Predicting Variables of Violent Behavior in Schizophrenic Patients

Song, Hyun Joo; Min, Sung Kil; Ko, Rye Won; Lee, Nam Hoon; Kim, Tae Kyeung; Oh, Kyung Ja

Abstract

OBJECTIVES
This study aims to investigate the variables predicting violent behavior in schizophrenic patients.
METHODS
Subjects were 65 schizophrenic patients diagnosed according to DSM-IV. Schizophrenic symptom and the insight of patients were evaluated with the Postive and Negative Syndrome Scale (PANSS) and the Scale of Unawareness of Mental Disorder (SUMD). Violent behaviors were evaluated with the Modified Overt Aggression Scale (MOAS). The State Trait Anger Expression Inventory (STAXI) and the Symptom Checklist-90-Reversion (SCL-90-R) self-reported measures were used for the evaluation of anger state & trait, and general psychopathology. Statistic method used in analized variables were stepwise multiple regression and discriminant analysis as well as t-test.
RESULTS
Anger out style, hostility and trait anger variables were extracted as the significant predictors in stepwise multiple regression (37%). Emotion and attitude factors related to anger were more powerful predictor than symptom factors. These three variables could discriminate between violent group and nonviolent group in 72.4%.
CONCLUSION
Violent behavior of schizophrenic patients could be predicted more accurately by stable factors such as attitude and emotion related to anger than by symptom factors.