Présenter d’abord au clinicien expert une échelle construite à partir du test Rorschach, appelée index de contenu traumatique (TCI) par les psychologues américains codifiant le test selon la méthode Exner ou R-PAS, qui n’a jamais fait l’objet d’une publication en langue française. Montrer ensuite au lecteur que cet index est facilement transposable dans le système de cotation Rorschach de l’École de Paris et mettre sa validité à l’épreuve des faits dans le diagnostic de l’abus sexuel chez l’enfant.

Cinquante enfants dont les dossiers d’expertise psychologique pour abus sexuel ont été réexaminés par deux experts ayant conclu à une forte probabilité de la véracité de cet abus. L’index Rorschach de contenu traumatique TCI (Armstrong-Loewenstein, 1990 traduit dans le système de cotation de l’École de Paris pour l’étude portant sur ces 50 sujets âgés de 5 à 12 ans) a été comparé par deux praticiens, en cotation à double insu, avec celui de 50 enfants tout-venant n’ayant jamais consulté. Les échantillons ont été appariés quant au NSE parental.

Le seuil de différenciation relatif à un index égal ou supérieur à 0,25 (soit 25 % au moins de réponses renvoyant à un contenu traumatique) proposé dans la validation anglo-saxonne permet de différencier statistiquement les deux groupes de notre population française en identifiant 72 % des enfants du groupe clinique contre seulement 6 % de faux-positifs dans le groupe témoin.

L’index TCI est pertinent dans la clinique de l’expertise psychologique. Des pistes de recherche sont suggérées.

The article has two aims. The first objective is to present to French clinicians working in forensic field a translation of an index built and validated using Rorschach Inkblot method by Armstrong and Loewenstein (1990), called Trauma Content Index (TCI). TCI was codified inside the Comprehensive System and integrated in recent R-PAS. The Rorschach indicators that constitute TCI, namely responses related to blood, anatomy, sex, aggressive movement and morbid content can be easily translated in French Parisian School system. For us, these indicators are highly relevant because they refer, according to our conception, to the attack of the image of the body and the representation of self which are the first indicators to be degraded in case of sexual abuse. The second objective is to study if American validated TCI cutoff to identify sexual abuse has the same usefulness when applied to a cohort of French children suspected of sexual abuse than in American sample and using Parisian School Rorschach codifying system.

We compared TCI of 50 children aged from 5 to 12 evaluated with Rorschach Inkblot A methodology in double-blinded-scoring where the two experts were both convinced of probability child sexual abuse diagnosis. We have obtained a high fidelity score of interrater agreement (0.92). These data were compared to TCI index of 50 children matched control group with samples stratified according to socio-economic level of parents as well as age. Chi square method was used to compare the two groups.

A TCI cutoff of 0.25 allows us to differentiate statistically the two groups (Chi²: 19.01; P<0.001). Such a cutoff enables us to identify 72 % of the clinical group and only 6 % of the control group as false positives, showing the interest of TCI index applied to children suspected of sexual abuse. We agree with point of view issued from Kamphuis-Kugeares and Finn research (2000) upon adults in their own samples (n=100). That study compared Rorschach records of non-dissociative outpatients with histories of (a) definite sexual abuse, (b) suspected but unconfirmed sexual abuse or (c) non-sexual abuse. They conclude that, although TCI was strongly associated with the presence and severity of sexual abuse, the TCI is one factor among many that can be used to assess the validity of clients’ claims of past sexual abuse. French and South America studies on children with projective tests (Condamin 2006, Mazoyer-Roques, 2014, Scortegagna and de Villemor-Amoral 2012) are more qualitative oriented than Anglo–Saxons researches (Leifer et al., 1991, Kikuchi et al., 2010), but they end up at the same conclusion. That is the reason why we compared our two groups with as second projective tool: the ant's tales test (Royer 1978, de Tychey 2010) used to study child's body image whose usefulness was recently suggested to identify child sexual abuse (Kabuth et al., 2017). Joined use of TCI and indicators issued from that tool allow us to identify 94 % (instead of 72 % with TCI single use) of our clinical group.

TCI and projective tools are useful in forensic field to undertake child sexual abuse diagnosis. It would be important for future research to make comparative studies to approach TCI cutoff scores derived from subjects confronted to various post-traumatic situations and their psychiatric disorders forms. So it would perhaps possible to highlight specific cutoff values to child sexual abuse diagnosis.