Opening the 'black box' of performance-based financing in the healthcare sector of western Uganda

In response to frustration about the slow (but still significant) improvement of the health care sectors in low- and middle-income countries (LMICs), some 20 years ago several international and domestic policy entrepreneurs put forward the idea of Performance-Based Financing (PBF). PBF can be defined as: a supply-side reform package that is guided towards improved performance … by using performancebased financial incentives for health providers and … most or all of the following elements: a separation of functions (purchasing, regulating, providing, verifying health care services), spending autonomy for the health facilities, strict monitoring and verification of services delivered, community involvement, performance-based planning and accountability arrangements. The introduction of PBF for health sectors in LMICs has sparked a lively debate; however, a thorough understanding of PBF and its theoretical basis is lacking, which jeopardizes this debate. This PhD dissertation therefore sets out to unravel the theory of PBF by studying a PBF intervention by BTC/Enabel in Western Uganda. We devised a methodological strategy that consisted of a combination of realist evaluation research and systems thinking, or, more specifically, causal loop diagramming (CLD). This strategy focused on the mechanisms initiated by the implementation of a PBF intervention, rather than solely looking at the intervention outcomes. This was considered a way of ‘opening the black box’, by which we distinguished seven mechanisms that may be triggered by the intervention and that, in combination, constitute the programme theory: financial incentivisation, non-financial incentivisation, management, knowledge and saliency, financial accessibility, patient feedback and the workload mechanism. In order to analyse these mechanisms, we conducted before and after case studies in two districts of Western Uganda: Kasese and Kyenjojo. Data was collected before the intervention and two years later (after one year of implementation). In total, we surveyed 175 health workers; organized 59 semi-structured and 11 unstructured interviews with health workers; conducted 16 key-informant interviews with high-level officials within the MoH, BTC/Enabel, Catholic and Protestant medical bureaus and key stakeholders at the district level; consulted relevant policy documents; and made observations at 16 health facilities during both baseline and end line study. The results indicate that a range of barriers either impede the triggering of some of the mechanisms or reduce their impact. The many delays, the lack of a coherent communication strategy at different levels of the health system that would adequately and, in a timely fashion, inform health facilities and workers on the intervention itself and on the reasons for the delays, and the lack of a rationalized coverage plan are the most significant barriers. However, we identified three mechanisms as particularly important: the management mechanism, entailing a more active health unit management committee, with more investment in the work environment; the financial accessibility mechanism, meaning that lower user fees lead to more patients being able to attend the facility; and the knowledge and saliency mechanism, enhancing awareness about Ugandan Clinical Guidelines. This research increases insight into the most important barriers and mechanisms at play in the specific BTC/Enabel PBF intervention. Furthermore, it was an opportunity to test an innovative methodological strategy and to learn more about its strengths and limitations. We believe that addressing these limitations will help us to further increase our understanding of PBF theory.