TrialsgrambetaRCTs are beautiful

Jauhar CBTp

2014

Meta-Analysis of CBT for Psychosis with Bias Assessment

Across 52 RCTs of CBT for psychosis, CBTp produced small reductions in positive and overall symptoms relative to active and inactive controls.

Result

Primary endpoint: Symptom-level effect sizes

Overall symptoms g=−0.33 (95% CI −0.47 to −0.19); positive g=−0.25 (−0.37 to −0.13); negative g=−0.13 (−0.25 to −0.01). Masking moderated: overall −0.62 unmasked vs −0.15 masked (p=0.001); positive −0.57 vs −0.08 (p<0.001). Control intervention did NOT moderate effect.

Original publication
Subtopic
Meta-analysis
Design
Meta-analysis with bias assessment
Enrolled
34 patients
Relevance
Important: one of several pillars.
Source
Jauhar S, et al. Br J Psychiatry. 2014;204:20-29.doi 10.1192/bjp.bp.112.116285

Discussion & critique

Calibrates expectations: CBTp's specific effect is modest after methodological adjustments.