Diagnosis of schizophrenia: a critical review of current diagnostic systems.

The data relevant to the evaluation of six systems for diagnosing schizophrenia are reviewed. They are summarized in terms of the reliability, predictive validity, specificity, and comprehensiveness of each system. Unfortunately, none, of these systems (Schneider's First-rank Symptoms, New Haven Schizophrenia Index, Flexible System, Feighner Criteria, Research Diagnostic Criteria, and DSM-III) have established construct validity. It is noted therefore that they are all, in a sense, arbitrary. Choosing one over another cannot be data-based. Because the elevation of any one diagnostic system to an official status is thought to be premature, clinicians and researchers alike are advised to exercise caution and openmindedness in their use of DSM-III. There is as yet no evidence that its criteria for schizophrenia are either less arbitrary or better (in identifying a group of "true" schizophrenics) than those of other systems or DSM-II.