[Homicide by mentally ill: clinical and criminological analysis].

The present study analysed the characteristics of homicide in internees according to the Social Defence system in Belgium. The Social Defence system was inaugurated in 1930, following the ideas of Adolphe Prins, a Belgian specialist in the criminal law. The Social Defence system concerns those offenders who are considered as mentally ill at large. The concept of mental illness encompasses the classical notion of "dementia" but also those people with mental unbalance as personality disordered offenders and mentally retarded persons. In the present study, we considered all those internees who committed a homicide or a homicide attempt and whose cases were examined by the review board between September 1998 and June 2000. We collected 99 cases and studied the age at the time of the offence (mean: 32.2 years), their diagnoses, the status of victims. These 99 murderers committed 111 "facts", a fact being a murder or murder attempt on one or several person(s) without arrest between the different phases of the commitment. These facts provoked 132 victims (72 men and 60 women); 61.36 % deceased. There was no significant difference in the characteristics of homicide versus homicide attempt. The results showed that, in our population, 59.6 % of the patients endorsed a diagnosis of psychosis (37 cases of paranoid schizophrenia, 2 schizophrenias of other types, 7 schizo-affective disorders, 1 autistic disorder, and 12 delusional disorders). Other axis I disorders were 3 intermittent explosive disorders, 2 major depressive disorders, 2 paraphilias and 1 bipolar disorder. The 32 offenders with no axis I major mental diseases presented such diagnoses, alone or comorbid, as antisocial personality disorder (n = 17), borderline personality disorder (n = 8), paranoid personality disorder (n = 4), and schizoid personality disorder (n = 2), 14 were mentally retarded and 5 presented a cerebral damage or an epilepsy. The age of the offender at the first homicide was not significantly different between the different diagnoses: group 1 : schizophrenia-type group (29.63 years) ; group 2 : comorbid diagnoses of schizophrenia and cluster B personality disorder (31.64 years) ; group 3 : cluster B personality disordered people (without psychosis) (27.90 years) ; and group 4 : the mixed group of residual diagnoses (32.63 years). Only the persons with a delusional disorder (group 5) significantly committed their homicide at an older age (47,06 year). We found no significant differences between group 2 (54.55 %), 3 (69.23 %), and 4 (50 %) in the proportion of offenders having substance problems. The group 1 (schizophrenia-type without comorbid personality disorder) presented significantly less problems with substances (13.83 %) than the three former groups. The group 5 (20.0 % prevalence) exhibited only a significant difference with group 3. We distinguished instrumental and emotional violence. Instrumental violence was more represented in the facts committed by group 2, 3 and 4 versus group 5. Group 1 differed also significantly from groups 2 and 3. The status of the victim(s) was divided in: 1) members of the family; 2) specifically known persons (outside the family); 3) specifically chosen victims (chosen, searched or followed because of their status, gender, profession, social role); 4) opportunity victims (victims present at the time and having sufficient characteristics to be attacked, e.g. being a woman, suspected to have some money); and 5) accidental victims. We demonstrated in our sample that victims of murderers in their family or specifically known were more frequently victims of pure psychotic offenders (groups 1 + 5) than of offenders of the mixed group (groups 2, 3 and 4) (21.97% vs 10.61%; 16,67% vs 13.64 %, respectively). On the contrary, specifically chosen (2.27 % vs 8.33 %) and opportunity victims (3.03 % vs 11.36 %) were more frequently attacked by the "mixed group". Accidental murders were almost only committed by psychotics (10.61 % vs 0.76 %), often in the context of a spree murder. In conclusion, we discuss that, from such a biased sample, the interest resides in the study of the comparison of the homicidal behaviour between psychotics (schizophrenia-type or delusional disordered), "mixed" and non-psychotics in terms of age at murder (older in delusional disorder), substance problems (more problems in non-psychotics), motivation (more emotional offences in psychotics) and statutes of victims (more family and known victims in psychotics). We are encouraged to precise our data on a larger sample and a longer period in future studies.