Prevalence and precursors of the use of restraint and seclusion in a private psychiatric hospital: comparison of child and adolescent patients.

The use of restraint and seclusion is highly regulated in psychiatric inpatient settings. However, the majority of studies of restraint and seclusion are based on public hospitals serving adult patients, with some limited data available on adolescents and children. This paper presents prospectively collected data on restraint and seclusion over a 2-year period at a private psychiatric hospital whose patients include large numbers of both adolescents and pre-adolescent children. 2 years of restraint and seclusion data were analyzed on a total of 2,411 unique patients. Types of seclusion included in-room seclusion on the treatment unit and off-unit seclusion in a separate seclusion annex. Restraints consisted solely of short term (<15 min) and longer term (>14 min) manual restraints. The use of IM medication was also recorded. The precipitants of these events were examined. These included physical and verbal threats, stabbing or throwing objects, attempts to elope, attempts to hurt one's self or another, or property destruction. Out of 2,411 child and adolescent in-patients admitted during the period under review, only 703 (29%) experienced restraint or seclusion. Among these, the modal number of events per patient was one (n = 156), but the maximum number of occurrences was 163. Child patients had a much higher frequency of events (n = 396, 53%) than adolescents (n = 307, 19%). There were notable differences in the types of seclusion events, with children typically experiencing in-room seclusion on the unit. When age was examined as a continuous variable, younger patients had a higher prevalence of restraint and seclusion, significantly more restraint and seclusion, and these restraint and seclusion events were significantly shorter than those seen in older patients. Multiple other potential determinants of these events were examined, including diagnosis, symptom severity at admission, age, and gender, but none of these predicted these events. Restraint and seclusion events were more common for children and less so for adolescents, with robust age effects for the likelihood of any seclusions, the number of seclusions and restraints, and the duration of seclusions and restraints. Patients who experienced restraint or seclusion typically required it only once during their hospitalization. Only age was found to be a predictor of the restraint and seclusion variables. Given these findings, it appears that management of agitated behavior in children and adolescents may be a qualitatively different phenomenon. Future research should be directed at understanding the determinants of high frequency agitated behavior and developing alternatives to seclusion or restraint.