Validity in the diagnosis of personality disorders: The use of the LEAD standard.

This study examined the reliability, stability, and clinical and predictive validity of Axis II diagnosis in 40 patients with major depression. Subjects were assessed with the Personality Disorder Examination (PDE) and the Personality Assessment Form (PAF). The clinical validity of these measures was determined by examining their agreement with consensual diagnoses formulated using the LEAD (Longitudinal, Expert, All Data) standard proposed by Spitzer (1983). Interrater reliability for categorical decisions was excellent with the PDE but only fair with the PAF; reliability improved on the PAF when assessed using continuous scores. Stability of diagnoses over a 6-month period was moderate with both the PDE and PAF. Concordance among the 3 methods was limited, with unanimous agreement on 20of 40cases(50%) at intake and!8of31 cases(58%) at follow up. Patients with personality disorders identified by the LEAD standard consensus (and to a lesser degree, by the PDE) had worse outcomes at the 6-month follow-up assessment.