommunicationtechnology.Availableat: https://www. destatis.de/EN/Themes/Society-Environment/Income-ConsumptionLiving-Conditions/Equipment-Consumer-Durables/Tables/ liste-equipment-households-information-communicationtechnology-germany.html#55714.AccessedDecember11,2023.

31.KishL.AProcedureforobjectiverespondentselectionwithinthe household. JAMA. 1949;44(247):380–7.

32.HäderM.(2012).Dataqualityintelephonesurveysviamobileand landlinephone.InHäderS,HäderM,KühneM,(Eds), Telephone SurveysinEurope:ResearchandPractice(247-62).Heidelberg, Germany:SpringerBerlin.

33.MartsolfGR,SchofieldRE,JohnsonDR,etal.Editorsandresearchers beware:calculatingresponseratesinrandomdigitdialhealthsurveys. HealthServRes. 2013;48(2Pt1):665–76.

34.AmericanAssociationforPublicOpinionResearch.Standarddefinitions - finaldispositionsofcasecodesandoutcomeratesforsurveys. 2016.Availableat: https://aapor.org/wp-content/uploads/2022 /11/Standard-Definitions20169theditionfinal.pdf

AccessedDecember11,2023.

35.StatisticalOfficeforHamburgandSchleswig-Holstein.(2020). Hamburg StatisticalYearbook2019/2020.Hamburg,Germany:StatisticalOffice forHamburgandSchleswig-Holstein.

36.FederalStatisticalOffice.Educationlevelofthepopulationresultsofthemicrocensus2019;2020.Availableat: https://www. destatis.de/EN/Themes/Society-Environment/EducationResearch-Culture/Educational-Level/Tables/educationalattainment-population-germany.html

AccessedDecember11,2023.

37.KolenikovS.Calibratingsurveydatausingiterativeproportional fitting (raking). StataJ 2014;14(1):22–59.

38.HalbeslebenJRBandWhitmanMV.Evaluatingsurveyqualityinhealth servicesresearch:adecisionframeworkforassessingnonresponse bias. HealthServRes. 2013;48(3):913–30.

39.UnwinM,KinsmanL,RigbyS.Whyarewewaiting?Patients’ perspectivesforaccessingemergencydepartmentserviceswithnonurgentcomplaints. IntEmergNurs. 2016;29:3–8.

40.MessickS.Validityofpsychologicalassessment:validationof inferencesfrompersons’ responsesandperformancesasscientific inquiryintoscoremeaning. AmPsychol. 1995;50:741–49.

41.PelikanJ,GanahlK,VandenBrouckeS,etal.(2013).Measuringhealth literacyinEurope:introducingtheEuropeanhealthliteracysurvey questionnaire(HLS-EU-Q).In:OkanO,BauerU,Levin-ZamirD, PinheiroP,SørensenK(Eds), InternationalHandbookofHealth Literacy(115-38).Bristol,England:PolicyPress.

42.TaberKS.TheUseofCronbach’salphawhendevelopingandreporting researchinstrumentsinscienceeducation. ResSciEduc. 2018;48(6):1273–96.

43.vondemKnesebeckO,KoensS,SchäferI,etal.Publicknowledge aboutemergencycare – resultsofapopulationsurveyfromGermany. FrontPublicHealth. 2022;9:2204.

44.KaiserHF.Theapplicationofelectroniccomputerstofactoranalysis. EducPsycholMeas. 1960;20(1):141–51.

45.IBMCorp.IBMSPSSStatisticsforWindows,Version27.0.2019. Availableat: https://www.ibm.com/products/spss-statistics AccessedDecember11,2023.

46.FitzGeraldG,TolooGS,AitkenP,etal.Publicuseandperceptionsof emergencydepartments:apopulationsurvey. EmergMedAustralas. 2015;27(4):336–42.

47.KleinJ,vondemKnesebeckO.Inequalitiesinhealthcareutilization amongmigrantsandnon-migrantsinGermany:asystematicreview. Int JEquityHealth. 2018;17(1):160.

48.KeizerE,ChristensenMB,CarlsenAH,etal.Factorsrelatedtoout-ofhourshelp-seekingforacutehealthproblems:asurveystudyusingcase scenarios. BMCPublicHealth. 2019;19(1):33.

49.KümpelL,OslisloS,CantuRR,etal.Exploringtheviews oflow-acuityemergencydepartmentconsultersonaneducational interventionandgeneralpractitionerappointmentservice: aqualitativestudyinBerlin,Germany. BMJOpen. 2023;13(4):e070054.

50.Nwanaji-EnweremJC,Smith-WilkersonM,GordonB,etal.Universal precautionsplus:physician-directedstrategiesforimprovingpatient healthliteracyintheemergencydepartment. WestJEmergMed. 2023;24(1):110–3.

51.SørensenK,PelikanJM,RöthlinF,etal.HealthliteracyinEurope: comparativeresultsoftheEuropeanhealthliteracysurvey(HLS-EU). EurJPublicHealth. 2015;25(6):1053–8.

52.NutbeamDandLloydJE.Understandingandrespondingtohealth literacyasasocialdeterminantofhealth. AnnuRevPublicHealth. 2021;42:159–73.

53.BaierN,GeisslerA,BechM,etal.Emergencyandurgentcaresystems inAustralia,Denmark,England,France,GermanyandtheNetherlands – analyzingorganization,paymentandreforms. HealthPolicy. 2019;123(1):1–10.

54.VandenHeedeKandVandeVoordeC.Interventionstoreduce emergencydepartmentutilisation:areviewofreviews. HealthPolicy. 2016;120(12):1337–49.

55.CowlingTEandGunningEJ.AccesstogeneralpracticeinEngland: political,theoretical,andempiricalconsiderations. BrJGenPract. 2016;66(650):e680–e682.

56.DutwinDandLavrakasP.Trendsintelephoneoutcomes,2008-2015. SurveyPractice. 2016;9(3).

ORIGINAL RESEARCH

SupportforThrombolyticTherapyforAcuteStrokePatients onDirectOralAnticoagulants:Mortality andBleedingComplications

PaulKoscumb,MD*

LukeMurphy,MD*

MatthewTalbott,DO*

ShivaNuti,BS*

GeorgeGolovko,PhD†

HashemShaltoni,MD‡

DietrichJehle,MD*

SectionEditor:RichardLucarelli,MD

*UniversityofTexasMedicalBranch,DepartmentofEmergencyMedicine, Galveston,Texas

† UniversityofTexasMedicalBranch,DepartmentofPharmacologyandToxicology, Galveston,Texas

‡ UniversityofTexasMedicalBranch,DepartmentofNeurology,Galveston,Texas

Submissionhistory:SubmittedApril10,2023;RevisionreceivedOctober25,2023;AcceptedDecember11,2023

ElectronicallypublishedApril8,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.18063

Background: Alteplase(tPA)istheinitialtreatmentforacuteischemicstroke.CurrenttPAguidelines excludepatientswhotookdirectoralanticoagulants(DOAC)withintheprior48hours.InthispropensitymatchedretrospectivestudywecomparedacuteischemicstrokepatientstreatedwithtPAwhohad receivedDOACswithin48hoursofthrombolysistothosenotpreviouslytreatedwithDOACs,regarding threeoutcomes:mortality;intracranialhemorrhage(ICH);andneedforacutebloodtransfusions(asa markerofsignificantbloodloss).

Methods: UsingtheUnitedStatescohortof54healthcareorganizationsintheTriNetxdatabase,we identified8,582strokepatientstreatedwithtPAonDOACswithin48hoursofthrombolysisand46,703 strokepatientstreatedwithtPAnotonDOACssinceJanuary1,2012.Weperformedpropensityscore matchingondemographicinformationandsevenpriorclinicaldiagnosticgroups,resultinginatotalof 17,164acutestrokepatientsevenlymatchedbetweengroups.Werecordedmortalityrates,frequencyof ICH,andneedforbloodtransfusionsforeachgroupovertheensuing7-and30-dayperiods.

Results: PatientstreatedwithtPAonDOACshadreducedmortality(3.3%vs7.3%;riskratio[RR]0.456; P < 0.001),fewerICHs(6.8%vs10.1%;RR0.678; P < 0.001),andlessriskofmajorbleedingas measuredbyfrequencyofbloodtransfusions(0.5%vs1.5%;RR0.317; p < 0.001)at7dayspost thrombolytic,thanthetPApatientsnotonDOACS.Findingsfor30dayspost-thrombolyticsweresimilar/ statisticallysignificantwithlowermortalityrate(7.2%vs13.1%;RR0.550; P < 0.001),fewerICHs(7.6% vs10.8%;RR0.705; P < 0.001),andfewerbloodtransfusions(0.9%vs2.0%;RR0.448; P < 0.001).

Conclusion: AcuteischemicstrokepatientstreatedwithtPAwhoreceivedDOACswithin48hoursof thrombolysishadlowermortalityrates,reducedincidenceofICH,andlessbloodlossthanthosenoton DOACs.OurstudysuggeststhatprioruseofDOACsshouldnotbeacontraindicationtothrombolysisfor ischemicstroke.[WestJEmergMed.2024;25(3)399–406.]

INTRODUCTION

IntheUnitedStates,strokeremainscommon,withthe estimatedriskofstrokeoveranindividual’slifetime

approachingoneinfour.Ischemicstrokerepresents87%of acuteinsultswithintracerebralhemorrhageand subarachnoidhemorrhagemakinguptheremaining

balance.1,2 Forpatientswhoareeligible,themainstayof treatmentofischemicstrokeisrestorationofblood flowvia thrombolyticsand/orthrombectomy.Alteplase(tPA)is currentlytheonlythrombolyticapprovedforuseinacute ischemicstrokebytheUSFoodandDrugAdministration.3

Thromboembolismfromatrial fibrillationisafrequent causeofischemicstrokeandbecomesparticularlyprevalent withaging.IntheFraminghamstudy,atrial fibrillation representeda1.5%riskofstrokeinthe50–59agegroupand rosetoa23.5%riskinthe80–89agegroup.4 Asa preventativemeasure,mostpatientswithatrial fibrillation aretreatedwithanticoagulation,whichcaneffectively reducetheriskofstrokebyapproximatelytwo-thirdswhen comparedtoplacebo.5–9 Patientswithvalvularatrial fibrillationshouldbetreatedwithoralvitaminKantagonists (VKA).10,11 However,directoralanticoagulants(DOAC), whichincludedirectthrombininhibitorsandfactorXa inhibitors,havedemonstratednon-inferioritytoVKAsinthe preventionofacuteischemicstrokeinpatientswithnonvalvularatrial fibrillation.12–15

Themostrecent2019updatetothe2018AmericanHeart Association(AHA)guidelinesontPAexcludespatientswith concomitantusageofDOACswithin48hoursofthelast doseintake,unlesscoagulationparametersareobtainedand normal.Coagulationparametersincludetestssuchas activatedpartialthromboplastintime(aPTT),prothrombin time(PT),plateletcount,thrombintime(TT),ordirect factorXaactivityassay.16 Recentdatahassuggestedthatthe useofDOACsmaynotincreasetheriskofsymptomatic intracerebralhemorrhageevenintheabsenceofreversal agents,17–20 butthishasnotyetledtoachangeinguidelines. Promptadministrationofthrombolyticstopatientswith acuteischemicstrokehasthepotentialtoleadtoclot breakdownand,ideally,restoredcerebralperfusion.Patients withacuteischemicstrokewhoaretreatedwith thrombolyticshaveimprovedneurologicaloutcomeatthree months21,22 andhavealowerriskoflong-termmortality23; therefore,itiscriticaltoidentifythemaximumnumberof patientsthatcansafelyreceivethisintervention.Theuseof DOACshasrapidlyincreasedinthepastdecadebothfor primarystrokepreventioninpatientswithatrial fibrillation, andinthetreatmentofvenousthromboembolism.24,25 Therefore,thisrepresentsalargecohortofpatientswhomay bedeprivedofthepotentialbenefitsofreperfusiontherapy. WesoughttoassesswhetherpatientsreceivingDOACs whoalsoreceivedtPAwouldhaveanincreasedriskof adverseevents.Weevaluatedmortalityandrateof hemorrhagicconversionacrossalarge,retrospectivedataset. Duetothenatureofthehealthcaredataset,wewerenotable toassessforseverityofallotherbleedingdirectly.Thus,we evaluatedforwhetherpatientsrequiredbloodtransfusionas asurrogatemarkerforclinicallysignificantbleeding. Weanalyzedthesethreeoutcomesat7and30days postthrombolytic.

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

WhiletPAistheinitialtreatmentforischemic stroke,guidelinesexcludepatientstaking directoralanticoagulants(DOAC)dueto theoreticalriskofbleeding.

Whatwastheresearchquestion?

DopatientstakingDOACswhoreceive tPAhaveworseoutcomesthanthosenot takinganticoagulants?

Whatwasthemajor findingofthestudy?

WithDOACtherewasreducedmortality (3.3%vs7.3%;RR0.456;P < 0.001)and intracerebralhemorrhage(6.8%vs10.1%; RR0.678;P < 0.001)

Howdoesthisimprovepopulationhealth?

UseofDOACsisincreasing;thisshouldnot preventpatientstakingDOACsfrom receivingthebene fi tsofreperfusiontherapy forischemicstroke.

METHODS

TriNetXisaglobalfederatedhealthresearchnetwork providingde-identifiedaccesstoretrospectiveelectronic healthrecords(diagnoses,procedures,medications,lab values,genomicinformation)fromapproximately91million patientsin54largehealthcareorganizations(HCO)within theUnitedStates.26 InthisstudyweusedtheUS CollaborativeNetworktoidentifypatientswhoweretreated withtPAforacutestroke.Twocohortswereidentifiedfor thisstudywithinthegroupofpatientstreatedwithtPAon thedayoforwithinonedayofthediagnosisofacutestroke: Cohort1consistedofacutestrokepatientstreatedwithtPA whohadreceivedaDOAConthedayoforonedaypriorto theirpresentationwithacutestroke;andCohort2consisted ofactuestrokepatientstreatedwithtPAwhohadnot receivedaDOACwithinsevendaysofthediagnosis ofstroke.

Weidentifiedstrokepatientsofallethnicities,races,and gendersusingInternationalClassificationofDiseases,10th modification(ICD-10)codeI63(CerebralInfarction,1.49 millioncases).Inthedatabasetherewere565,835patients whoweretreatedwithtPAusingRxNORM:8410.The datasetwaslimitedtothosepatientswhoseindexevent occurredonorafterJanuary1,2012,andwhohadthe diagnosisofcerebralinfarctionsoastoexcludepatients

treatedwithtPAforacutecoronarysyndrome,pulmonary embolism,etc.Wethengeneratedtwocohorts.InCohort1, patientshadreceivedaDOAC(edoxaban,dabigatran, apixaban,orrivaroxaban)onthedayoforonedaypriorto theirstrokeandthrombolyticusingRxNORM1599538, 1037042,1364430,or1114195,resultingin8,582patients. Cohort2wasdefinedasstrokepatientstreatedwithtPAwho weredocumentedtonotbeonDOACsintheprior7days, resultingin46,703patients.

Tocontrolforpotentiallyconfoundingriskfactorsforthe measuredoutcomes,weperformedpropensityscore matchingbasedontheageatstrokediagnosis,race, ethnicity,gender,presenceofhypertensivediseases(ICD-10 codesI10-I16),diabetesmellitus(E08-E13),acutekidney failureandchronickidneydisease(N17-N19),overweight status/obesity(E66),heartfailure(I50),cardiacarrest(I46), andischemicheartdiseases(I20-I25).Weusedthebalanced cohorttoolinTriNetXformatching.26

Weperformedtheoutcomeanalysisbetweenthetwo cohortsforthreeevents:death(vitalstatus:deceased); nontraumaticintracranialhemorrhage(ICH)andblood transfusions(CurrentProceduralTerminologycode36430). NontraumaticICHwasdefinedasnontraumatic subarachnoid – ICD-10codeI60;nontraumatic intracerebralhemorrhage – ICD-10:I61,ornontraumatic acutesubduralhemorrhage – ICD-10:I62.01.Ratesofblood transfusionwereusedasamarkerofsignificantbloodloss postthrombolyticadministration.Alltestedoutcomes occurredonorafterthedayofdiagnosisofstroke.We measuredoutcomesoveraperiodof7and30dayspost thrombolytics.Patientswhohadtheoutcomeatthetimeof orpriortothedesignatedtimewindowweresubsequently excludedfromtheanalysis.26

Weperformedunivariateanalysisusingthemeasure-ofassociationtoolinTriNetX,whichcomparesoutcomes withinthedesignatedtimeframesforeachcohortreported bothasriskratios(RR),oddsratios,95%confidence intervals(CI)oftheseratiosandriskdifferenceasa P -value. Weobtainedde-identifiedpatientdatafromtheTriNetXUS CollaborativeNetworkdatabaseonNovember4,2022,and weperformedthedataanalysesonthesamedate.We reportedouroutcomesasRRswith95%CIsandrisk differences.TheTriNetXplatformprovidesaccessto aggregatedcountsandstatisticalsummariesofde-identified patientrecords.Noprotectedhealthinformationor personaldataisavailabletotheplatformusers;therefore, thisprojectwasexemptfrominstitutionalreview boardreview(www.trinetx.com).26 Ourmanuscript followsSTROBEguidelinesforobservational cohortstudies.27

RESULTS

Weidentified91,1707,410patientsintheTriNetXUnited StatesCollaborativeNetworkfrom54academicmedical

centers/healthcareorganizations.InCohort1ofpatients treatedwithtPAonDOACswithin48hoursofthrombolysis forstrokeweidentified8,582patients.InCohort2ofpatients treatedwithtPAforstrokedocumentedtonotbeonDOACs inthepriorsevendays,therewere46,703patients.After propensityscorematchingonbasicdemographic informationandsevenpriorclinicaldiagnosticgroups associatedwithmortalitytherewereatotalof17,164acute strokepatientsevenlymatchedbetweentheDOACandnoDOACgroups.

Mostofthedemographicgroupsexceptforgender(male/ female)werestatisticallydifferentbetweenthetwocohorts beforematching.Allpre-existingmedicalconditions associatedwithmortalitywerestatisticallydifferentbetween cohorts.Afterpropensitymatchingmostofthedemographic groupsorpre-existingmedicalconditionswerestatistically differentbetweenthecohorts. Tables1 and 2 presentthe demographiccharacteristicsandpre-existingconditionsin Cohorts1and2beforeandafterpropensitymatching. TriNetXreportsinfrequenteventswithoutcomesthatare ≤10as10;sothedifferencebetweenthetwocohortsmay havebeenslightlygreaterfortheNativeAmericanand Hawaiiandemographicgroupswherethenumberinthe DOACgroupislistedas10.

Weexcludedpatientsiftheyhadanoutcomeatthetimeof orpriortothedesignatedindexeventbasedonwhatis recordedinthehealthrecords.Theriskanalysisforthe mortalityoutcomehad193patientsexcludedfromCohort1 (DOAC)and171patientsfromCohort2(no-DOAC).These exclusionsarenecessarywhenthetimingofanoutcome diagnosisisuncertainoroccursbeforethetimewindow. Theseexclusionsarealsoinpartnecessarywhentheoutcome andindexeventoccurswithinhoursofeachother,asthe TriNetXdatabasedoesnotalwayshavethedegreeof granularitytodistinguishwhicheventoccurred first.The DOACsweremuchlessfrequentlyusedmorethan10years ago(introducedapproximately13yearsago);thus,we eliminatedthisperiodfromthetreatmentandcontrolgroups toavoidconfounding.

PatientstreatedwithtPAonDOACswerefoundtohave reducedmortality(3.3%vs7.3%;RR0.456; P < 0.001), lowerincidenceofICH(6.8%vs10.1%;RR0.678; P < 0.001),andlessriskofmajorbleedingasmeasured byfrequencyofbloodtransfusions(0.5%vs1.5%;RR0.317; P < 0.001)atsevendayspostthrombolytic,thanthetPA patientswhohadnotbeenonDOACS.Wefoundsimilar statisticallysignificant findingswithlowermortalityrate (7.2%vs13.1%;RR0.550; P < 0.001),lowerincidenceof ICH(7.6%vs10.8%;RR0.705; P < 0.001),andfewerblood transfusions(0.9%vs2.0%;RR0.448; P < 0.001)at30days aftertheadministrationofthethrombolyticagentinthe 10-yeardataset.Thisinformationregardingthepatient outcomesatsevendays(Table3)and30days(Table4)post thrombolyticisbelow.The95%CIsfortheRRofdeath,

Table1. Cohort1(N = 8,582)andCohort2(N = 46,703)characteristicsbeforepropensityscorematching. Demographics CohortMean

ICH,andsignificantotherbleedingarealsopresentedin thesetables.

DISCUSSION

Inthislarge,multicenter,propensity-matched, retrospectivestudy,patientswithischemicstrokewho receivedtPAandhadreceivedDOACswithintwodaysof

thrombolyticswerefoundtohavesignificantlylowerriskof death,ICH,andbleedingwhencomparedtopatientswho receivedtPAwithoutpriorDOACs.These findings werestatisticallysignificantatboth7and30days post-thrombolytic.Thisissignificantbecausecurrentstroke guidelinesrecommendagainstadministrationof thrombolysisinpatientswhohavetakenaDOACwithin

Table2. Cohort1(N = 8,582)andCohort2(N = 8,582)characteristicsafterpropensityscorematching.

48hoursofpresentationduetoatheoreticalincreasedriskof bleeding.AlthoughpriorstudieshaveindicatedthatDOAC usemaynotincreasetheriskofintracerebralhemorrhage,to ourknowledgeourstudyisthe firsttosuggestadecreased riskofbleedingaswellasadecreasedriskofdeath.

Recommendationsforwithholdingthrombolyticsfrom patientsusingDOACsintheabsenceoflabtestingsuchas

activatedpartialthromboplastintime(aPTT),prothrombin time(PT),plateletcount,thrombintime(TT_,ordirect factorXaactivityassaywere firstinstitutedin2013basedon consensusopinionwithlimited-to-nodata(ClassIIIC recommendation).28 Theserecommendationsmayhavebeen prudentatthetimeasDOACswerearelativelynovelclassof medication,anddatasurroundingtheirusewasonly

Table3. Outcomesat7Daysafterpropensityscorematching.

OutcometPA + DOAC* (n = 8,582)tPA-DOAC(n = 8,582)RiskRatio(95%CI‡)Probability(p)

*DirectOralAnticoagulant.

‡ConfidenceInterval. †IntracranialHemorrhage.

Table4. Outcomesat30Daysafterpropensityscorematching.

OutcometPA + DOAC*(n = 8,582)tPA-DOAC(n = 8,582)RiskRatio(95%CI‡)Probability(p)

Mortality7.2%13.1%0.550(.500,0.604) <

BloodTransfusion0.9%2.0%0.448(0.341,0.590)

beginningtoemerge.However,theclinicalutilityofmany commoncoagulationparametersinpatientsonDOACsis limited.TheAPTT,PT,andTTarereadilyavailablebut poorlysensitiveandspecificformonitoringofDOACsand shouldnotbeusedquantitativelytoevaluatethedegreeof anticoagulationeffect.Bycontrast,plasmadrug concentrationandanti-factorXaassaysmayquantifythe degreeofanticoagulationbutarenotalwaysavailableand mayrequirespecializedlaboratorysend-outtesting.29 Atour institution,diluteTTisonlyperformedtwiceweekly,and anti-factorXaforDOACsisnotperformedbyour hematologylabdespiteourinstitutionbeinga comprehensivestrokecenter.Wesuspectthatthisissimilar tomanycomprehensivestrokecentersnationwide.Awaiting theaboveresultsisproblematicsincetPAadministrationis timesensitive;therefore,patientsarefunctionallyexcluded fromreceivingtPAduetothetimerequiredtoobtain thesestudies.

IthasbeenknownfornearlyadecadenowthatDOACuse isnon-inferiortowarfarininthepreventionofacuteischemic strokeinpatientswithnonvalvularatrial fibrillation.12–15 However,whileadequateanticoagulationdecreasestherisk ofstroke,itdoesnotcompletelynegatetherisk,withthe estimatedannualriskofischemicstrokedespiteoral anticoagulationbeingapproximately1–2%.30 Weconjecture thatperhapsthiscohortofpatients,whilestillexperiencing ischemicstroke,hasrelativelymilderstrokeswithsmaller thrombusburden,whichputsthesepatientsatlowerriskof deathregardlessoftPAadministration.Furthermore,these patientsmayhavesmallerareasofinfarction,whichmayput thematadiminishedriskofhemorrhagicconversiondueto lessvolumeoffragiletissue.

OurresultsfavoringDOACsareinlinewithotherdata thatalsosuggestthatDOACsareassociatedwithlowerrisk

offatalordisablingstrokewhencomparedtocoumadin.31 Furtherstudyinthisareacouldseektostratifypatientsby initialNationalInstitutesofHealthStrokeScalescoreor volumeofinfarctiononimagingtoconfirmthesehypotheses. Thesehypothesesdonotfullyexplainthethirdoutcome demonstratingalowerrateofbloodtransfusionasa surrogateforextracranialbleeding.Itispossiblethatless debilitatingstrokesamongpatientsonDOACscouldplace lessmetabolicstressonthebodyandresultinalessglobally criticalconditionofthepatient.Itisknown,forexample, thatlengthofintensivecareunitstayincreasestheriskof gastrointestinalbleeding.32 IfpatientsreceivingDOACs havesmallerstrokesandlessseverediseaseresultingin decreasedmultisystemorganfailure,thenthismayaccount forthedifferencesseeninourdataset.Itisalsopossiblethat thelimitedhalf-lifeofDOACmedicationsallowsfora reductionintheirbleedingeffectevenwithatimeof abstinencelessthan48hours.Themaximalhalf-lifeof rivaroxabanisapproximately12hours,andthatofapixaban issimilar.Warfarin’smeanhalf-lifeontheotherhandis40 hours,andthiscanvarywidely.

