Journal of Consulting and Clinical Psychology · 1999 · 41 citations · 0 references
MedicinePsychiatryPsychologyForensic PsychiatryPsychiatric AssessmentRisk AssessmentViolenceViolence ManagementViolence Risk AssessmentViolence Risk ManagementClient Violence Risk AssessmentInpatient Violence RiskPredictive ValidityPredictors Of ViolenceRisk Assessment ValidityViolence MeasurementViolence Risk EvaluationViolence Risk Prediction
This study evaluated the predictive validity of the HCR‑20 and the PCL:SV for violence risk among psychiatric patients. The study coded files of 193 civilly committed patients and followed them for an average of 626 days. Receiver‑operator‑characteristic analyses showed the HCR‑20 had strong predictive validity (AUC .76–.80) and that individuals scoring above its median were 6–13 times more likely to be violent, while the PCL:SV had more variable AUCs (.68–.79) and regression confirmed that the HCR‑20 added incremental validity and that only its subscales predicted violence.
This study evaluated the predictive validity of the HCR-20 (Historical, Clinical, and Risk Management) violence risk assessment scheme and the Psychopathy Checklist: Screening Version (PCL:SV). Files of 193 civilly committed patients were coded. Patients were followed up in the community for an average of 626 days. Receiver operating characteristic analyses with the HCR-20 yielded strong associations with violence (areas under curve [AUCs] = .76-.80). Persons scoring above the HCR-20 median were 6 to 13 times more likely to be violent than those scoring below the median. PCL:SV AUCs were more variable (.68-.79). Regression analyses revealed that the HCR-20 added incremental validity to the PCL:SV and that only HCR-20 subscales predicted violence. Implications for risk assessment research, and the clinical assessment and management of violence, are discussed.