Abstract
<title>Abstract</title> <p>Background Health promotion interventions are increasingly recognised as essential strategies for non-communicable disease (NCD) prevention, particularly in low- and middle-income countries (LMICs) where the NCD burden is greatest. However, the successful translation of health promotion programmes into practice depends not only on their design but on how faithfully and effectively they are implemented in real-world community settings. Process evaluation, defined as the systematic assessment of how an intervention is delivered, adopted, and sustained, is critical for understanding implementation fidelity, contextual barriers, and the mechanisms driving programme outcomes. The RE-AIM framework (Reach, Effectiveness, Adoption, Implementation, and Maintenance) provides a structured, multi-dimensional approach to process and outcome evaluation of health promotion interventions. Despite its utility, rigorous RE-AIM-guided, mixed-methods process evaluations of community-based health promotion programmes remain scarce in LMIC settings. This study applies the RE-AIM framework to evaluate the Happy Village Plus (HVP) initiative, a volunteer-led, multi-sectoral community health promotion programme in Sri Lanka, to generate evidence on its reach, effectiveness, adoption, implementation, and early maintenance. Methods A convergent mixed methods process evaluation was conducted across six Grama Niladari (GN) divisions in Sri Lanka’s two most densely populated districts (Colombo and Gampaha). Quantitative implementation data were drawn from 2,192 Kobo Toolbox-recorded activities and analysed descriptively using SPSS and STATA. Qualitative data were obtained from six Health Promotion Officers’ (HPO) logbooks (one per site) and 29 semi-structured key informant interviews (KIIs) with stakeholders including HPOs, volunteers, community leaders, and policymakers. Qualitative data were analysed using inductive thematic analysis in NVivo 15. Findings were integrated and triangulated in accordance with Medical Research Council (MRC) guidelines for evaluating complex interventions. Results HVP delivered 1,005 community-level and 1,187 process-level interventions across six sites, of which 714 community-level interventions directly targeted diet and physical activity, though stakeholder participation varied significantly (8.51%–37.57%). Qualitatively, reach was constrained by systemic community distrust and unhealthy social norms, which were mitigated through locally trusted volunteers. Stakeholder adoption was highest among the Health Promotion and Food Authority (HPFA; 79.83% of deliveries) and community members (engaged in 95% of interventions), while formal public health officers remained under-engaged in direct delivery. Implementation was sustained through multi-sectoral collaboration and culturally sensitive programme adaptation despite material resource constraints. Effectiveness was qualitatively facilitated by community health literacy, primary health care system integration, and the socio-economic context of Sri Lanka’s 2022 economic crisis. Early maintenance was evidenced by community-initiated monthly BMI monitoring activities and institutional commitments to integrate HVP into routine health services. Conclusions HVP demonstrated feasibility and adaptability as a volunteer-led, multi-sectoral NCD prevention model in a resource-constrained LMIC setting. These findings offer transferable lessons on the role of volunteer credibility, multi-sectoral collaboration, and context-sensitive adaptation in sustaining community NCD prevention, and provide a replicable mixed-methods RE-AIM evaluation framework for comparable LMIC settings.</p>