TherateofICHseeninthisstudyisslightlyhigherthan wasseeninpriorstudies,forexample,intheNINDStrial wherethesymptomaticICHratewasfoundtobe6.4%in patientstreatedwiththrombolytics.Thiswasdefinedas “ any CT[computedtomography]-documentedhemorrhagethat wastemporallyrelatedtodeteriorationinthepatient’ s clinicalconditioninthejudgmentoftheclinicalinvestigator” within36hoursoftreatment.21 Ourstudyevaluatedfor outcomesat7and30daysfollowingtreatment,andthemore expansivetimeframemaypartiallyexplaintheincreased hemorrhagerateseeninourstudy.Additionally,ourstudy includedallICH,notonlysymptomaticICH.Institutional strokeprotocolsmaymandateforaroutineCTbrainforall

patientstreatedwiththrombolyticsregardlessofwhether theyweresymptomatic,andthismayhavecapturedpatients includedinourstudywhowouldhavebeenexcluded fromothers.

LIMITATIONS

Theretrospectivecohortdesignofthisstudymakes establishingcausalitydifficult.However,toourknowledge ourstudyismuchlargerthananyotherintheliterature evaluatingoutcomesaftertPAinpatientstakingDOACsvs thosenottakingDOACs.Thesizeofourstudy,combined withthepropensitymatchingthatweperformedonour sample,givesitmorepowertoevaluatefordifferencesin outcomes.Furthermore,thegeneralizabilityofourstudyis increasedbythenumberofinstitutionsqueriedbyTriNetX.

Weperformed1:1propensitymatchingforage,gender, race,hypertensivediseases,ischemicheartdisease,diabetes mellitus,acuteandchronickidneyfailure,heartfailure, overweightstatus/obesity,andpriorcardiacarrest,asthese areknownriskfactorsformortality.Itisimportanttonote thatalthoughweselectedmultiplepotentiallyconfounding variablesformatching,avariablethatwedidnotinclude couldhaveconfoundedtherelationshipbetweendrug treatmentandmortality.Authorsthathavebeencriticalof propensitymatchingtechniquesacknowledgethatmost potentialdeficienciesofthistechniqueareminimizedin largerdatasetssuchasthisone.

Additionally,weusedbloodtransfusionasasurrogate markerofsignificanthemorrhage,althoughthiscannotbe determinedtobesecondarytothethrombotic administration.Thisoutcomewasinvestigatedoveraperiod of7and30dayspost-thrombotictreatment,whichshould limitconfoundingduetoothercausesofbleeding.

Wedidnotexcludewarfarinusefromourtwopatient populations.Itispossible,butunlikely,thatthepresenceof warfarinuseinsomenon-DOACpatientcohortmayhave skewedtowardhigherratesofICH.However,2019AHA guidelinesspecificallyrecommendagainsttreatmentof ischemicstrokeswiththrombolyticsinpatientswithan internationalnormalizedratio >1.7,whichwouldexclude mostpatientstherapeuticonwarfarin.Inaddition,patients takingwarfarinareunlikelytobeconcurrentlytreatedwith aDOAC.

Weareunsurewhythepatientsincludedinthisstudywere treatedwiththrombolytics,contrarytocurrentstroke guidelines.Itispossiblethatthesewerepatientswho, throughcoagulationassayssuchasdirectfactorXa,were foundnottobeanticoagulated.However,duetothetimeconsumingnatureofthesestudiesandtheneedtoadminister thrombolyticsinatime-sensitivemannerandtheirlimited availabilityevenatcomprehensivestrokecenterssuchasour own,webelieveitisunlikelythatthesepatientswereincluded inasystemicmannerorinlargenumbers.Wewereunableto evaluatewhethertherewasaclusteringofpatientson

DOACswhoreceivedthrombolyticsatcertaincenters,asthe TriNetXprivacypolicydoesnotallowustoviewthis informationduetothede-identifiednatureofthedataset. Thede-identifiednatureofthedatasetalsorequiredustobe dependentonaccurateinputoftheelectronichealthrecordto evaluatewhetherornotthepatientwastakingaDOAC.

CONCLUSION

Inthislarge,retrospective,multicenterstudy,patients takingDOACswhoreceivedtPAforacuteischemicstroke hadareducedriskofdeath,lowerincidenceofICH,and decreasedbloodlossincomparisontothosewhoreceived tPAandwerenottakingDOACs.Ourstudyaddstothe increasingevidencethatDOACuseshouldnotbea contraindicationtothrombolyticsintheinitialtreatmentof acuteischemicstroke.Thestrokeguidelinesshouldbe updatedtoreflectthese findings.

AddressforCorrespondence:PaulKoscumb,MD,Universityof TexasMedicalBranch,DepartmentofEmergencyMedicine, 301UniversityBlvd.,Galveston,TX77555-1173.Email: pakoscum@utmb.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Thisresearchwassupportedbythe InstituteforTranslationalSciencesattheUniversityofTexas MedicalBranch,supportedinpartbyaClinicalandTranslational ScienceAward(UL1TR001439)fromtheNationalCenterfor AdvancingTranslationalSciencesattheNationalInstitutesofHealth (NIH).Thecontentissolelytheresponsibilityoftheauthorsanddoes notnecessarilyrepresenttheofficialviewsoftheNIH.Thereareno conflictsofinteresttodeclare.

Copyright:©2024Koscumbetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.GBD2016LifetimeRiskofStrokeCollaboratorsFeiginVL,NguyenG, etal.Global,Regional,andCountry-SpecificLifetimeRisksofStroke, 1990and2016. NEnglJMed.2018;379(25):2429–37.

2.TsaoCW,AdayAW,AlmarzooqZI,etal.HeartDiseaseandStroke Statistics-2022Update:AReportFromtheAmericanHeartAssociation [publishedcorrectionappearsin Circulation.2022;146(10):e141]. Circulation.2022;145(8):e153–639.

3.PotlaNandGantiL.Tenecteplasevs.alteplaseforacuteischemic stroke:asystematicreview. IntJEmergMed.2022;15(1):1.

4.WolfPA,AbbottRD,KannelWB.Atrial fibrillationasanindependentrisk factorforstroke:theFraminghamStudy. Stroke.1991;22(8):983–8.

5.PetersenP,BoysenG,GodtfredsenJ,etal.Placebo-controlled, randomisedtrialofwarfarinandaspirinforpreventionof

thromboemboliccomplicationsinchronicatrial fibrillation.The CopenhagenAFASAKstudy. Lancet.1989;1(8631):175–9.

6.StrokePreventioninAtrialFibrillationStudy.Finalresults. Circulation 1991;84(2):527–39.

7.BostonAreaAnticoagulationTrialforAtrialFibrillation InvestigatorsSingerDE,HughesRA,etal.Theeffectoflow-dose warfarinontheriskofstrokeinpatientswithnonrheumaticatrial fibrillation. NEnglJMed.1990;323(22):1505–11.

8.ConnollySJ,LaupacisA,GentM,etal.CanadianAtrialFibrillation Anticoagulation(CAFA)Study. JAmCollCardiol.1991;18(2):349–55.

9.EzekowitzMD,BridgersSL,JamesKE,etal.Warfarinintheprevention ofstrokeassociatedwithnonrheumaticatrial fibrillation.VeteransAffairs StrokePreventioninNonrheumaticAtrialFibrillationInvestigators [publishedcorrectionappearsin NEnglJMed.1993;328(2):148]. NEnglJMed.1992;327(20):1406–12.

10.ConnollySJ,KarthikeyanG,NtsekheM,etal.Rivaroxabaninrheumatic heartdisease-associatedatrial fibrillation. NEnglJMed 2022;387(11):978–88.

11.EikelboomJW,ConnollySJ,BrueckmannM,etal.Dabigatranversus warfarininpatientswithmechanicalheartvalves. NEnglJMed 2013;369(13):1206–14.

12.ConnollySJ,EzekowitzMD,YusufS,etal.Dabigatranversuswarfarin inpatientswithatrial fibrillation[publishedcorrectionappearsin NEnglJ Med.2010;363(19):1877]. NEnglJMed.2009;361(12):1139–51.

13.PatelMR,MahaffeyKW,GargJ,etal.Rivaroxabanversuswarfarinin nonvalvularatrial fibrillation. NEnglJMed.2011;365(10):883–91.

14.GrangerCB,AlexanderJH,McMurrayJJ,etal.Apixabanversus warfarininpatientswithatrial fibrillation. NEnglJMed 2011;365(11):981–92.

15.GiuglianoRP,RuffCT,BraunwaldE,etal.Edoxabanversuswarfarinin patientswithatrial fibrillation. NEnglJMed.2013;369(22):2093–104.

16.PowersWJ,RabinsteinAA,AckersonT,etal.GuidelinesfortheEarly ManagementofPatientswithAcuteIschemicStroke:2019Updateto the2018GuidelinesfortheEarlyManagementofAcuteIschemic Stroke:AGuidelineforHealthcareProfessionalsfromtheAmerican HeartAssociation/AmericanStrokeAssociation[published correctionappearsin Stroke.2019;50(12):e440-1]. Stroke 2019;50(12):e344–418.

17.ShahjoueiS,TsivgoulisG,GoyalN,etal.Safetyofintravenous thrombolysisamongpatientstakingdirectoralanticoagulants:a systematicreviewandmeta-analysis[publishedcorrectionappearsin Stroke.2020;51(3):e59]. Stroke.2020;51(2):533–41.

18.XianY,FederspielJJ,HernandezAF,etal.Useofintravenous recombinanttissueplasminogenactivatorinpatientswithacute ischemicstrokewhotakenon-vitaminKantagonistoralanticoagulants beforestroke. Circulation.2017;135(11):1024–35.

19.SeiffgeDJ,HooffRJ,NolteCH,etal.Recanalizationtherapiesinacute ischemicstrokepatients:impactofpriortreatmentwithnoveloral anticoagulantsonbleedingcomplicationsandoutcome. Circulation 2015;132(13):1261–9.

20.SuzukiK,AokiJ,SakamotoY,etal.LowriskofICHafterreperfusion therapyinacutestrokepatientstreatedwithdirectoralanti-coagulant. JNeurolSci.2017;379:207–11.

21.NationalInstituteofNeurologicalDisordersandStrokert-PAStroke StudyGroup.Tissueplasminogenactivatorforacuteischemicstroke. NEnglJMed.1995;333(24):1581–7.

22.HackeW,KasteM,BluhmkiE,etal.Thrombolysiswithalteplase 3to4.5hoursafteracuteischemicstroke. NEnglJMed 2008;359(13):1317–29.

23.SchmitzML,SimonsenCZ,HundborgH,etal.Acuteischemicstroke andlong-termoutcomeafterthrombolysis:nationwidepropensityscorematchedfollow-upstudy. Stroke.2014;45(10):3070–2.

24.Sindet-PedersenC,PallisgaardJL,StaerkL,etal.Temporaltrendsin initiationofVKA,rivaroxaban,apixabananddabigatranforthetreatment ofvenousthromboembolism. SciRep.2017;7(1):3347.

25.YuAYX,MaloS,SvensonLW,etal.Temporaltrendsintheuseand comparativeeffectivenessofdirectoralanticoagulantagentsversus warfarinfornonvalvularatrial fibrillation. JAmHeartAssoc 2017;6(11):e007129.

26.MurphyL,HillTP,PaulK,etal.Tenecteplaseversusalteplaseforacute stroke:mortalityandbleedingcomplications. AnnEmergMed 2023;82(6):720–8.

27.vonElmE,AltmanDG,EggerM,etal.TheStrengtheningtheReporting ofObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies. JClinEpidemiol 2008;61(4):344–9.

28.JauchEC,SaverJL,AdamsHPJr,etal.GuidelinesfortheEarly ManagementofPatientswithAcuteIschemicStroke:aguidelinefor healthcareprofessionalsfromtheAmericanHeartAssociation/ AmericanStrokeAssociation. Stroke.2013;44(3):870–947.

29.ConwaySE,HwangAY,PonteCD,etal.Laboratoryandclinical monitoringofdirectactingoralanticoagulants:whatcliniciansneedto know. Pharmacotherapy.2017;37(2):236–48.

30.StaerkL,GerdsTA,LipGYH,etal.Standardandreduceddosesof dabigatran,rivaroxabanandapixabanforstrokeprevention inatrial fibrillation:anationwidecohortstudy. JInternMed 2018;283(1):45–55.

31.CostelloM,MurphyR,JudgeC,etal.Effectofnon-vitamin-Koral anticoagulantsonstrokeseveritycomparedtowarfarin:ameta-analysis ofrandomizedcontrolledtrials. EurJNeurol.2020;27(3):413–8.

32.KumarS,RamosC,Garcia-CarrasquilloRJ,etal.Incidenceandrisk factorsforgastrointestinalbleedingamongpatientsadmittedtomedical intensivecareunits. FrontlineGastroenterol.2017;8(3):167–73.

ORIGINAL RESEARCH

Patient-relatedFactorsAssociatedwithPotentiallyUnnecessary

TransfersforPediatricPatientswithAsthma:

ARetrospectiveCohortStudy

GregoryA.Peters,MD*†‡

RebeccaE.Cash,PhD,MPH*†

ScottA.Goldberg,MD,MPH*‡

JingyaGao,MS†

LilyM.Kolb,MPH§

CarlosA.CamargoJr.,MD,DrPH*†∥

*HarvardMedicalSchool,Boston,Massachusetts

† MassachusettsGeneralHospital,DepartmentofEmergencyMedicine, Boston,Massachusetts

‡ BrighamandWomen’sHospital,DepartmentofEmergencyMedicine, Boston,Massachusetts

§ TheDonaldandBarbaraZuckerSchoolofMedicineatHofstra/Northwell, Hempstead,NewYork

∥

HarvardT.H.ChanSchoolofPublicHealth,DepartmentofEpidemiology, Boston,Massachusetts

SectionEditor: DavidThompson,MD

Submissionhistory:SubmittedJuly12,2023;RevisionreceivedDecember8,2023;AcceptedDecember28,2023

ElectronicallypublishedApril9,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.18399

Background/Objective: Asthmaisacommonchronicmedicalconditionamongchildrenandthemost commondiagnosisassociatedwithinterfacilitytransportsforpediatricpatients.Asmanyas40%of pediatrictransfersmaybeunnecessary,resultinginpotentialdelaysincareandunnecessarycosts.Our objectivewastoidentifythepatient-relatedfactorsassociatedwithpotentiallyunnecessarytransfersfor pediatricpatientswithasthma.

Methods: WeusedpatientcaredatafromtheCaliforniaDepartmentofHealthCareAccessand Informationpatientdischargeandemergencydepartment(ED)datasetstocaptureEDvisitswherea pediatricpatient(age2–17years)presentedwithasthmaandwastransferredtoanotherEDoracute carehospital.Theoutcomeofinterestwasapotentiallyunnecessarytransfer,definedasavisitwhere lengthofstayaftertransferwas <24hoursandnoadvancedserviceswereused,suchasrespiratory therapyorcriticalcare.Patient-relatedcharacteristicswereextracted,includingage,gender,race/ ethnicity,primarylanguage,insurancestatus,andclinicalcharacteristics.First,weuseddescriptive statisticstocomparenecessaryvsunnecessarytransfers.Second,weusedgeneralizedestimating equationsaccountingforclusteringbyEDtoestimateoddsratios(OR)andidentifyfactorsassociated withpotentiallyunnecessarytransfers.

Results: Atotalof4,233pediatricEDpatientsweretransferredwithadiagnosisofasthma,including461 (11%)transfersthatmetcriteriaaspotentiallyunnecessary.Medianagewas12years(interquartile range7–15),and46%werefemale.Factorsassociatedwithincreasedoddsofpotentiallyunnecessary transferwhilecontrollingforkeyfactorsincludedyoungerage(eg,2–5years,OR2.0,95%confidence interval[CI]1.4–2.9),malegender(OR1.4,95%CI1.1–1.7),andHispanicethnicity(OR1.6,95%CI 1.2–2.1),whilemultiplehospitalizationsforasthmaperyearwasassociatedwithdecreasedodds(OR 0.2,95%CI0.1–0.4).

Conclusion: Severalpatient-relatedfactorswereassociatedwithincreasedordecreasedoddsof potentiallyunnecessarytransfersamongpediatricpatientspresentingtotheEDwithasthma.These factorscanbeconsideredinfutureworktobetterunderstand,predict,andreduceunnecessarytransfers andtheirnegativeconsequences.[WestJEmergMed.2024;25(3)407–414.]

INTRODUCTION

Background

Asthmaisacommonchronicmedicalconditionamong children,1 affecting7.5%oftheoverallpediatricpopulation2 withpeakprevalenceinyoungteenagers(12–14years)at nearly11%.3 Childrenwithasthmaexacerbationsaccount forapproximately650,000emergencydepartment(ED) visitsintheUSannually,andmanyofthesevisitsresultin hospitaladmission,includingviainterfacilitytransferby emergencymedicalservices(EMS)toanotherhospital.4 Indeed,asthmaisthemostcommonprimarymedical diagnosisassociatedwithinterfacilitytransportforpediatric patients.5 Interfacilitytransfersaretypicallyinitiatedby emergencyphysicians,citinganeedforahigherlevelofcare (ie,criticalcare),recommendationofspecialtyservices(eg, pediatricpulmonology),orcapacity-relatedlimitations(ie, currentavailabilityofbedsorotherresources).Despitethe commonplacenatureofpediatrictransfersforasthmainthe ED,thereisnopriorliteraturetosupportpolicymakersand otherstakeholdersthatincludeemergencyphysicians andadministratorswhennavigatingthisroutine decision-makingprocess.

Importance

Priorworkhasshownthatmorethan40%of undifferentiatedpediatrictransferswereeitherdischarged directlyfromthereceivingEDorwithin24hoursofdirect admissionupontransfer,6 andthatonlyone-quarterofall pediatrictransfersarecompletedtoprovideahigherormore specializedlevelofcaretothepatient.7 Theseoutcomesare importantbecauseinterfacilitytransfersareassociatedwith misseddosesofmedication,prolongedtimetoinitiationof inpatientcare,andasubstantial financialburdenonfamilies andtaxpayers.7,8 Withtheserisksandcostsinmind,arecent studyof1.7millionpediatrictransfersintheUSreportedthat only12.3%ofallpediatrictransfersmetcriteriafora medicallynecessarytransfer,demonstratingthelimited directbenefittopatientcareinmanycasesretrospectively.9 Moreover,socioeconomicallyvulnerablepopulationsare disproportionallyaffectedbyasthma,10–14 andthe disproportionate financialburdenofinterfacilitytransferson underservedruralpatientshasbeenpreviouslydescribed,8 indicatingtheimportantlikelyhealthequityimplicationsof thistopic.Theseconsiderationsunderscoretheneedfor improvedguidanceforpolicymakerswhencontemplating theroutinepracticeofinterfacilitytransferofpediatric patientswhopresenttotheEDwithasthma.

GoalsofthisInvestigation

Weaimedtodescribethepatient-relatedfactors associatedwithpotentiallyunnecessaryinterfacilitytransfer ofpediatricpatientspresentingtotheEDwithasthma.Our ultimategoalinthisworkistostimulatediscussionand futureresearchregardingthecharacteristicsofpatientsmost

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Asthmaisthemostcommondiagnosis associatedwithinterfacilitytransferof pediatricpatients.Transfersentailcosts, delaysincare,andresourcestrain.

Whatwastheresearchquestion?

Whichpatient-relatedfactorsareassociated withunnecessarytransferforpediatric asthmaexacerbations?

Whatwasthemajor findingofthestudy?

Youngerage(OR2.0,95%CI1.4-2.9)and Hispanicethnicity(OR1.6,95%CI1.2 – 2.1) wereassociatedwithunnecessarytransfer.

Howdoesthisimprovepopulationhealth?

Severalpatient-relatedfactorswere associatedwithincreasedoddsofunnecessary transfer,whichcancausepreventablestrain onfamiliesandhealthcaresystems.

likelytoexperiencetheconsequencesofunnecessarytransfer andtodevelopinterventionstoreduceunnecessarystrainon patientsandtheirfamilies,EMS,andhospitalresources.

METHODS

StudyDesignandDataSource

Weusedaretrospectivecohortstudytoanalyzepatient careandhealthcareadministrativedatafromasampleof pediatricpatientswhopresentedtotheEDwithasthma.The primarysourceofdataforthisstudywastheCalifornia DepartmentofHealthCareAccessandInformation (HCAI),fromwhichwereceivedanon-publicversionofthe EmergencyDepartmentandAmbulatorySurgery(EDAS) DataandPatientDischargeData(PDD)datasets.The HCAIcompilesitsdataviamandatorystandardized collectionfromalllicensedhospitalsthroughoutthestateof California.TheHCAIorganizesdataasuniqueencounters betweenapatientandhealthcarefacility,suchthateach recordcorrespondstoonepatientencounteratagiven facility(eg,aninterfacilitytransferwouldgeneratetwo recordsforthatpatient).VisitstotheEDthatresultinsamehospitaladmissionareincludedinthePDD,whereasall otherEDvisits,includingthosethatresultininterfacility transfer,areincludedintheEDAS.CombiningEDASand PDDforagivenyearprovidesafulldatasetofallunique, unduplicatedEDvisitsinCaliforniawithinthatyear.The HCAIdatasetsaresubjecttostandardizedqualityassurance

procedures.MoredetailedinformationaboutHCAIcanbe foundonitswebsite.15

ThisstudyfollowedtheStrengtheningtheReportingof ObservationalStudiesinEpidemiology(STROBE) reportingguidelineforcross-sectionalstudies16 andwas approvedbytheMassGeneralBrighamHumanResearch CommitteeandtheCaliforniaCommitteefortheProtection ofHumanSubjects.

StudyPopulation

WeextractedallEDvisitsfromtheEDASandPDD datasetsfromHCAIduring2016–2019.Weincludedall patientsaged2–17yearstospecificallystudypediatric patients,andweexcludedpatientsyoungerthantwoyearsof agebecausewheezingatthisyoungageismorelikely attributabletoatransientconditionsuchasbronchiolitis ratherthanachroniconesuchasasthma.17,18 Weincluded pediatricpatientswhopresentedtotheEDandhada diagnosisofasthma(ie,InternationalClassificationof Diseases,Rev10,codesJ45,J98.01,andR06.2)forany documenteddischargediagnosis.Finally,weincluded pediatricasthmaEDvisitsthatresultedinaninterfacility transferfromtheED,regardlessoftheinitialcaresettingat thereceivingfacility(eg,EDtoED,EDtoinpatient floor, EDtointensivecareunit,etc).Insummary,the finalstudy sampleincludedpatientsaged2–17yearswhopresentedto anEDinCaliforniaduring2016–2019,werediagnosedwith asthma,andweretransferredviaEMStoanotherfacility.

Measures

Theprimaryoutcomemeasureusedforthisstudywas potentiallyunnecessaryinterfacilitytransfer.Thismeasure wasdesignedtocapturepatienttransfersthat withinthe limitationsofthisdatasource werenotassociatedwith clearretrospectiveindicationsthatthepatientreceived clinicallynecessaryservicesthatrequiredtransfer.We definedpotentiallyunnecessarypediatrictransferbasedon recentliterature(includingallchiefcomplaintcategories),9 wherethetransferdidnotresultinadispositionofdeathor admission >24hours,involvesedationoradvancedimaging (definedasanyimagingstudyapartfromplainradiographs), orincuroperatingroomorcriticalcarecharges.Weadded respiratorytherapyasanadditionalmarkerofnecessary transferinthecaseofasthma,inadditiontoservicescaptured undercriticalcaresuchaspositive-pressureventilation.All remainingtransferswereconsiderednecessary.

Moregenerally,weidentifiedinterfacilitytransfersby findingtwoencountersassociatedwithasingleunique patientidentifierwithinonedayofeachotherperencounter date,wheretheencountersoccurredattwoseparate hospitals,andwheredispositionatthesendingfacilitywas designatedasatransfer.Additionalvariablesofinterest includedsociodemographiccharacteristicsanddetailsto describepatients’ medicalhistoryandhealthcareutilization

relatedtoasthma.Demographicdataincludedpatientage, gender,race,ethnicity,primarylanguage,residence urbanicity,andinsurancestatus.Wealsoincludedthe numberofEDvisitsandhospitaladmissionseachpatient hadperyearwhereasthmawaslistedasthedischarge diagnosis.Finally,wecalculatedthenumberofcomplex chronicconditionsfromeachpatient’spastmedical historyperthePediatricComplexChronicCondition version2system(includingtechnologydependenceand organtransplantation).19,20

StatisticalAnalysis

First,weuseddescriptivestatisticstocomparepediatric transfersforasthmathatmetvsdidnotmeetcriteriafor potentiallyunnecessarytransfer.Comparisonsbetween groupsweremadeusing t -tests,Wilcoxonrank-sumtests,or chi-squaretestsasappropriate.Weusedgeneralized estimatingequations(GEE)accountingforclusteringby facilitytocalculateadjustedoddsofunnecessarytransfer, estimatedwithbinominaldistribution,logitlinkfunction, workingindependencecorrelation,androbuststandard errors.Covariatesincludedthepatient-relatedfactors describedabovethatwereincludedaprioribasedonprior literatureandsubstantivereasoning.Weperformedall statisticalanalysesusingStataversion15.0(StataCorp, CollegeStation,TX).

