Predicting violence in psychiatric inpatients: a prospective study with the HCR-20 violence risk assessment scheme

Abstract

This prospective longitudinal study explores the predictive validity of the HCR-20 and the Psychopathy Checklist: Screening Version (PCL:SV) in a mentally disordered high-risk population. Files of 78 inpatients were coded and patients were followed up while in hospital for 12 months by nursing staff who recorded violent incidents using the Modified Overt Aggression Scale. Receiver operating characteristic analyses yielded moderate to strong associations between HCR-20 and violence (AUCs = .69–.77). PCL:SV AUCs were lower and more unstable (.61–.70). Regression analyses revealed that both tools were good violence predictors in the short-term, but only HCR-20 and particularly the clinical subscale contributed to prediction beyond this period. Patients scoring above the HCR-20 mean were 2.1–2.5 times more likely to be violent than those scoring below the mean. The results suggest that the HCR-20's Spanish adaptation is a useful tool for predicting the likelihood of inpatient violence in civil psychiatric wards. Implications for risk assessment research are discussed.

Keywords:

• HCR-20
• institutional aggression
• mental disorder
• risk assessment
• prospective