The use of restraints on a general psychiatric unit.

A 9-month prospective study of all patients restrained on the acute psychiatric units of two general hospitals and a group of nonrestrained controls revealed significant demographic and diagnostic differences between restrained and nonrestrained patients. Restraint occurred more often in young, unmarried, seriously ill men with a previous history of violent behavior and previous psychiatric treatment. These characteristics were coupled with inadequate neuroleptic treatment and a perception of inadequate power authority (low staffing levels of male orderlies). There was a significantly higher frequency of a history of previous violence in restrained patients compared to nonrestrained patients on the same unit. These data have important clinical and legal implications for persons involved in making predictions of "dangerousness."