L’étude évalue la prévalence du trouble de la personnalité sadique (TPS, DSM III-R) et ses relations avec les critères de la psychopathie selon la PCL-R (Hare, 1991) en milieu psychiatrique sécuritaire. Le TPS est défini à partir de 8 items décrits dans l’annexe du DSM III-R. Ces critères ont été évalués à partir de l’analyse des dossiers cliniques et pénaux ainsi que des informations cliniques collatérales (kappa = 0,80, n = 10). L’évaluation des scores de psychopathie a été réalisée selon les recommandations de Hare. La population est composée de 76 patients masculins internés en Belgique à la suite d’un délit violent (âge moyen = 36,14). Les résultats indiquent une prévalence du TPS de 25 % (n = 19). Parmi les huit items du TPS, les items 1 (use de la violence pour dominer autrui) et 6 (effraie autrui pour obtenir ce qu’il veut) présentent les plus faibles sensibilités. Les corrélations entre les items du TPS et ceux de la PCL-R n’indiquent pas une association significative entre les deux entités diagnostiques. La comparaison des scores de psychopathie (score total, scores aux facteurs narcissique et antisocial) entre patients « sadiques » (n = 19) et « non sadiques » (n = 57) n’est pas significative. L’ensemble des résultats questionne plutôt la validité du TPS selon le DSM III-R auprès d’une population psychiatrique délinquante.

The population was composed of 76 male patients (mean age = 36.14). All of them having committed a violent offence indexed in their institutional file : (1) sexual offences on children ; (2) rapes of adult women ; (3) homicide offence ; and (4) assaults and batteries. TPS was defined by the following 8 diagnostic criteria as described in DSM III-R : 1) has used physical cruelty or violence for the purpose of establishing dominance in a relationship ; 2) humiliates or demeans people in the presence of others ; 3) has treated or disciplined someone under his or her control unusually harshly ; 4) is amused by, or takes pleasure in, the psychological or physical suffering of others ; 5) has lied for the purpose of harming or inflicting pain on others 6) gets other people to do what he or she wants by frightening them 7) restricts the autonomy of people with whom he or she has a close relationship ; 8) is fascinated by violence, weapons, martial arts, injury, or torture. These criteria were assessed from (a) clinical and institutional files and (b) clinical collateral informations. TPS assessment was conducted by two ­trainees in clinical psychology (kappa = 0.87 ; n = 20). The assessment of psychopathy was conducted according to the guidelines of the Hare psychopathy checklist manual (PCL-R, 1991, 2003) : coding of clinical and institutional files and semi-structural clinical interviews. The PCL-R is mainly composed by 2 factors : factor 1 « Emotional detachment » describing the core psychological component of psychopathy, and factor 2 « Chronically antisocial factor » reflecting behavioral instability and antisocial life style. The total cut-off score for the inclusion of the diagnosis was 25. The prevalence of TPS in the population was 25 % (n = 19) and is congruent with the large range described in the literature (0.5 to 33 %). The most frequent criteria were 6 (gets other people to do what he or she wants by frightening them), 1 (has used using physical cruelty or violence for the purpose of establishing dominance in a relationship) and 3 (has treated or disciplined someone under his or her control unusually harshly). The most sensible criteria were : 7 (restricts the autonomy of people with whom he or she has a close relationship), 8 (major interest for violence) and 4 (pleasure in the psychological or physical suffering of others). The most specific criteria were : 3 (has treated or disciplined someone under his or her control unusually harshly), 6 (gets other people to do what her or she wants by frightening them), 4 (takes pleasure in the psychological or physical suffering of others) and 1 (has used physical ­cruelty or violence for the purpose of establishing dominance in a relationship). As concerns psychopathy, the mean of factor 1, factor 2 and the PCL-R total scores were 7.40, 9.08 and 18.67. Thus, 38 % of patients were considered as « low psychopaths», 36 % were considered as « moderate psychopaths» and 26 % were considered as « high psychopaths». In spite of few significant positive correlations between some TPS and PCL-R criteria, TPS diagnosis was not significantly correlated with factors 1, factor 2, nor with total score of the PCL-R. The mean psychopathy total score did not differ between sadistic and non-sadistic patients. Moreover, a two ways ANOVA comparing PCL-R factors 1 and 2 did not reveal any differences between sadistics and non-sadistics. Again, these comparisons did not support hypothesis of a strong association between TPS and psychopathy.