Sources of discrepancy in the comparison of a lay-administered diagnostic instrument with clinical diagnosis

Abstract

This article presents a comparison of the diagnostic results from a lay-administered diagnostic interview for depression with a clinical diagnosis in a community sample. The purpose of the comparison is to assess reasons for the discrepancies between the two diagnostic approaches. The majority of the discrepancies result from inconsistent episode reports at the two interviews. We propose a number of possible explanations for the inconsistencies. An empirical analysis that identifies predictors of consistent vs. inconsistent reports supports a strong influence for respondent recall error. We discuss the implications of our analysis for previous comparisons of lay-administered and clinical diagnostic instruments.