Actuarial risk assessment in sexually motivated intimate-partner violence.

Résumé

The present study is the first independent cross-validation of the Ontario Domestic Assault Risk Assessment (ODARA) and the Domestic Violence Risk Appraisal Guide (DVRAG) using an incarcerated high-risk sample (N = 66) of offenders released from the Austrian Prison System who have committed at least one sexually motivated offense against their actual or former intimate partners. The mean follow-up period was approximately 55 months. Both instruments showed evidence for their reliability and predictive accuracy, supporting the cross-cultural transferability of these risk assessment instruments. For the prediction of domestic violence recidivism, ODARA and DVRAG yield good predictive accuracy (area under the receiver operating characteristic curve, AUC = .71), and for general criminal and general violent recidivism, both instruments exhibit moderate effect sizes (AUC = .66-.71). Also, the results provide evidence for the discriminant validity of the ODARA. When examining the association between individual ODARA items and recidivism, only a few items were found to be related to domestic violence recidivism. The integration of the Psychopathy Checklist-Revised (PCL-R) does not add any incremental predictive accuracy to the ODARA, suggesting that ODARA items capture antisocial and psychopathic traits sufficiently even in incarcerated high-risk offenders.