Racial and Ethnic Disparities in EMS Use of Restraints and Sedation for Patients With Behavioral Health Emergencies

Abstract

Importance: Emergency medical services (EMS) clinicians commonly care for patients with behavioral health emergencies (BHEs), including acute agitation. There are known racial and ethnic disparities in the use of physical restraint and chemical sedation for BHEs in emergency department settings, but less is known about disparities in prehospital use of restraint or sedation.

Objective: To investigate the association of patient race and ethnicity with the use of prehospital physical restraint and chemical sedation during EMS encounters for BHEs.

Design, setting, and participants: This nationwide retrospective cohort study used data from EMS agencies across the US that participated in the 2021 ESO Data Collaborative research dataset. Emergency medical services encounters among patients aged 16 to 90 years with a primary or secondary impression, sign or symptom, or protocol use associated with a BHE from January 1 to December 31, 2021, were included. Statistical analysis was conducted from July 2023 to March 2024.

Exposures: Patient race and ethnicity, which was categorized as Hispanic, non-Hispanic Black, non-Hispanic White, non-Hispanic other (American Indian or Alaska Native, Asian, Hawaiian Native or Other Pacific Islander, other, or multiracial), and unknown.

Main outcomes and measures: The primary outcome was administration of any physical restraint and/or chemical sedation (defined as any antipsychotic medication, benzodiazepine, or ketamine).

Results: A total of 661 307 encounters (median age, 41 years [IQR, 30-56 years]; 56.9% male) were included. Race and ethnicity were documented as 9.9% Hispanic, 20.2% non-Hispanic Black, 59.5% non-Hispanic White, 1.9% non-Hispanic other, and 8.6% unknown race and ethnicity. Restraint and/or sedation was used in 46 042 (7.0%) of encounters, and use differed across racial and ethnic groups (Hispanic, 10.6%; non-Hispanic Black, 7.9%; non-Hispanic White, 6.1%; non-Hispanic other, 10.9%; unknown race and ethnicity, 5.9%; P < .001). In mixed-effects logistic regression models accounting for clustering by EMS agency and adjusted for age, gender, urbanicity, and community diversity, patients who were non-Hispanic Black had significantly greater odds of being restrained or sedated across all categories compared with non-Hispanic White patients (eg, any restraint and/or sedation: adjusted odds ratio [AOR], 1.17 [95% CI, 1.14-1.21]; physical restraint: AOR, 1.22 [95% CI, 1.18-1.26]). There was no significant difference in adjusted odds of any restraint and/or sedation use for the remaining racial and ethnic groups compared to non-Hispanic White patients. Clustering was associated with agency-level variation in restraint or sedation use (intraclass correlation coefficient, 0.16 [95% CI, 0.14-0.17]).

Conclusions and relevance: This nationwide retrospective cohort study of EMS encounters for patients with BHEs found differences in the use of prehospital restraint and/or sedation by patient race and ethnicity and an agency-level association with variation in restraint and/or sedation use. These data may inform improvements to protocols and training aimed at equitable care for BHEs.

Figure 1.. Flow Diagram for Emergency Medical Services (EMS) Encounters Included in Study Population

Figure 2.. Adjusted Odds of Any Restraint and/or Sedation Use by Patient Race and Ethnicity Among Emergency Medical Services Encounters for Behavioral Health Emergencies

Models accounted for clustering by emergency medical services agency and adjusted for age, gender, urbanicity, and community diversity (Social Vulnerability Index racial and ethnic minority status theme). AOR indicates adjusted odds ratio.

Figure 3.. Adjusted Odds of Secondary Outcomes by Patient Race and Ethnicity Among Emergency Medical Services Encounters for Behavioral Health Emergencies

Models accounted for clustering by emergency medical services agency and adjusted for age, gender, urbanicity, and community diversity (Social Vulnerability Index racial and ethnic minority status theme). AOR indicates adjusted odds ratio.