Objective: Use of restraint in acute psychiatric wards is highly controversial. Knowledge is limited about the characteristics of patients who are restrained and the predictors of use of restraint. This study examined whether restrained patients differed from nonrestrained patients in demographic, clinical, and medicolegal variables and to what extent the variables predicted use of restraint.
Methods: A two-year retrospective case-control design was used. The sample comprised all restrained patients (N=375) and a randomly selected control group of nonrestrained patients (N=374) from three catchment-area-based acute psychiatric wards in Norway. Data sources were restraint protocols and electronic patient files.
Results: The restrained patients were significantly younger and more likely to be men, to reside outside the wards' catchment areas, and to have an immigrant background. Restrained patients also had more admissions and longer inpatient stays than nonrestrained patients and were more likely to be involuntarily referred and to have one or more of the following ICD-10 diagnoses: a substance use disorder, schizophrenia or a related psychotic disorder, and bipolar disorder. Binary logistic regression analyses, adjusting for age, gender, immigrant background, and catchment area, indicated that the number of admissions, length of stay, legal basis for referral, and diagnosis each independently predicted the use of restraint. No interactions were found.
Conclusions: Use of restraint was predicted by multiple admissions, long inpatient stays, involuntary admission, and serious mental illness. Identifying patients at risk may inform the development of alternatives to restraint for these patients.