Psychiatric nurses' thoughts and feelings about restraint use: a decision dilemma.

Patients continue to be physically restrained in psychiatric in-patient units. Studies concerned with staff-related variables have suggested that the emotional reactions of professionals to violent or potentially violent patients may influence their use of restrictive measures. However, no research existed that described psychiatric nurses' thoughts and feelings while they were involved in restraint situations nor what effects their thoughts and feelings had on their decision to restrain. Therefore, an ethnographic qualitative study was conducted in order to describe systematically nurses' thoughts and feelings toward restraint use in the in-patient psychiatric setting. The conceptual approach guiding the study was Etzioni's (1992) theoretical work on the role of normative-affective factors in decision making. Following ethical approval of the study, ethnographic interviews were conducted with six nurses from an in-patient psychiatric unit who had participated in a situation involving the physical restraint of a patient. The analysis of the nurses' thoughts and feelings revealed that the restraint situation represented a decision dilemma for them. This overall finding was supported by four themes: (1) the framing of the situation: the potential for imminent harm; (2) the unsuccessful search for alternatives to physical restraints; (3) the conflicted nurse; and (4) the contextual conditions of restraint. The results indicated that restraint use is more complex than is currently conveyed in the literature in that normative-affective factors influenced nurses' restraint decisions. The findings advance our understanding of why restraints continue to be used in psychiatric units. Further research is necessary to examine the findings in other settings and with a larger and more diverse population in order to draw definitive conclusions about the continued use of physical restraints in the care of patients on psychiatric units in hospitals.