[Schizophrenia and violence, incidence and risk factors: a Tunisian sample]

Schizophrenia appears to be the mental pathology the most associated with violence. The aim of this study is to show the incidence and the different risk factors of violence among schizophrenics.

We have compared a group of 30 violent schizophrenic inpatients with another group of 30 nonviolent schizophrenic inpatients hospitalised during the same period. These two groups have been matched according to age and gender. The comparison concerned: sociodemographic parameters, family and personal psychiatric history, legal antecedents, social insertion, clinic, Clinical Global Impressions (CGI), Global Impairment Scale (GIS) and Positive And Negative Syndrome Scale (PANSS) scores for admissions, familial support and insight, compliance to treatment, administered treatments, and awareness degree.

Violent schizophrenics represent 18.07% of all hospitalisations and 26.08% of schizophrenic patients. When compared to violent schizophrenic patients, nonviolent schizophrenic patients have a better socioeconomic level (77% versus 43%), better professional adaptation (67% versus 10%) and familial support (60% versus 10%), better insight (87% versus 23%) and therapeutic control (70% versus 17%). Differences are significant. We found significantly more personal antecedents of inflicted violence within violent schizophrenics (50% versus 13%), more addictive behavior (53% versus 13%), and more paranoid and indifferentiated forms (87% versus 47%) than in nonviolent schizophrenics. The average of CGI scores was significantly higher within violent schizophrenics (5.27+/-0.8 versus 3.77+/-0.5). Conversely, the average of EGF scores was lowest (37.6+/-6.5 versus 47.8+/-5.6). The comparison of PANSS scores revealed that violent schizophrenic subjects are characterised by the existence of more positive signs and more general symptoms (34.4+/-4.7 versus 20.2+/-4.5; 55.1+/-11.4 versus 46.1+/-6.9). Violent schizophrenics are characterised by higher neuroleptic doses (2375+/-738 mg/d versus 1610+/-434 mg/d). Differences here are also significant.

Addictive behaviour seems to considerably increase the risk of turning to violence. Thus in our study, 53% of violent patients showed an addictive behaviour. These results have also been reported by other authors. It is obvious that alcohol and drug abuse double the risk of violence among schizophrenic subjects. Psychotic decompensation and rich symptomatology increase the violent potential among the schizophrenics. In our study, the PANSS scores were higher among violent subjects. Nonviolent schizophrenic subjects have a lesser symptomatology of psychiatric disorders and a better outcome as shown by the CGI and EGF scores. In our study, the group of violent subjects needed higher neuroleptic doses and were noncompliant. Compliance permits the acquisition, and then maintains, the stability of the mental status and plays an essential role in decreasing dangerousness. In fact, violent schizophrenics exhibit low insight, implying diminished awareness of the legal implications of their acts, and are little aware of their illness and its dangerousness. In our study, we noted better familial support among nonviolent subjects. According to the literature, violent schizophrenics are characterised by a particularly hostile and rejecting familial environment.

Awareness of these factors will allow us to provide improved prevention of violence within schizophrenic subjects.