DSM-III personality disorders: Diagnostic overlap and internal consistency of individual DSM-III criteria

Abstract

We rated the presence or absence of every DSM-III personality criteria in a cohort of 131 non-psychotic subjects. Ratings were based on the Structured Interview for DSM-III Personality Disorders (SIDP) which was administered to each patient and a knowledgable informant. Diagnositc overlap of the personality disorders (PD) was examined. Sensitivity, specificity, predictive value—positive and negative, and interrater reliability was calculated for each criteria item of each personality disorder. Problems in reliability of DSM-III personality disorders can be traced to specific criteria within those disorders that are associated with low reliability and/or low predictive value positive when compared to the other criteria used to define the personality disorder. Improvements are suggested. We were unable to demonstrate a significantly greater enhancement of reliability among criteria structures using a polythetic classification over those using monothetic classification despite a trend favoring the former. This study agrees with others that find a great deal of overlap between borderline PD and histrionic PD as defined in DSM-III. Passive-aggressive PD is very rare in the absence of some other PD. Paranoid and schizoid personality disorders were also rare though this may reflect the fact that individuals with these disorders rarely seek treatment.