Abstract
Current concepts of schizophrenia and recent changes in diagnostic practice, multidimensional outcome assessment, and predictive methodology are reviewed. The highly successful results of earlier studies are not borne out in recent work which indicates that only about 8–40% of the outcome variance can be predicted using multivariate analysis of symptoms or groups of symptoms. Other influences during the illness which may significantly modify the outcome include the economic situation, life events, social class, the family, extrinsic handicaps, migration and culture. These factors have been inadequately considered in prognosis research, and are probably not considered sufficiently in therapy.