To evaluate factors associated with the use of restrain and its duration among psychiatrically hospitalized adolescents.

Retrospective, chart-based analysis of patients treated during a 2-year period in Department of Adolescent Psychiatry with average number of patients of 22.6±4.7 and an average admission rate of 320/year.

There were 543 events of restrain during a 2 year period in 164 patients (27/10000 patient days).Restrain was used during 413 (28%) out of 1462 of 12-hour-long analyzed duties. Median duration of restrain per patient was 240 (IQR 90-465) minutes. Factors associated with the risk for restrain in multivariate model were: nurse workload (OR 1.17; 95%CI: 1.09-1.26), fraction of female patients in the unit (OR 3.26; 95%CI: 1.06-10.07), weekday duty (OR 1.67; 95%CI: 1.30-2.13), spring or autumn season (OR 1.51; 95%CI: 1.19-1.93). Child and adolescent psychiatrist on duty (beta= -0.10; p< 0.001) and greater nurse workload (beta= -0.11; p< 0.01) were associated with shorter duration of restrain per patient. Night-time duties were associated with longer use of restrain than day-time ones: 270.0 (25%-75% 92.5-558.8) vs 225.0 (25%-75% 90.0-360.0) minutes. Similarly weekend 292.5 min (120.0-555.0) vs weekday 206.3 min (90.0-427.5) or spring/autumn 295.0 min (25%-75% 105.0-510.0) vs summer/winter 180.0 min (25%-75% 75.0-408.8) duties were associated with highly statistically significant (p< 0.01) longer duration of restrain per patient.

Health service-related factors are important predictors of the use and duration of restrain. Adequate number of well-qualified nursing staff should be consider as means to counter this problem.