Figure1. Inclusionandexclusioncriteriausedtodevelopthe final studysample.Statisticsregardingemergencydepartmentvisitsand hospitaladmissionsdonotrepresentpatientsbeingincludedor excluded;insteadtheyprovidecontexttoaidinunderstandingthe relativerateofpatienttransferswithinthesample. ED,emergencydepartment.

RESULTS

Fromaninitialsampleof3,709,523pediatricencounters, atotalof4,233patientswithasthmaweretransferredfrom

anED(eachincludingapairoftwoencounters,oneatthe sendingfacilityandasecondatthereceivingfacility; Figure1).Amongthissample,461(11%)metcriteriaas

Table1. Descriptivecharacteristicsandcomparisonofnecessaryandpotentiallyunnecessaryinterfacilitytransfers.

FactorOverallNecessaryUnnecessary

Totaln4,2333,772(89)461(11)

Ageinyears

Mean(SD)10.8(4.8)11.0(4.8)9(4.6) <

Agecategory,n(%) <0.001

Child(2–5years)843(20)718(19)125(27)

Schoolage(6–12years)1,451(34)1,249(33)202(44)

Teen(13–17years)1,939(46)1,805(48)134(29)

Gender,n(%)

Female1,952(46)1,789(47)163(35)

Male2,281(54)1,983(53)298(65)

Race/ethnicity,n(%)

Non-HispanicWhite1,041(25)961(26)80(17)

Non-HispanicBlack918(22)835(22)83(18)

Hispanic1,724(41)1,491(40)233(51)

Non-Hispanicother518(12)455(12)63(14)

English3,874(92)3,467(92)407(88)

Spanish322(8)275(7)47(10)

Otherormissing37(1)30(1)NR

EDvisitswithouttransferforasthmaperyear

Mean(SD)2.1(1.9)2.1(2.0)1.6(1.6) <0.001 >2EDvisitperyear,n(%)1,068(25)983(26)85(18)

Admissionsforasthmaperyear

Mean(SD)1.2(0.8)1.2(0.8)0.6(0.7) <

>1admissionperyear,n(%)688(16)669(18)19(4) <0.001

Anycomplexchroniccondition†,n(%)463(11)425(11)38(8)0.050 Patientresidenceurbanicity,n(%)

Rural83(2)78(2)NR

Urban4,131(98)3,679(98)452(99) Missing 1915NR

Insurancestatus,n(%)

Public2,514(59)2,235(59)279(61) Private1,548(37)1,385(37)163(35) Self-pay/other/missing171(4)152(4)19(4)

Columnpercentagesshown.Percentagesmaynotsumto100duetorounding.

*T-testorchi-squaretestasappropriate.

†Complexchronicconditionsdefinedusingversion2definitionfromFeudtneretal,2014,(https://bmcpediatr.biomedcentral.com/articles/10. 1186/1471-2431-14-199),adaptedfromKurowskietal,2014.(https://pubmed.ncbi.nlm.nih.gov/25039935/).Notably,theversion2definition includestechnologydependenceandtransplantbutdoesnotaddthesetothetotal.Inthisanalysis,technologydependenceandtransplant areincludedintotalcomplexchroniccondition.

ED,emergencydepartment; NR,notreported(duetodatareportingrestrictions).

potentiallyunnecessary.Themeanageofallpediatric patientswithatransferforasthmawas10.8years(SD4.8), 47%werefemale,and41%wereHispanic(Table1).Patients withapotentiallyunnecessarytransferwereyounger(9vs 11years, P < 0.001),moreoftenmale(65%vs53%, P < 0.001),andmoreoftenhadHispanicethnicity (51%vs40%, P < 0.001).Intermsofclinicalcharacteristics,

Table2. Oddsofunnecessarytransferusinggeneralizedestimating equationstoaccountforclusteringbyhospital.

Agecategory

Child(2–5years)2.011.39,2.91

Schoolage(6–12years)2.031.50,2.73

Teen(13–17years)1.00(referent)

Gender

Female1.00(referent)

Male1.391.15,1.70

Race/ethnicity

Hispanic1.00(referent)

Non-HispanicBlack0.630.45,0.86

Non-HispanicWhite0.630.48,0.83

Non-Hispanicother0.880.65,1.20

Primarylanguage

English1.00(referent)

Spanish1.180.82,1.70

Otherormissing1.690.74,3.87

EDvisitsforasthmaperyear

0–21.00(referent)

≥30.940.68,1.28

Admissionsforasthmaperyear

0–11.00(referent)

≥20.220.13,0.36

Complexchroniccondition*

None1.00(referent)

≥10.790.55,1.15

Patientresidenceurbanicity

Rural0.720.35,1.58

Urban1.00(referent)

Insurancestatus

Public0.980.75,1.28

Private1.00(referent)

Self-pay/other/missing1.020.55,1.89

*Complexchronicconditionsdefinedusingversion2de finition.In thisanalysis,technologydependenceandtransplantareincluded intotalcomplexchroniccondition.

AOR,adjustedoddsratio; CI,confidenceinterval; ED,emergencydepartment.

patientswhometcriteriaforpotentiallyunnecessarytransfer lessoftenhadacomplexchroniccondition(8%vs11%, P = 0.05)andovertheprioryearexperiencedfewerED visits(1.6vs2.1, P < 0.001)andhospitaladmissions (0.6vs1.2, P < 0.001)forasthma.

UsingtheGEEmodeltoaccountforclusteringbyhospital (Table2),wefoundthatyoungeragegroupswereassociated withincreasedadjustedoddsofpotentiallyunnecessary transfer(ie,age2–5years,oddsratio[OR]2.01,95% confidenceinterval[CI]1.39–2.91,comparedtoage 13–17years).Malegender(OR1.39,95%CI1.15–1.70)was alsoassociatedwithincreasedoddsofpotentially unnecessarytransfer.Noassociationswerefoundwith insurancestatus,residenceurbanicity,orprimarylanguage. PatientsofHispanicethnicity,comparedtonon-Hispanic Whitepatients,hadincreasedoddsofpotentially unnecessarytransfer(OR1.59,95%CI1.21–2.10).(Note thatHispanicethnicitywasusedasthereferentin Table2 giventhatthiswasthelargestracial/ethnicgroupinthis sample.)Twoormorehospitaladmissionsforasthmaper yearwasassociatedwithdecreasedoddsofpotentially unnecessarytransfer(OR0.22,95%CI0.13–0.36),whereas noassociationswerefoundwithEDvisitsforasthmaor absenceofanycomplexchronicconditions.

DISCUSSION

Usingacomprehensive,statewidedatasetofEDvisitsand admissions,wefoundseveralpatientcharacteristics associatedwithpotentiallyunnecessarytransferofpediatric patientswhopresenttotheEDwithasthma.These findings describethepatient-levelcharacteristicsassociatedwith elevated(orreduced)oddsofpotentiallyunnecessary transfer,whichcaninformpolicymakersandED administratorstoconsidersubpopulationswithelevatedrisk ofunnecessarytransferwhendevelopingfuturestudiesand policiesrelatedtothetransferofpediatricpatientswith asthma.Potentiallyunnecessarytransfersmarkcaseswhere patientsdonotshowevidenceofthebenefitsoftransfer,such asahigherlevelofcareoraccesstoaspecialistbutdo experiencetherisksandcostsassociatedwithtransfer.

Therateofpotentialunnecessarytransferamongthis cohortwas11%,whichismuchlowerthanreported previouslyinstudiesofundifferentiatedpediatricpatients, includingratesofone-in-twotonearlynine-tenths.9,21,22 However,inpriorliteraturethediagnosticcategoryof respiratoryemergencieshadthegreatestnumberof transferredpediatricpatientsandwastheonlydiagnostic categoryassociatedwithdecreasedoddsofdirectdischarge homefromtheED,whichmayatleastpartiallyexplainwhy weobservedalowerrateofpotentiallyunnecessary transfer.21 Perhapsrespiratoryemergenciesarerelativelyless likelytobequicklydischargedfromthereceivingfacility comparedtootherdiagnosesbecausetheyaremorelikelyto involveanobservationperiod(eg,continuousoxygen

saturationmonitoring,gradualreductioninfrequencyof respiratorytreatments),orperhapsbecauseemergency physicianshavemorecomfortwithdecision-makingfor thesetransfersduetotheirmorecommonplacenature. Similarly,theincreasedprevalenceofobservationunitsmay havecontributedtothis findingcomparedtoearlierstudies whenthisoptionwaslessavailable.

Demographicfactorsassociatedwithincreasedoddsof unnecessarytransfersincludedyoungerage,malegender, andHispanicethnicity.Youngerpatientswereassociated withincreasedoddsofpotentiallyunnecessarytransfer, whichwasconsistentwithpriorreportsthatfound particularlyhighriskamongpreschool-agepatients.9,22 One possibleexplanationforthis findingcouldsimplybe relativelyinferiorcomfortamongclinicianscaringfor youngerpatients.Increasedoddsofpotentiallyunnecessary transferamongmalepatientswasnotexpectedaprioriand wasnotfoundintheliteratureamongundifferentiated pediatricpatients,wherefemalepatientsaremoreoften foundtobeathigherriskwhengender-baseddifferencesare found.9 Priorresearchhassuggestedthatmalepediatric patientstendtohavegreaterprevalenceandillnessseverityof asthmabeforepuberty,incontrasttogreaterprevalenceand severityamongfemalesafterpuberty,withsomeindication thatsexhormonesmayplayarole.23,24 Furtherresearch focusedonthetransferofpediatricpatientswithasthmawill beneededtodeterminewhethergender-relateddifferences arewidespread,andifso,whatmayaccountfor thisdisparity.

Hispanicethnicitywasalsofoundtobeassociatedwith increasedoddsofpotentiallyunnecessarytransfer.One possiblecontributingfactortothis findingmighthavebeen languagebarriers;however,evenaftercontrollingfor primarylanguagespoken,wefoundanindependent associationwithHispanicethnicity.Hispanicethnicityhas previouslybeenreportedtobeassociatedwithincreasedodds ofunnecessarytransfer,butthereasonsforthisremain unclear.9,25 Giventheknown financialcostsandmedical risksthatcanbeassociatedwithinterfacilitytransfer7,8 and prior findingsthatHispanicpatientsincurgreatercosts associatedwiththeirasthma-relatedcareingeneral,26 we encouragefurtherworktoevaluatethistrendtogainabetter understandingofhowtominimizedisparitiesintheburden ofundueriskassociatedwithpotentiallyunnecessary interfacilitytransfer.

Wefoundnoassociationwithpatientresidenceurbanicity orwithinsurancestatus,incontrasttopriorresearchthat foundincreasedriskofpotentiallyunnecessarytransfer amongurbanpatientsandthosewithpublichealth insurance.9 Incontrasttoour findings,importantpriorwork hashighlightedthatruralpatientsaremorelikelyto experiencepotentiallyunnecessarytransferbecausetheir nearesthospitalstendtobelessresourced,especiallyfor pediatriccare,comparedtourbanpatientswhoaremore

likelytoresideingeographicproximitytomoreresourced centersthatarelesslikelytotransferpediatricpatients.8 CharacteristicsspecifictothestateofCaliforniamayaccount forthesedifferences,althoughratesofpublicinsuranceand urbanresidencewerequitesimilarinthiscohortcomparedto priorwork.Alternatively,ourfocusonasthmacouldbethe explanatoryfactor,giventhatasthmaismorecommonin urbansettings(particularlyinthepresenceofotherfactors thattendtoaffectsuchareas,suchasenvironmentaland housing-relatedinsults)andisconsideredanambulatory care-sensitiveconditionshowntobemodulatedbychanges ininsurancestatusamongpopulations.14,27 Takentogether, asolid,generalizableconclusionregardingthepotential associationbetweenpotentiallyunnecessarytransferand urbanicityorinsurancestatusremainsdifficulttoreach.

Finally,regardingclinicalpatient-levelfactors,wefound thatmultiplehospitaladmissionsforasthmaperyearwere associatedwithdecreasedoddsofpotentiallyunnecessary transfer,whereasnoassociationwasfoundwith ≥3EDvisits orthepresenceofacomplexchroniccondition.The finding thatmultiplehospitaladmissionsforasthmawasassociated withdecreasedriskisnotunexpected.Theincreasedriskof mortalityassociatedwithrecenthospitalizationforasthma notonlysuggestsgreaterlikelihoodofincreasedclinical severityamongthiscohortbutisalsoawelldocumentedand widelytaughtpieceofevidencethatdirectlyfactorsinto clinicaldecision-making,whichmakesitlesslikelyforsuch patientstomeetcriteriaforpotentiallyunnecessary transfer.28,29 However,thesamelineofworkthattendsto indicatetheassociationbetweenhospitalizationsand mortalityoftenhighlightssimilarassociationswithEDuse, albeitanintuitivelyweakerassociationgiventhatchildren whovisittheEDandareadmittedpresumablyhavemore severeasthmathanthosewhoareinsteaddischarged.

Priorresearchhasshownthatcomplexchronicconditions tendtobeassociatedwithlongerlengthofstayandgreater resourceutilizationwhentransferredcomparedtonot transferredforadmission,suggestingthatcomplexpediatric patientstendtobeappropriatelytransferred.However,we didnot finddecreasedriskamongthiscohort,perhapsdueto over-triageinsomecaseswhereanasthmaexacerbationwas relativelymild,buttransferwasneverthelesspursuedoutof concernforpoorreserveandlikelyclinicaldeteriorationorto provideahigherlevelofspecializedcare,perhapsinsome caseswithteamstowhomthepatientisknown.

LIMITATIONS

Therearelimitationsassociatedwiththisstudy.First,the HCAIdataincludesamixofadministrativeandclinicaldata sourcedfrompatienthealthrecords.Thus,definitionsofkey variables,suchasrace/ethnicityandevenincludingthe diagnosisofasthma,aresubjecttoinformationbiasand misclassification.Mostnotably,theprimaryoutcome measureusesacompositedefinitionpreviouslyestablishedin

theliteraturebutslightlyadaptedforthepurposesofastudy focusedonasthma.Thisdefinitionofpotentiallyunnecessary transferisausefultoolforretrospectiveresearchbutisfar fromperfectandlikelyincludessomedegreeoferror associatedwithmisjudgmentintheabsenceofmore detailedcase-by-caseinformation.Moreover, appropriatenessofdischargewithin24hourswasnot assessed,suchasthoroughdocumentationofreturnvisits. Similarly,definingtheinclusioncriteriaforthisstudyrelied uponassumptionsandtrade-offs,suchasincludingpatients withanyactivediagnosisofasthmaexacerbation,rather thanprimarydiagnosis,toovercomethelimitationsofthis datasetdespitetheriskofincludingpatientswithother primarydiagnoses.

Second,theHCAIdatasetisrestrictedtothestateof California.Californiaisalarge,heavilypopulated,and diversestate,whichmakes findingsfromsamplesofits residentsmorenationallyrelevantthanthosefrommost states.However,thegeneralizabilityofthese findingsto otherstatesandnationallyisunclear.Third,themodelsused inthisanalysisfocusonpatient-levelcharacteristics, thereforeneglectingthemanyotherfactorsthathavebeen reportedtoplayintothedecisionforinterfacilitytransfer, includingpatientvolumesandhospitalcapacities,hospitalrelatedfactorssuchasresourcesavailable,andcommunityrelatedfactorssuchastheavailabilityofoutpatient physicians.Fourth,weusedaretrospectivecohort design,whichlimitstheinterpretationof these findings.

Fifth,datawasfrom2016–2019,priortotheCOVID-19 pandemic.TheCOVID-19pandemichadwidespreadand complexeffectsontheUShealthcaresystem,includingthe managementofrespiratoryconditionsandthereallocation ofpediatricandcriticalcareresourcesthroughoutsystems; thus,useofdatapriortothepandemicmightmoreclosely relatetocurrentconditionsastheUShealthcaresystem continuestoadaptandfurthernormalize.Ongoingresearch intotheeffectsoftheCOVID-19pandemiconthe interfacilitytransferofpediatricpatientswithrespiratory complaintswillprovideadditionalinformationtobetter evaluatethisassessment.

CONCLUSION

UsingastatewidedatasetofEDvisitsandadmission,we foundthatyounger,Hispanic,andmalechildrenwho presentedtotheEDwithasthmahadhigheroddsof experiencingpotentiallyunnecessaryinterfacilitytransfer. Patientswithmultiplehospitaladmissionsforasthmawithin theprioryearwerefoundtohavedecreasedoddsof potentiallyunnecessarytransfer.Importantnextstepsinthis lineofinvestigationincludestudiestargetedatdiscrepancies betweenthese findingsandpriorresearch,investigationof the financialcostsassociatedwithunnecessarytransferof pediatricpatientswithasthma,andanalysisofthehealthcare

systems-relatedfactorsassociatedwithpotentially unnecessaryinterfacilitytransfers.Theseinsightscanbe consideredbypolicymakersandEDadministratorsto identifysubpopulationsofpatientsthataremorelikelytobe impactedbynewinterventionsortoinformfuturestudies concernedwithdisparitiesindelaysincareor financialcosts associatedwithunnecessarytransfer.Findingsfromthis studywillneedvalidationthroughamorerigorous prospectivestudytoconfirmthepatientcharacteristicsthat mightbeassociatedwithincreasedriskofpossiblyavoidable transfersandthepotentialconsequencesassociated withthem.

AddressforCorrespondence:CarlosACamargo,MDDrPH, MassachusettsGeneralHospital,DepartmentofEmergency Medicine,125NashuaSt.,Suite920,Boston,MA02114. Email: ccamargo@partners.org

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.ThisresearchwasfundedbytheSociety forAcademicEmergencyMedicineFoundation.Therearenoother conflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Petersetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.HealthResources&ServicesAdministration.NationalSurveyof Children’sHealth:NSCHDataBrief.2020.Availableat: https://mchb. hrsa.gov/sites/default/files/mchb/about-us/nsch-data-brief.pdf AccessedJune12,2021.

2.CDC.Table4–1CurrentAsthmaPrevalencePercentsbyAge,United States:NationalHealthInterviewSurvey,2018.2019.Availableat: https://www.cdc.gov/asthma/nhis/2018/table4-1.htm AccessedDecember25,2020.

3.CDC.MostRecentNationalAsthmaData.2023.Availableat: https:// www.cdc.gov/asthma/most_recent_national_asthma_data.htm AccessedFebruary2,2021.

4.CDC.Asthmaemergencydepartment(ED)visits2010–2018.2021. Availableat: https://www.cdc.gov/asthma/asthma_stats/ asthma-ed-visits_2010-2018.html.AccessedJuly6,2021.

5.RosenthalJL,HiltonJF,TeufelRJ,etal.Profilinginterfacilitytransfers forhospitalizedpediatricpatients. HospPediatr. 2016;6(6):345–53.

6.LiJ,MonuteauxMC,BachurRG.Interfacilitytransfersofnoncriticallyill childrentoacademicpediatricemergencydepartments. Pediatrics. 2012;130(1):83–92.

7.StolteE,IwanowR,HallC.Capacity-relatedinterfacilitypatient transports:patientsaffected,waittimesinvolvedandassociated morbidity. CJEM. 2006;8(4):262–8.

8.MohrNM,HarlandKK,ShaneDM,etal.Potentiallyavoidablepediatric interfacilitytransferisacostlyburdenforruralfamilies:acohortstudy. AcadEmergMed. 2016;23(8):885–94.

9.RichardKR,GlissonKL,ShahN,etal.Predictorsofpotentially unnecessarytransferstopediatricemergencydepartments. Hosp Pediatr. 2020;10(5):424–9.

10.WrightRJandSubramanianSV.Advancingamultilevelframework forepidemiologicresearchonasthmadisparities. Chest. 2007;132(5Suppl):757S–769S.

11.FornoEandCeled´onJC.Healthdisparitiesinasthma. AmJRespirCrit CareMed. 2012;185(10):1033–5.

12.GoldDRandWrightR.Populationdisparitiesinasthma. AnnuRev PublicHealth. 2005;26(1):89–113.

13.Bryant-StephensT.Asthmadisparitiesinurbanenvironments. JAllergy ClinImmunolGlob. 2009;123(6):1199–206.

14.PetersGA,OrdoobadiAJ,CashRE,etal.AssociationofAffordable CareActimplementationwithambulanceutilizationforasthma emergenciesinNewYorkCity,2008–2018. JAMANetwOpen. 2020;3(11):e2025586.

15.HCAI.DataResources.Availableat: https://hcai.ca.gov/data/ data-resources/.AccessedMay1,2023.

16.vonElmE,AltmanDG,EggerM,etal.TheStrengtheningtheReporting ofObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies. AnnInternMed. 2007;147(8):573–7.

17.MartinezFD,WrightAL,TaussigLM,etal.Asthmaandwheezinginthe firstsixyearsoflife. NEngJMed 1995;332(3):133–8.

18.NASEMSO.NationalModelEMSClinicalGuidelinesVersion3.0.2022. Availableat: https://nasemso.org/wp-content/uploads/National-ModelEMS-Clinical-Guidelines_2022.pdf.AccessedApril9,2022.

19.MurtaghKurowskiE,ByczkowskiT,Grupp-PhelanJM. Comparisonofemergencycaredeliveredtochildrenandyoungadults

withcomplexchronicconditionsbetweenpediatricand generalemergencydepartments. AcadEmergMed. 2014;21(7):778–84.

20.FeudtnerC,FeinsteinJA,ZhongW,etal.Pediatriccomplexchronic conditionsclassificationsystemversion2:updatedforICD-10and complexmedicaltechnologydependenceandtransplantation. BMC Pediatr. 2014;14(1):199.

21.PeeblesER,MillerMR,LynchTP,etal.Factorsassociatedwith dischargehomeaftertransfertoapediatricemergencydepartment. PediatrEmergCare. 2018;34(9):650.

22.GattuRK,TeshomeG,CaiL,etal.Interhospitalpediatricpatient transfers factorsinfluencingrapiddispositionaftertransfer. Pediatr EmergCare. 2014;30(1):26.

23.ChowdhuryNU,GunturVP,NewcombDC,etal.Sexandgenderin asthma. EurRespirRev. 2021;30(162):210067.

24.FuseiniHandNewcombDC.Mechanismsdrivinggenderdifferencesin asthma. CurrAllergyAsthmaRep. 2017;17(3):19.

25.OsenHB,BassRR,AbdullahF,etal.Rapiddischargeaftertransfer:risk factors,incidence,andimplicationsfortraumasystems. JTraumaAcute CareSurg. 2010;69(3):602.

26.GlickAF,TomopoulosS,FiermanAH,etal.Disparitiesinmortalityand morbidityinpediatricasthmahospitalizations,2007to2011. Acad Pediatr. 2016;16(5):430–7.

27.GrantTLandWoodRA.Theinfluenceofurbanexposuresand residenceonchildhoodasthma. PediatrAllergyImmunol. 2022;33(5):e13784.

28.ChangY-L,KoH-K,LuM-S,etal.Independentriskfactorsfordeathin patientsadmittedforasthmaexacerbationinTaiwan. NPJPrimCare RespirMed. 2020;30(1):1–8.

29.D’AmatoG,VitaleC,MolinoA,etal.Asthma-relateddeaths. Multidiscip RespirMed. 2016;11:37.

EXPERT COMMENTARY

PublicHealthInterventionsintheEmergencyDepartment: AFrameworkforEvaluation

ElisabethFassas,MD,MSc

KyleFischer,MD,MPH

StephenSchenkel,MD,MPP

JohnDavidGatz,MD

DanielB.Gingold,MD,MPH

SectionEditor:TonyZitek,MD

UniversityofMarylandSchoolofMedicine,Baltimore,Maryland

Submissionhistory:SubmittedMay21,2023;RevisionreceivedDecember14,2023;AcceptedDecember15,2023

ElectronicallypublishedMarch29,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18316

Emergencydepartments(ED)intheUnitedStatesserveadualroleinpublichealth:aportalofentryto thehealthsystemandasafetynetforthecommunityatlarge.PublichealthofficialsoftentargettheED forpublichealthinterventionsduetotheperceptionthatitisuniquelyabletoreachunderserved populations.However,undertimeandresourceconstraints,emergencyphysiciansandpublichealth officialsmustmakecalculateddecisionsinchoosingwhichinterventionsintheirlocalcontextcould providemaximalimpacttoachievepublichealthbenefit.Weidentifyhowdecisionsregardingpublic healthinterventionsareaffectedbyconsiderationsofcost,time,andavailablepersonnel,andfurther considertheroleoflocalcommunityneeds,healthdepartmentgoals,andpoliticalenvironment.We describeasampleofED-basedpublichealthinterventionsanddemonstratehowtouseaproposed frameworktoassessinterventions.Wepositaseriesofquestionsandvariablestoconsider:local diseaseprevalence;abilityoftheEDtoperformtheintervention;relativeefficacyoftheEDvscommunity partnershipsastheprimaryinterventionlocation;andexpectedoutcomes.Inusingthisframework, cliniciansshouldbeempoweredtoimprovethepublichealthintheircommunities.[WestJEmergMed. 2024;25(3)415–422.]

