An analysis of evolutions in maternal health policies and implementation adaptations in Uganda during the MDG period (2000-2015)
Throughout the Millennium Development Goals (MDG) period 2000-2015, Uganda persistently experienced high maternal mortality. Notwithstanding the several maternal health policy reforms undertaken, Uganda did not achieve its MDG 5 target by 2015. This failure to achieve policy targets has been explained mainly from the epidemiological, interventional and health systems perspectives. The processes that gave rise to such maternal health policies remain underexplored in Uganda. This study examined maternal health policy formulation and implementation in Uganda to underscore how elite interests, policy design and implementation adaptations by street-level bureaucrats influenced policy failure. A three-phase, mixed methods retrospective study was conducted at national and sub-national levels. Study I and II analysed the evolutions of national maternal health policies from 2000 – 2015 (MDG period). Thirty Key Informant Interviews and two Focus Group Discussions were conducted with purposively selected national policymakers. Maternal health policy (n=16) and national programme performance and/or survey (n=21) documents were reviewed. Study III analysed implementation of Emergency Obstetric Care (EmOC) and was conducted in Iganga, Luwero and Masindi districts. Eight doctors and seventeen Midwives who provided EmOC services in Uganda’s public health facilities during the MDG period participated in in-depth interviews. Elite, policy mix and street-level bureaucracy theoretical lenses were applied using thematic, content and descriptive analytical approaches. The findings reveal that fourteen maternal health policy shifts were introduced during the MDG period. The shifts in Maternal Health Policies were driven by a small group of powerful national elites mostly in favour of their covert political and economic interests rather than the stated goal of reducing maternal mortality. Although the policy shifts progressively achieved 87.5% comprehensiveness with respect to the three delays that cause maternal death, they lacked coherence and consistency within the instruments. This arose from policy layering caused by incremental policy change, which led to complex mixes, distortions in existing synergies within and across policies rendering them ineffective in design. Consequently, ideal upstream EmOC policies could not be fully operationalised at the frontline because of incomplete and unreliable package of inputs, supplies, inadequate workforce size and skills mix. Instead, street-level bureaucrats oftentimes improvised to sustain EmOC services. This led to delivery of incomplete and inconsistent EmOC service packages, which were unresponsive to obstetric emergencies. As a result, policies yielded incidental benefit for maternal health as mortality persisted and all national targets missed. Maternal policies should be reformed to reflect actions that address the causes of persistent maternal mortality while ensuring that they are underpinned by a unifying theory of change and responding to bottlenecks in frontline implementation systems. The 3Cs (Coherence, Comprehensiveness and Consistency) framework developed and applied in this study is a useful approach, which policy analysts can use to systematically assess policy design to guide policy reform processes in future. All actors should be held accountable for policy performance.