February 1980

Diagnostic Systems and Prognostic Validity

Author Affiliations

University of Maryland School of Medicine Baltimore, MD 21228
Yale Psychiatric Institute New Haven, CT 06510
Theoretical Statistics and Mathematics Branch National Institutes of Health Bethesda, MD 20014

Arch Gen Psychiatry. 1980;37(2):228-229. doi:10.1001/archpsyc.1980.01780150118013

To the Editor.—  we were impressed with the report by Kendell et al in the Archives (36:25-31, 1979), comparing six diagnostic systems according to their prognostic validity. It helped clarify an issue relevant to our 12-point flexible system that we had left unsettled, and it provides an opportunity to comment on two other points.Using two-year and five-year outcome data, we previously reported that several well-known sets of diagnostic criteria using cross-sectional sign and symptom criteria were weak predictors of outcome in patients classed as schizophrenic.1.2 Findings reported by Kendell et al are consistent in most respects. Our 12-point flexible diagnostic system showed a trend (P <.10) toward predictive significance in our study; hence, that Kendell et al found this system to be a significant predictor of symptomatic course in their study suggests the borderline significance in our data may actually reflect some prognostic power. However, it must be

First Page Preview View Large
First page PDF preview
First page PDF preview