Abstract
BACKGROUND: Cognitive behavioural therapy has been used for schizophrenia, but to which extent it is effective is still controversial. Results of existing meta-analyses are of difficult interpretation, because they mainly present effect sizes in the form of standardized mean differences between intervention and control groups based on rating scales, which are of unclear clinical meaning. No meta-analysis has considered the number of patients responding to treatment yet. Based on this ground, we present the first meta-analysis examining the response rates of patients with schizophrenia and positive symptoms to cognitive behavioural therapy. METHODS: We searched multiple databases for randomized controlled trials on psychological interventions of schizophrenia including patients with positive symptoms, and included for this analysis the studies on cognitive behavioural therapy (last search: January 2018). We applied a validated imputation method to calculate the number of responders from rating scales for the outcomes overall symptoms and positive symptoms, based on two criteria, at least 20% and at least 50% reduction from baseline on PANSS or BPRS total scores. Data were pooled in a single-group summary meta-analysis using R software. Additionally, several potential moderators of response to cognitive behavioural therapy were examined by subgroup and meta-regression analyses. The protocol has been registered in PROSPERO (CRD42017067795). RESULTS: We included 33 studies with a total of 1142 participants receiving cognitive behavioural therapy. On average, 44.5 and 13.2% of the patients reached a 20% (minimally improved) and 50% (much improved) reduction of overall symptoms. Similarly, 52.9 and 24.8% of the patients reached a 20%/50% reduction of positive symptoms. Subgroup and meta-regression analyses revealed a better treatment response in overall symptoms for patients that were not treatment resistant and in studies with researchers' allegiance. Of borderline significance was the better response in studies employing expert therapists and in patients that were more severely ill at baseline. Blinding of outcome assessor, number of sessions, treatment duration, age and gender were not significant moderators of response. CONCLUSIONS: Our findings suggest that adding cognitive behavioural therapy to pharmacotherapy brings about a minimal improvement in overall symptoms among 44.5% of its recipients. Several study and patients characteristics can moderate response rates.