INTRODUCTION

Emergencydepartments(EDs)intheUnitedStatesservea dualroleinpublichealth:aportalofentrytothehealth system,andasafetynetforthecommunityatlarge. Clinically,itspositionisclear;theEDprovidesunscheduled acutecare,regardlessofapatient’sabilitytopay.Givenits functionasasafetynetforpeoplelackingconsistentaccessto care,however,theEDisoftenidentifiedforpotentialpublic healthinterventionsduetoaperceptionthatithasaunique abilitytoreachunderservedpopulations.Unsurprisingly,the fieldofemergencymedicine(EM)hastakenonthischallenge andpioneeredanumberofeffectivepublichealth interventions,rangingfromcommunityviolenceprevention1 totreatmentofopioidusedisorder.2

Onestudy3 identified43conditionsproposedinthepeerreviewedliteratureforED-basedpublichealthscreening and/orintervention.Giventhelogisticalimprobabilityofany departmentemployingallproposedinterventions,clinicians

mustmakecalculateddecisionsabout which interventionsto deployand how toimplementthemsuccessfully. Unfortunately,thereisalackofevidence-basedguidancein theEMliteratureonhowEDsshouldprioritizeand implementsuchinterventionssoastomaximallybenefitthe publichealthoftheirlocalcommunity.Thesedecisionsare increasinglyimportantgiventhegrowingstressanddemands alreadyplacedonEDsaroundthecountry.Annualpatient volumeshaveincreasedsubstantially.Patientacuityis gettingmorecomplex.Emergencydepartmentboardinghas becomeanationalcrisis.4 Giventhesignificantresource limitationsoftheEDfromthesetypesoffactors,anypublic healthinterventionbeyondcoreclinicalcaremusthavea clearroleintheEDsetting.

Inthispaper,weproposeaframeworkgroundedin implementationscienceprinciplesforEDstoprioritize interventionsthatmaximizepublichealthbenefitsandreview thekeyelementsofsuccessfulimplementation.Wepresent

thisfromtheperspectiveofourownexpertise:EDmedical directorswhohaveimplementednumerousED-basedpublic healthinterventions;anemergencymedicalservicesmedical directorworkingonpopulationhealthprojects;publichealth researchersandadvocates;publicpolicyexperts;and emergencyphysicians.Werecognizethattheconversation regardingED-basedpublichealthinterventionsis challengingandaffectedbymanyconsiderationsboth internalandexternaltotheED,butwebelievesuccessis possiblewiththerightapproach.

PROPOSAL

Thevolumeofpotentialpublichealthproposals necessitatesaframeworkfordeterminingwhicharemost meaningfullydeployedasinterventionsinaspecificED.As eachadditionalpublichealthscreeningorinterventiontakes timewithinthecontextofanEDvisit,thereisatangiblecost totheindividualpatientassociatedwithparticipationin publichealth-focusedinterventions.Prioritizationis challengingforEDadministrators,asproposedinitiatives rarelyarisebya fixedprocessbutratherfromaconstellation offactors:acutepublichealthemergencies;issuesoflongstandingconcernwithindividualinterestorexpertisefroma frequentlychangingphysicianandnursingstaff;strategic initiativesfromhospitalsystems;andoftenchanging prioritiesfromlocalpublichealthdepartmentsorpolitical

leaders.Considerationsoffunding,time,andcapacityto providetheinterventionwith fidelityareoftenincomplete. Moreover,interventionsmaybeimplementedwithoutaplan forrigorousevaluationtojustifytheircontinuedpresence. Giventhesechallenges,asystematicapproachtodecisionmakingmaymaximizehealthoutcomes.

Inthiscontext,weprovideaframeworkforconsidering themeritsofconductingaparticularinterventionwithinan EDvisit.Asnoidealframeworkyetexists,wehaveadapted constructsfromtheConsolidatedFrameworkfor ImplementationResearch(CFIR).Thisimplementation frameworkwasoriginallypublishedin2009,7 representing thecumulativeknowledgeofimplementationscienceatthe time.Itisa “pragmaticstructure” foreffective implementationofprogramsandsystemschange precisely whatisneededforenactingeffectivepublichealthprograms intheED.Wedidnot findallconstructsofCFIRpertinentto determiningtheappropriatenessofanew,ED-basedpublic healthintervention.Thoseconstructsdeemedmostrelevant, byauthorconsensus,areoutlinedin Table1 asamodified frameworkforconsideringthemeritsofapotential intervention.Theframeworkwepresentisthusa commentary,basedonourexperienceinEMandpublic healthadministration.

TheCFIRgroupsimplementationscienceconstructs across fivedomains(intervention,process,individuals,inner

Table1. Recommendedconsiderationsforimplementingnewemergencydepartment-basedpublichealthinterventions(Consolidated FrameworkforImplementationResearchmodel).

CFIRmajor domains RelevantCFIR constructsQuestionstoconsider

Intervention characteristics

1.Evidencestrength andquality

• Hastheproposedinterventionshowneffectivenessinpatient-centered outcomesintheEDsetting?

• Ifnot,hastheinterventionshownbenefitthatislikelytotranslatetotheED setting?

• Howstrongistheevidencebase?

2.Relativeadvantage

3.Adaptability

4.Trialability

5.Complexity

• AretherelocationsotherthananED,eitherinthehospitalorinthecommunity thatmaybeamorepatient-centricinterventionsite?

• Cananyoftheselocationsperformthisinterventionmoreeasily,efficiently, cheaply,oreffectively?

• Willthelocalcontextrequireanydeviationsfromtheestablishedprogram model?Ifso,howcouldthesedifferencesimpactefficacy?

• Doestheproposedinterventionhavethe flexibilitytoevolve,asnecessary,after initialimplementation?

• Whatisthetimelineoftheintervention?Isthereaclearendpoint?

• Willitbepossibletoendtheinterventionifnoteffective?

• Whatchallengesmightarisetomaintaining fidelitytotheestablishedprogram model?

• Whatarepossibleunintendedadverseeffectsoftheinterventionfornonparticipants?Arecostsshared,orarespecificpopulationsdisproportionately harmed?

• Aretherehealthequityconsiderations?

(Continuedonnextpage)

Table1. Continued.

CFIRmajor domains RelevantCFIR constructsQuestionstoconsider

Externalcontext*1.Patientneedsand resources

• Whatisthelocalprevalenceofthetargetedconditioninthegeneralpopulation? TheEDpopulation?

• IsthetargetedpopulationmostreadilyaccessiblewithintheED?Arethere alternativeandpotentiallymorepatient-centeredlocations?

• HowdoestheconditionaffectlocalEDutilization,includingreturnvisitsand hospitalization?

Organizational characteristics**

2.Externalnetworking

3.Peerpressure

4.Externalpolicyand incentives

1.Culture

2.Compatibility

3.Relativepriority

4.Leadership engagement

5.Availableresources

6.Accesstoknowledge andinformation

Characteristicsof individualsinvolved

1.Knowledge&beliefs abouttheintervention

2.Individualstageof change

3.Otherpersonal attributes

Processof Implementation

1.Planning

2.Opinionleaders

3.Champions

4.Executing

5.Reflectingand evaluating

• ArethereeffectivesystemsinplacetocontinuecareafterEDdischarge?

• Howmighttheabsence,change,orlossofexternalpartnersaffectthe intervention?

• Howdoestheengagementofothersintheareaaffecttheneedforthe interventionandthepotentialforefficacy?

• Whatstakeholdersorpolicymakersareencouragingimplementation?

• Forprogramsrelyingonexternalfunding,whatisthelong-termstabilityofthis funding?

• Doestheintervention fitwithintheorganizationalmissionoftheED?

• Doestheintervention fitwithintheorganizationalmissionofthehospital?

• Howdoesthisintervention fitwithintheexistingworkflowoftheED?

• HowwouldtheinterventionalterEDperformancemetrics?

• WhatessentialEDprocessesmightbeimpactedbytheintervention?For example,willthroughputbereduced,waittimesincreased,ortriageburdened?

• Whatotherprogramsmayneedtobesacrificedforimplementation?

• Doexpectedbenefitsoutweighpotentialdisruption?

• Istherebuy-infrombothEDandhospitalleadership?

• IstherebandwidthwithintheEDleadershipfortheintervention?

• WilladditionalresourcesberequiredtoaccomplishtheinterventionintheED? Howmightthoseresourcesbemadeavailable?

• Arethereadditionaloutsideresourcesthatthatcouldbebroughttobear?

• Isthisaconditioninwhichemergencyclinicianshavespecificexpertise?

• Whatsourcesofpublichealthexpertisecanbetappedwithinthedepartment?

• Whatadditionaltrainingortechnicalexpertisemightbeaccessed?

• AretheassumptionssupportingimplementationintheEDvalid?

• Arefront-linestaffmotivatedtoparticipateintheintervention?

• Whatcultural,religious,orpoliticalconcernsmaystaffhaveaboutthe intervention?

• Howwilltheplanbedevelopedanddisseminated?

• Howmuchtimeisneededtodevelopanimplementationplanandformulate alliances?

• Whatsupportoroppositionwillimplementationhavefromopinionleaders?

• Howisaprojectchampiongoingtobeidentified?

• Wouldthatchampionhavethebandwidth,expertise,andinfluencetoovercome obstaclestotheintervention?

• Whatistheprocessforcontinuedmonitoringandimprovement?

• Whatwillbetheprocessforevaluationofinterventioneffectiveness?

*TheoriginalCFIRmodelwordingcalledexternalsetting “outersetting.” Thelanguagewaschangedforclaritywhenweadaptedtheframework.

**TheoriginalCFIRmodelcalledorganizationalcharacteristics “innersetting.” Thelanguagewaschangedforclaritywhenweadapted theframework.

CFIR,ConsolidatedFrameworkforImplementationResearch; ED,emergencydepartment.

setting,outersetting)thatcanassistsystematicassessmentof opportunitiesandbarrierstosuccessfulimplementation. Manyofthesearewellsuitedtobeconsideredevenearlierin theimplementationprocess,asaninitialassessmentofvalue andappropriateness.Theseareposedaspriorityquestionsin Table1.Wefurtherexplorethisproposedframeworkby discussingitsapplicationtoseveralestablishedand experimental,ED-basedinterventions.Theseexamplesare meanttoberepresentativeofbenefitsandchallengesthat mayaccompanytheimplementationofcertaininterventions. Theyarenotmeanttobecomprehensive.

CASEEXAMPLES

Table2 listsmany(butnotall)proposedpublichealth interventionsintheEDaccordingtolevelofacceptanceand penetrance.Someinterventionshavebecomesoengrainedin theEDworkflowthattheynolongerareperceivedas “public health” interventions.Tetanusvaccines,aswellasscreening forsexuallytransmittedinfections,fallunderthiscategory. Below,weexploretheproposedframeworkusingindividual interventionsascasestudiesasaguidefromwhichtoexplore theproposedquestions.Eachexamplewasselectedto

highlightmajorconsiderationsrequiredtodeployand maximizepublichealthbenefits,andeachvariesintheextent towhichtheinterventionisacceptedandimplementedin EDsthroughoutthecountry.Weconsidertheoveralldisease prevalenceandimpactoftheinterventionsasitrelatesto futureEDutilization.Weexplorewhethertheinterventionis typicallyintegratedwith,runsparallelto,orisseparatefrom theworkflowofanEDvisit.Similarly,weexaminetheability andappropriatenessofperformingtheinterventionsby consideringboth financialcostsandrequisiteresources.

HIVScreening

TheUSCentersforDiseaseControlandPrevention endorsedED-basedscreeningforundiagnosedHIVin2001,8 buttheserecommendationshavenotrisentothelevelof officialguidelinesorqualitymetrics.Suchprogramshavethe potentialtotestlargepopulationsandmay findindividuals whodonothaveaccesstotraditionaltestingprograms.

Multiplestudieshaveexaminedhowtobest fitHIV screeningintoexistingEDworkflowordevelopparallel workflows. 9 Frequentquestionsincludewhichpatientsto test(universalvssymptomsvsrisk-basedscreening);who

Table2. Selectiveoverviewofthecurrentstateofemergencydepartment‒basedpublichealthinterventions.

Levelof acceptanceConceptSelectexamplesNotes

EstablishedAcceptedinterventionsthatare well-integratedintheEDsetting

SupportedInterventionsforwhichimplementation iscontextdependentbasedon,for example,localepidemiology,local resources,andcommunitypriorities.

• Sexuallytransmitteddiseasetesting

• Tetanusvaccination

• Bloodpressurescreening

• Smokingandtobaccoscreening

• Intimatepartnerviolencescreening

• Substanceusescreening, intervention,andreferraltotreatment

• HIVscreeningandreferralfor treatment

• HepatitisAandCscreeningand referralfortreatment

• Naloxoneprovisionforsubstance useandoverdose

• BuprenorphineinitiationintheEDfor opioidusedisorder

• Communityviolenceintervention programs

• Depressionscreeningandreferral

Typicallycodifiedbycurrentfederal guidelinesorrecommendations,such asTheJointCommission,The CentersforMedicareandMedicaid Servicesrequirementsor reimbursement,USPreventive ServicesTaskforcerecommendations.

Potentiallywidelydiscussedinthe emergencymedicineliterature,these aretypicallynon-regulated interventionsthatmaybethetopicsof grantsorregionalimplementation. Nationalguidelinesmaybesupportive butnotnecessarilywithinthe EDsetting.

ExperimentalInterventionsarediscussedor implementedatasmallnumberof selectdepartments,often experimentalorotherwiseresearch oriented.

• HepatitisAvaccination

• Earlypregnancylinkagetocare

• Dementiascreening

• Naloxoneprovisionforallopioid prescriptions

• COVID-19vaccination

• Screeningforhousinginsecurityand otherhealth-relatedsocialneeds

ED,emergencydepartment; COVID-19,coronavirusdisease2019.

Potentiallygrantfunded,thesemay alsobeindividualdepartmental projectsorthesubjectsoftrials.Well establishedguidancewithinoroutside theEDisrare.

shouldinitiatescreening(counselororclinician);andthe operationalneedsofsuchprograms.9 Arecentlarge, randomizedtrialcomparinguniversalscreeningagainsttwo typesoftargetedscreeningshowedsimilareffectivenessin identifyingnewcases,butwithlowerresourceexpenditureof targetedscreeningprograms.9 Researchdemonstratingthat clinician-basedtestingresultsinlowerscreeningrates suggeststhepotentialbenefitofdedicatingadditionalstaffing andfundstosuchinitiativestomaximizeeffectiveness. Operationalchallengesmayfurthercomplicateeffortsto establishED-basedHIVscreenings,includingpoorlinkage tocare,10,11 lowwillingnesstotestamongmarginalized populations,12,13 andlackofculturalcompetency surroundingtestinginitiatives.14,15 Factorssuchaslower HIVincidence,improvedcommunityawarenessandriskmitigation,increasedtestingduringroutinemedicalcare, fewerregulatorybarrierstoHIVscreeninginotherlocations, moreeffectiveanti-retroviralmedication,anddecreased stigmaofthediseasemayalsohavechangedthebenefitof ED-basedprogramssincetheywere firstdevelopedmore than20yearsago.

IntimatePartnerViolenceScreening

Intimatepartnerviolence(IPV)refersto “physical violence,sexualviolence,stalkingandpsychological aggressionbyacurrentorformerpartner” andaffectsan estimatedoneinfourwomenandonein10men nationwide.16 ScreeningforIPVinwomenofreproductive agemayhelpamelioratephysicalandpsychologic sequelae.17,18 TheUSPreventiveServicesTaskForce (USPSTF)providesaGradeBrecommendationthat “cliniciansscreenforIPVinwomenofreproductiveageand provideorreferwomenwhoscreenpositivetoongoing supportservices.”

Screeningforphysicalinjurycouldbereadilyintegrated intoanED’sexistingassessmentofacuteinjuries.However, forcomplaintswithlessobviousconnectionstoIPV,suchas mentalhealthconditionsexacerbatedbyIPVexposure, integrationofscreeningmaybehardertodefineor standardizeintheabsenceofuniversalscreeningprotocols. Inpractice,universalscreeningisoftendeployedwhile collectingpatientinformationonamyriadofothervariables (eg,pastmedicalhistory,medicationhistory,suicide screening),andmaybeproneto “clickfatigue,” whereinthe screener,taskedwithcompilingalargeamountofdataina shortamountoftime,isunabletoperformthescreening questionswiththeintended fidelity.19 Patientsalsocanbe fatiguedbytimespentscreeningforconditionsnotrelatedto theirchiefcomplaintandmaybereluctanttodivulge sensitiveinformationinthissetting.However,focused screeningofhigh-riskpopulationsmaymisspatientsandis pronetobias.

TheexistingevidencebasecitedbytheUSPSTFincludes 30studies,includingthreerandomcontrolledtrials(RCT),

whichyieldednuancedresultshighlightingthenecessityof bothcomponentsofscreeningandrobustintervention.As Feltneretalreportintheirconclusion: “Althoughavailable screeningtoolsmayreasonablyidentifywomenexperiencing IPV,trialsofIPVscreeninginadultwomendidnotshowa reductioninIPVorimprovementinqualityoflifeover3to 18months.”20 Thishighlightsthechallengeoftranslating positivescreensintopositivehealthoutcomes.Practicing clinicianswillrecognizethatinterveningtoprotectvictimsof IPVischallengingwhenpatientspresentexplicitlywiththis complaint,letalonewhenpatientsmaybeunwillingor unabletodivulgesymptomsofabuse.Closerelationships withcommunityresourcesequippedtoassistvictimsofIPV arenecessarytoensureeffectiveness,whichrequires substantialandsustainedadministrativesupport.

CommunityViolenceInterventionPrograms

GunviolenceintheUnitedStatesremainsanintractable publichealthproblem,with2020recording19,384 homicides.21 Inresponse,hospitalshaveimplemented hospitalviolenceinterventionprograms(HVIP)inEDsand wards.22 Theseprogramsusewhatisdescribedasa “golden moment” ofopportunitywhenpatientsareinthehospitalto fosterclose,long-termcarerelationshipsbetweenculturally competentviolencepreventionprofessionalsandpatients. Thisincludesthecreationofcomprehensiveneeds assessments,deliveryofcasemanagementservices,longtermpeersupport,mentalhealthservices,andaddressing socialdeterminantsofhealthasrootcausesofviolence.

InitialstudiesofHVIPshavedemonstratedpromising resultswithdecreasedinjuryrecidivismandimproved intermediateoutcomessuchasdeliveryofmentalhealth services.23 However,toachievetheseoutcomes,significant commitmentisrequiredbyEDs,includingbuy-infrom multiplehospitaldepartments,communitypartners,and internalprogramchampions.Thecostsofhiringspecially trainedstaffaresignificant,astimeandexpertisetoperform thisinterventionisoftenoutsidethetypicalworkloadof emergencyclinicians.Manyprogramsrequireanannual budgetofgreaterthan$300,000.Thisfundinghas historicallybeenchallenging,althoughrecentdevelopments allowforreimbursementthroughtheMedicaidprogramina minorityofstates.24

HepatitisAVaccination

HepatitisAvirus(HAV)isavaccine-preventable transmissibleinfectionwiththepotentialforlong-term,fatal liverdisease.Asinglevaccinedosageisupto98%effectiveat preventingtransmission.25 Consequently,ED-basedHAV vaccinationhasthepotentialtolimitlong-termsequelaein thoseathighestriskofcontractingtheillness.Still,the processofidentifyingtheseat-riskindividualsrelieson simplescreeningquestionsthatareoftenincorporatedinto standardhistory-takinginstrumentsandpracticesinthe

emergencycontext.Storageandprovisionofvaccinescan leverageexistinghospitalpharmacyandnursingprotocols. Whileat-riskgroups,includingindividualsexperiencing homelessnessorusingintravenousdrugs,menwhohavesex withmen,andthosewhohavebeenincarcerated26 frequently receivehealthcareintheemergencysetting,therearealso outpatientclinicsandother,non-healthcareentities(eg, homelessshelters,nightclubs,jails,substanceusetreatment facilities)tailoredtoservethispopulation.Targetingthese communitysitesmayachievebetterpenetranceofthe interventionforunderservedpopulationmorequickly atlowercost,givenmodestenrollmentof ED-basedprograms.27

COVID-19VaccineAdministration

VaccinationeffortsbasedintheEDwerealsobolsteredby thepresumptiveviewthattheEDpatientpopulationmight nothavereadyaccesstovaccinationoutsidetheED,28 aswell asbyadesireonthepartofmanystaffmemberstotakepart inanationaleffortofclearimport.29 AnED-based vaccinationseemedtobeanobviousextensionofhospitalbasedvaccinationprograms.TheEDvaccinationcould leverageresourcessuchasreadyaccesstopharmacyand freezers,arelativelysmallpoolofstaffwhocouldbetrained toadministervaccines,andculturalcompetencyinoffering vaccinations.Absenttheseconsiderationswasanassessment ofresourceandvaccineavailabilityinthesettingofCOVID19-relatedstaffingshortages.Themultipledosingregimen forCOVID-19addedcomplexityandrequiredaseparate workflowwithintheEDcontextandrequiredfollow-upthat wassometimesnotpossiblewithintheEDsetting. Additionally,muchwasunknownaboutwhethertheED offeredvaccinationtoanewordifferentpopulationorwas redundanttootherhospital,state,orlocalcommunity efforts.Withvaryingdiseaseincidenceandincreasing vaccinationrates,therewaslikelyashortwindowtorealizea modestbenefitfortheintervention.

DISCUSSION

Emergencyphysiciansarecommittedtoimprovingpublic healthoutcomes,asevidencedbythe2009and2021Society forAcademicEmergencyMedicineconsensus conferences30,31 andthedevelopmentofseveralpost-EM fellowshipsinrecentyearscommittedtopublichealthand publicpolicy.32 Emergencydepartmentshaveembraced manypublichealthtaskssuchasscreening,surveillance,and interventionsoutsidethetraditionalscopeofemergencycare. Withlimitedtimeandresources,notallpublichealthprojects canbeundertaken.Tomaximizepublicbenefit,caremustbe takentoselectinterventionsthathavethelargestimpact whilemaintainingintegritytotheED’scoreclinicalmission. Whileemergencyphysicianstakeprideinthemantra, “anyone,anything,anytime,” wemustrecognizethatsome resourcesmaybebetterspentoutsidethewallsoftheED.

Thisdoesnotmeanabandoningcertainpatientpopulations, butratherbringingtheskillsofemergencyphysiciansbeyond thewallsoftheEDthroughavarietyofcreativeways,such ascollaborationswithpublichealthornonprofit organizations,leveragingemergencymedicalservices experienceandconnectionstodevelopmobileintegrated healthprograms,33 ordeployingthetacticsof “streetmedicine.”

Additionally,emergencyphysiciansshouldconsidernot justhowtheprogramdesignaffectsthatsingleconditionbut howadaptabletheinterventionisforaspecificdepartment andavailableresources.Considerahypothetical interventionthatmayhave90%sensitivityforuniversal screening,butonly70%fortargetedscreening.Depending onthedifferenceinstafftimebetweenthetwo,implementing thelowersensitivitytargetedapproachmayinfactallowthe sameEDtodeployaninterventionforanadditionalpublic healthconcernwiththemarginalresourcesneededfor universalscreening,thusmaximizingoverallbenefit.

Screeningprogramsthatcollectdatabutdonotprovide aninterventioninresponsetopositivescreensareunlikelyto beimpactful.Wepositthatthehighestvaluescreening programshaveappropriatesensitivityandspecificityfor theirtargetcondition,arecosteffective,andareactionable. Thevalueofascreeningprogramshouldbeassessedbased onthepatientpopulationmostinneedofthisscreening,the effectivenessofapossibleintervention,andtheproposed rationaleorrelativeadvantagefordoingitintheED. Additionally,buy-inforaninterventionisnecessaryfrom stakeholdersacrossmultiplelevelsoftheorganization: hospitalandEDleadership,physicians,nursing,andstaff. Failuretoobtainsupportfromleadershipallocating resourcesorstaffcarryingouttheinterventioncandamage moraleandlimitprogramefficacy.

Interventionsforpositivescreeningresults,whetherfor chronicinfectiousdiseaseorhealth-relatedsocialneeds,may needtobeprovidedoutsidetheED.Therefore,robust externalnetworksbetweentheEDandoutpatientclinicsand socialservicesarethemostimportantpartofascreeningand referralprogram.MostEDsenthusiasticallyembrace additionalresourcestocoordinatecarefortheirmost vulnerablepatients,withorwithoutformalizedscreening programs.Thus,intheplanningprocessEDsshouldensure thereissignificantbuy-infrompotentialexternalpartners,so thatanyscreeningimplementedhastangibledownstream effects.Manymaybepublicclinicsornonprofit organizationsthatmaythemselvesbeunderfundedand understaffed,necessitatingexternalfundingthatshouldbe equitablydistributedbetweenstakeholders.External partnersoftenbenefitfromachampionpointofcontactin theEDtoadvertise,monitor,andcoordinate referralpathways.

PatientopennesstoacceptinganinterventionintheEDis alsoanimportantfactorinanED-basedintervention.What

expectationsdopatientsbringintotheED?Forexample,a patientsufferinganankleinjurymaynotwanttoanswer questionsabouttheirmaritalsexualpracticesorsmoking habitswhileawaitingtheresultsofaradiograph.Such questionsmaybeperceivedasirrelevanttothestatedreason forthevisit,andthepatientmay findtheminvasive oralienating.

