Accuracy of the Historical, Clinical and Risk Management Scales (HCR-20) in predicting violence and other offenses in forensic psychiatric patients in Brazil.

Résumé

BACKGROUND: Assessing the risk of violence is a complex task. In Latin America it is often based on clinical criteria that are not very objective or structured. HCR-20 has been used to increase the accuracy of this exam.AIMS: The aim of this study was to examine the predictive validity of the Historical, Clinical and Risk Management Scales (HCR-20) violence risk assessment scale on a sample of Brazilian male forensic psychiatric inpatients.METHOD: A concurrent prospective cohort design was used. The cohort was selected among the population of inpatients in Unit D (N=68) at Instituto Psiquiátrico Forense Mauricio Cardoso (IPF), Brazil. For the baseline assessment the following instruments: HCR-20-Assessing Risk for Violence, Version 2, and Hare Psychopathy Checklist, Revised (PCL-R) were used. During the one-year follow up, episodes of violent and/or anti-social behavior were assessed, and recorded on the Yudofsky's Overt Aggression Scale (OAS) and Tengström et al.'s Follow-Up Questionnaire. The accuracy of HCR-20 and PCL-R to predict violent and/or anti-social behavior was assessed.RESULTS AND CONCLUSIONS: For the whole cohort, the mean total score of PCL-R was 13.54 and of HCR-20 it was 23.32. The rate of recidivism in the twelve month follow up was 73.5%. Outstanding among the risk factors explored for their predictive efficacy are scale HCR-20 and subscale H for any event, and scale HCR-20 for a violent event. The predictive efficacy of scales HCR-20 and PCL-R was greater for any antisocial event than for a violent event. By taking into account the possibility of recidivism and the probability of recidivism accumulated over time, instruments HCR-20 and PCL-R behaved as expected. In all these explorations, the instruments significantly differentiated the group of the sample that recidivated earlier.