Cognitive-behavioral therapy for medication-resistant symptoms.

Cognitive-behavioral therapy for psychosis is described. It draws on the cognitive models and therapy approach of Beck and colleagues, combined with an application of stress-vulnerability models of schizophrenia and cognitive models of psychotic symptoms. There is encouraging evidence for the efficacy of this approach. Four controlled trials have found that cognitive-behavioral therapy reduces symptoms of psychosis, and there is some evidence that it may contribute to relapse reduction. Studies that have examined factors that predict treatment response are reviewed. There is preliminary evidence that a good outcome is partially predicted by a measure of cognitive flexibility or a "chink of insight." People who present with only negative symptoms may show poorer outcome. However, there is no evidence that intelligence or symptom severity is associated with outcome. Implications for selecting patients and for optimal duration of treatment are discussed. Finally, the importance of taking account of the heterogeneity of people with psychosis, so that individual treatment goals are identified, is discussed.