Identifyingaliteraturebaseforproposedinterventions thatshowsbenefittopatient-centeredoutcomes(eg, improvedbloodpressure,reducedmortality),orpopulationbasedoutcomes(e.g.,fewercommunityoverdosesor shootings),isanoptimalstandardforconsidering implementationofpublichealthinterventionintheED. Observationalstudieswithoutwell-matchedcontrolsare oftensubjecttoselectionbiasandregressiontothemean. Rigorousevaluationmethodologythatisolatestheeffectof theinterventiononmeaningfuloutcomes,suchasRCTs,is preferredtoidentifythemostimpactfulinterventions. Ideally,implementedinterventionswillcontinuemonitoring andevaluationofkeymetricstoensurelocalefficacy.When thatevidenceisabsent,wehopethatthisframeworkcan informthedecision-makingprocessanalogoustothewaywe makeclinicaldecisionsintheabsenceofrobustevidence.

Emergencydepartmentsareintimatelyfamiliarwiththe waysinwhichsocialneedsdrivehealthcareutilizationand outcomes.However,disparitiesinpopulation-basedhealth outcomesarenotdrivenprimarilybylackofquality emergencycare,butbydisparitiesinbroadersocial determinantsofhealth.Thesedisparitiesareunlikelytobe amelioratedbyaone-timeinterventionwithintheED context.Thus,emergencyphysiciansmustconsider implementingpublichealthprogramsnotasaone-time isolatedinterventionbutratherasthebeginningprocessof long-term,transformative,structuralchangeofthe healthcareandsocialservicessystemsasawhole.34

CONCLUSION

Emergencycliniciansandstaffcaredeeplyaboutthe publichealthofthecommunitiestheyserve.Tomaximize publichealthbenefit,emergencyphysiciansfacechallenging decisionsregardingwhichpublichealthinterventionshold themostpotentialforimpact,aswellasthewaytheyare deployed.Localdynamicswillinformdecision-making the balanceofbenefitsandharmsmaydifferonaccountof context-specificcircumstances.Manyproposed interventionscouldalsobeimplementedeffectivelyinsome settingsbutnotinothers.Giventhatthereisno “onesize fits all” approach,wehaveproposedaframeworkgroundedin implementationsciencetoassesspotentialinterventionsina systematicmannertomaximizepublichealthintervention withoutdetractingfromtheED’scorefunction.Itiscritical touseaguidingframeworktoproperlyevaluateefficiency, feasibility,localcontext,andcostbeforedeploymentofany ED-basedpublichealthintervention.

AddressforCorrespondence:JohnDavidGatz,MD,Universityof MarylandSchoolofMedicine,DepartmentofEmergencyMedicine, 110SouthPacaSt.,6thFloor,Suite200,Baltimore,MD21201. Email: jgatz@som.umaryland.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Dr.Gatzwassupportedbyaresearch grantfromtheNationalFoundationofEmergencyMedicineduring thistime.Therearenootherconflictsofinterestorsourcesoffunding todeclare.

Copyright:©2024Fassasetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.ZunLS,DowneyL,RosenJ.TheeffectivenessofanED-basedviolence preventionprogram. AmJEmergMed. 2006;24(1):8–13.

2.D’OnofrioG,O’ConnorPG,PantalonMV,etal.Emergencydepartmentinitiatedbuprenorphine/naloxonetreatmentforopioiddependence:a randomizedclinicaltrial. JAMA. 2015;313(16):1636–44.

3.KelenGD.Publichealthinitiativesintheemergencydepartment:notso goodforthepublichealth? AcadEmergMed. 2008;15(2):194–7.

4.MareinissDP.Theimpendingstorm:COVID-19,pandemicsandour overwhelmedemergencydepartments. AmJEmergMed. 2020;38(6):1293–4.

5.BabcockIrvinC,WyerPC,GersonLW.Preventivecareinthe emergencydepartment,PartII:Clinicalpreventiveservices–an emergencymedicineevidence-basedreview.SocietyforAcademic EmergencyMedicinePublicHealthandEducationTaskForce PreventiveServicesWorkGroup. AcadEmergMed. 2000;7(9):1042–54.

6.RhodesKV,GordonJA,LoweRA.Preventivecareintheemergency department,PartI:Clinicalpreventiveservices–Aretheyrelevantto emergencymedicine?SocietyforAcademicEmergencyMedicine PublicHealthandEducationTaskForcePreventiveServicesWork Group. AcadEmergMed. 2000;7(9):1036–41.

7.DamschroderLJ,AronDC,KeithRE,etal.Fosteringimplementationof healthservicesresearch findingsintopractice:aconsolidated frameworkforadvancingimplementationscience. ImplementSci. 2009;4:50.

8.BransonBM,HandsfieldHH,LampeMA,etal.Revised recommendationsforHIVtestingofadults,adolescents,andpregnant womeninhealth-caresettings. MMWRRecommRep. 2006;55(RR-14):1–CE4.

9.EscuderoDJ,BahamonM,PanakosP,etal.Howtobestconduct universalHIVscreeninginemergencydepartmentsisfarfromsettled. J AmCollEmergPhysiciansOpen. 2021;2(1):e12352.

10.MenonAA,Nganga-GoodC,MartisM,etal.Linkage-to-caremethods andratesinU.S.emergencydepartment-basedHIVtestingprograms:a

systematicliteraturereviewbriefreport. AcadEmergMed. 2016;23(7):835–42.

11.EhrenkranzPD,AhnCJ,MetlayJP,etal.Availabilityofrapidhuman immunodeficiencyvirustestinginacademicemergencydepartments. AcadEmergMed. 2008;15(2):144–50.

12.MerchantRC,SeageGR,MayerKH,etal.Emergencydepartment patientacceptanceofopt-in,universal,rapidHIVscreening. Public HealthRep. 2008;123Suppl3(Suppl3):27–40.

13.HaukoosJS,HopkinsE,ByynyRL,etal.Patientacceptanceofrapid HIVtestingpracticesinanurbanemergencydepartment:assessmentof the2006CDCrecommendationsforHIVscreeninginhealthcare settings. AnnEmergMed. 2008;51(3):303–9,309.e1.

14.MerchantRC,LiuT,ClarkMA,etal.FacilitatingHIV/AIDSandHIV testingliteracyforemergencydepartmentpatients:arandomized, controlled,trial. BMCEmergMed. 2018;18(1):21.

15.TorresM.RapidHIVscreeningintheemergencydepartment. Emerg MedClinNorthAm. 2010;28(2):369–80.

16.CentersforDiseaseControlandPrevention.Intimatepartnerviolence. 2021.Availableat: https://www.cdc.gov/violenceprevention/ intimatepartnerviolence/index.html.AccessedAugust30,2022.

17.CDC.Preventingintimatepartnerviolence.CentersforDiseaseControl andPrevention.UpdatedOctober11,2022. https://www.cdc.gov/ violenceprevention/intimatepartnerviolence/fastfact.html AccessedSeptember2,2022.

18.USPreventativeServicesTaskForce.FinalRecommendation Statement – IntimatePartnerViolence,Elderabuse,andAbuseof VulnerableAdults:Screening.2018.Availableat: https://www. uspreventiveservicestaskforce.org/uspstf/recommendation/ intimate-partner-violence-and-abuse-of-elderly-and-vulnerableadults-screening.AccessedSeptember2,2022.

19.CollierR.RethinkingEHRinterfacestoreduceclickfatigueand physicianburnout. CMAJ. 2018;190(33):E994–5.

20.FeltnerC,WallaceI,BerkmanN,etal.Screeningforintimatepartner violence,elderabuse,andabuseofvulnerableadults:evidencereport andsystematicreviewfortheUSPreventiveServicesTaskForce. JAMA. 2018;320(16):1688–1701.

21.GramlichJ.WhatthedatasaysaboutgundeathsintheU.S.Originally published2019.LastupdatedApril26,2023.Availableat:~https://www. pewresearch.org/fact-tank/2022/02/03/what-the-data-says-aboutgun-deaths-in-the-u-s/#:~:text=The%2019%2C384%20gun% 20murders%20that,75%25%20increase%20over%2010%20years AccessedOctober10,2022.

22.BonneS,HinkA,ViolanoP,etal.Understandingthemakeupofa growing field:aCommitteeonTraumasurveyofthenationalnetworkof hospital-basedviolenceinterventionprograms. AmJSurg. 2022;223(1):137–45.

23.StrongBL,ShipperAG,DowntonKD,etal.Theeffectsofhealthcarebasedviolenceinterventionprogramsoninjuryrecidivismandcosts:a systematicreview. JTraumaAcuteCareSurg. 2016;81(5):961–70.

24.ZavalaC,BuggsSA,FischerKR.StatesshoulduseMedicaidtosupport violenceinterventionefforts. JTraumaAcuteCareSurg. 2022;92(2):e25–7.

25.CentersforDiseaseControlandPrevention. ViralHepatitis:HepatitisA 2020.Availableat: https://www-cdc-gov.proxy-hs.researchport.umd. edu/hepatitis/hav/index.htm.AccessedSeptember20,2022.

26.KaighC,BlomeA,SchreyerKE,etal.Emergencydepartment-based hepatitisAvaccinationprograminresponsetoanoutbreak. WestJ EmergMed. 2020;21(4):906–8.

27.JamesTL,AschkenasyM,EliseoLJ,etal.ResponsetohepatitisA epidemic:emergencydepartmentcollaborationwithpublichealth commission. JEmergMed. 2009;36(4):412–6.

28.RodriguezRM,TorresJR,ChangAM,etal.TheRapidEvaluationof COVID-19VaccinationinEmergencyDepartmentsforUnderserved PatientsStudy. AnnEmergMed. 2021;78(4):502–10.

29.WaxmanMJ,MoschellaP,DuberHC,etal.EmergencydepartmentbasedCOVID-19vaccination:Wheredowestand?. AcadEmergMed. 2021;28(6):707–9.

30.BernsteinSLandD’OnofrioG.Publichealthintheemergency department:AcademicEmergencyMedicineConsensusConference executivesummary. AcadEmergMed. 2009;16(11):1037–9.

31.SchoenfeldEM,LinMP,Samuels-KalowME.Executivesummaryofthe 2021SAEMConsensusConference:Frombedsidetopolicy:advancing socialemergencymedicineandpopulationhealththroughresearch, collaboration,andeducation. AcadEmergMed. 2022;29(3):354–63.

32.SocietyforAcademicEmergencyMedicine.FellowshipsDirectory. Availableat: https://member.saem.org/SAEMIMIS/SAEM_Directories/ Fellowship_Directory/SAEM_Directories/P/FellowshipMap.aspx AccessedMay7,2022.

33.GingoldDB,LiangY,StryckmanB,etal.Theeffectofamobile integratedhealthprogramonhealthcarecostandutilization. Health ServRes. 2021;56(6):1146–55.

34.Samuels-KalowME,BoggsKM,CashRE,etal.Screeningforhealthrelatedsocialneedsofemergencydepartmentpatients. AnnEmerg Med. 2021;77(1):62–8.

ORIGINAL RESEARCH

Trauma-informedCareTraininginTraumaandEmergency Medicine:AReviewoftheExistingCurricula

CeceliaMorra,BA*

KevinNguyen,BS*

RitaSieracki,MLS*

AshleyPavlic,MD,MA†

CourtneyBarry,PsyD,MS‡

SectionEditor:MuhammadWaseem,MD

*MedicalCollegeofWisconsin,Milwaukee,Wisconsin

† MedicalCollegeofWisconsin,DepartmentofEmergencyMedicine, Milwaukee,Wisconsin

‡ MedicalCollegeofWisconsin,DepartmentofPsychiatryandBehavioralMedicine, Milwaukee,Wisconsin

Submissionhistory:SubmittedJuly7,2023;RevisionreceivedDecember29,2023;AcceptedJanuary4,2023

ElectronicallypublishedApril8,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem

DOI: 10.5811/westjem.18394

BackgroundandObjectives: Greaterlifetimeexposuretopsychologicaltraumacorrelateswitha highernumberofhealthcomorbiditiesandnegativehealthoutcomes.However,physiciansoftenarenot specificallytrainedinhowtocareforpatientswithtrauma,especiallyinacutecaresettings.Ourobjective wastoidentifyimplementedtrauma-informedcare(TIC)trainingprotocolsforemergencyand/ortrauma servicephysiciansthathavebothsufficientdetailthattheycanbeadaptedandoutcomedataindicating positiveimpact.

Methods: WeconductedacomprehensiveliteraturesearchinMEDLINE(Ovid),Scopus,PsycInfo,Web ofScience,CochraneLibrary,Ebsco’sAcademicSearchPremier,andMedEdPORTAL.Inclusion criteriawereEMandtraumaserviceclinicians(medicaldoctors,physicianassistantsandnurse practitioners,residents),adultand/orpediatricpatients,andtrainingevaluation.Evaluationwasbased ontheKirkpatrickModel.

Results: Wescreened2,280uniquearticlesandidentifiedtwodifferenttrainingprotocols.Results demonstratedthetrainingincludedpatient-centeredcommunicationandinterprofessionalcollaboration. Onecurriculumdemonstratedthattargetedoutcomeswereduetothetraining(Level4).Bothcurricula receivedoverallpositivereactions(Level1)andillustratedbehavioralchange(Level3).Neitherwere foundtospecificallyillustratelearningduetothetraining(Level2).

Conclusion: Study findingsfromourreviewshowapaucityofpublishedTICtrainingprotocolsthat demonstratepositiveimpactandaredescribedsufficientlytobeadoptedbroadly.Currenttraining protocolsdemonstratedanincreasingcomfortlevelwiththeTICapproach,integrationintocurrent practices,andreferralstotraumainterventionspecialists.[WestJEmergMed.2024;25(3)423–430.]

INTRODUCTION

Greaterpsychologicaltraumaexposurewithinone’ s lifetimecorrelateswithanincreasednumberofhealth comorbiditiesandnegativehealthoutcomes.1 Childhood exposurestotraumaarelinkedtoincreasedhealthrisksin adulthoodforsubstanceusedisorder,depression,obesity, heartdisease,cancer,andmore.Experiencingtraumais oftenthoughtofasarareoccurrence,butthefoundational adversechildhoodexperiences(ACE)studyhasshownhow

commonandpervasivetraumaticeventsarewithintheUS. Thestudyinvestigateddifferentcategoriesofchildhood traumathatincludedphysical/sexual/emotionalabuse, parentalincarceration,andparentaldruguse.Morethan halfoftheparticipantsreportedatleastoneACE,and25% reportedmorethantwocategoriesofACEs.From 2011–2015,thestateofWisconsinrantheBehavioralRisk FactorSurvey,whichfoundthat57%ofthe25,518adult participantsreportedoneormoreACEs.2

ManystudiesrecommendscreeningforACEsinthe emergencydepartment(ED),butthishasnotbecome commonpractice.3 TheACEquestionnaireremainsthemost commontoolusedforsuchscreening1;however,morerecent researchhassuggestedthattrauma-informedcare(TIC) shouldbeappliedinallpatientinteractionsbecausepatients withahistoryoftraumainfrequentlyclassifythemselvesas such.4 Practicingwiththeassumptionthateachpatienthas experiencedsomeformoftraumaallowsthehealthcareteam toavoidre-traumatization,orthere-experiencingofaprior traumawhenexposedtoanewtraumaticevent,andto delivercompassionateandpatient-centeredcare,whichisa criticalpieceinTIC.5–7

Becauseapatient’ s firstcontactwiththehealthcaresystem isoftenintheacutecaresetting,itiscrucialthatthese cliniciansareequippedwiththeappropriateresourcesand knowledgetoprovideTIC.5,8,9 Thisencouragesthemtousea moremindfulapproachtoassessingpatients.Studieshave indicatedthat11–61%ofEDpatientspresentwithatrauma historyand20%ofpatientsatadmissionreportsuffering fromacuteemotionaldistress.10,11 Intheseacutecare settings,therearemultiplescenariosinwhichretraumatizationcanoccur.Forexample,althoughitisnot surprisingto findthatrestraintuseonapatientcanbe harmful,studiessuggestthatevenaroutinephysicalexam withoutverbalcuescanunintentionallyre-traumatizea patient.12,13 Sucheventscancausepatientstowithdrawfrom thehealthcareinteractionanddecision-making,whichleads toaportrayalofpatientnon-adherence.Furthermore, patientswithtraumahistoryarelesslikelytoseekouta primarycarephysician,insteadrelyingontheEDfor treatment.14,15 Thus,ifTICisnotpracticedinthesesettings, long-termhealthoutcomesareimpairedandmorbidityis increased.Itisessentialthatmedicalstaffbetrainedin trauma-informedpracticestoprovidehigh-qualitycareand promotehealing.

TheTICpyramidoutlines fiveoverarchingprinciples: 1)patient-centeredcommunicationandcare; 2)understandingthehealtheffectsoftrauma; 3)interprofessionalcollaboration;4)understandingyour ownhistoryandreactions;and5)screening,including universaltraumaprecautionsandtrauma-specificstrategies (Figure1).16 The firsttwoprinciplesareuniversal precautionsthatfostertrustandrapportandcanbeused withoutestablishingapatient’straumahistory.The remainingthreeprinciplesarespecificforwhentheclinician knowsthepatienthasexperiencedtrauma.

ThepositiveimpactofTICinbothacutecareandprimary caresettingshasbeenwelldocumentedinthe literature.9,17–23 TheuseofTIChasbeenassociatedwith improvedchildhoodandfamilyadjustmentduringperiodsof increasedadversity,enhancedhealthoutcomes,increased satisfactionwithcare,andbettermentalhealthoutcomes,

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Manypatientswhoaccessemergencycare haveahistoryofpsychologicaltrauma.Best practicesrecommendatrauma-informedcare (TIC)approach.

Whatwastheresearchquestion?

WhatTICtrainingprotocolshaveshown apositiveimpactforemergencyand traumaclinicians?

Whatwasthemajorquantitative findingofthe study?Majorcomparisonwithp-valueand confidenceinterval.

OnlytwoTICcurriculaintheliteratureshow positiveimpactandarereproducible.

Howdoesthisimprovepopulationhealth?

Thelimitedexistingcurriculashowthat targetedTICtrainingincreasesclinicianuse ofTICpracticesandimprovespatient outcomesandsatisfaction.

withdecreasedsubstanceabuseratesandreducedposttraumaticstressdisordersymptoms.22

PracticingTICcanalsodecreasethepsychologicaland emotionalburdenonthehealthcareteam.10,23 Frequent occupationalexposuretothetraumaexperiencesofothersis consideredsecondarytraumaandisthoughttohavea cumulativeeffectonclinicianwell-being,resultingingreater distressovertime.24 Theimpactoftheseexperienceshasbeen describedasclinicianburnout,compassionfatigue,and

Figure1. Trauma-informedcarepyramidadaptedfrom Rajaetal(2015). Universaltraumaprecautions*;trauma-specificcare.**16

secondarytraumaticstress.Clinicianburnouthasbeen showntoleadtopoorsleep,distraction,anddefensiveness, amongotherphysicalandpsychologicalramifications.23 Thesereactionsimpairaclinician’sabilitytodelivercare, increasingthelikelihoodofmedicalerrorsandmaking patientsfeellesssafe.23 PracticingTICallowsthehealthcare teamtonotonlyidentifyandattendtoapatient’sprior trauma,mitigatenewtraumaduetocurrentmedicalcare, andbetterunderstandthereactionsandbehaviorsofpatients andtheirfamilies,butalsoencouragesthemedicalteamto recognizetheirownhistorywithtrauma.

OuraiminthisscopingreviewwastoidentifyTICtraining protocolsdesignedforUSEDsandtraumaservicesthathave demonstratedpositiveimpactinordertodevelopanew trainingprotocoltobeusedinthesesettings.Identifying implementedtrainingwillassistclinicians,healthcare practices,andteachingprogramsinterestedinimproving knowledgeandclinicalpracticetoaddresstrauma. Reviewingexistingtrainingprotocolswillfacilitate adaptationanddevelopmentoffuturetrainingtofurther improvepatientcare.

METHODS

LiteratureSearch

WeusedthePRISMAExtensionforScopingReviews (PRISMA-ScR)checklistasareportingguideforthis review.25 Furthermore,wemodeledthispaperfroma previouspublishedpaperonascopingreviewonTICwithin theprimarycaresetting.21 Acomprehensiveliteraturesearch wasdevelopedbyamedicallibrarianandpeerreviewedusing thePRESSguideline.26 Searcheswereconductedin MEDLINE(Ovid),Scopus,PsycInfo,WebofScience, CochraneLibrary,Ebsco’sAcademicSearchPremier,and MedEdPORTAL,andthesearcheswereconductedtwice. SearcheswerelimitedtoEnglishlanguagearticles.Therewas norestrictiononyearorstatusofpublication;weincluded articlesthroughNovember24,2021,inthesearch.Search strategieswerecreatedusingmedicalsubjectheadings (MeSH)andkeywordscombinedwithdatabase-specific advancedsearchtechniques.MeSHtermsandkeywords wereidentifiedtorepresenttrauma-informedapproach trainingforemergencyandtraumacareclinicians.Thefull searchstrategyfromOvidMedlineisfurtherdetailedin Table1.Wedownloadedatotalof6,786resultsfromthe literaturesearchesintoEndNote,andduplicatearticleswere removed;2,280uniquepublicationswereuploaded intoRayyan(RayyanSystemsInc,Boston,MA; https://www.rayyan.ai/)forscreening(Figure2).

StudySelection

Alltheresultswerescreenedbythreeindependent reviewerstodetermineeligibilityforthisreview.The first phaseofscreeningwasablindedtitle/abstractreview conductedinRayyan,andpotentiallyrelevantarticleswere

movedtothesecondphaseofscreeningforthefulltextofthe publications.Conflictswereresolvedwithgroupdiscussion andconsensus.Finalanalysisincludedidentificationof specifictrainingprotocolsfromeachofthearticles.

EvaluationCriteria

AlthoughmanypapersreferenceTICtraining,we specificallysoughttrainingprotocolsthatweredescribedto thelevelthattheycouldbeduplicatedandthathadbeen evaluatedwithaminimaldegreeofrigor.Studieswere selectediftheymetthefollowingcriteria:thepopulation includedemergencyortraumaserviceclinicians(medical doctors,nurses,residents,nursepractitionersandphysician assistants);studydesigninvolvedTICtrainingforemergency ortraumacliniciansandincludedevaluationofthetraining, andthesettingwasaUSEDortraumahospital environment.OnlyarticleswritteninEnglishwereincluded. Weevaluatedtrainingprotocolsbasedonwho participated,modeandlengthoftraining,evaluation methods,results,andKirkpatricklevels(Table2).The KirkpatrickModel,developedin1959,remainsthemost commonmethodforevaluatingtheimpactoftraining programsandisprimarilyusedtoassessmedicaltraining.As awell-establishedtoolforevaluation,theKirkpatrickModel iswidelyconsideredtobeavalidandreliabletoolthatcanbe implementedwitheasetomeasuretheeffectivenessof trainingonaparticulartargetgoal.Themodelusesfour levelsoftrainingevaluation:Level1:Reaction how favorable,engaging,andrelevanttrainingistothe participants’ jobs;Level2:Learning didparticipants acquiretheintendedknowledge,skills,attitude,confidence, andcommitmentthroughparticipation;Level3:Behavior willparticipantsapplywhattheylearnedinpractice;and Level4:Results aretargetedoutcomes(changesinclinician behaviorsandimprovedpatientoutcomes)duetotraining.27

RESULTS

Afterreviewing2,280uniquearticles,weincluded16 articlesforfull-textreview.Ofthe16articles,onlytwowere includedinthe finalanalysis.Fourteenarticleswereexcluded fortargetingtheincorrectpopulation,23,28 havingatraining centerlocationoutsidetheUS,29,30 notdescribingtheTIC trainingcurriculum,31–39 orlackingevaluationofthe curriculum.40 Theincludedarticleshighlightdifferent trainingprotocols,oneaddressingthetreatmentofagitation andoneencouragingclinicianreferrals.6,7 Botharticlescover subjectmatterrelatedtopatient-centeredcommunication, usein-personlearningmethodologies,includingdidactic sessionsorroleplays,andaddressinterprofessional collaborationaspartofthetraining.6,7 Thesetwoarticles detailthedevelopmentofTICcurriculaforemergencyand traumacliniciansfromtheirdesigntotheirimpact,providing comprehensiveinsightthatwillbeabletoinformthe developmentoffuturetrainingprotocols.

Table1. OvidMEDLINEsearchstrategy(throughNovember24,2021).

1(traumainformedor(acesoradversechild*event*oradversechild*experience*)).mp.

2trauma.ti.ortrauma.ab.ortraumatiz*.mp.ortraumatis*.mp.orretraumatis*.mp.orretraumatiz*.mp.

3expstress,psychological/orpsychologicalstress*.mp.orstressfulevent*.mp.orstressfulexperience*.mp.orexplifechange events/orlifechang*event*.mp.

4expRESILIENCE,PSYCHOLOGICAL/orresilien*.mp.orcoping.mp.orcope.mp.orcoped.mp.

