Diagnostic systems and prognostic validity.

To the Editor.— we were impressed with the report by Kendell et al in theArchives(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