A Scoping Review of Current Social Emergency Medicine Research

Introduction Social emergency medicine (EM) is an emerging field that examines the intersection of emergency care and social factors that influence health outcomes. We conducted a scoping review to explore the breadth and content of existing research pertaining to social EM to identify potential areas where future social EM research efforts should be directed. Methods We conducted a comprehensive PubMed search using Medical Subject Heading terms and phrases pertaining to social EM topic areas (e.g., “homelessness,” “housing instability”) based on previously published expert consensus. For searches that yielded fewer than 100 total publications, we used the PubMed “similar publications” tool to expand the search and ensure no relevant publications were missed. Studies were independently abstracted by two investigators and classified as relevant if they were conducted in US or Canadian emergency departments (ED). We classified relevant publications by study design type (observational or interventional research, systematic review, or commentary), publication site, and year. Discrepancies in relevant publications or classification were reviewed by a third investigator. Results Our search strategy yielded 1,571 publications, of which 590 (38%) were relevant to social EM; among relevant publications, 58 (10%) were interventional studies, 410 (69%) were observational studies, 26 (4%) were systematic reviews, and 96 (16%) were commentaries. The majority (68%) of studies were published between 2010–2020. Firearm research and lesbian, gay, bisexual, transgender, and queer (LGBTQ) health research in particular grew rapidly over the last five years. The human trafficking topic area had the highest percentage (21%) of interventional studies. A significant portion of publications -- as high as 42% in the firearm violence topic area – included observational data or interventions related to children or the pediatric ED. Areas with more search results often included many publications describing disparities known to predispose ED patients to adverse outcomes (e.g., socioeconomic or racial disparities), or the influence of social determinants on ED utilization. Conclusion Social emergency medicine research has been growing over the past 10 years, although areas such as firearm violence and LGBTQ health have had more research activity than other topics. The field would benefit from a consensus-driven research agenda.


INTRODUCTION Background and Importance
In 1848 Rudolph Virchow declared social problems to be "largely within the jurisdiction" of physicians. 1,2 Emergency physicians serve as safety net providers and are often on the front line of epidemics, natural disasters, and civil unrest. 3 The emergency department (ED) is a unique place to identify and intervene in social issues, as patients often present with complaints directly influenced by social determinants of health (SDOH), 4 and EDs serve patients who have limited access to care. 5 As a result, the field of social emergency medicine (EM) has developed to examine and influence social factors in the context of acute healthcare needs. The scope of social EM is immense, including domains from housing insecurity to substance use, to gun and intimate partner violence, and many others. Many domains within social EM are known to influence emergency care utilization and health outcomes.

Goals of This Investigation
While prior systematic reviews have examined the existing literature with a specific focus on material needs, there is a need to characterize the literature examining the broader field of social factors, including non-material factors -such as language, exposure to violence, and immigration status -known to influence emergency care and outcomes. 6 The primary aims of this scoping review were to understand and map the breadth of current literature for various social EM topics and categorize the type of research that exists for each topic, in order to identify potential areas where future social EM research efforts should be directed.

METHODS
This review was informed by the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines for scoping reviews. We identified 11 content areas based on a previously published systematic review of patients' social and economic needs, including housing needs, employment needs, education and literacy, financial insecurity, personal safety (including intimate partner violence, human trafficking, firearms, child abuse, and elder abuse), and food insecurity. 7 Additional topic areas were added based on author consensus, including lesbian, gay, bisexual, transgender, and queer (LGBTQ) health, language, immigration, incarceration, and transportation needs. Two final search terms ("social determinants of health" and "social emergency medicine training"), were added in consultation with a research librarian to ensure inclusion of publications that address more than one topic, as well as educational research.
We conducted a comprehensive literature search using a combination of Medical Subject Heading (MeSH) terms and phrases pertaining to topic areas (eg, "homelessness," "housing instability"). We restricted studies to those conducted in the US or Canada. We used the PubMed database for our searches, with the exception of the "Social Emergency Medicine Training" search, which also used the MedEd Portal database. For searches that yielded fewer than 100 total publications, we used the PubMed "similar publications" tool to expand the search and ensure no relevant publications were missed. Criteria for inclusion were as follows: (1) published in English; (2) conducted in the US or Canada through July 31, 2020; and (3) deemed relevant to social EM. Studies were considered relevant to social EM if they focused on social factors in the context of acute healthcare needs; therefore, we included the following criteria: 1) the study population consisted of ED patients or emergency clinicians; 2) the study or intervention occurred in the ED; or (3) ED utilization or outcomes were defined as a primary outcome.
Once a publication was deemed to meet inclusion criteria we extracted additional information such as title, PubMed ID, year of publication, and study design type (original observational or interventional research, systematic review, or commentary) into a standardized data collection form. We further catalogued observational and interventional publications by setting (single center, multicenter regional, and multicenter national). For each publication, study objectives (eg, defining prevalence, evaluating an educational intervention) were also recorded. For search results in each topic area, two co-investigators independently assessed each study for inclusion and relevance to social EM, Any discrepancies in relevance or categorization were reviewed and reconciled by a third reviewer. We also classified publications classified as relating to pediatric populations if they included children or adolescents (≤ 21 years) or if they were conducted in pediatric EDs.

