Pham T, Rémy S, Dailliet A, Lienard L
Unité d'Observation et de Traitement, Prison de Mons, Belgique.
Encephale. 1998 May-Jun;24(3):173-9.
This study examines risks of violent behavior among 62 offenders in a belgian psychiatric security institution. Two variables were assessed as predictor of the subjects' history of violent behavior recorded in their medical files: (1) subjective classification as "violent" vs "non-violent" by care taking personnel; and (2) scores on Hare's (1991) Revised Psychopathy Checklist (PCL-R). The study assessed also the score of the Michigan Actuarial Scale of Violent Behavior (MARSS) which was not considered as predictor. The PCL-R, the subjective classification and the MARSS were found to be highly intercorrelated. Regression analysis suggest that the PCL-R is a significant predictor of violent behavior. The use of this instrument for the prediction of violent behavior is thus encouraged.
本研究调查了比利时一家精神病安全机构中62名罪犯的暴力行为风险。评估了两个变量作为受试者医疗档案中记录的暴力行为历史的预测指标：（1）护理人员将其主观分类为“暴力型”与“非暴力型”；（2）哈雷（1991年）修订的精神病态检查表（PCL-R）得分。该研究还评估了密歇根暴力行为精算量表（MARSS）的得分，该量表未被视为预测指标。结果发现，PCL-R、主观分类和MARSS之间高度相关。回归分析表明，PCL-R是暴力行为的重要预测指标。因此，鼓励使用该工具来预测暴力行为。
