Involuntary hospitalization and violent behaviors: medical act or social control? A 3-Year Retrospective Analysis.

BACKGROUND: The present retrospective study is aimed at assessing the clinical and psychopathological correlates of violent behaviors in a sample of acute involuntary committed inpatients. SUBJECTS AND METHODS: Involuntary inpatients were retrospectively assessed for the presence of violent behaviors. Patients with and without overt hetero-aggressive behaviors were compared according to socio-demographic, clinical and psychopathological features. A stepwise backward logistic regression was performed in order to assess the variables most associated with the perpetration of violent acts. The sample of violent patients was then divided in two subgroups on the basis of the presence/absence of a serious mental illness (SMI). Bivariate analyses were performed between SMI and non-SMI violent patients. RESULTS: In the present sample of 160 inpatients, 88 (55%) perpetrated violent acts. Subjects who performed violence presented a higher rate of mood stabilizers prescription (p=0.038). The PANSS-excited component was positively associated with violent behaviors (p=0.027, Odds Ratio (OR)=1.14, Confidence Interval (CI) 1.01-1.28), whilst the PANSS-depressed/anxiety factor displayed a negative association (p=0.015, OR=0.78, CI 0.64-0.95). Violent inpatients diagnosed with SMI presented higher rehospitalization rate (p=0.009), longer length of stay (p=0.005), more frequent long-acting injectable antipsychotics prescription (p<0.001) and a higher positive symptoms severity as measured by the PANSS-positive factor (p=0.049). CONCLUSIONS: The clinical population of acute psychiatric inpatients performing violent behavior represents a specific and heterogeneous subgroup of patients for which prevention and treatment strategies should be addressed.