Despite growing literature, few studies have explored the implementation of policy interventions to reduce maternal and perinatal mortality in low- and middle-income countries (LMICs). Even fewer studies explicitly articulate the theoretical approaches used to understand contextual influences on policy implementation. This under-use of theory may account for the limited understanding of the variations in implementation processes and outcomes. We share findings from a study exploring how a health system-level policy intervention was implemented to improve maternal and child health outcomes in a resource limited LMIC. Our qualitative multiple case study was informed by the Normalization Process Theory (NPT). It was conducted across eight districts and among ten health facilities in Uganda, with 48 purposively selected participants. These included health care workers located at each of the cases, policy makers from the Ministry of Health, and from agencies and professional associations. Data were collected using semi-structured, in-depth interviews to understand uptake and use of Uganda's maternal and perinatal death surveillance and response (MPDSR) policy and were inductively and deductively analyzed using NPT constructs and subconstructs. We identified six broad themes that may explain the observed variations in the implementation of the MPDSR policy. These include: 1) perception of the implementation of the policy, 2) leadership of the implementation process, 3) structural arrangements and coordination, 4) extent of management support and adequacy of resources, 5) variations in appraisal and reconfiguration efforts and 6) variations in barriers to implementation of the policy. The variations in sense making and relational efforts, especially perceptions of the implementation process and leadership capacity, had ripple effects across operational and appraisal efforts. Adopting theoretically informed approaches to assessing the implementation of policy interventions is crucial, especially within resource limited settings.
Read full abstract