Prediction of restraints among youth in a psychiatric hospital: Application of translational action research

This study extends a translational action research program by applying a theoretically based measure of risk in predicting incidents of restraint among children and adolescents in a secure psychiatric hospital. Youth inpatients (N = 149, aged 5-17 years) were assessed at intake for the presence of selected individual and contextual risk factors, and their involvement in critical incidents was tracked (i.e., number of episodes in which restraint was applied) for the remainder of their hospitalization. Models including history of aggression or history of previous placements as well as combined models including several individual and contextual factors significantly predicted the likelihood of a youth becoming involved in at least one restraint. Unique predictors of restraint involvement included history of aggression against adults and history of previous psychiatric hospitalizations. None of the variables assessed predicted the extent of a youth's restraint involvement. The implications of these findings are discussed with respect to future research and empirically informed practice with high-risk youth.