B. Gouvernet

Validation d’une version française de la forme brève de l’inventaire des troubles de la personnalité pour le DSM-5 (PID-5 BF) de Krueger

OBJECTIVE Since the publication of the DSM-5 (APA, 2013), the dimensional conception of the personality disorders is co-existing with the classical categorical paradigm. Tools have been proposed for the evaluations of five big pathological factors to be explored further according to the APA (negative affectivity, detachment, antagonism, disinhibition, psychoticism). Despite numerous works using these questionnaires (30 works in 3 years according to Al-Adjani et al., 2015), none of them have yet been translated into French. Also, the main objective of the paper is to present a French translation of the Personality Inventory for DSM -5 by Kruegger et al. (2013) in its brief form of 25 items (PID-5 BF). METHOD To reach this goal, we have employed the classic translation-retranslation method (Vallerand, 1989) and tested the consistence and the validity of this French version among a non-clinical sample (n=216) of young adults (age=31.4, SD=4.8), in joining some other questionnaires in their short forms to study the external validity of the PID-5 about the psychological distress (SCL-10, Nguyen, 1983), the categorical diagnosis of personality disorders (SAPAS, Moran et al., 2003) and the classical Big Five dimensions of the personality (BDI 10, Ramamstedt and John, 2007). The internal consistency of this translation has been studied through the classical outcomes on factor analysis for the dimensional repartitions of the items in 5 scales and Cronbachs alpha for the consistency of each found dimensions. The external validity has been explored by studying Pearsons correlations between the outcomes on each dimension of the PID-5 BF and both the clinical dimensions of SCL-10, personality dimensions of the BFI-10 or personality disorders (SAPAS). RESULTS Factor analysis led to the same repartition of the 25 items as the original versions. Each of the dimensions is consistent enough (α>.65) to be taken into account as clinically significant. The items of the French version of the PID-5 BF follow the expected repartitions in 5 dimensions, which are consistent enough. Although their mean scores are significantly not different from the outcomes found by Krueger with the PID-5 200 items among another non-clinical population (n=264), one cannot say that is enough to ensure the external validity of our translation, for it uses neither the same tools nor sample. A comparison with a French translation of the PID-5 would be more significant. However, the external validity of the French version seems to be significant enough. Global score on the PID-5 is correlated both to the Global Severity Index of the SCL-10, which reflects global psychological distress, and SAPASs score, which evaluates the suspicion of personality disorder. The clinical validity of the PID-5 is confirmed by the relationships between negative affectivity and anxiety or depression or antagonism and hostility, although the clinical scale of the SCL-10, with one item by dimension, is less sensitive than the complete original version in 90 items (DeRogatis, 1974). PID-5 score and domains are also correlated with the Big Five personality dimension and global score of personality disorders which led us to think that it is coherent with the evaluation of personality suffering (r=.34) and dimensions. The links between negative affectivity and neurosism (r=.48) or between desinhibition and extraversion (r=.32) or the negative correlation between psychoticism and conscientiousness (r=-0.16) are consistent with the expectations related both to the descriptions of the domains by the DSM and outcomes on the comparisons between PID-5 200 item scales and NEO-PI or BFI 45 items. DISCUSSION This translation offers enough consistency and validity to be used in future studies. This could lead us to either continue studying a more representative general population or testing its validity in focusing on a clinical sample where personality disorders are prevalent, such as homeless men or substance users. As soon as a French version of the PID-5 200 items is published, one can compare the outcomes on PID-5 BF and PID-5 to lead to estimations of personality disorders and pathological domains among French populations and explore personality disorders throughout a dimensional paradigm instead of syndromic perspective. One can also see whether the items that have been kept for each dimension are as saturated in the French version as in the original one. Among general populations, comparisons with clinical distress, syndromic personality disorders or dimensional aspect of personality could be done with complete versions of PID-5, Symptom Check-list, Personality Disorders Questionnaires or Big Five Inventory; therefore, the brief forms of any questionnaire could be used among any people whose psychological distress or side effects impaired their attention and concentration.