Abstract
To the Editor.— The National Institute of Mental Health Diagnostic Interview Schedule (DIS) was described by Robins et al in the April 1981 issue of the Archives (1981;38:381-389). The DIS was designed for use by lay interviewers. Computer programs have been developed to classify subjects diagnostically. The computer rules are based on three sets of criteria (DSM-III, Feighner criteria, and Research Diagnostic Criteria [RDC]). I am fully sympathetic to the need for a procedure suitable for use in large-scale epidemiological studies that will yield indexes of different types and severities of mental disorder rather than simple frequencies of symptoms or levels of impairment. However, I have serious reservations about the ability of the DIS to produce computerized diagnoses that will be equivalent to diagnoses made by clinicians using the RDC or DSM-III criteria as they originally were designed to be used. The article refers to evidence of the validity of

This publication has 0 references indexed in Scilit: