Health Facility Committees (HFCs) made of elected community members are often presented as key for improving the delivery of services in primary health-care facilities. They are expected to help Health Facility (HF) staff make decisions that best serve the interests of the population. More recently, Performance-Based Financing (PBF) advocates have also put the HFC at the core of health reform, expecting it to hold HF staff into account for the HF performances and development. In Burundi, a country where PBF is implemented nationwide, a randomised control trial was implemented in 251 health facilities where the HFC had been largely inactive in recent years. A random sample of 168 HFCs was trained on their roles and rights, with a subset also given information about the performance of their HF (using PBF indicators) and the PBF approach in general. The interventions, taking place in 2011-2013, made the HFCs better organised but largely failed to generate any effect on HF management and service delivery. Nested qualitative analysis reveals important tensions between nurses and HFC members that often prevent further change at the HF. In the HFs that received both the training and information interventions, this tension appeared exacerbated: the turnover of chief nurses was significantly higher as the HFCs exerted pressure to remove them. This situation was more likely to happen if the HFC had already received training before the interventions, thereby suggesting that repeated training empowers committees. Overall, the results provide rare rigorous evidence on HFCs, suggesting that more attention needs to be paid to the socio-economic and cultural contexts in which they operate. They also invite to caution when discussing the role of HFCs as a possible watchdog in PBF schemes.

Several unresolved questions remain, concerning the integration of vertical programmes and the sustainability of the system given the considerable costs, since funding is not yet fully ensured by the State and its partners.

Community governance, the direct (co-)management of public services by community members, is a popular approach to improve the quality of, and access to, healthcare services–including in so-called ‘fragile’ states. The effectiveness of such approach is, however, debated, and scholars and practitioners have emphasised the need to properly reflect on the contextual features that may influence social accountability interventions. We study a randomised intervention during which community-elected health facility committee members were trained on their roles and rights in the co-management of primary healthcare facilities. 328 publicly-funded health facilities of Burundi and Sud Kivu in DR Congo were followed over a period of one year. In Kivu, but not in Burundi, the intervention strengthened the position of the committee vis-à-vis the health facility nurses and affected the management of the facility. HFC members mostly focused on improving the elements most accessible to them: hiring staff and engaging in basic construction and maintenance work. Using survey data and interviews, we argue that part of the discrepancy in results between the two contexts can be explained by differences in health facilities’ management (whether they primarily depend on a local church or more distant authorities) as well as different local histories of relationship to public service providers. The former affects the room available for change, while the latter affects the relevance of the citizens’ committee as an acceptable way to interact with healthcare providers. No effect was found on the perceived quality of and access to services, and the committees, even when strengthened, appear disconnected from the citizens. The findings are an invitation to re-think the conditions under which bottom-up accountability mechanisms such as citizens committees can be effective in ‘fragile’ settings.