A network meta-analysis of psychological interventions for schizophrenia and psychosis: Impact on symptoms.

BACKGROUND:Evidence for the effectiveness of psychological interventions for schizophrenia/psychosis is growing, however there is no consensus on the psychological intervention most likely to reduce symptoms. METHODS:A network meta-analysis was conducted to identify all randomised controlled trials (RCTs) of psychological interventions for adults with schizophrenia/psychosis. A systematic review of the literature using MEDLINE, PsycINFO, EMBASE and CENTRAL led to an analysis of 90 RCTs with 8440 randomised participants across 24 psychological intervention, and control groups. Psychological interventions were categorised and rated for treatment fidelity and risk of bias. Data for total symptoms were extracted and network meta-analysis, using a frequentist approach, was undertaken using Stata SE v15 to compare the direct and indirect evidence for the effectiveness of each psychological intervention. FINDINGS:Psychological interventions were more likely to reduce symptoms than control groups, and one intervention, mindfulness-based psychoeducation, was consistently ranked as most likely to reduce total symptoms. Subgroup analyses identified differential effectiveness in different settings and for different subgroups. INTERPRETATION:Mindfulness-based psychoeducation was consistently ranked as most likely to reduce symptoms; however all studies were based in China. More RCTs in a variety of cultural contexts would help to elucidate whether these findings generalise internationally. A number of psychological interventions could potentially be more effective than interventions recommended by NICE guidelines, such as CBT and family therapy, and additional RCTs and meta-analyses are needed to generate more conclusive evidence in this regard.