Community Engagement and Involvement for Lassa Fever Control in Sierra Leone: A Comparative Realist Evaluation

Background: Lassa fever remains a persistent public health threat in Sierra Leone, requiring community participation alongside biomedical interventions. Community Engagement and Involvement (CEI) can strengthen trust, surveillance, accountability, health-seeking behaviour, and local ownership, but its effectiveness may vary according to contextual conditions. Objective: This study examines how CEI operated across contrasting contexts and how contextual conditions shaped mechanisms and surveillance-related outcomes in Lassa fever control. Methods: A realist-informed convergent mixed-methods design was employed in Lower Bambara Chiefdom (Panguma), Kenema District, and Njaluahun Chiefdom (Segbema), Kailahun District. Quantitative data were collected from 130 respondents (74 Lower Bambara; 56 Njaluahun) using a structured questionnaire, complemented by key informant interviews and focus group discussions. Descriptive statistics, χ² tests, non-parametric comparisons, and Firth penalised logistic regression were used. Qualitative data were thematically analysed and integrated abductively with quantitative findings to refine the programme theory. Results: The distribution of reported cooperation-enabling factors differed between chiefdoms, with the main cooperation-enabling factors varying (χ²=10.7, df=3, p=0.014), while reasons for improved cooperation showed a stronger difference (χ²=24.6, df=4, p<0.001). Mutual trust was reported by 55.4% in Lower Bambara versus 19.6% in Njaluahun. Reciprocal accountability was prominent in both sites (64.9% vs. 60.7%). Lower Bambara had substantially greater odds of effective surveillance reporting (AOR=139.72, 95% CI: 20.57–949.08, p<0.001), although the wide confidence interval indicates substantial statistical uncertainty. Funding shortages were the leading CEI challenge (53.8%), while sustained funding was the most important condition for sustainability (46.9%). Qualitative findings showed that trusted mediation, bidirectional communication, community ownership, leadership alignment, and institutional responsiveness were described as important processes through which CEI operated, whereas mistrust and fragmented institutional relationships constrained these processes. CEI effectiveness appeared to be context-dependent. Conclusion: The comparative findings suggest that where contextual conditions supported credible relationships, trust could be linked to stronger community ownership, reciprocal accountability, and cooperation with surveillance, whereas weaker relational and institutional conditions were associated with more conditional forms of participation. The refined programme theory therefore proposes that the contribution of CEI to surveillance-related outcomes depends on the interaction between contextual conditions and relational mechanisms rather than on CEI exposure alone.