RESULTS
Our search strategy identified 1571 publications, of which 590 publications in 18 categories were classified as relevant to social EM. Depiction of search strategy and classification process are in Figure 1. The study designs of included publications were as follows: 58 (10%) interventional publications; 410 (69%) observational publications; 26 (4%) systematic reviews; and 96 (16%) commentaries. Publication years ranged from 1968 to 2020, with 402 (68%) eligible articles published since 2010. Results are summarized in Figure 2. Study objectives within each topic are summarized in the Table. Figures 3A through 3D show study type by year for select topics with the largest number of studies (firearms, intimate partner violence, child abuse, and housing/homelessness).

Child Abuse
We identified 114 relevant publications: 71 observational studies 70-141 ; 12 interventional studies 142-153 ; three review publications 154-156 ; and 28 commentary publications ( Figure  3B). 157-184 There were several common objectives among the observational studies. Twenty-two (31%) observational publications focused on determining incidence/prevalence of child abuse in different settings (single EDs, specific geographic areas, nationwide), and characterizing cases of child abuse. 70 Child abuse cases were often categorized by demographic characteristics, such as age, gender, race, and insurance status, as well as injury patterns. Nineteen (26%) studies focused specifically on injury patterns of abused children, and the likelihood of child abuse among patients presenting with fractures, head trauma, and oral injuries. 71,75,81,82,86,91,95,97,99,101,107,[109][110][111]113,115,117,119,127 About 22 (31%) studies focused specifically on child sexual assault cases, 70,79,80,96,98,106,112,116,118,120,123,125,130,132,133,141 with six of these studies looking at sexually transmitted infection (STI) and pregnancy testing, STI prophylaxis, and the use of sexual assault nurse examiners. 77,78,121,122,140,150 Two of three review publications focused on screening, 154,155 with one publication focusing on improving the ED workflow for suspected or confirmed child abuse cases. 156 Other common study objectives included examining and amending the ED workflow for child abuse cases, developing screening protocols, and understanding provider knowledge and training with regard to child abuse in the ED. A plurality (42%) of the interventional studies involved evaluations of educational interventions for ED providers meant to improve child abuse screening and recognition. 142,144,147,149,152 Three (25%) interventional studies focused on child sexual assault. 142,144,150

Elder Abuse
We identified 31 relevant publications: 16 observational studies 185-215 ; three review publications 187,212,214 ; and 12 commentary publications. 185,186,192,195,197,198,200,201,205,210,211,215 Common objectives among the observational studies included the following: developing and testing screening tools (N = 5, 31%) 194,196,202,204,206 ; ED utilization by abused patients (N = 2, 13%) Shah et al. (13%) 191,208,213 ; patient characteristics (N = 2, 13%) 202,209 ; provider awareness and perspectives on elder abuse (N = 2, 13%) 188,203 ; and prevalence of elder abuse (N = 1, 6%). 189 There was a lack of interventional studies regarding educational interventions or the use of screening tools. All 12 commentary publications from the 1990s to 2019 served to raise awareness about elder abuse in the ED and ways to identify and combat it.

Human Trafficking
We identified 19 relevant publications: four interventional studies 336-339 ; seven observational studies 79,340-345 ; one systematic review 346 ; and seven commentary publications. 347-353 All publications were published after 2012. Of the seven observational studies, three (43%) related to screening tools to identify patients experiencing sex trafficking. 340,342,344 Two (25%) focused on patient characteristics, 79,345 one was a case report (13%), 343 and the other study focused on emergency nurses' perspectives (13%). 341 All four interventional studies looked at the efficacy of educational modules on ED staff in better understanding the issue of human trafficking in the ED and being better able to identify human trafficking victims in the ED. The systematic review was of existing human trafficking screening tools in the ED. Seven studies (37%) focused specifically on child sex trafficking victims in the ED. 340,[344][345][346]352,353

