Publication: Optimising and personalising task-shared psychosocial interventions for common mental disorders: a Bayesian component network meta-analysis of individual participant data

This study investigated which specific components of psychosocial interventions delivered by non-specialists are most effective for treating depression and anxiety. Using a systematic review and Bayesian individual participant data component network meta-analysis of 34 randomized controlled trials (10,612 participants), the authors dismantled interventions into their active treatment elements and compared their incremental benefits. Strengthening social support, behavioural activation, and problem management emerged as the most effective components, while relaxation and, less clearly, cognitive reframing appeared detrimental when added to other elements. Component effects varied systematically by baseline symptom severity and other sociodemographic characteristics, highlighting the importance of personalisation. Individuals with lived experience contributed to interpreting the findings, and a freely accessible web application allows clinicians and researchers to generate personalised effect estimates for different client profiles.