Clinical and psychometric validation of the psychotic depression assessment scale.

BACKGROUND:Recent studies have indicated that the 11-item Psychotic Depression Assessment Scale (PDAS), consisting of the 6-item melancholia subscale (HAM-D6) of the Hamilton Depression Rating Scale and 5 psychosis items from the Brief Psychiatric Rating Scale (BPRS), is a valid measure for the severity of psychotic depression. The aim of this study was to subject the PDAS, and its depression (HAM-D6) and psychosis (BPRS5) subscales to further validation. METHODS:Patients diagnosed with psychotic depression at Danish psychiatric hospitals participated in semi-structured interviews. Video recordings of these interviews were assessed by two experienced psychiatrists (global severity rating of psychotic depression, depressive symptoms and psychotic symptoms) and by two young physicians (rating on 27 symptom items, including the 11 PDAS items). The clinical validity and responsiveness of the PDAS and its subscales was investigated by Spearman correlation analysis of the global severity ratings and the PDAS, HAM-D6, and BPRS5 total scores. The unidimensionality of the scales was tested by item response theory analysis (Mokken). RESULTS:Ratings from 39 participants with unipolar psychotic depression and nine participants with bipolar psychotic depression were included in the analysis. The Spearman correlation analysis indicated that the PDAS, HAM-D6 and BPRS5 were clinically valid (correlation coefficients from 0.78 to 0.85, p<0.001) and responsive (correlation coefficients from 0.72 to 0.86, p<0.001) measures of psychotic depression. According to the Mokken analysis, all three scales were unidimensional. CONCLUSIONS:The clinical validity, responsiveness and unidimensionality of the PDAS and its subscales were confirmed in an independent sample of patients with psychotic depression.