Post-traumatic stress disorder assessment with a structured interview: Reliability and concordance with a standardized clinical interview

Abstract

Epidemiologie studies of post' traumatic stress disorder (PTSD) have used the PTSD module of the NIMH Diagnostic Interview (DIS) in its various editions and modifications. Although the diagnoses of numerous disorders made by the DIS or the WHO - Composite International Diagnostic Interview (C1DI), which is modelled on the DIS, have been compared to clinical diagnoses, little is known about the performance of these instruments in diagnosing PTSD. In this study, we examine the test-retest reliability of a modified version of the PTSD section of the DIS-IV and the CIDl 2.1 and compare it with an independently conducted clinical interview in the 1996 Detroit Area Survey of Trauma, an epidemiological study of a representative sample of 2181 persons. A blind readministration of the structured interview was conducted by a lay interviewer 12-18 months after the initial interview, on 32 respondents classified as PTSD cases in the initial interview and on 23 non-cases who reported exposure to trauma. The clinical reappraisal was conducted blindly by two psychiatric social workers, using the Clinician Administered PTSD Scale for DSM-/V (CAPS-DX). The data were weighted to adjust for the oversampling of cases and the differential probabilities of selection of traumatic events across respondents with different numbers of events. The test-retest consistency of the structured interview was a kappa of 0.62 and an odds ratio of 42.5. The comparison of the structured interview with the clinical reappraisal showed agreement in 81% of the assessed sample. Positive predicted value was 0.75, negative predictive value was 0.97, and the odds ratio was 94.8 (all weighted values). Discrepant cases were mostly 'false positives' and, of these, the majority were subthreshold cases missing only one symptom in the CAPS-DX.