5expAdaptation,Psychological/or(psychological*adj5adapt*).mp.or(emotional*adj5adjust*).mp.orexpemotionaladjustment/ 6expStressDisorders,Post-Traumatic/orposttraumaticstressdisorder*.mp.orposttraumaticstressdisorder*.mp.orptsd.mp.or posttraumaticneuros*.mp.orposttraumaticneuros*.mp.or(moral*adj5injur*).mp. 7expsocialsupport/orsocialsupport*.mp.orsocialnetwork*.mp.

8expselfcare/orselfcare.mp.

9wellbeing.mp.orexp “QualityofLife”/orqol.mp.orqualityoflife.mp.orlifequality.mp.

102or3or4or5or6or7or8or9

11patientcentered*.mp.orexpPatient-CenteredCare/orpatientfocused*.mp.ormedicalhome*.mp.orclientcentered*.mp.

12exp “DeliveryofHealthCare,Integrated”/or(behavioraladj5healthadj5integrat*).mp.or(behaviouraladj5healthadj5integrat*). mp.or(integratedadj5care).mp.

1311or12

1410and13

151or14

16expeducation/orexpcurriculum/orexpeducation,professional/orexpeducation,medical/orcurricul*.mp.ored.fs. 17(educat*ortrain*ororientat*orlectur*orteach*orworkshop*orpre-postorimplement*orassessment*).mp. 18expsimulation/orsimulat*.mp.orscreen*.mp.

19expTEACHING/orexpTEACHINGMATERIALS/orexplectures/

20expEducation,Medical,Continuing/orcontinuingmedicaleducat*.mp.orcme.mp.

21expHealthPersonnel/edorinterprofessionaleducat*.mp.

22expprogramdevelopment/or(program*adj5develop*).mp.

23expqualityimprovement/or(qualityadj5improv*).mp.

24expEvaluationStudiesasTopic/or(researchadj5evaluat*).mp.or(program*adj5evaluat*).mp.

2516or17or18or19or20or21or22or23or24

26AdvancedTraumaLifeSupportCare/orexpemergencymedicine/oremergencynursing/orexpEmergencyService,Hospital/

27(emergencyoremergicenter*oremergencycenter*ortraumaservice*ortraumaunit*ortraumacenter*oradvance*trauma*).mp.

28(emergencydepartment*oremergencyhospitalservice*oremergencyoutpatientunit*oremergencyroom*oremergencyunit*or emergencyward*orhospitalemergencyservice*oremergencyservice*oremergencynurs*oremergencyphysician*or emergencymedicine).mp.

2926or27or28

3015and25and29

31limit30toEnglishlanguage

FivePointsofTrauma-InformedCare

OneoftheTICtrainingprotocols,entitledtheFivePoints ofTIC,wasimplementedforLevelItraumacenter clinicians.7 CliniciansandstafffromthedepartmentsofEM, pediatrics,surgery,andsocialwork,aswellasmedical studentsandnurses,amongothers,participatedintraining thatconsistedofa90-minuteworkshop,facilitatedbya pediatricianandformerpatients.Thismodeloutlined five pillarstoguidecliniciansandaidfamiliesaffectedbytrauma

orviolentinjury:safety;screening;understandingcontext; avoidingre-traumatization;anddischargeplanning. Additionally,thetrainingfocusedonpromotingapatient’ s senseofsafety,whichcanhelpimprovetheirhealingand establishtrustbetweenclinicianandpatient.7 Thisincludes factorssuchasprivacy,aconsistentanddependable clinician,andasoothingenvironment.

Withinthisworkshop,participantsdiscussedcorrelating clinicalcases,complextrauma,andahospital-basedviolence

interventionprogram(VIP).TheVIPincludedtrauma interventionspecialistswhocouldprovidecrisisintervention, support,andpsychoeducationontrauma.Next,participants reviewedtheFivePointsofTICanddiscussedpatientcases. Followingthecases,theyheldapatientpanelandVIPpanel, wherepatientswereabletosharetheirexperienceswith trauma-sensitivecommunicationskillsandhealing.

TheKirkpatricklevelshighlightedforthisprotocolinclude Levels1and3.Participantscompletedpre-andpostworkshopsurveysassessingcomfortwiththeFivePointsof TIC.Resultsdemonstratedanincreaseincomfortlevelswith TIC(P < .001)forattendings,residents,fellows,andmedical students,withmedicalstudentshavingthehighestincreasein comfortlevels(Level1).Additionally,behavioralchangewas directlyassessed,withVIPreferralsfromphysicians significantlyincreasingfrom7.3%in2014to47.8%ofpatients referredin2018followingthecourse(P < .001)(Level3). Theseresultsdemonstratethatasaresultoftraining,therecan beanimprovementinTICcomfortandfamiliaritywithTIC approaches,leadingtosubstantivechangeinpractice.

BETAProject

AnotherTICtrainingprotocolwascompletedbynurses, andlaterallstaff,intheED.6 Participantscompleted

ManagementoftheAgitatedPatientintheEmergency Departmenttraining,partoftheBestPracticesinEvaluation andTreatmentofAgitation(BETA)project,whichfocuses onevidence-basedguidelinesandnon-pharmacological interventionstominimizeuseofrestraintsandseclusion whencaringforagitatedpatients.De-escalationtechniques, environmentalmodifications,andsensoryapproachesare thefoundationofthisapproach.

Thefour-hourtrainingconsistedofdidacticsimulations androleplay.Beyondstaffeducation,theprotocolalso calledforthedevelopmentofnewclinicalprocessesand ongoingmonitoringandfeedback.BasedontheKirkpatrick Modeloftrainingevaluation,learningassociatedwiththis protocolincludedLevels1,3,and4.Followingcompletion ofthetraining,resultsindicatedthatthenursesfoundit valuableandabletobeeasilyintegratedintotheirpractice (Level1).Participantsreportedimprovedconfidenceand satisfactionwithmanagingaggressivepatients(Level1). Therewasalsoasignificantreductioninrestraintuseinthe ED,demonstratingthatabehavioralchangeandimproved outcomescanoccurthroughprovidingstaffwithTIC knowledgeandtheskillstoaddressunderlyingcausesof patientbehaviors(Level3and4).

DISCUSSION

Thisreviewhighlightstheneedforcontinueddevelopment andevaluationofoutcomesofTICtrainingsforemergency andtraumaservicephysicians.Althoughonlytwocurricula wereidentifiedthatmettheinclusionandexclusioncriteria establishedforthisreview,severalstudieshighlightedthe importanceofTICtraining(Hawkins,Fisher).9,34 These studiesdonot,however,includespecificcurriculathatwere usedtotrainemergencyandtraumaservicephysicians.To promotetheliteratureonthistopicandaidinstitutions strivingtobringTICtotheirEMortraumaservices,itis importanttonotonlyidentifythetrainingcurriculaavailable foremergencyandtraumaserviceclinicians,buttoevaluate the effectiveness oftheTICtraining.

Priortodesigningandimplementingatrainingprotocol,a needsassessmentcanbeconductedtodeterminethespecific deficitswithinaninstitutionorpractice.21,41 Thisisastep thatwasnotindicatedinthecurrentincludedresultsandmay beanimportantpiecepriortocreatingacurriculum.6,7 To createthemostimpactfulcurriculum,theneedsofthe clinicians,patients,andcommunitiesmustbeunderstood. First,thisinvolvessurveyingclinicianattitudesandbeliefs aboutTIC,aswellasspecificknowledgeofwhatTIC encompassesanditsroleinbuildingtrustwithinthemedical system.8 Second,thisinvolvesaskingclinicianswhattheyfeel theymayneedinTICtrainingandtheoutcomestheyare hopingfor.

Theneedsasperceivedbyphysiciansonatraumaservice maydifferdramaticallyfromthoseasperceivedby outpatientprimarycarephysicians.21 Furthermore,aneeds

Figure2. Sourceselectionprocess.

Table2. Summaryoftwoapproachestotrauma-informedcaretrainingforemergencyandtraumaphysicians.

Kirkpatrick Modellevel oflearning ColeR (2014) Twonewnursesand fourrecentgraduate nurses(pilotgroup) atanacute-care, LevelIItrauma center.Expandedto includeallEDstaff

SourcePopulation Training methodologyEvaluationResults

Four-hourpilot trainingcourse:read 2BETAProject articlesandcomplete homeworkbefore workshop;didactic sessionsand roleplay

• Post-test

• Classevaluation bypilotgroup

McNamara M,etal (2020)

318cliniciansand hospitalstaff membersattwo LevelIpediatric traumacenters

90-minuteworkshops from2015–2018plus patientexpertadvice andapanel discussion,followed byimplementationof “FivePointsofTIC” curriculum

• Pre-/post-workshop surveys

• Trackedreferralsto theViolence Intervention Program(VIP)from 2014–2018

1.Mostparticipantsfoundthe trainingvaluableandintegrable intotheirexistingpractice.

2.Initially,15–20episodesof restraintorseclusionpermonth decreasedto0episodes.

3.Overallbehavioralhealth seclusion/restrainthoursreduced from38.5h/mo.(August2011)to 0h/mo.(September2013)with overallshorterepisodeduration andimproveddocumentation compliance.

1.Increasedreferralsto VIPfromphysicians (P < 0.001;7.3%to47.8%).

2.Decreasedprobabilityofpatients beingidentifiedonlybyVIPstaff (P < 0.001;62.1%to23.4%).

3.Self-reportedcomfortwithTIC afterworkshopsimprovedby21% (P < 0.001).

ED,emergencydepartment; BETA,BestPracticesinEvaluationandTreatmentofAgitation; h/mo,hours/months; TIC,trauma-informedcare.

assessmentwouldpromoteunderstandingofanyTIC approachesthatarealreadybeingimplemented(whetheror nottheyareexplicitlyrecognizedasTIC)withintheEDor traumaservicesetting.Finally,thisneedsassessmentwould focusonaddressingconcernsoftheuniquepatient populationsthattheclinicianscarefor.EvenacrossEDsand traumaservices,theremaybemarkeddifferencesinpatient populationsandcommunityresourcesalreadyavailable, whichmayimpactwhatisemphasizedinahospital-based TICtraining.

Whendevelopingatrainingprotocol,outcomeshave indicatedthatevenmoderatetrainingimprovestheabilityof thehealthcareteamtoprovideTIC;however,moreintensive protocolsarecorrelatedwithimprovedresults.35 Protocols suchastheFivePillarsofTICandtheBETAProject,which usein-personworkshopswithcase-baseddiscussions,roleplay,andsimulationsversusdidacticsalone,showgreater clinicianproficiencyassociatedwithimprovedpatientreportedoutcomesandphysiciancomfortlevels.6,7,35

Althoughtheliteratureregardingtheoutcomesof implementingTICtrainingforemergencyandtrauma servicephysiciansislimited,researchonthedevelopmentof suchtrainingprogramssuggeststhattrainingand simulationsshouldencourageamultidisciplinaryapproach, mirroringtherealityoftheenvironment.41 Thismethodhelps bothtoidentifysystem-levelconditionsthatmightimpact

• Reaction – Level1

• Behavior – Level3

• Results – Level4

• Reaction – Level1

• Behavior – Level3

thedeliveryofTIC,suchasorganizationalissues,andto highlightanysocialdynamicsorauthorityhierarchiesthat coulddiscourageteammembersfromvoicingconcerns.

AsnotedintheincludedarticlesbyCole(2014)and McNamaraetal(2020),thesuccessofaTICprotocolcanbe evaluatedthroughpre-andpost-trainingsurveysor evaluationstogaugetheimpactofthecourseonhealthcare cliniciansandtheirpractice.6,7,35 Metricsthatincludereferral tooutsideresources,involvementofsocialworkers,and patientsatisfactioncanbeusedtotracksuccessful implementationofTICmethodsasillustratedthroughthe BETAProject.7 Thesecanbemonitoredbymonthlyor quarterlychartauditsandpatientsurveys.6 Additionally, long-termevaluationofbehavioralchanges,knowledgeand beliefs,andcomfortwithprovidingTICshouldbetrackedto monitortheimpactofthetrainingprogram.

Finally,futureresearchshouldemphasizethewaysin whichTICcanimprovehealthcarecosts,clinician satisfactionandwell-being,andlong-termhealthoutcomes forpatientsaffectedbytraumaticexperiences,including reducedre-traumatization,decreasedhealthcareutilization, improvedmentalandphysicalhealthoutcomes,and decreasedsubstanceuse.19 Existingevidencesuggeststhat recognizingtrauma’simpactonpatientbehaviorandhealth allowsclinicianstoavoidunnecessaryinterventions, decreasereadmissions,andimprovehealthoutcomes.5

Additionally,evaluatingTICpracticestoreduceclinician burnoutcouldlimitstaffturnoverandassociatedrecruitment andtrainingcosts.10,23,24 Thisdata,alongwithmorerobust datafromemergencyandtraumaservicesthathave implementedTICprotocols,iscriticalinultimately providingthemostconsiderateandappropriatecare forpatients.

LIMITATIONS

Thisreviewislimitedinthatonlytwoarticleswerefound tomeetinclusioncriteria.Whiletherewasmoreavailable researchonTICtrainingevaluationswithinmentalhealth andprimarycaresettings,theuniquenatureofEDsand traumaserviceswarrantedstrictinclusioncriteria,which resultedinanarrowselectionofliterature.Inthesesettings, patientsarefacedwithunfamiliarphysiciansandfast-paced interactions,andthereisevidenceindicatingthatalarge proportionofpatientsinthesesettingsreportahistoryof traumaandoftenrelyonacutecareforallhealthcare needs.11,12,15,16 Anenvironmentemphasizingempathyand safetyisparamountinTIC,especiallyinthese departments.7,10 Theprimaryintentionofincluding evaluationsofTICtrainingonlyinUShealthcarefacilities wastoaccountfordifferencesinhealthcaresystemresources andinvestmentintrainingcomparedtoothercountries. ExcludedwerearticlesdiscussingthepotentialofcertainTIC trainingandpracticeswithoutevaluationofeffectiveness thatwouldinformfuturecurriculadevelopment.Withthe necessarycriteriathatwereestablishedforarobustreview, the finalresultsyieldedlimiteddatafordeterminingthemost optimalfeaturesofaTICtrainingprotocol.

CONCLUSION

Ourreviewdemonstratesaconsiderablepaucityinthe literatureregardingimplementedandevaluatedtraumainformedcarecurriculaforemergencyandtraumaservice clinicians.However,theexistingtrainingprotocols demonstratethat,withtargetedtraining,cliniciansbecome morecomfortablewithTICandcanintegrateaspectsofTIC intocurrentpractices.

AddressforCorrespondence:CeceliaMorra,BA,MedicalCollegeof Wisconsin,8701W.WatertownPlankRd.,Milwaukee,WI53226. Email: cmorra@mcw.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Morraetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons

Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.FelittiVJ,AndaRF,NordenbergD,etal.Relationshipofchildhood abuseandhouseholddysfunctiontomanyoftheleadingcausesof deathinadults.TheAdverseChildhoodExperiences(ACE)Study. AmJ PrevMed. 1998;14(4):245–58.

2.ChildAbuseandNeglectPreventionBoard.AdverseChildhood ExperiencesinWisconsin:2011–2015BehavioralRiskFactorSurvey Findings.2018.Availableat: https://preventionboard.wi.gov/ Documents/ACE-Brief_2018FINAL.pdf.AccessedFebruary23,2022.

3.BinnieV,LeBrocqueR,JessupM,etal.Adultfrequentpresentationto emergencydepartmentsandadversechildhoodexperiences:ascoping review. AustralasEmergCare. 2021;24(4):264–79.

4.GuestH.Aconceptanalysisoftrauma-informedcare. NursForum. 2021;56(4):1000–7.

5.CochranCB.Infusingtheprinciplesoftrauma-informedcareinto emergencynursing:acomprehensiveapproachtochangepractice. JForensicNurs. 2019;15(4):206–13.

6.ColeR.Reducingrestraintuseinatraumacenteremergencyroom. NursClinNorthAm. 2014;49(3):371–81.

7.McNamaraM,CaneR,HoffmanY,etal.Traininghospitalpersonnelin trauma-informedcare:assessinganinterprofessionalworkshopwith patientsasteachers. AcadPediatr. 2021;21(1):158–64.

8.HawkinsBE,CoupetEJr,Saint-HilaireS,etal.Trauma-informedacute careofpatientswithviolence-relatedinjury. JInterpersViolence. 2022;37(19–20):NP18376–93.

9.FischerKR,BakesKM,CorbinTJ,etal.Trauma-informedcarefor violentlyinjuredpatientsintheemergencydepartment. AnnEmergMed. 2019;73(2):193–202.

10.BodyR,KaideE,KendalS,etal.Notallsufferingispain: Sourcesofpatients’ sufferingintheemergencydepartmentcallfor improvementsincommunicationfrompractitioners. EmergMedJ. 2015;32(1):15–20.

11.OphuisRH,OlijBF,PolinderS,etal.Prevalenceofpost-traumatic stressdisorder,acutestressdisorderanddepressionfollowingviolence relatedinjurytreatedattheemergencydepartment:asystematicreview. BMCPsychiatry. 2018;18(1):311.

12.BonnerG,LoweT,RawcliffeD,etal.Traumaforall:apilotstudyofthe subjectiveexperienceofphysicalrestraintformentalhealthinpatients andstaffintheUK. JPsychiatrMentHealthNurs. 2002;9(4):465–73.

13.U.S.DepartmentofHealthandHumanServices,SubstanceAbuseand MentalHealthServicesAdministration.SAMHSA’sConceptofTrauma andGuidanceforaTrauma-InformedApproach.2014.Availableat: https://store.samhsa.gov/sites/default/files/d7/priv/sma14-4884.pdf AccessedFebruary28,2022.

14.HargreavesMK,MoutonCP,LiuJ,etal.Adversechildhoodexperiences andhealthcareutilizationinalow-incomepopulation. JHealthCare PoorUnderserved. 2019;30(2):749–67.

15.KoballAM,RasmussenC,Olson-DorffD,etal.Therelationship betweenadversechildhoodexperiences,healthcareutilization,costof careandmedicalcomorbidities. ChildAbuseNegl. 2019;90:120–6.

16.RajaS,HasnainM,HoerschM,etal.Trauma-informedcareinmedicine: currentknowledgeandfutureresearchdirections. FamCommunity Health. 2015;38(3):216–26.

17.BruceMM,Kassam-AdamsN,RogersM,etal.Traumaproviders’ knowledge,views,andpracticeoftrauma-informedcare. JTrauma Nurs. 2018;25(2):131–8.

18.CorbinTJ,PurtleJ,RichLJ,etal.Theprevalenceoftraumaand childhoodadversityinanurban,hospital-basedviolenceintervention program. JHealthCarePoorUnderserved. 2013;24(3):1021–30.

19.CorbinTJ,RichJA,BloomSL,etal.Developingatrauma-informed, emergencydepartment-basedinterventionforvictimsofurbanviolence. JTraumaDissociation. 2011;12(5):510–25.

20.CunradiCB,CaetanoR,AlterHJ,etal.Adversechildhoodexperiences areassociatedwithat-riskdrinking,cannabisandillicitdrugusein femalesbutnotmales:anemergencydepartmentstudy. AmJDrug AlcoholAbuse. 2020;46(6):739–48.

21.GundackerC,BarryC,LaurentE,etal.Ascopingreviewoftraumainformedcurriculaforprimarycareproviders. FamMed. 2021;53(10):843–56.

22.MeléndezGuevaraAM,LindstromJohnsonS,ElamK,etal.Culturally responsivetrauma-informedservices:amultilevelperspectivefrom practitionersservingLatinxchildrenandfamilies. CommunityMent HealthJ. 2021;57(2):325–39.

23.WeissD,Kassam-AdamsN,MurrayC,etal.Applicationofaframework toimplementtrauma-informedcarethroughoutapediatrichealthcare network. JContinEducHealthProf. 2017;37(1):55–60.

24.RobinsPM,MeltzerL,ZelikovskyN.Theexperienceofsecondary traumaticstressuponcareprovidersworkingwithinachildren’s hospital. JPediatrNurs. 2009;24(4):270–9.

25.TriccoAC,LillieE,ZarinW,etal.PRISMAextensionforscopingreviews (PRISMA-ScR):checklistandexplanation. AnnInternMed. 2018;169(7):467–73.

26.McGowanJ,SampsonM,SalzwedelDM,etal.PRESSpeerreviewof electronicsearchstrategies:2015guidelinestatement. JClinEpidemiol. 2016;75:40–6.

27.AlsalamahAandCallinanC.TheKirkpatrickModelfortraining evaluation:bibliometricanalysisafter60years(1959–2020). Industrial andCommercialTraining. 2021;54(1):36–63.

28.NandiM,PuranamS,Paccione-DyszlewskiM,etal.Makinguniversal trauma-informedhealthcareareality:apilotinitiativetotrainfuture

providers. TheBrownUniversityChildandAdolescentBehaviorLetter. 2018;34(12):1–6.

29.DearieV,CharlestonR,HallA,etal.Evaluationofemergency departmentstafftrainingtousetrauma-informedcareformentalhealth consumers. IntJMentHealthNurs. 2015;24(0):14–5.

30.HallA,McKennaB,DearieV,etalT.Educatingemergencydepartment nursesabouttraumainformedcareforpeoplepresentingwithmental healthcrisis:apilotstudy. BMCNurs. 2016;15:21.

31.AgboolaIK,CoupetEJr,WongAH. “Thecoatsthatwecantakeoffand theoneswecan’t”:theroleoftrauma-informedcareonraceandbias duringagitationintheemergencydepartment. AnnEmergMed. 2021;77(5):493–8.

32.GargA,PandaP,MalayS,etal.Ahumantraffickingeducational programandpoint-of-carereferencetoolforpediatricresidents. MedEdPORTAL. 2021;17:11179.

33.GoldstonDBandAsarnowJR.Qualityimprovementforacutetraumainformedsuicidepreventioncare:introductiontospecialissue. Evid BasedPractChildAdolescMentHealth,2021;6(3):303–6.

34.HawkinsBE,CoupetEJr,Saint-HilaireS,etal.Trauma-informedacute careofpatientswithviolence-relatedinjury. JInterpersViolence. 2022;37(19-20):NP18376–93.

35.DarnellD,ParkerL,EngstromA,etal.EvaluationofaLevelItrauma centerprovidertraininginpatient-centeredalcoholbriefinterventions usingtheBehaviorChangeCounselingIndexratedbystandardized patients. TraumaSurgAcuteCareOpen. 2019;4(1):e000370.

36.Traumacaresystemsdevelopment,evaluation,andfunding.American CollegeofEmergencyPhysicians. AnnEmergMed. 1999;34(2):308.

37.AlpertEJ.Ajustoutcome,or ‘just’ anoutcome?Towardstraumainformedandsurvivor-focusedemergencyresponsestosexualassault. EmergMedJ. 2018;35(12):753–4.

38.IrvinN.Commentary:Usingatrauma-informedcareframeworkto addresstheupstreamanddownstreamcorrelatesofyouthviolence. AnnEmergMed. 2019;74(5S):S55–8.

39.DennisBM,NolanTL,BrownCE,etal.Usingachecklisttoimprove familycommunicationintraumacare. AmSurg. 2016;82(1):59–64.

40.AuerbachM,ButlerL,MyersSR,etal.Implementingfamilypresence duringpediatricresuscitationsintheemergencydepartment:familycenteredcareandtrauma-informedcarebestpractices. JEmergNurs. 2021;47(5):689–92.

41.ArmeniaS,ThangamathesvaranL,CaineAD,etal.Theroleofhighfidelityteam-basedsimulationinacutecaresettings:asystematic review. SurgJ(NY). 2018;4(3):e136–51.

RelationshipofBeta-HumanChorionicGonadotropintoEctopic PregnancyDetectionandSize

DuaneM.Eisaman* NicoleE.Brown†

SarahGeyer‡

SectionEditor:PatrickMeloy,MD

*UniversityofPittsburgh † ChathamUniversity ‡ UniversityofPittsburghMedicalCenter

Submissionhistory:SubmittedJuly18,2023;RevisionreceivedJanuary10,2024;AcceptedJanuary19,2024

ElectronicallypublishedMay3,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.18396

Introduction: Ectopicpregnanciesareasignificantcauseofmorbidityandmortalityinthe firsttrimester ofpregnancy.Hospitalprotocolsrequiringaspecificbeta-humanchorionicgonadotropin(β-hCG)levelto qualifyfordiagnostictesting(pelvicultrasound)candelaydiagnosisandtreatment.Inthisstudywe soughttodeterminetherelationshipbetween β-hCGlevelandthesizeofectopicpregnancywith associatedoutcomes.

Methods: Weperformedaretrospectivecasereviewofpatientsdiagnosedwithectopicpregnancyinan urban,academicemergencydepartmentspecializinginobstetricalcare,fromJanuary1, 2015–December31,2017.Variablesextractedincludedpresentation,treatment,adverseoutcomes, andratesofrupture.

