Évaluation du sadisme sexuel sévère dans un population médico-légale. Validation francophone de la Sexual Sadism Scale (SESAS)

Objectives: Sexual sadism is associated with a high risk of sexual violence and general recidivism in sex offenders. However, its evaluation encounters vague diagnostic criteria and uses idiosyncratic methods that require evaluators to infer the individual's motivations and sadistic fantasies. The Sexual Sadism Scale (SESAS) is a cumulative scale that is based exclusively on elements of the crime scene. Both inter-rater agreement and scale reliability of this one-dimensional scale comprising 11 items is high. The items are coded dichotomously based on offenders's files. An individual would be classified as likely meeting the diagnostic criteria if at least 4 of the criteria are present. The objective of our study was the French validation of the SESAS. Methods: From the sex offenders’ population in our High Security Hospital “Les Marronniers”, 62 participants were randomly selected. Participants signed a consent form. Inter-rater agreements were analyzed from two evaluators. Offender's files contained data concerning the offense(s), psychiatric expertises carried out for the courts in charge of the follow-up of the participants, and the psychiatric and social reports concerning the evolution of the patients. Results have been anonymized. Statistical analyzes were performed using the Statistical Package for Social Science (SPSS) software. In relation to the hypotheses of our study, the following analyzes were performed: Pearson's correlation coefficient (inter-rater agreement), Kappa coefficient (inter-rater agreement on items and inter-rater agreement on the established diagnosis), Alpha of Cronbach (Internal Consistency) and Principal Component Analysis (PCA) Results: Of the 62 participants included, three met the diagnostic criteria for sexual sadism (4 or more criteria, 4.84%). The distribution of scores indicates that the 95% of the sample had 2 items or less, and the remaining 5% had 6 or more items. Of the 11 items in SESAS, the Pearson's correlation coefficient (r) is significant (r = 0.76, p < 0.001) and indicates a satisfactory positive association. Alpha of Cronbach is very satisfactory (α = 0.86). The items in the scale tend to represent a good internal consistency, meaning that the scale is in adequacy with the object of its measurement: the latent profile of sexual sadism. From the Principal Component Analysis, A two-factor structure can not be retained. Conclusions: From this study, we were able to propose the French version of the SESAS as a useful scale in the evaluation of sexual sadism in a forensic population. We confirm the dimensional aspect of sexual sadism, as it tends to be described in the DSM-5 which distinguishes paraphilia from paraphilic disorder. Moreover, we show here, following the original studies, that an evaluation of sexual sadism based on the crime scene behavioral indicators allows to limit the bias of the idiosyncratic approaches, marked by the inference of individual's sadistic motivations and fantasies.