Lesbian, Gay, Bisexual, Transgender, and Queer Health
We identified 22 relevant publications: 14 observational studies 354-367 ; one interventional study 368 ; and seven 369-375 commentary publications. Of these, 21 (95%) were published after 2014. Of the observational studies, five (36%) focused on patient provider attitudes toward sexual orientation and gender identity data collection in the ED, 354,355,361,362,365 and six (43%) focused on the care of transgender patients in the ED, 356,358-360,367,370 with many surveying experiences of discrimination among transgender patients. 358,360,363,367 Four (29%) observational publications focused on LGBTQ health competency training by emergency care providers. 356,364,366,368,373 One (7%) publication broke down intimate partner violence prevalence in the ED by the sexual orientation of patients. 357 The commentary publications centered on the same themes.
The single interventional publication used pre/post data to evaluate the efficacy of an ED competency training in LGBTQ health. 368

Immigration
We identified 24 relevant publications: 376-399 20 observational studies 376-395 ; one interventional study 399 ; and three commentary publications. 396-398 All observational publications investigated ED utilization in immigrant vs non-immigrant groups, with some specifically assessing Latino populations. Two publications (10%) studied the fear of ED utilization among Latino populations. 376,386 The single interventional study assessed a texting-based intervention of Latino families as a means to reduce ED utilization while increasing well-care and vaccine adherence. 399

Incarceration
We identified 11 relevant publications: eight observational studies 400-407 ; two interventional studies 408,409 ; and one commentary publication. 410 Of the observational studies, five (63%) publications centered on ED utilization after release from prison. 400,[403][404][405]407 Both interventional publications focused on models of care for recently released prisoners. Of all publications, three (38%) focused on pediatric populations. 401,402,404 Language We identified 32 relevant publications 411-442 ; 26 observational studies 411-436 ; three interventional studies 437-439 ; one review publication; and two commentary publications. The observational research spanned a broad range of topic areas covering many parts of ED care, including triage (8%), 419,438 history of present illness collection (4%), 416 management of care (4%), 411 interpreter utilization and need (12%), 415,428,429 ED resource utilization (15%), 423,424,431,433 length of stay (8%), 427,430 the discharge process (15%), 417,420-422 and follow-up care (8%). 432,435 Of the interventional studies, one examined the role of the patient's preferred language in the success of a drinking intervention. 437 Another looked at the efficacy and efficiency of a bilingual, kiosk-based selftriage system compared to a nurse. 438 The third publication investigated the effectiveness of a bilingual medical history

Shah et al.
Review of Current Social EM Research questionnaire. 439 The review and commentary pieces described the language barriers patients face in the ED 440 and utilization of interpreter services. 441,442 Of all publications, 12 (38%) focused on pediatric populations. 411,417,[420][421][422][425][426][427]430,431,433,438 Literacy We identified 34 relevant publications 443-475 : 25 observational studies 443-466 ; four interventional studies 467-470 ; three review publications 473-475 ; and two commentary publications. 471,472 Of the observational studies, 11 (41%) examined health literacy screening and patients' understanding of discharge instructions, 443,445,451-453,461-463,466,474,476 eight (30%) investigated the relationship between health literacy and ED utilization, 447,448,455,457,458,460,469,473 and 10 (37%) focused on the literacy of the parents of pediatric patients. 446,448,454,457,458,[467][468][469]473,475 One study focused on ways to improve a patient's understanding of the clinical encounter with improved communication tools for physicians or teach-back strategies with patients. 450 All four of the interventional studies involved educational interventions for parents of pediatric patients. 467-470

Food Insecurity
We identified 29 relevant publications 25 observational studies 549-574 ; two interventional studies, 575,576 ; and two commentary publications. 577,578 Objectives among observational studies included the following: food insecurity prevalence (27%) 555,[560][561][562]564,565,567 ;ED utilization (19%) 549,550,552,570,571 ; screening (8%), 553,572 ; and cost of care (12%). 550,559,561 Four (15%) publications explored the health consequences of food insecurity. 561,565-567 Five (19%) publications focused the intersection between diabetic patients, food insecurity, and presentation to the ED. 552,554,559,563,568 Three publications (12%) also focused specifically on Supplemental Nutrition Assistance Program benefits running out near the end of the month, 569 and the impact on patients with diabetes 568 and hypertension. 556 Nine (35%) observational studies focused on pediatric populations. 549,[553][554][555]560,562,566,573,574 One interventional study was a randomized controlled trial of two screening methods, 576 and the other was a program to improve access to food for pediatric ED patients. 575

Transportation
Two relevant publications were identified, 579,580 both of which were observational and published in 2019. One publication compared proximity of freestanding EDs and hospital EDs to public transit in three different metro areas. 579 The second discussed ridesharing services as alternative options to ambulances for stable psychiatric patients to reach the emergency department. 580

