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

Abstract

Physical restraints are occasionally used in the emergency department (ED) to manage acute agitation to prevent self-harm and to protect staff, with a reported frequency of 0.5-1.4%. There is limited information comparing the demographic factors associated with the use of restraints for behavioral versus medical indications. Prior studies have also identified an association between socioeconomic and demographic factors and the use physical restraints, raising concern for the presence of implicit bias. The objective of this study was to identify the socioeconomic and demographic characteristics of patients placed in physical restraints for behavioral versus medical indications and to assess the variables that may increase the risk of patients being placed in restraints specifically for behavior indications. A cross sectional study of adult ED patients at a Level 1 Trauma tertiary academic medical center during March 2019 through February 2021 was performed to identify patients placed in physical restraints for medical or behavioral indications. Comparisons between behavioral and medical restraint groups were assessed using the t- test (for age) and the Fisher exact test (for sex, race, ethnicity, emergency severity index (ESI), insurance status, time of arrival, age over 50, psychiatric admission, and medical admission). Factors associated with behavioral restraint use compared to medical restraint use were assessed by a multivariable logistic regression analysis. Odds ratio (OR) and their 95% confidence intervals were reported. A two-sided 0.05 significance level was used throughout. SAS versions 9.4 (SAS Institute, Cary, North Carolina) software was utilized for statistical analysis. Among the 172, 708 ED visits during the study period, 1, 165 (0.67%) involved the use of restraints. Behavioral restraints were applied in 612 patients (53.5%) and medical restraints in 553 (47.5%). Overall physical restraint use was similar between white and non-white patients (0.74% and 0.71% respectively, Fisher Exact OR 1.04, p=0.55). However, significant differences were found among age, sex, race, ESI, insurance status, psychiatric and medically hospitalized patients in further analysis between the behavioral and medical restraint groups (p<0.0001). Factors that were identified as significantly increasing the risk of receiving behavioral restraints in particular included male sex, insurance and race; the c-statistics for the logistic model was 0.864. The demographic and socioeconomic characteristics of patients placed in restraints for medical indications are different than those placed in restraints for behavioral indications. Although race was identified as a factor increasing the risk for behavioral restraints use in the ED, the overall use of physical restraints was similar in white versus non-white patients suggesting against the presence of implicit bias against non-white patients.View Large Image Figure ViewerDownload Hi-res image Download (PPT)