[Validity of the International Personality Disorder Examination (IPDE) questionnaire in a sample of prison inmates].

OBJECTIVE To discover the validity indices and establish the most adequate cut off point when using the IPDE questionnaire on a sample of prison inmates. MATERIALS AND METHODS A transversal study was carried out on a correlatively selected sample of 100 inmates at two prisons in Madrid. Evaluation instruments, a questionnaire for demographic, prison and toxicological data and a complete interview and IPDE assessment questionnaire (version DSM-IV) were utilised as well as a conditional probability study of the IPDE questionnaire with different cut off points based on the use of the IPDE interview as the "gold standard". RESULTS The cut off point of 3 or more non-coincident answers showed low specificity (2.5%) for the presence of one or more personality disorders, and low sensitivity to antisocial (56.7%) and borderline (58.8%) personality disorders. Discusion: The IPDE questionnaire of little use amongst the studied prison population when the habitual reference standards were applied due to the very high number of false positives that were produced. The best validity indices for identifying one or more personality disorders are obtained with a probable cut off point being equal to 4 or more answers that do not coincide with those expected. The IPDE questionnaire was of no great benefit for the inmates in this study because, even when using the habitual cut off point of 3 or more non-coincident questions, sensibility to antisocial and borderline personality disorders, which are the most common PDs amongst the sample group, was found to be low.