To the Editor.
— Although it is important to note the publication of empirical data concerning the validity of DSM-HI, Helzer et al (Archives 1981;38:791-797) may have too readily dismissed the tautology inherent in using outcome as the validating criteria for the DSM-III and Feighner definitions of schizophrenia.A side-by-side examination of the Feighner and DSM-III criteria demonstrates the similarity between them, a fact confirmed by Helzer et al, who found almost complete concordance. Thus, in light of the numerous previous studies confirming the predictive validity of the Feighner criteria, it is no surprise that DSM-III shows similar predictive validity. The data appear to indicate, however, that for both definitions predictive validity derives only from the six-month duration requirement. Thus, DSM-IIT "schizophreniform disorder" (DSM-III schizophrenia minus the sixmonth duration) is less predictive than the generic diagnosis of psychosis used to assemble the entire Netherne sample.