Financial Insecurity
We identified two relevant publications, both of which were observational. 581,582 Both publications focused on the financial burden for patients of specific chief complaints in the ED, including atopic dermatitis and orthopedic injuries. One publication looked specifically at the pediatric population. 582

Education
Two relevant publications were identified, both of which were observational studies. One publication explored the association between educational attainment and patterns of ED use in patients with sickle cell disease, 583 and the other focused on the relationship between educational attainment and likelihood of receiving opioids for pain management in the ED. 584

Employment
We identified three relevant publications: two observational studies 585,586 ; and one systematic review. 587 The systematic review broadly examined social and demographic characteristics influencing ED use, and included unemployment as one of many variables. Of the observational studies, one correlated unemployment rates and trauma admissions in New Orleans, 586 and the other correlated ED visits with areas experiencing "economic hazard," which included unemployment rate. 585

Social Determinants of Health (SDOH)
We identified seven relevant publications 3,6,555,588-593 : five observational studies 3,588-590,593 ; one review publication 6 ; and one commentary publication. 592 There were no interventional studies. Three (60%) of the observational publications focused on the SDOH of specific populations -dialysis patients, 588 patients with sickle cell disease, 589 and patients who inject intravenous drugs 593 -and the relationship with ED utilization. Another publication focused on predicting ED visits using SDOH measures. 590 Two publications (29%) focused on pediatric populations. 555,590

Review of Current Social EM Research
Shah et al.

Social Emergency Medicine Training
A total of three relevant publications were identified: one educational intervention 594 ; and two commentary publications. 591,595 The education intervention assessed the impact of a longitudinal curriculum for fourth-year medical students on their EM clerkship rotation. 594 The commentary publications discussed the incorporation of SDOH into various aspects of EM training.

DISCUSSION
We identified 590 publications in 18 categories relevant to social EM, demonstrating a high degree of interest in social EM topics. Despite the large and growing number of relevant publications across categories, only 58 publications (10%) were interventional studies. In most topic areas, observational studies have already done a thorough job of describing and characterizing disparities by social identity and circumstance. For example, while a large number of studies looked at ways to effectively screen patients for things like interpersonal violence, health literacy, and human trafficking, there were few publications following up on outcomes for patients who screened positive. Even fewer interventional studies examined patient-oriented outcomes; most interventional studies were educational in nature, with outcomes such as clinician awareness and effectiveness of screening. The dearth of interventional studies examining patient outcomes underscores a need for funding to support testing and dissemination of potential interventions, given that observational studies are more feasible and less resource-intensive than interventional studies.
Topics with the most published research included gun violence, child abuse, intimate partner violence, and housing/homelessness; these four categories combined constituted 63% of all relevant publications. There were several topic areas in which the literature base has grown rapidly in recent years, including gun violence and LGBTQ health. Topics such as elder abuse and incarceration have been the topic of few publications in the last five years, suggesting possible stagnation in these areas. About one third of the relevant publications included were related to the pediatric ED. We found very little research in the following eight topic areas: transportation, financial insecurity, education, employment, incarceration, racism, and legal needs, possibly because they may have been traditionally perceived as less directly related to clinical care and may thus have received less attention.
Prior literature has examined the scope of EM research focused on material needs; our study also examines nonmaterial social risk factors for health outcomes. While the acknowledgment of the interplay between social factors and patients' acute health care needs and outcomes has existed in medical literature for decades, terminology such as "social emergency medicine" is more recent and has increased following a consensus conference about the field. 7

LIMITATIONS
There were several limitations to our review. First, we largely used only the PubMed database, which may have left out relevant publications; however, we systematically searched PubMed, and a majority of biomedical publications are indexed in PubMed. All our search terms were specific to EM, which may have also left out research relevant to EM conducted in related settings or fields. We limited our search to "title only" rather than "title and abstract," which may have also omitted relevant publications; however, after attempting both "title only" and "title and abstract" searches, we found "title only" searches to have much higher relevance. We also did not conduct a detailed analysis of publication quality, given that we set out to complete a scoping review rather than a systematic review; however, publication quality would have been difficult to assess across the diversity of topic areas given the vast array of topics and study designs. We maximized reliability by using two independent reviewers for each topic area, with a third reviewer who reconciled any differences in opinion regarding relevance or publication inclusion.

CONCLUSION
Social emergency medicine research has accelerated in recent years. Numerous observational studies and commentary publications have defined and characterized problems relevant to social EM, and several educational interventions have demonstrated ways to improve provider awareness of different social EM topics. However, based on our review, there is a dearth of social EM research focused on patient-centered interventions. A consensus-driven research agenda should be pursued to accelerate patient-centered interventions aimed at social factors that influence acute healthcare and outcomes.