Violence Predictors in Psychiatric Inpatients / 신경정신의학

ABSTRACT

OBJECTIVES: The goal of this study was to predict the factors related to psychiatric inpatients' violence in a mental hospital. This was a follow-up study to assess the psychopathology of patients and the quality of the initial therapeutic alliance between the patient and the therapist, as a predictor of the risk of violent behavior. METHODS: The subjects were 105 psychiatric inpatients admitted to St. Andrews' Neuropsychiatric Hospital from January 1998 to November 1998, but alcohol dependent patients and conduct disorder patients were excluded. Within the first five hospital days, we rated past history of violence, patients' psychopathology using Brief Psychiatric Rating Scale and quality of initial therapeutic alliance. We also evaluated violent behaviors using Overt Aggression Scales during the two weeks before admission and during the two weeks of hospital stay and compared the 67 non-violent ingatients with 38 violent ingatients on the utility of socio-demographics variables, past history of violence, utility of the initial therapeutic alliance and psychopathology in evaluating the risk of violent behaviors. Multiple logistic analysis was conducted to analyze risk factors for violent behaviors. RESULTS: 1) The violent inpatients significantly were more closely related with the previous history of violent behavior. 2) The weaker the initial therapeutic alliance were, the significantly higher violent behavior exhibited. 3) Violent behavior was related significantly with higher BPRS-agitation-excitement factor score and higher BPRS-hostile-suspicious factor score. Violent behavior also was related with lower BPRS-withdrawal-retardation factor score. 4) According to the multiple logistic analysis, the predictors of violent behavior were past history of violence and initial therapeutic alliance and BPRS-agitation-excitement factor. CONCLUSION: The results suggest that initial therapeutic alliance and the specific psychopathology ane useful in evaluating patients' risk for violence.