[Violent behaviour of patients in institutions].

OBJECTIVE

The assessment of violence risk has impact on psychiatrists' decisions not only in forensic psychiatry but concerns many aspects of civil psychiatric units, too. Within the last decade, many studies on violence have been published. The results are misinterpreted sometimes.

METHOD

A review about studies on in-patient violence is given.

RESULTS

In-patient violence and violence in the community should strictly be separated concerning predictors and predictions. Predictors of community violence both for persons with major mental disorders and persons without mental illness are criminal history, male gender, younger age, and substance abuse. Psychopathological and clinical variables play a minor role. In contrast, in-patient violence is closely related to the severity of psychopathological symptoms. More detailed determinations of risk-related symptoms yield conflicting results due to inevitable problems of sample selection.

CONCLUSIONS

The assessment of violence risk has to consider the different risk constellations for violence in institutions and in the community. Predictions of violence by individual risk variables are limited because situation-related factors are important, too.