Implementation of the user fee exemption policy for caesarean sections in Benin

Introduction
Achieving universal health coverage will require that states have the right capacity to develop and implement appropriate policies that address the most urgent problems. One essential dimension is health financing. Currently, still little is known on how health financing policies are translated into reality for beneficiaries in low-and middle-income countries. The story of the user fee exemption policies (UFEP) in West-Africa provides a good case. Many governments introduced UFEP, promising to reduce catastrophic health expenditure, increase access to services, promote equity, and promote social cohesion. However, such policies are inconsistently reaching their objectives, and little is known of the causes of the failing implementation. With this study, I set out to advance the understanding of how health financing policies (in casu the UFEP for caesarean section introduced in 2009 in Benin) was developed and implemented, with which effect, how, under which circumstances and why.
Methods
I adopted realist evaluation and based my study on the realist research cycle. I developed my initial programme theory based on a literature review using an adapted version of the realist template for systematic reviews. This programme theory had a multi-level structure, including passages from the national to the meso-and micro-level of the health system. I tested this preliminary programme theory in three case studies. Qualitative and quantitative data were collected through in-depth interviews, observation of policy formulation and implementation practices, data extraction from health information records, review of policy and programme documents, and a review of digital media content related to the policy. These data were collected at national, regional and district level, and in seven hospitals. All this allowed me to track the evolution of the policy from 2005 to 2018.
Results
I confirmed that a national health policy passes through three transformation steps as it ‘journeys’ from the primary policy intention to become a programme (administration passage), that has to be adopted at a meso-level (adoption passage) before it is micro-implemented by providers in interaction with the patients or beneficiaries (micro-implementation passage). At each passage, outcomes can vary from the worst (no achievement of the policy intention or negative non-intended consequences) to the best scenario(overachieving implementation outcome). These implementation outcomes can be explained by how actors agree, feel incentivised or forced (through bottom-up and/or top-down pressure) to comply with the policy. I also found mechanisms like trust and self-efficacy at work. The triggering of these causal mechanisms happens in specific configurations of context, actors and policy attributes. Important context components include the legal frame in which the actors navigate, the organisational culture, the ownership status of the hospital, the level of organizational autonomy, the financial priorities of the hospital, the rural or urban nature of the location and the level of organisational trust. Regarding the actors, the background of the policymakers and policy implementers, their engagement vis-à-vis the policy, their perception of their room of manoeuvre, their sensitivity to the voice of communities, and the actual expression of the voice of communities and beneficiaries matter. In terms of the policy, I found that the clarity of the policy intention, the alignment of the policy instruments with the problem it addresses, the perception of the policy intention among policymakers, implementers and beneficiaries, the coherence in the articulation of the policy instruments (procedures and resources) is important. I present in more detail how these factors relate and need to be aligned to sustain (or not) an appropriate implementation of the policy.
Discussion
My multi-level framework of policy development and implementation allowed me to analyse the implementation and its outcomes by system-level and to explore the roles of specific groups of actors. It helped to better understand the configurations of policy attributes, context, actors, and mechanisms that explain the observed implementation outcomes of the UFEP in Benin. My study also points to the importance of the policy implementation capacity of a health system and to how this can be assessed and strengthened through practice and evaluation. Realist evaluation provides ontological, epistemological, and methodological frames for such systematic systemic and continuous learning. In West-Africa, this may have a large impact in terms of strengthening health systems and creating a more favourable space for explanatory and contextually relevant evidence to nourish policies.
Conclusion
If we want to achieve Universal Health Coverage, effectively supporting long-lasting transformational changes in health systems will be essential. Key elements will be centring policymaking and implementation processes more on local actors to empower them and organise continuous contextual learning; to that end realist approaches can contribute.