Non-specialist delivered psychosocial care after war and interpersonal violence: a systematic review and meta-analysis

Background: The effectiveness of non-specialist provider-delivered care for mood and anxiety disorders has been established, but effects in violence-exposed populations remain unclear. We examine the net effectiveness of psychosocial treatments delivered by non-specialists to adults experiencing distress following interpersonal and war-related violence. Methods: In this systematic review and meta-analysis, we searched MEDLINE/PubMed, Embase, PsycINFO, Global Health, Cochrane Library, Latin American and Caribbean Health Sciences Literature, and SciELO from Jan 1, 2000, to June 1, 2025. Studies were included if they were psychosocial treatment randomised controlled trials for adults exposed to violence; additionally, the treatment had to be delivered by a non-specialist provider. Pairs of authors double-screened and double-extracted data by applying a codebook developed for this study. We extracted information on data, population characteristics, and four primary outcomes (depression, post-traumatic stress disorder, anxiety, and impairment symptom severity), and used standardised checklists to assess risk of bias. We pooled study-level data in pairwise meta-analyses using a random-effects model to examine subgroup differences by national setting, treatment setting, provider type, violence type, and treatment focus. Our study did not involve people with lived experience of mental health conditions. This protocol was registered a priori with PROSPERO (CRD42022306099). Findings: We screened 54 748 abstracts and identified 45 eligible studies conducted with adults exposed to violence and reporting psychosocial distress. The sample included 9431 participants (mean age 37·9 years; 95% CI 37·7 to 38·1); reporting on sex or gender and ethnicity differences was inconsistent across the included studies. Sample sizes differed by outcome (N=9431 for PTSD; N=9060 for depression; N=8983 for impairment; and N=6545 for anxiety), altogether totalling 34 019 unique data-points. Overall, non-specialist-delivered treatments outperformed control conditions in reducing anxiety (standard mean difference -0·44, 95% CI -0·57 to -0·32; p<0·0001), depression (-0·41, -0·51 to -0·31; p<0·0001), post-traumatic stress disorder (-0·34, -0·44 to -0·24; p<0·0001), and impairment symptoms (-0·34, -0·47 to -0·22; p<0·0001). A similar pattern of effects was seen in the sensitivity analyses (-0·30 [-0·49 to -0·10] to -0·55 [-0·79 to -0·30]; p<0·0001). The strongest treatment effects emerged among refugees. Treatment effects were replicated in most subgroups with a few exceptions: all outcomes for veterans were non-significant, depression and anxiety severity among interpersonal violence survivors did not significantly improve, and transdiagnostic treatments showed a clear benefit across all outcomes. Most studies had some bias (k=44 [98%]) and variability between studies for the main outcomes ranged from I2=77-84%. Interpretation: Small to moderate clinical benefits for non-specialist-delivered care were observed across a broad set of treatments for diverse populations exposed to violence worldwide. These comprehensive analyses can inform psychosocial programming related to setting, providers, types of violence, and treatment foci, to combine treatment and implementation approaches for specific settings and populations. Funding: None.