Download Full Text (232 KB)

Publication Date



Introduction:Hospitalized children are at risk for agitation and behavioral dysregulation, which may escalate to unsafe behaviors. If less restrictive de-escalation strategies fail to mitigate serious safety risks to self or others, mechanical restraints may be used on a limited basis. Restraint use, however, is problematic due to associated negative child mental health outcomes and injury risk. Pediatric studies of restraint use outside of dedicated psychiatric facilities are lacking. Specifically, clinical and demographic features associated with restraint use are poorly described. A broader understanding of restraint use among hospitalized children can guide alternative strategies for management and restraint mitigation efforts.

Objective: To describe clinical and demographic features of children associated with mechanical restraint use for unsafe behaviors.

Design/methods: We conducted a single-center, retrospective cohort study of patients 5-20 years of age hospitalized during 2017-2021 at our free-standing children’s hospital, which has no associated psychiatric or behavioral health units. We identified restraint encounters by electronic medical record query of physician orders and nursing documentation. Relative risk of restraint was modeled as a function of patient and hospitalization characteristics using a mixed Poisson working model with a random intercept to adjust for clustering of hospitalizations within patients; length of stay was included an exposure variable.

Results: We included 29,964 patients (46,853 encounters), including 221 patients (271 encounters) with associated restraint use. Based on model results, restrained patients tend to be older and to have higher sex- and age-adjusted weight z-scores, and were more likely to be male, have public insurance, and receive general pediatric care vs. surgery or specialty care (Table 1). Risk of restraint was lowest among Hispanic patients and highest among Black patients. Nearly two-thirds (63%) of patients experiencing restraint had a coded behavioral health diagnosis at some visit during the study period.

Conclusions: We observed strong associations between specific clinical and demographic characteristics and the use of restraints during hospitalizations, and our results suggest racial disparities in restraint use. Our results may inform interventions to reduce unnecessary restraint use and improve health equity. Multicenter studies describing health disparities in restraint use, as well as variation in restraint use and associated clinical outcomes are needed.

Document Type


Characteristics Of Hospitalized Children Associated With Restraint Use At A Free-Standing Children’s Hospital