Objective: Increasingly, epidemiological studies are employing computerized and highly standardized interviews, such as the Composite International Diagnostic Interview (CIDI-Auto), to assess the prevalence of psychiatric illness. Recent studies conducted in specialist units have reported poor agreement between experienced clinicians' and CIDI-Auto diagnoses. In this study we investigated the concordance rate between clinicians and the CIDI-Auto for the diagnosis of six anxiety disorders and two mood disorders, whereby the CIDI-Auto was treated as the 'gold standard'.
Method: Subjects were 262 patients who were assessed by a clinical psychologist or psychiatrist and completed the CIDI-Auto at a tertiary referral unit for anxiety and mood disorders. Agreement between the clinicians' diagnoses and the diagnoses generated by the CIDI-Auto according to both DSM-IV and ICD-10 codes, were examined by kappa statistics. Sensitivity and specificity values were also calculated.
Results: Agreement between clinicians and the CIDI-Auto (DSM-IV) ranged from poor for social phobia and posttraumatic stress disorder (kappa < 0.30) to moderate for obsessive- compulsive disorder (OCD; kappa = 0.52). Agreement between clinicians and the CIDI-Auto (ICD-10) ranged from poor for major depressive episode (kappa = 0.25) to moderate for OCD (kappa = 0.57). With the CIDI diagnosis treated as the gold standard, clinicians' diagnoses showed low sensitivity (kappa < 0.70) for all the disorders except for OCD (for ICD-10), but high specificity (kappa > 0.70) for all the disorders.
Conclusion: Poor agreements between experienced clinicians and the CIDI-Auto were reported for anxiety and mood disorders in the current study. A major implication is that if diagnosis alone directed treatment, then patients could receive different treatments, depending on whether the computer, or a clinician, made the diagnosis.