Treatment outcome evaluation methodology with schizophrenics

Recent developments in research and treatment of schizophrenia have important implications for assessment of outcome of behavioral interventions. Major measurement domains were reviewed, and a number of general issues were discussed. Diagnosis, derived from structured interviews, has become a sine qua non for well-designed studies. Self-report does not have a major place as an outcome measure with this population, except in the context of interviewer rating scales. Behavioral observation is also used infrequently, primarily because of cost and clinical utility. In contrast, data derived from the family, either through behavioral assessment or interview, play an increasingly important role. Perhaps the most critical problem for evaluating behavioral interventions is the fact that medication and other components of compound treatment programs will invariably account for a significant portion of the overall outcome variance. Several suggestions are provided for dealing with this and other, related issues.