Agreement between self-report and semistructured interviewing in the assessment of personality disorders

This paper presents the results of a study comparing two instruments for the assessment of personality disorders according to the DSM-III-R and the ICD-10 classification systems: the International Personality Disorder Examination (IPDE), a semi-structured clinical interview, and the ''Questionnaire on Personality Traits'' (VKP), a self-report questionnaire. IPDE interviews and VKP self-reports were administered to 108 psychiatric patients and 'normals'. The VKP appeared to overestimate the prevalence of personality disorders when compared with the IPDE: prevalence of personality disorders is about 2.5 times higher when assessed by self-report than by interview. The mean Kappas for DSM and ICD disorders were low, respectively 0.25 and 0.26. These mean Kappas were only calculated if a positive DSM and ICD disorder base rate is at least 5%. The sensitivity of the VKP is 100% for eight of the 22 disorders. For 15 out of 22 disorders, the specificity of the VKP is reasonable to high (greater than or equal to 0.80). The probable and positive diagnoses of the VKP cover 87% of the positive IPDE diagnoses. On the dimensional level high correlations were found between similar scales of the IPDE and the VKP, and low correlations between scales that measure different concepts, pointing to reasonable convergent and discriminant validity. Based on the results of this study it is concluded that the VKP is a suitable screening instrument for most DSM-III-R and ICD-10 personality disorders. Copyright (C) 1996 Elsevier Science Ltd.