A Meta-Analysis of CBT for Medication-Resistant Psychosis

A team of BC psychologists has performed the first meta-analysis of cognitive-behavioral therapy for medication-resistant psychosis. The 16 published studies that met their inclusion criteria comprised 12 trials and 639 individual patients. Medication resistance was defined as inadequate response of positive symptoms to at least one medication at adequate dose and duration, or treatment with clozapine. All the trials entailed assignment to either individual CBT for 10 to 24 sessions, or to a control intervention such as treatment as usual, psychoeducation or befriending. Four trials lacked masked raters. Outcome measures were typically PANSS or BPRS, and most studies had a follow-up assessment 3 to 18 months after completion of treatment. Based on pooled outcome data, effect size was derived with Hedge’s g.

For improving positive symptoms, the effect size of CBT compared to control intervention at the end of treatment was 0.47. At follow-up 3-18 months after treatment, the effect size was 0.41. Among studies with an outcome measurement for general psychopathology, such as depression and anxiety, the effect sizes were 0.52 at treatment end and 0.40 at follow-up. According to the researchers, excluding studies without masked raters did not significantly change the effect sizes.

This meta-analysis yielded a medium effect size for time-limited CBT in medication-treated patients with residual positive symptoms. The results suggest improvement may be maintained beyond a year. These were not necessarily treatment-resistant patients as typically defined, although some were on clozapine. Research on other important outcomes such as hospital admission, psychosocial functioning, and suicide would help determine the place of CBT in managing treatment-resistant psychosis.


Burns AM, Erickson DH, Brenner CA. Cognitive-behavioral therapy for medication-resistant psychosis: a meta-analytic review. Psychiatr Serv. Published online 1 Apr 2014. Abstract