Comparison of Diagnostic Interview Schedule to psychiatrist diagnoses of alcohol use disorder in psychiatric inpatients

Abstract

This study provides data about the extent of alcohol use disorders among general adult psychiatric inpatients. The accuracy of alcohol use disorder diagnoses given by the lay-administered Diagnostic Interview Schedule (DIS) and by staff psychiatrists, as compared to each other, was investigated. From consecutive admissions to a private psychiatric hospital, 55 patients with alcohol use disorders were identified by trained research assistants using the DIS (n = 162). A comparison of DIS diagnoses to clinicians' diagnoses revealed that 66 patients (40.7% of all admissions) were given an alcohol diagnosis by the DIS or clinician, 35 patients (21.6%) by DIS and clinician, 20 (12.3%) by DIS only, and 11 (6.8%) by clinician only. The two diagnostic approaches were also compared using several accuracy measures (sensitivity and specificity ratios, percentage of agreement, and kappa). With psychiatrists' diagnoses as reference, the DIS sensitivity ratio was 76 and specificity ratio 83. There were 12.3% false positive and 6.8% false negative diagnoses assigned by the DIS. Kappa was .56 and percentage of agreement 80.9%. Recomputing accuracy measures for the psychiatrist, using the DIS as the reference, clinician assessment sensitivity was 64 and specificity 90. Compared to each other, the DIS "overdiagnosed" and the staff psychiatrists "underdiagnosed" by about 1 in 20 cases. Among those alcohol use diagnoses upon which both clinician and DIS agreed, alcohol abuse was the predominant diagnosis. Among all patients with an alcohol diagnosis, the main nonsubstance abuse diagnoses (as assigned by the clinician) were: major depressive, dysthymic, bipolar, schizophrenic, and personality disorders.