Abstract
IMPORTANCE: There is currently no evidence-based method to identify the forces that may drive someone to attempt suicide. OBJECTIVE: To examine whether cognitive behavior therapy (CBT) interventions that incorporate a narrative assessment are associated with a greater reduction in suicide attempts than comparable interventions without this component. DATA SOURCES: Studies were identified through version 25.0.1 of the Metapsy Suicide Prevention Database, which includes randomized clinical trials published up to April 2025 retrieved through PubMed, Embase, Web of Science, Scopus, and Cochrane Central, including unpublished studies and references from relevant articles. STUDY SELECTION: Included studies were randomized clinical trials of CBT interventions reporting suicide attempts as an outcome. Studies using waiting list controls, reporting only suicidal ideation, or lacking information on the type of assessment used were excluded. DATA EXTRACTION AND SYNTHESIS: Two reviewers independently extracted data and assessed risk of bias using the Cochrane Risk of Bias 2 tool. Meta-analyses were conducted using 3-level models with robust variance estimation. Relative risks (RRs) were pooled using the Mantel-Haenszel method. Analyses followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses reporting guidelines. Data were pooled using a random-effects model. MAIN OUTCOMES AND MEASURES: The primary outcome was the incidence of suicide attempts. Interventions were grouped by presence or absence of a narrative assessment component. RESULTS: Twenty-three studies with 3262 participants met inclusion criteria. CBT interventions including a narrative assessment were associated with a significantly reduced risk of suicide attempt compared with controls (RR, 0.68; 95% CI, 0.53-0.87; 1764 participants across 14 studies), whereas CBT interventions without this component were not associated with risk of suicide attempt (RR, 1.17; 95% CI, 0.63-2.20; 1498 participants across 9 studies). Subgroup comparison indicated a significant difference between groups (Q1 = 7.27; P = .007; I2 = 86%). Studies without a narrative assessment had significantly younger participants, a lower event rate, and slightly higher risk of bias. CONCLUSIONS AND RELEVANCE: In this systematic review and meta-analysis, CBT interventions including a narrative assessment were associated with a reduced risk of suicide attempt, while CBT without this component did not have an association with risk of suicide attempt. Age differences between study populations may partly explain this finding, as interventions in younger populations often show smaller and more inconsistent effect sizes. These results suggest that a narrative assessment may be a simple and effective way to capture the forces that lead to suicide attempts and to direct interventions toward their prevention. Causal conclusions require direct head-to-head trials.