Characteristics of Patients Restrained in the Emergency Department and Evaluation for Disparities in Care

Kaytlena Stillman; James Mirocha; Joel Geiderman; Sam Torbati

doi:10.1016/j.jemermed.2023.05.008

Characteristics of Patients Restrained in the Emergency Department and Evaluation for Disparities in Care

J Emerg Med. 2023 Nov;65(5):e393-e402. doi: 10.1016/j.jemermed.2023.05.008. Epub 2023 Jun 4.

Authors

Kaytlena Stillman¹, James Mirocha², Joel Geiderman¹, Sam Torbati¹

Affiliations

¹ Department of Emergency Medicine.
² Biostatistics and Bioinformatics Research Center, Cedars-Sinai Medical Center, Los Angeles, California.

PMID: 37722949
DOI: 10.1016/j.jemermed.2023.05.008

Abstract

Background: Physical restraints are used in the emergency department (ED) for agitated patients to prevent self-harm and protect staff. Prior studies identified associations between sociodemographic factors and ED physical restraints use.

Objectives: The primary objective was to compare characteristics of ED patients receiving physical restraints for violent and nonviolent indications vs. patients who were not restrained. The secondary objective was to compare rates of restraint use among ED providers.

Methods: This was a single-center cross-sectional study of adult ED patients from March 2019 to February 2021. Factors compared across groups were age over 50 years, gender, race, ethnicity, insurance, housing, primary language, Emergency Severity Index, time of arrival, mode of arrival, chief complaint, and medical admission. Odds ratios were reported. Rates of emergency physician restraint orders were compared using the chi-square test.

Results: Restraints were used in 1228 (0.9%) visits. Younger age, male gender, "unknown" ethnicity, self-pay or "other" nonprivate insurance, homelessness, arrival by first responders, and medical hospitalization were associated with increased odds of restraint. Black patients had lower odds of any restraint than White patients (odds ratio 0.93; 95% confidence interval 0.79-1.09) and higher odds of violent restraint than White patients, although not significant (odds ratio 1.55; 95% confidence interval 0.95-2.54). ED providers had significant differences in total and violent restraint use (p-values < 0.0001 and 0.0003, respectively).

Conclusion: At this institution, certain sociodemographic characteristics were associated with receiving both types of physical restraint. Emergency physicians also differed in restraint-ordering practice. Further investigation is needed to understand the influence of implicit bias on ED restraint use.

Keywords: emergency department; health care disparities; physical restraint.