509 - Associations between Child Protective Services and Restraint Use in a Pediatric Emergency Department

Publication Number: 509.3812

Isaac V. Faustino, Yale School of Medicine, New haven, CT, United States; Max J. Rolison, Yale School of Medicine, New Haven, CT, United States; Ambrose Wong, Yale University School of Medicine, New Haven, CT, United States; Gunjan Tiyyagura, Yale School of Medicine, New Haven, CT, United States

Background: Approximately 6-10% of children presenting to emergency departments (EDs) for psychiatric care are physically restrained. While previous studies have identified factors like race and ethnicity to be associated with a higher likelihood of physical restraint use, the impact of child protective services (CPS) involvement on this relationship is unknown.

Objective: To examine associations between physical restraint use and children with CPS involvement.

Design/Methods: We conducted a retrospective cohort study using data abstracted from electronic health records. All pediatric ED encounters from a tertiary care children’s hospital with behavioral chief complaints between Jan 2021 and Oct 2023 were included for analysis. Use of restraints was defined as a physical restraint order in the encounter’s electronic health record, while length of stay referred to the time between ED arrival to departure. CPS involvement was defined as the child either being in active CPS custody or having a new CPS referral made during the encounter. We used a Cox regression model to determine the hazard of a child being restrained with CPS involvement. We adjusted the model for age, legal sex, race and ethnicity. Adjusted linear regression models were used to determine the difference in time to restraint amongst those restrained and the length of stay with and without CPS involvement.

Results: We analyzed a total of 6,288 unique ED encounters with behavioral health chief complaints, of which 913 (14.5%) had CPS involvement. The mean age was 12.9 years (SD: 2.8 years) with 2638 (42.0%) males. Restraints were ordered in 115 (1.8%) encounters. Unadjusted and adjusted hazard ratios of being restrained with CPS involvement were 2.2 (95% confidence interval, CI: 1.5, 3.2, p< 0.001) and 1.8 (95% CI: 1.2, 2.7, p=0.007) respectively.
Amongst those who were restrained, CPS was involved in 35 (30.4%) encounters. Time to restraint was longer by 2.3 hours (95% CI: 1.0, 5.0, p=0.04) with CPS involvement. Length of stay was longer by 2.1 hours (95% CI: 1.4, 3.2, p< 0.001) with CPS involvement.

Conclusion(s): Among children presenting to the ED with behavioral chief complaints, CPS involvement was associated with a higher hazard of physical restraint use and an increase in length of stay. Among those restrained, CPS involvement was associated with an increase in time to restraint. These findings suggest current systems may provide suboptimal care for children with CPS involvement. Future work should be done to clarify how structural forces lead to increased restraint use and length of stay, both of which are markers of poorer ED-based care quality.