Abstract
Decentralisation is expected to make health systems more responsive by moving planning and decision-making closer to communities, yet participation does not automatically improve services. This study examined the relationship between local leaders’ participation in planning and decision-making and health service delivery in Buhweju District, Uganda. A cross-sectional mixed-methods design combined questionnaires, key-informant interviews and focus-group discussions. Of 242 targeted participants, 237 provided usable survey responses (97.9%). Planning was measured with six five-point Likert items (Cronbach’s alpha = .771). Four binary service indicators—medicines, health-worker availability, waiting time and patient-provider relationships—were averaged to form an index ranging from 0 to 1 (alpha = .748). Planning participation was rated very highly (item means 4.54–4.67; composite mean = 4.61), but only 46.0% reported usual medicine availability, 41.4% health-worker availability and 40.5% acceptable waiting time; 78.5% reported good patient-provider relationships. Planning was inversely associated with service delivery (r = −.282, p < .001) and remained the only significant predictor in a four-predictor model (standardised beta = −.269, p < .001; R² = .084). Qualitative accounts described active collaboration but limited translation of priorities into service improvements. Because the data are cross-sectional, the coefficient is not causal. It may reflect reactive planning, restricted decision space or implementation constraints. Decentralised systems therefore need participation that is linked to funded decisions, implementation responsibility and follow-up.