Associations of Restraint and Seclusion With Race and Ethnicity on an Adolescent Inpatient Psychiatry Service

There is limited work examining the association of race and ethnicity with restraint and seclusion (R/S) in pediatric inpatient psychiatric units. The present study assessed risk of R/S by race and ethnicity based on a retrospective review of electronic medical records (EMRs) from an adolescent inpatient psychiatric service. Demographic, diagnostic, and R/S data were analyzed for all 1,865 admissions of 1,327 patients from an adolescent unit at a child and adolescent psychiatric hospital from June 2018 to June 2021. R/S occurred in 459 of the admissions. For the purpose of patient privacy and statistical analysis, race was grouped into the following: Black or African American, other (American Indian or Alaskan Native, Asian, multiracial, other), and White. Patients identified as unknown were not included in the analysis. A binary logistic regression with a repeated subject effect regressed R/S onto race and adjusted for age, gender, and length of stay (LOS). There was an overall significant association of R/S and race (χ22 = 16.81, p < .001), but not ethnicity. In a regression model adjusted for age, gender, and LOS, patients identified as Black or African American were at significantly higher risk of R/S compared with patients identified as White (odds ratio = 1.66, p = .036). There was no significant difference in risk of R/S between patients identified as White vs other. Younger age and longer LOS were also significantly associated with R/S. These findings highlight a critical health care disparity related to race on an inpatient adolescent psychiatry service. There is likely a combination of individual and systemic factors leading to discriminatory practices in the use of R/S. Future work will assess potential associations with diagnosis and child welfare involvement and will examine additional characteristics of R/S. Focus groups held with hospital and community stakeholders will guide next steps to address these findings.