Testing four diagnostic systems for schizophrenia.

Abstract

Four diagnostic systems for schizophrenia--New Haven, Feighner et al, Research Diagnostic Criteria, and DSM-III--were tested for reliability, comprehensiveness, concordance with established clinical use, specificity, and predictive validity. The sample consisted of 400 patients admitted to Chestnut Lodge, Rockville, Md, 330 of whom received long-term follow-up assessment by interview an average of 15 years after discharge. Diagnostic system criteria were applied retrospectively to abstracted medical records. Findings replicated recent studies on schizophrenia diagnosis and extended the generalizability of current nosologic systems to chronically ill patients. Results also indirectly supported the validity of applying these systems to medical records. This study demonstrated advantages and disadvantages of each system.