Results: Weidentified519uniqueectopicpregnancies.Ofthoseectopicpregnancies,22.9%presented withevidenceofruptureonultrasound,and14.4%showedevidenceofhemodynamicinstability(pulse >100beatsperminute;systolicbloodpressure <90millimetersofmercury;orevidenceofsignificant bloodloss)onpresentation.Medicalmanagementoutcomeswereasfollows:of177patientswho receivedsingle-dosemethotrexate,14.7%failedmedicalmanagementandrequiredsurgical intervention;of46whoreceivedmulti-dosemethotrexate,36.9%failedmedicalmanagementand requiredsurgicalintervention.Ultimately,55.7%ofpatientsrequiredoperativemanagementoftheir ectopicpregnancy.Mean β-hCGlevelatinitialpresentationwas7,096milli-internationalunitsper milliliter(mIU/mL)(SD88,872mIU/mL)withamedianof1,289mIU/mL;50.4%ofectopicpregnancies presentedwith β-hCGlevelslessthanthestandarddiscriminatoryzoneof1,500mIU/mL.Additionally, 44%ofthepatientswhopresentedwithevidenceofrupturehad β-hCGlevelslessthan1,500mIU/mL. Comparisonofsizeofectopicpregnancy(basedonmaximumdimensioninmillimeters)to β-hCGlevels revealedaveryweakcorrelation(r = 0.144, P < .001),anddetectionofectopicpregnanciesby ultrasoundwasindependentof β-hCGlevels.

Conclusion: Levelsof β-hCGdonotcorrelatewiththepresenceorsizeofanectopicpregnancy, indicatingneedfordiagnosticimagingregardlessof β-hCGlevelinpatientswithclinicalsuspicionfor ectopicpregnancy.Almostone-sixthofpatientspresentedwithevidenceofhemodynamicinstability, andapproximatelyonequarterofpatientspresentedwithevidenceofrupturerequiringemergent operativemanagement.Ultimately,morethanhalfofpatientsrequiredanoperativeprocedureto definitivelymanagetheirectopicpregnancy.[WestJEmergMed.2024;25(3)431–435.]

INTRODUCTION

Ectopicpregnanciesarenotarareoccurrence,affecting approximately2%ofallpregnancies.1–3 Thereisno centralizedsystemtomonitorratesofectopicpregnancy; thus,thetrueincidenceislikelyhigherthanthisestimate.2 Ectopicpregnancyisaleadingcauseofmorbidityand mortalityinthe firsttrimesterofpregnancy.Medical managementrequiresmorefromapatientintermsoffollowup(multipleblooddraws,ultrasounds,andappointments) whencomparedtosurgicalmanagement;between12–24%of patientswillfailmedicalmanagementandultimatelyrequire surgicalmanagement.4

Theclassicpresentationofanectopicpregnancyis unilateralpelvicpainwithvaginalbleedinginthepresenceof apositivepregnancytest.Riskfactors(presentin50%of thosewithectopicpregnancy)includepriorectopic pregnancy;historyofpelvicinflammatorydiseaseorpelvic surgery;assistedreproductivetechnologyforconception;age >35years;tobaccouse;intrauterinediethylstilbestrol exposure;andpresenceofanintrauterinedeviceatthetime ofconception.1,2 About96%ofectopicpregnancieswill occurwithintheadnexa;rarerlocationsincludethecervix, cesareansectionscars,ovaries,andabdominalcavity.1,3,5

Diagnosticworkupforsuspectedectopicpregnancy includesbloodworktocheckforbeta-humanchorionic gonadotropin(β-hCG)levels,Rhtype,hemoglobinlevel,and transvaginalultrasoundtovisualizelocationofthepregnancy (intra-vsextrauterine).Thepublisheddiscriminatoryzonefor β-hCGlevels(1,500–4,000milli-internationalunitsper milliliter[mIU/mL])canbeusedtoaidforcorrelationto expectedultrasound findings.Patientswitha β-hCG >3,500 mIU/mLshouldhave findingsonultrasoundthatdemonstrate thelocationofthepregnancy(intrauterine[gestationalsac plusyolksacwithintheendometrialcavity]vsectopic[lackof intrauterinepregnancywithanextrauterinemasswith sonographiccharacteristicsconsistentwithectopic pregnancy]).1 Whilethediscriminatoryzoneishelpfulto determinewhenanintrauterinegestationshouldbeseenon ultrasound,signsofanectopicpregnancymaybevisibleat significantlylower β-hCGlevels.6 Inthisstudywesoughtto determinewhether β-hCGlevelscorrelatewiththesizeofan ectopicpregnancyaswellastherateoftreatmentfailureof ectopicpregnancy.

METHODS

Weperformedaretrospectivecasereviewofpatientsseen inanurban,academicEDhousedinatertiary-carefacility specializinginobstetricalcare.Casesoccurredbetween January1,2015–December31,2017,forpatientswhowere diagnosedwithanectopicpregnancy.Thisstudywas approvedbytheinstitutionalreviewboard;participant consentwasnotrequired.Thecriterionforinclusionwasa diagnosedectopicpregnancyinthechartand/oron ultrasound,identifiedbysearchingbillingcodesfor “Ectopic

PopulationHealthResearchCapsule

Whatdowealreadyknowaboutthisissue?

Beta-humanchorionicgonadotropin ( β -hCG)levelsandthediscriminatory zonecutoffsareusedtodetermine ultrasound-orderingalgorithmsat manyhospitals.

Whatwastheresearchquestion?

Do β -hCGlevelscorrelatewithsize ofanectopicpregnancyand/orrateof treatmentfailure?

Whatwasthemajor findingofthestudy?

50.4%ofectopicpregnancies,and44%who hadrupture,had β -hCGlevelslessthan1,500 mIU/mL.Theectopicpregnancy β -hCGlevels onlyveryweaklycorrelated(r = 0.144, P < .001)withectopicsize.

Howdoesthisimprovepopulationhealth?

β -hCGlevelsshouldnotbeafactorinordering transvaginalultrasoundinapatientwith suspectedectopicpregnancy.

Pregnancy,”“EctopicPregnancy,Other,”“Abdominal Pregnancy,”“TubalPregnancy,”“EctopicPregnancy, Nonspecific, ” and “OvarianPregnancy.” Thissearchyielded 1,265visitsduringtheresearchperiodwith519uniquecases ofectopicpregnancy(whereeachcasecouldhavemultiple associatedvisits).Patientsofallageswereincludedinthe studyiftheymettheinclusioncriterion.Exclusioncriteria waslossofpatienttofollow-upafterinitialpresentation(thus makingitimpossibletoconfirm finaldiagnosisofectopic pregnancyand/oroutcomesrelatedtotreatment)orifthe patientwasultimatelyfoundtohaveadiagnosisotherthan ectopicpregnancy(ie,intrauterinepregnancy).Forthe519 casesofectopicpregnancy,dataextractedfromeachchart includedthefollowingvariables:presentation(including initial β-hCGlevelsandultrasound findings[sizeofectopic pregnancy,evidenceofrupture,and/orfetalheartbeat]); treatment(expectant,medical,and/orsurgical);and treatmentoutcomes(successful,failure,orrupture).Data wasextractedbytwoindividuals;theprincipalinvestigator (xx)reviewedtheextracteddataevery fivecasestoensure consistencyofdataextraction.Wecompareddatausing standardmeans,thePearsoncorrelationcoefficient,and Student t -testingusingExcel(MicrosoftCorp,Redmond, WA).Forchartswheredatawasincomplete(forexample, missingultrasoundevidenceofectopicpregnancy),we

includedthechartintheanalyseswheredatacouldbe extractedandweexcludeddatafromtheanalyseswhereit wasmissing.

RESULTS

Thedatasetcontained519uniqueectopicpregnancies presentingtotheEDduringathree-yearperiod(Table).The averageageofthesubjectsatpresentationwas29.39years withnearequalracialdistributionbetweenself-reported BlackandWhite.Evidenceofectopicrupturewaspresentin 22.9%ofcasesonpresentation,and14.4%hadevidenceof hemodynamicinstability.Aheartbeatwasdetected9.1%of thetime,whichpercurrentAmericanCollegeof ObstetriciansandGynecologists(ACOG)recommendation

Table. Ageofpatientinyears,reportedrace,rateofectopicrupture atpresentation,unstableatpresentation(heartrate >100bpm, systolicbloodpressure <90mmHg,ordocumentationofinstability), rateofmethotrexatefailureaftertreatment,rateofliveectopicat presentation, β-hCGlevelsforvisualizedectopic, β-hCGlevelsfor rupturedvsnon-rupturedectopic,andgreatestdiameterof visualizedectopic.

VariableassessedFindings

Age(Years)29.39(SD4.42)

Race(Mayincludemorethanone)

Rupturedatpresentation22.9%

Unstableatpresentation14.4%

Methotrexatefailurerate

Liveectopicatpresentation9.1%

β-hCGlevels(mIU/mL)

Ruptured7,005(SD8,949)

Non-ruptured7,609(SD8,773)

T-test P = 0.51

β-hCGforvisualizedectopicpregnancy(mIU/mL;n = 456)

Mean4,964(SD12,217)

Median1,209

Greatestdiameterofvisualizedectopicpregnancy (mm;n = 456)

Mean24(SD14.3)

Median21

β-hCG,beta-humanchorionicgonadotropin; mlU/ML,milliinternationalunitspermilliliter; mm,millimeter.

Figure. β-hCGlevelinrelationtothegreatestdimensionofectopic pregnancy(n = 456).Pearsoncorrelationbetween β-hCGleveland greatestdimensionwasr = 0.144(P < .001),demonstratingavery weakcorrelation.

β-hCG,beta-humanchorionicgonadotropin; mm,millimeters; mIU/mL, milli-internationalunitspermilliliter.

isarelativecontraindicationfortreatmentwith methotrexate.1 Subjectsreceivingsingle-dosemethotrexate treatmenthadafailurerateof14.7%.Subjectswhoreceived multi-dosemethotrexate(generallybecausetheydidnot meetACOGrequirementsforsingle-dosemethotrexate treatment)hadafailurerateof36.9%.The β-hCGlevels forbothrupturedandnon-rupturedectopicpregnancies weresimilar.

Greatestdimensionofectopicpregnancy(diameterin millimeters[mm])wascorrelatedwith β-hCGlevel(milliinternationalunitspermilliliter[mIU/mL])foreachpatient (Figure).Weexcludedfromthisanalysispatientswhodidnot haveanextrauterinemassvisualizedonultrasoundas dimensionsofectopicsizewereunknown.Mean β-hCGfor thisgroupof456subjectswas4,964mIU/mL(SD12,217; 95%confidenceinterval[CI]3,793–6.135),andmedian β-hCGwas1,209mIU/mL;meangreatestdiameterwas 24mm(SD14.3;95%CI22.69–25.31)andmedianwas 21mm.ThePearsoncorrelationcoefficient(r-value)ofthis groupwasr = 0.144(P < .001),indicatingaveryweak correlationbetween β-hCGlevelandgreatestdimensionof ectopicpregnancy.WealsocalculatedthePearson correlationfor β-hCGcomparedtoectopicvolume(inmm3); theresultswerenotstatisticallysignificant.

DISCUSSION

Nearly25%ofpatientswithanectopicpregnancyinthis studypresentedwithevidenceofrupture,andabout15%of patientspresentedwithevidenceofhemodynamicinstability, bothscenariosrequiringemergentsurgicaltreatment.The ectopicrupturerateseeninthisstudywasgreaterthanprior reportedratesof15%,although,interestingly,lessthan10% ofrupturedectopicpregnanciesinthestudyrequiredablood transfusion(datanotshown),whichissimilartoprevious publisheddataof8.7%.7,8

Themajorityofvisualizedectopicpregnanciesinthe currentstudyhad β-hCGlevelsbelowthetraditional1,500 mIU/mLdiscriminatoryzone,withthelowest β-hCGlevelin thestudybeing9mIU/mL(whichhadsonographicevidence ofanectopicpregnancy).The β-hCGlevelsweresimilar betweenrupturedandnon-rupturedectopicpregnancies. Priorstudieshavealsodemonstratedlow β-hCGlevelsin ectopicpregnancies,withlevelsaslowas <10mIU/mL documented.9 Whileectopicpregnanciestypicallyhave lower β-hCGlevelscomparedtointrauterinepregnancies, thepresenceofverylow β-hCGlevelsindocumentedectopic pregnancycasesisstriking.9 Onestudyfoundthat41%of ectopicpregnancieshada β-hCGlevel <2,000mIU/mLat thetimeofdiagnosis,andapproximately9%(18of204cases) hada β-hCGlevelunder100mIU/mL.9 These findings combinedwithourresults(50.4%ofcaseshada β-hCG <1,500mIU/mL,and8.5%hadalevel <100mIU/mL) emphasizetheneedtoconsiderandworkupsuspected ectopicpregnanciesfullyatthetimeofpresentation, regardlessofserum β-hCGlevel,toavoidmissingthe diagnosisofectopicpregnancy.Thus,discriminatorycutoffs of β-hCGlevelsshouldnotbeadeterminingfactorwhen orderingtransvaginalultrasonographytoevaluatefor ectopicpregnancy.10

Wefoundaveryweakcorrelationbetween β-hCGandsize ofectopicpregnancy,whenlookingatgreatestdimension, andnocorrelationof β-hCGtovolumeofectopicpregnancy. These findingsareinstarkcontrasttopriorstudies,where β-hCGlevelsandectopicpregnancyvolumewerefoundtobe stronglycorrelated.9 Therefore,thisstudyindicatesthat β-hCGlevelsmayhavelesspredictivevalueforestimating thesizeorvolumeofanectopicpregnancythanpreviously thought,furtherstrengtheningtheneedfortransvaginal ultrasoundattainmentregardlessof β-hCGlevel.

Post-methotrexatefailureorruptureisnotuncommon; about20%ofpatientsinthecurrentstudyultimately requiredsurgicalmanagement(higherthanthe10.2% reportedinpreviousstudies).11 Thisincreasedrateofpostmethotrexatefailuremightberelatedtostudylocation (tertiary-carecenterforobstetricalcare),patientdelayin seekingcare,orhigherrateofcasesnecessitatingmultipledosemethotrexatetreatment(duetotreatmentatatertiarycarecenter),althoughanyconclusionregardingthese variablesisdifficult.Patientsreceivingmethotrexatefor ectopicpregnancywithcomplaintsofneworworsening abdominalpainorincreasedvaginalbleedingshouldbe evaluatedforpotentialectopicpregnancyrupture.

LIMITATIONS

Thestudytookplaceatasingle-siteacademiccenter, wherethepopulationoftheEDcouldhavebeenskeweddue tolocation(urban),referralcenterstatus(tertiarycarefor obstetricalcare),andhigheracuitylevelofcarecomparedto manyhospitals.Inaddition,patientencounterstookplace

between2015–2017usingACOGguidelinespublishedin 2008;however,currentACOGguidelines(releasedin2018) aresimilartothoseusedduringthetimeframeofpatientcare; thus,anypotentialeffectonthedatawasnegligible.1,12

CONCLUSION

Inthesettingofapositivepregnancytest,pelvicpainand/ orvaginalbleedingshouldpromptacompleteworkupfor ectopicpregnancytoincludeRhstatus,hemoglobin, β-hCG level,andtransvaginalpelvicultrasound(regardlessof β-hCGlevel).Cliniciansmustconsidertheongoingriskof ectopicpregnancyruptureaftermethotrexatetreatment. Finally,patientsmaybelosttofollow-upandhavean untreatedectopicpregnancy,whichcanleadtosignificant morbidityand/ormortality.

AddressforCorrespondence:NicoleE.Brown,MSChatham University,ChathamEastside166,WoodlandRoad,Pittsburgh, PA15232.Email: nbrown1@chatham.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Eisamanetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1.AmericanCollegeofObstetriciansandGynecologists.ACOG PracticeBulletin:TubalEctopicPregnancy. ObstetGynecol. 2018;131(3):e91–103.

2.MannLM,KreiselK,LlataE,etal.Trendsinectopicpregnancy diagnosesinUnitedStatesemergencydepartments,2006–2013. MaternChildHealthJ. 2020;24(2):213–21.

3.NaveedAK,AnjumMU,HassanA,etal.Methotrexateversusexpectant managementinectopicpregnancy:ameta-analysis. ArchGynecol Obstet. 2022;305(3):547–53.

4.AiobA,YousefH,AbuShqaraR,etal.Riskfactorsandpredictionof ectopicpregnancyrupturefollowingmethotrexatetreatment:a retrospectivecohortstudy. EurJObstetGynecolReprodBiol. 2023;285:181–5.

5.EbnerF,VargaD,SorgF,etal.Treatmentcostevaluationof extrauterinegravidity:aliteraturereviewofmedicalandsurgical treatmentcosts. ArchGynecolObstet. 2015;291(3):493–8.

6.Sobecki-RauschJ,MadrigalJM,CavensA,etal.Evaluationof adherenceratesfollowingectopicpregnancytoinformmanagement decisionsinapublichospitalsystem. JGynecolObstetHumReprod. 2021;50(9):102180.

7.CulliferRM,HuynhTQ,PacisMM,etal.Riskfactorsforblood transfusioninwomenrequiringsurgicalmanagementofectopic pregnancy. JMinInvasiveGynecol. 2020;27(7):S134.

8.MullanyK,MinneciM,MonjazebR,etal.Overviewofectopicpregnancy diagnosis,management,andinnovation. JWomen’sHealth(Lond). 2023;19:17455057231160349.

9.SaeedUandMuzharN.CorrelatingserumbetahCGlevelswith transvaginalsonographicfeaturesofectopicpregnancy. JRawalpindi MedColl. 2017;21(1):64–7.

10.HirschlerLEandSotiV.Theutilityofmonitoringbeta-human chorionicgonadotropinlevelsinanectopicpregnancy. Cureus. 2023;15(1):e34063.

11.SendyF,AlShehriE,AlAjmiA,etal.Failurerateofsingledose methotrexateinmanagementofectopicpregnancy. ObstetGynecolInt. 2015;2015:902426.

12.AmericanCollegeofObstetriciansandGynecologists.ACOG PracticeBulletin:TubalEctopicPregnancy. ObstetGynecol. 2008;131(3):613–5.

ORIGINAL RESEARCH

PrevalenceandCharacteristicsofEmergencyDepartmentVisits byPregnantPeople:AnAnalysisofaNationalEmergency DepartmentSample(2010–2020)

CarlPreiksaitis,MD*

MonicaSaxena,MD,JD*

JiaqiZhang,MS†

AndreaHenkel,MD,MS†

SectionEditor:StephenMeldon,MD

*StanfordUniversitySchoolofMedicine,DepartmentofEmergencyMedicine, Stanford,California † StanfordUniversitySchoolofMedicine,DepartmentofObstetricsandGynecology, Stanford,California

Submissionhistory:SubmittedMarch18,2023;RevisionreceivedOctober5,2023;AcceptedJanuary9,2024

ElectronicallypublishedApril2,2024

Fulltextavailablethroughopenaccessat http://escholarship.org/uc/uciem_westjem DOI: 10.5811/westjem.60461

Introduction: Thenumberandcharacteristicsofpregnantpatientspresentingtotheemergency department(ED)hasnotbeenwelldescribed.Ourobjectiveinthisstudywastodeterminethe prevalenceandcharacteristicsofpregnantpatientspresentingtoEDsintheUSbetween2010–2020.

Methods: Wecompletedaretrospective,cross-sectionalstudyofpatientencountersathospital-based EDsintheUSfrom2010–2020.UsingtheEDsubsampleoftheNationalHospitalAmbulatoryMedical CareSurvey(NHAMCS)weidentifiedEDvisitsforfemalepatientsaged15–44years.Wedefineda subsampleoftheseasvisitsforpregnantpatientsusingdischargediagnosiscodesspecificto pregnancy.Wecomparedthispopulationofpregnantpatientvisitstothosefornon-pregnantpatients andcomputedpointestimatesfornationallyweightedvalues.Multivariablelinearregressionwasusedto determinefactorsindependentlyassociatedwithpregnantpatientvisits.

Results: The2010–2020NHAMCSdatasetincluded255,963EDvisits.Ofthesevisits59,080werefor femalepatients15–44yearsold,and6,068ofthosevisitswereforpregnantpatients.Pregnantpatients accountedfor3%(95%confidenceinterval[CI]2.7–3.2)ofallEDvisitsand8.6%(95%CI8–9.3)ofall visitsamongfemalepatients15–44years.Weightingtoanationalsample,thisequatesto2.77million pregnantpatientspresentingforEDvisitsannually.PregnantpatientsweremorelikelytobeBlack, Hispanic,ortousepublicinsurance.

Conclusion: PregnantpatientsmakeupasignificantnumberofEDvisitsannuallyandaremorelikelyto bepeopleofcolororpubliclyinsured.Interventionstoaddresstheeffectsofchangingabortionlegislation onemergencymedicinepracticemaybenefitfromconsiderationthatcertainpopulationsofpregnant peoplearemorelikelytopresenttotheEDforcare.[WestJEmergMed.2024;25(3)436–443.]

INTRODUCTION

Background

InJune2022,theUSSupremeCourtruledon Dobbsv JacksonWomen ’ sHealthOrganization anddeterminedthat thereisnoconstitutionalrighttoabortion,allowing individualstatestolegislateabortionrestrictions.1 This decisionhasmultipleanticipatedimplicationsforemergency clinicians,includinganincreaseinpregnantpatients

presentingtotheemergencydepartment(ED)asaresultof barrierstocare,complicationsofself-managedabortions,or delayedpresentationofemergentdiagnosisduetofearof legalrepercussions.2

Importance

UseoftheEDishighamongpregnantpatients,with studiesshowingthatapproximately35%ofthesepatients

willvisittheEDatleastonceduringtheirpregnancy.3,4 Thesepatientsaremorelikelytobeofracialandethnic minorities,publiclyinsured,andhavebarrierstoprenatal careaccess.3,5,6 Lessisknown,however,aboutthetotal populationofpregnantpatientswhopresenttotheED.A secondaryanalysisofthe2006–2016NationalHospital AmbulatoryMedicalCareSurvey(NHAMCS)identified thattherewereapproximately900,000visitstotheEDfor earlypregnancyloss,butthetotalpopulationofpregnant patientswasnotdescribed.5

Oldercohortstudiesdisagreeonthepregnancyrate amongreproductive-capablefemalepatients,reporting valuesrangingfrom2.3–33%.7–9 Thesestudiesreportthat manypregnanciesare firstidentifiedintheED,buttherateof incidentalpregnancyinmorerecentyearsandhowoften thesepatientsareprovidedwithcounselinghasnotbeen described.10 Infact,beforetherepealof RoevWade there wasanidentifiedneedforfurtheremergencyphysician traininginpatient-centeredreproductivehealthcare.5,10 Withincreasedlegalrestrictions,theneedforemergency medicinepolicyandphysicianeducationhasneverbeen greater.2,11 Todothissuccessfully,wemusthaveabetter understandingofthepopulationthatwillbeaffectedbythese changes:pregnantpeople.Thishasnotbeenrecently reportedintheliterature,whichledustoundertake thisstudy.

GoalsofthisInvestigation

Theprimaryobjectiveofourstudywastoidentifythe prevalenceofandcharacterizepregnantpatientspresenting toUSEDsbetween2010–2020.

METHODS

StudyDesignandDataSource

Wecompletedaretrospective,cross-sectionalstudyof patientencountersathospital-basedEDsintheUSfrom 2010–2020.Thisperiodwasselectedas2010sawthepassage oftheAffordableCareAct,whichhadmanyeffectsfor expandinghealthcareandcontraceptivecoveragefor women. 12 DatawasfromthepubliclyavailableED subsampleoftheNHAMCS.13 TheNHAMCSisasurvey conductedannuallybytheNationalCenterforHealth Statistics(NCHS),apartoftheUSCentersforDisease ControlandPrevention(CDC).14

TheNHAMCSusesathree-stageprobabilitysampling designinanefforttoprovidearepresentativesampleofall EDsinthecountry.14 First,112geographicprobability samplingunits,determinedbyvariouscounties,towns,and citiesarechosen.Theseareselectedtoberepresentativeof differentgeographicalregionsandurbanandruralareas. Withinthesesamplingunits,450–500short-stayhospitals (averagelengthofstayfewerthan30days)aresampledto ensureadiversityofhospitalsizeandtype.Finally,EDsthat provideunscheduledcare24hoursaday/7daysaweekat

thesesampledhospitalsareselectedforinclusion.Each includedEDhasvisitdatarecordedoverarandomly assigned4-weekperiod.Duringthisperiod,datafrom selectedvisitsisabstractedfromthechartandenteredintoan electronicformbytrainedcensustakers.

FurtherdescriptionoftheNHAMCSsurveyisavailable ontheNCHSwebsite.14 TheNHAMCSproducesadataset thatreflectsabroadspectrumofEDvisits.Usingthesurvey’ s samplingdesign,eachdataentryisassignedaweightto accountfortherelativecontributionofthatentrytothe largersample.Asaresult,eachdatapointinthedataset representsavaryingnumberofactualvisits,dependingonits assignedweight.

StudyPopulation

Todetermineourstudypopulation,weidentifiedthetotal EDvisitsforNHAMCSbetween2010–2020forallpatients. Visitswereselectedforwomenofreproductiveage,defined bytheCDCasbeingbetweenages15–44years.15 We acknowledgethatresearchsurroundingpregnancyoften assumescisgenderidentities,whichmaynotdescribepeople whoaretransgenderornon-binary.Weattemptedtouse languagethatisasinclusive;however,thedataanalyzedin thisstudyusesgenderlabelsthatcannotbechangedwhile remainingaccuratetothesourcematerial.

DefinitionofPregnancy

Wedefinedvisitsforpregnantpatientswithinourcohort asthosevisitsthathadanInternationalClassificationof DiseasesRevisions9or10(ICD-9and-10)diagnosiscode specifictopregnancyasoneofthedischargediagnoses(eg, ectopicpregnancy;excessivevomitinginpregnancy; pregnantstate,incidental).Specificdiagnosiscodesusedfor patientidentificationarelistedin Appendix1.Thesewere initially filteredbySPSSStatisticsv27(IBMCorporation, Armonk,NY)andwerethenhand-verifiedbyonestudy author(CP).WeexcludedencounterswithoutICDdiagnosis codesorcodedonlyas “elopement” or “leftwithout beingseen.”

PregnancyIdentifiedintheEmergencyDepartment

Asthishasimportantimplicationsforemergency clinicians,asecondarygoalofouranalysiswastoidentifyan estimateoftheincidenceofnewpregnancydiagnosisinthe ED thatis,visitswherepatientswere firstidentifiedas pregnantduringtheirEDvisit.Wedefinedasubsetof pregnantpatientvisitsas “incidentalpregnancy” through EDreason-for-visit(RFV)codesandwhetherpregnancywas testedforintheED.TheRFVcodesincludethechief complaint,aswellasothersymptomsormedicalproblems relatedtotheEDvisit.16 WeexaminedRFVcodesand excludedpatientvisitswithcodesthatsuggestedaprevious diagnosisofpregnancy(eg,1790.0Problemsandother conditionsrelatedtopregnancy;2735.0Diagnosed

complicationsofpregnancyandpuerperium).Withthegoal ofobtainingaconservativeestimate,wealsoexcluded patientvisitswithRFVcodesforvaginalbleeding.Patient visitsfromthisgroupthathadapregnancytestsentintheED madeupour “incidentalpregnancy” population.Toobtain evidenceofconstructvalidity,weexaminedtheICD-9and ICD-10dischargecodesforthe “incidentalpregnancy” populationtoensuretheywereconsistentwithanew pregnancydiagnosis.

CharacteristicsofEDVisits

Availabledemographicsincludedpatientage;race/ ethnicity(non-HispanicBlack,non-HispanicWhite, Hispanic,andother);payment/insurancestatus(private insurance,publicinsurance[Medicare,Medicaidorother state-basedprogram],self-pay,andotherorunknown insurance);andresidence(private,unhoused,orother).

Visitcharacteristicsincludeddayoftheweekthevisittook place(weekendorweekday),season(Fall,Winter,Spring,or Summer),andyearofvisit.Visitcharacteristicsalsoincluded hospitaladmission,whetherapregnancytestwassent,useof ultrasound,consultation,lengthofvisit,waittimetoseea clinician,andreturnvisitwithin72hours.Pregnancytest, ultrasounduse,hospitaladmission,consultation,andreturn visitwithin72hoursweredichotomousvariables,andreturn visitreferredtowhetherthepatientwasseeninthesameED intheprior72hoursforanyreason.Wedefinedwaittimeas thetimefromarrivalto firstcliniciancontact,andlengthof visitwasdefinedasthetimefromarrivaltodischarge.We analyzedbothvaluesascontinuousvariables.

Hospital-levelcharacteristicsincludedgeographicregion (Midwest,Northeast,South,andWest)andmetropolitan statisticalarea,reflectinganurbanvsrurallocationas definedbytheUSOfficeofManagementandBudget.

DataAnalysis

WefollowedtheStrengtheningtheReportingof ObservationalStudiesinEpidemiology(STROBE) reportingguidelines.17 WedescriptivelyanalyzedEDvisits forpregnantpatientsanddeterminedtheproportionofED visitsforpregnantpatientsamongallEDvisits(forwomen andmen)aswellasamongwomen15–44yearsold. Demographic,visit,andhospitalcharacteristicsof presentationsofpregnantpatientsamongwomenof reproductiveageweresimilarlyanalyzed.Wecomparedthe characteristicsofvisitsforpregnantreproductive-aged womentothosefornon-pregnantreproductive-agedwomen usingchi-squaredtestsforcategoricalfactorsandtwosample t -testforcontinuousvariables.Statistical significancewassetat P < 0.01asrecommendedby NHAMCSdocumentation.13

Wecomparedcharacteristicsofvisitsforpregnant patientstonon-pregnantpatientsusingmultivariablelogistic regression.Weexaminedunadjustedassociationsandthen

usedmultivariablelogisticregressionmodelstodetermine factorsindependentlyassociatedwithvisitsforpregnant patients.Modelsgeneratedoddsratios(OR)and95% confidenceintervals(CI).Statisticalcalculationswere completedwithSASOnDemandforAcademics(SAS,Inc, Cary,NC).

WeightedresultsrepresentativeofallEDvisitsroundedto thenearestthousandintheUSwerepresentedforanalysis unlessotherwisestated,asrecommendedbyNHAMCS.13,18 BasedonbestpracticesfortheuseofNHAMCSdatain research,weensuredthatallreportedestimateswerebased on >30unweightedrecords,hadarelativestandarderrorof <30%,anddidnotincludeanyitemswithanon-response rate >30%inouranalysis.14,18 TheNHAMCSimputesdata formissingvaluesinage,gender,race,andethnicityusinga model-basedsingle,sequentialregressionmethod.13 Race andethnicityhadthehighestaverageproportionofmissing valuesinourdataset(17%and21%,respectively); therefore,weperformedasensitivityanalysistoensure resultdurability.

Ethics

DatafromtheNHAMCSarede-identifiedandpublicly available.Useofthisdataforresearchpurposeshasbeen reviewedandapprovedbytheNationalCenterforHealth StatisticsEthicsReviewBoard(Protocol#2021-03).

RESULTS

The2010–2020NHAMCSdatasetincluded255,963visits (weightedn = 1,502,215,000,95%CI 1,342,435,000–1,661,995,000),including59,080visitsamong womenages15–44(weightedn = 353,012,000,95%CI 310,947,000–395,078,000).Atotalof6,068ofthesevisits wereforpregnantpatients(weightedn = 30,489,000,95%CI 26,117,000–34,861,000).Pregnantpatientvisitsaccounted for3.0%(95%CI2.7–3.2)ofallEDvisits.Thisequatesto 2.77millionpregnantpatientspresentingforEDvisits annually.Limitingthepopulationtowomenages 15–44years,pregnantpatientvisitsaccountedfor8.6%(95% CI8–9.3)ofallEDvisits.

Incidentalpregnancywasidentifiedin672patientvisits (weightedn = 4,056,000,95%CI3,323,000–4,789,000). Incidentalpregnancyvisitsaccountedfor13.3%(95%CI 12.7–13.7)ofallpregnantEDvisits.Annually,thisequates to368,000(95%CI352,000–379,000)visitswherepregnancy isdiagnosedintheED.Themajority(52%)oftheICD-9and 10codesforthesevisitswereforpregnancy-related complaints(eg,hyperemesisgravidarum,infectionofthe genitaltractinpregnancy),andtheremainderwere diagnosesofpregnancy(eg,encounterforsupervisionof normal firstpregnancy,pregnantstate,incidental),aligning withourassumptionthatthesevisitsrepresentednew pregnancydiagnoses.

StudyPopulationCharacteristics

Themedianageofwomenpresentingwithpregnancywas 25years(interquartilerange21–30years),withoverhalf (57.9%)betweentheagesof20–29years(Table1).Visitsfor pregnantpatientsweremorelikelytobeforBlack(31.6%vs 26.3%fornon-pregnantwomen)andHispanic(22.0%vs 15.5%)women.Pregnantpatientvisitsweremorelikelyto usepublicinsurancethannon-pregnantvisits.

EmergencyDepartmentVisitCharacteristics

Therewerenosignificantdifferencesinpresentationof pregnantpatientsbetweenweekdaysandweekendsoracross seasons.Thenumberofpregnantpatientspresentingtothe EDdidnotsignificantlyvaryacrossyears,evenwhen normalizedtototalpatientsinourpopulation.Regional

distributionofpregnantpatientvisitsdidnotdiffer.Pregnant patientsweremorelikelytopresenttoahospitalinanurban area(89%vs84.7%, P < 0.001).Only44.8%ofpatientvisits includedapregnancytest,and44%includedanultrasound; 17.5%ofpregnantpatientvisitsincludedapelvicexam. Sevenpercentofpregnantpatientvisitsresultedin hospitalizationvs4.7%ofnon-pregnantpatientvisits (P < 0.001);and11.5%ofpregnantpatientvisitsincluded evaluationbyaconsultingphysician.

Therewasnosignificantdifferenceinwaittimebetween pregnantvsnon-pregnantpatientvisits.TheEDvisits generallylastedlongerforpregnantpatients(33.6%over fourhoursvs24.5%, P < 0.001).Most(84.4%)patientswere seenbyanattendingphysician,withoutmeaningful differencesbetweengroupsthatwereseenbyotherclinicians.

Table1. Characteristicsoffemalepatientsaged15–44yearsoldpresentingtotheemergencydepartmentforcare,2010–2020,weighted andstratifiedbypregnancystatus.

Age25(21–30)28(22–36)

15–19years4,185,000(13.7)47,282,000(14.7)

20–29years17,640,000(57.9)124,098,000(38.5)

30–39years7,802,000(25.6)103,558,000(32.1)

40–44years862,000(2.8)47,586,000(14.8)

Race/ethnicity

Non-HispanicWhite13,199,000(43.3)178,581,000(55.4)

Non-HispanicBlack9,642,000(31.6)84,808,000(26.3)

Hispanic6,695,000(22.0)49,863,000(15.5)

Non-HispanicOther952,000(3.1)9,271,000(2.9)

Paymentsource

Privateinsurance8,039,000(26.4)96,150,000(29.8)

Publicinsurance15,197,000(49.8)133,820,000(41.5)

Self-pay3,289,000(10.8)46,446,000(14.4)

Other1,164,000(3.8)13,203,000(4.1)

Unknown2,800,000(9.2)32,905,000(10.2)

Residence 0.21

Privateresidence29,464,000(96.6)309,823,000(96.1)

Homeless76,000(0.2)1,556,000(0.5)

Other226,000(0.7)3,517,000(1.1)

Unknown723,000(2.4)7,628,000(2.4)

Visitcharacteristics

EDvisitday

Weekend8,087,000(26.5)86,105,000(26.7)

Weekday22,403,000(73.5)236,419,000(73.3) (Continuedonnextpage)

Table1. Continued.

Fall7,503,000(24.6)85,173,000(26.4)

Winter7,735,000(25.4)77,058,000(23.9)

Spring7,536,000(24.7)81,156,000(25.2)

Summer7,715,000(25.3)79,136,000(24.5)

Year

20102,453,000(8.0)17,134,000(5.3)

20112,716,000(8.9)32,089,000(9.9)

20122,363,000(7.8)30,316,000(9.4)

20132,362,000(7.7)30,350,000(9.4)

20142,570,000(8.4)32,656,000(10.1)

20153,028,000(9.9)31,333,000(9.7)

20163,073,000(10.1)31,538,000(9.8)

20172,902,000(9.5)30,809,000(9.6)

20182,710,000(8.9)27,058,000(8.4)

20193,558,000(11.7)31,182,000(9.7)

20202,743,000(9.0)28,059,000(8.7)

Hospitaladmittance2,147,000(7.0)15,048,000(4.7) <0.001

Pregnancytest13,665,000(44.8)87,455,000(27.1) <

Ultrasound13,423,000(44.0)18,889,000(5.9) <

72-hourrevisit1,293,000(4.2)12,776,000(4.0)0.03

Seenbyconsultant3,498,000(11.5)19,499,000(6.0) <

Lengthofvisit <

<1hr1,769,000(7.6)29,062,000(11.8)

1–2hr3,221,000(13.9)64,011,000(26)

2–4hr10,413,000(44.9)93,180,000(37.8)

>4hr7,789,000(33.6)60,415,0000(24.5)

Waittime

<30min15,538,000(59.5)172,983,000(61.6)

30min-1hr4,830,000(18.5)50,720,000(18.1)

1–2hr3,475,000(13.3)35,543,000(12.7)

>2hr2,280,000(8.7)21,565,000(7.7)

Hospitalcharacteristics

Geographicregion

Northeast4,241,000(13.9)51,007,000(15.8)

Midwest7,302,000(23.9)73,129,000(22.7)

South12,631,000(41.4)130,591,000(40.5)

West6,315,000(20.7)67,797,000(21.0)

Metropolitanstatisticalarea(MSA) <0.001

MSA25,040,000(89.0)247,480,000(84.7)

Non-MSA3,085,000(11.0)44,727,000(15.3)

Dataaren(%),median(interquartilerange). ED,emergencydepartment.

MultivariableAnalysis

UnadjustedandadjustedORsforassociationsofpatient demographicsandhospitalcharacteristicswithpresentation ofpregnantpatientscomparedtonon-pregnantpatientsto theEDarepresentedin Table2.Inthegeneratedmodel,age 20–29years,Hispanicethnicity,publicinsurancestatus,and metropolitanlocationweresignificantlyassociatedwith visitsforpregnantpatients.Theseresultsheldthrough sensitivityanalysestoensurethatimputationinthedataset didnotaffectour findings.

Table2. Bivariableandmultivariablelogisticregressionmodels.

Age

15–19yearsReferenceReference

Race/ethnicity

Non-HispanicWhiteReferenceReference

Residence

PrivateresidenceReferenceReference

EDvisitday Weekend0.99(0.90–

WeekdayReferenceReference

EDvisitseason

Geographicregion

West1.12(0.95–1.32)0.981.03(0.86–1.22)0.09

Metropolitanstatisticalarea(MSA) MSA1.47(1.22–1.76) <0.0011.41(1.18–1.67) <0.001

Non-MSAReferenceReference

OR,oddsratio; ED,emergencydepartment.

DISCUSSION

Inthisstudy,usingdataavailablefromNHAMCS,we estimatedthatthereare2.77millionEDvisitsforpregnant patientsannuallyintheUnitedStates.Mostcommonly, womenpresentingtotheEDwithpregnancyarebetweenthe agesof20–29years,publiclyinsured,andidentifyasBlackor Hispanic.Ofthesepregnantpatientvisits,weestimatethat 13.3%oftheseresultedinanewdiagnosisofpregnancyinthe ED,equivalenttoapproximately370,000pregnancies first identifiedintheEDannually.

Wefoundthat8.6%ofvisitsamongwomenbetweenages 15–44yearswereforpregnantpatients,whichgenerally alignswithpreviouslyreportedvalues;however,therewasa largeamountofvariabilityinreported figures. 7–9 Benson etalperformingasimilarevaluationforpatientspresenting forearlypregnancylossreportedapproximately900,000 visitsannually,whichwouldrepresent32%ofthe2.77 millionpregnantpatientvisitswedescribe.5 Thisishigher thantheoftenreportedrateof20%earlypregnancyloss, whichwesuspectisduetoearlypregnancylossbeinga commonreasonforEDpresentation.19

Ourdatashowsthatpregnantpatientsseekingcareinthe EDaremorelikelytobeBlack,Hispanic.orpublicly insured.Thesepopulationsarelesslikelytoreceiveroutine prenatalcareandhaveahigherratesofpregnancy-related morbidityandmortalitywhencomparedtoWhitepatients.20 Furthermore,unintendedpregnancyratesarehigherandthe rateofreferralfordesiredfamilyplanningservicesislowerin thesepatientgroups.21–23 Currentlymorethanonehalfof abortionsareamongwomenofcolordespitethesepatients experiencinggreaterbarrierstoaccessingfamilyplanning services.24 PatientsonMedicaidsimilarlyhavechallenges accessingabortioncareduetolimitedcoverageforthese services.24 The Dobbs decisionislikelytoexacerbatethese disparitiesandmaydisproportionatelyaffectpregnant patientspresentingtotheED.24 Futurestudiesshould investigatetheseimpactsaslegislationchangesandexamine howEDpresentationsandcaredifferbetweenstatesthat enactrestrictiveabortionlegislationcomparetostates withoutrestriction.

Ourresultssuggestthatalargeproportion(13.3%)of pregnantpatientswhoseekcareintheEDare firstdiagnosed duringtheirEDencounter.Basedonhistoricaldata,halfof thesepregnanciesareunplanned,andhalfwouldendin abortioninthepre-Dobbs era. 21 Discoveryofthese pregnanciesintheEDoffersanopportunityforcounseling andreferraltoavailableabortionservicesifdesired.Thisis especiallyimportantinstateswherestrictrestrictionson gestationalageforlegalabortionexist.Thesepatientsmay facebarrierstocareanddelaysincarefollowingED discharge,suggestingacriticalneedforcounselingand linkagetocareduringtheEDencounter.

Nationwideaccesstothesereproductivehealthcare servicesissupportedbytheAmericanCollegeofEmergency

Physicians.AlthoughtheEDistakingalargerrolein offeringthiscare,furtherresearchisrequiredtoidentifythe needsofthispopulation.25,26 Specifically,futurestudies coulddirectlymeasuretherateofnewpregnancydiagnosisin theED,determinecounselingpracticesamongemergency clinicians,andexaminehowthesepatientsarelinkedtocare ifapregnancyisundesired.Thisdata,alongwith comparisonsbetweenstateswithvaryingdegreesof legislationchange,couldhelpinformpolicychanges.

LIMITATIONS

ResultsarebasedondatafromtheNHAMCS,whichhas several,well-reportedlimitations.18 AlthoughtheNHAMCS makesgreateffortstoincludearepresentativesample,itis possiblethattheincludedvisitsarenotcompletely representativeofEDvisitsnationwide.Nevertheless,the NHAMCSisthelargestdatasettodatewithpopulationbasedestimatesofEDvisitsintheUS.Non-responseratefor itemsintheNHAMCSmayalsobiasresults;however,all ourvariablesofinteresthadnon-responseratesthatfell withinacceptablemargins,andthosewithhighernonresponserates(raceandethnicity)wereevaluatedwith sensitivitytestingtoensureimputedvaluesdidnot compromiseresults.

Wedefinedvisitswithpregnantpatientsinourpopulation bypregnancy-relatedICD-9and-10diagnoses,whichmay havebeenenteredinerrorfornon-pregnantpatients.Visits withanincidentalpregnancydiagnosiswerebasedontriage dataandpregnancytesting,whichmayhavemisclassified pregnanciesasincidentalorfailedtoidentifyotherincidental pregnanciesnotcaptured.Toobtainaconservativeestimate, weexcludedpatientpresentationsforvaginalbleeding, whichmayhaveraisedclinicianorpatientsuspicionof pregnancy.Duetothenatureofthedatasetweanalyzed,we werenotabletoprovidedefinitiveinformationabout completionofapreviouspregnancytestorultrasound,nor aboutthepatient’ssuspicionforpregnancy,whichwouldbe preferredmarkersforidentifyingnewpregnancydiagnoses.

Finally,wewereunabletoprovideinformationabout whetherthesepregnanciesweredesired,whetherpatientshad establishedcarewithanobstetrician,ortheoutcomesof thesepregnancies.

CONCLUSION

Ourstudyrevealsthatpregnantpatientsmakeup3%of EDvisitsannually.Givenrecentlegislativechanges concerningreproductivehealthcare,thesepatientscouldbe significantlyimpacted.TheED,oftenseenasthehealthcare system’ssafetynet,providescrucialcarethatmightnotbe availableelsewhere.Withthepossibilityofpregnantpatients turningmoreoftentotheEDforcare,thereisanurgentneed todevelopandimplementeducationalandpolicystrategies thatsupportthesepatientsinnavigatingtheincreasingly complexrealmoffamilyplanningservices.

AddressforCorrespondence:CarlPreiksaitis,MD,Stanford UniversitySchoolofMedicine,DepartmentofEmergencyMedicine, 900WelchRoad,Suite350,PaloAlto,CA94304.Email: cpreiksaitis@stanford.edu

ConflictsofInterest:Bythe WestJEMarticlesubmissionagreement, allauthorsarerequiredtodiscloseallaffiliations,fundingsources and financialormanagementrelationshipsthatcouldbeperceived aspotentialsourcesofbias.Noauthorhasprofessionalor financial relationshipswithanycompaniesthatarerelevanttothisstudy. Therearenoconflictsofinterestorsourcesoffundingtodeclare.

Copyright:©2024Preiksaitisetal.Thisisanopenaccessarticle distributedinaccordancewiththetermsoftheCreativeCommons Attribution(CCBY4.0)License.See: http://creativecommons.org/ licenses/by/4.0/

REFERENCES

1. Dobbsv.JacksonWomen’sHealthOrg.597U.S.2022.

2.Samuels-KalowME,AgrawalP,RodriguezG,etal.Post-Roe emergencymedicine:policy,clinical,training,andindividual implicationsforemergencyclinicians. AcadEmergMed. 2022;29(12):1414–21.

3.MalikS,KothariC,MacCallumC,etal.Emergencydepartmentusein theperinatalperiod:anopportunityforearlyintervention. AnnEmerg Med. 2017;70(6):835–9.

4.MatenchukBA,RosychukRJ,RoweBH,etal.Emergencydepartment visitsduringpregnancy. AnnEmergMed. 2023;81(2):197–208.

5.BensonLS,MagnussonSL,GrayKE,etal.Earlypregnancylossinthe emergencydepartment,2006–2016. JAmCollEmergPhysicians Open. 2021;2(6):e12549.

6.KilfoyleKA,VreesR,RakerCA,etal.Non-urgentandurgentemergency departmentuseduringpregnancy:anobservationalstudy. AmJObstet Gynecol. 2017;216(2):181.e1–7.

7.StengelCL,SeabergDC,MacLeodBA.Pregnancyintheemergency department:riskfactorsandprevalenceamongallwomen. AnnEmerg Med. 1994;24(4):697–700.

8.RamoskaEA,SacchettiAD,NeppM.Reliabilityofpatienthistoryin determiningthepossibilityofpregnancy. AnnEmergMed. 1989;18(1):48–50.

9.StroteJandChenG.Patientselfassessmentofpregnancystatusinthe emergencydepartment. EmergMedJ. 2006;23(7):554–7.

10.RobertsSCM,WingoE,KimportK.Aqualitativeexplorationofwomen’s experiencesdiscoveringpregnanciesintheemergencydepartment. ContraceptX. 2020;2:100024.

11.ACEPNow.Taskforceformedtoexamineemergencydept. reproductivehealth.2022.Availableat: https://www.acepnow.com/ article/task-force-formed-to-examine-emergency-dept-reproductivehealth/.AccessedNovember8,2022.

12.LeeLK,ChienA,StewartA,etal.Women’scoverage,utilization, affordability,andhealthaftertheACA:areviewoftheliterature. Health Aff(Millwood). 2020;39(3):387–94.

13.2020NHAMCSMicro-data filedocumentation.2020.Availableat: https://ftp.cdc.gov/pub/Health_Statistics/NCHS/ Dataset_Documentation/NHAMCS/doc20-ed-508.pdf AccessedNovember1,2022.

14.NAMCS/NHAMCS-AmbulatoryHealthCareDataHomepage.2022. https://www.cdc.gov/nchs/ahcd/index.htm AccessedNovember8,2022.

15.ZambranoLD,EllingtonS,StridP,etal.Update:characteristicsof symptomaticwomenofreproductiveagewithlaboratory-confirmed SARS-CoV-2infectionbypregnancystatus UnitedStates,January 22–October3,2020. MorbMortalWklyRep. 2020;69(44):1641–47.

16.SchneiderD,AppletonL,McLemoreT.Areasonforvisitclassification forambulatorycare. VitalHealthStat2. 1979;(78):1–63.

17.vonElmE,AltmanDG,EggerM,etal.TheStrengtheningtheReporting ofObservationalStudiesinEpidemiology(STROBE)statement: guidelinesforreportingobservationalstudies. AnnInternMed. 2007;147(8):573–7.

18.McCaigLFandBurtCW.UnderstandingandinterpretingtheNational HospitalAmbulatoryMedicalCareSurvey:keyquestionsandanswers. AnnEmergMed. 2012;60(6):716–21.e1.

19.ACOGPracticeBulletinNo.200:Earlypregnancyloss. ObstetGynecol. 2018;132(5):e197.

20.HillL,ArtigaS,RanjiU.Racialdisparitiesinmaternalandinfanthealth: currentstatusandeffortstoaddressthem.2022.Availableat: https:// www.kff.org/racial-equity-and-health-policy/issue-brief/racialdisparities-in-maternal-and-infant-health-current-status-andefforts-to-address-them/.AccessedNovember7,2022.

21.FinerLBandZolnaMR.DeclinesinunintendedpregnancyintheUnited States,2008–2011. NEnglJMed. 2016;374(9):843–52.

22.DehlendorfC,HarrisLH,WeitzTA.Disparitiesinabortionrates:apublic healthapproach. AmJPublicHealth. 2013;103(10):1772–9.

23.NobelK,LukeAA,RiceWS.Racialdisparitiesinpregnancyoptions counselingandreferralintheUSSouth. HealthServRes. 2023;58(1):9–18.

24.ArtigaS,HillL,RanjiU,etal.Whataretheimplicationsoftheoverturning ofRoev.Wadeforracialdisparities?Availableat: https://www.kff. org/racial-equity-and-health-policy/issue-brief/what-are-theimplications-of-the-overturning-of-roe-v-wade-for-racial-disparities/ AccessedNovember18,2022.

25.ACEPCouncilResolution24.Availableat: https://www.acep.org/ what-we-believe/actions-on-council-resolutions/councilresolution/ AccessedDecember3,2022.

26.PreiksaitisCandHenkelA.Theevolvingroleofemergencymedicinein familyplanningservices. CurrOpinObstetGynecol. 2023;35(6):484–9.