Risk of Psychiatric in-patient Violence: A Forensic Perspective

Objective: to assess the accuracy of Brøset violence checklist in predicting violence in psychiatric inpatients in a Pakistani setup. Study Design: Cross-sectional study. Place and Duration of Study: Department of Psychiatry, Mayo Hospital, Lahore Pakistan, from Mar to Sep 2021. Methodology: One Hundred and Sixty-Seven patients from either gender, admitted to an acute psychiatric ward, were included in the study and their scoring was done. Socio-demographic information was collected from patients’ files, and violence data and preventive measures were recorded on the Staff Observation Aggression Scale-Revised form by nursing staff for three days. SOAS-R score of 9 or more was declared as a violent incident. Results: Of the one hundred and sixty-seven patients, the mean age was 34.48±10.06 years. 109(65.3%) were males, and 58(34.7%)were females. Forty-nine episodes of violence were recorded. At a cutt off point of 2, Brøset violence checklist Sensitivity and Specificity were 69.4% and 92.4%, respectively. The corresponding PPV and NPV values were 79.1% and 92.4 %, respectively. ROC characteristics yielded an area under the curve of 0.86, showing good predictive accuracy. Conclusion: Brøset violence checklist has good predictive accuracy for the prediction of imminent violence in a psychiatric setting in Pakistan.