Complete delivery matters: Multilevel complier average causal effect analysis of universal mental health curricula in two randomized controlled trials in U.K. schools.

OBJECTIVE: Complete implementation is assumed in intention-to-treat analyses of universal, school-based mental health literacy curricula, but implementation data routinely reveal program delivery is often incomplete (or in extreme cases, no implementation occurs). Complier average causal effect estimation can rigorously determine whether this makes a difference to intervention outcomes. METHOD: We used multilevel complier average causal effect estimation to establish if intervention compliance modified treatment effects for universal, school-based mental health literacy interventions delivered in two large U.K. cluster randomized controlled trials (Interventions in Schools for Promoting Mental Wellbeing: Research in Education; Approaches for Wellbeing and Mental Health Literacy: Research in Education): Strategies for Safety and Wellbeing (k = 64 primary schools, N = 3,397 pupils, 91% compliers; k = 29 secondary schools, N = 3,692 pupils, 87% compliers), Mental Health and High School Curriculum Guide (k = 84 secondary schools, N = 4,879 pupils, 71% compliers; complete vs. incomplete delivery: effects on help-seeking intentions); and Youth Aware of Mental Health (k = 87 secondary schools, N = 5,408 pupils, 51% compliers; complete delivery vs. no delivery: effects on emotional difficulties). RESULTS: A compliance effect was observed for Youth Aware of Mental Health (d = -0.97), Mental Health and High School Curriculum Guide (d = 0.17), and Strategies for Safety and Wellbeing when implemented in secondary schools (d = 0.24) but not primary schools. CONCLUSION: This study provides important evidence that universal school-based mental health literacy curricula can serve as potentially efficacious programs to reduce emotional difficulties and enhance help-seeking behavior but that effects in secondary schools are largely contingent on, or amplified by, complete session delivery. To maximize impact, research must prioritize development of effective strategies and systems of support to overcome commonly identified barriers to implementation. (PsycInfo Database Record (c) 2026 APA, all rights reserved).