Promoting progress in child survival across four African countries: The role of strong health governance and leadership in maternal, neonatal and child health

Despite numerous international and national efforts, only 12 countries in the World Health Organization’s African Region met the Millennium Development Goal #4 (MDG#4) to reduce under-five mortality by two-thirds by 2015. Given the variability across sub-Saharan Africa, a four-country study was undertaken to examine barriers and facilitators of child survival prior to 2015. Liberia and Zambia were chosen to represent countries making substantial progress towards MDG#4, while Kenya and Zimbabwe represented countries making less progress. Our individual case studies suggested that strong health governance and leadership (HGL) was a significant driver of the greater success in Liberia and Zambia compared with Kenya and Zimbabwe. To elucidate specific components of national HGL that may have substantially influenced the pace of reductions in child mortality, we conducted a cross-country analysis of national policies and strategies pertaining to maternal, neonatal and child health (MNCH) and qualitative interviews with individuals working in MNCH in each of the four study countries. The three aspects of HGL identified in this study which most consistently contributed to the different progress towards MDG#4 among the four study countries were (1) establishing child survival as a top national priority backed by a comprehensive policy and strategy framework and sufficient human, financial and material resources; (2) bringing together donors, strategic partners, health and non-health stakeholders and beneficiaries to collaborate in strategic planning, decision-making, resource-allocation and coordination of services; and (3) maintaining accountability through a ‘monitor-review-act’ approach to improve MNCH. Although child mortality in sub-Saharan Africa remains high, this comparative study suggests key health leadership and governance factors that can facilitate reduction of child mortality and may prove useful in tackling current Sustainable Development Goals. We reviewed national policies and strategies issued between 2000 and 2013 and conducted key informant (KI) interviews in 2013 to explore eight content areas influencing child survival (WHO, 2006, 2007, 2010, 2012; Ban, 2010; WHO and PMNCH, 2011): (1) health care system (including HGL, structure, human resources for health, access & utilization, monitoring & evaluation and accountability), (2) national health strategies and policies, (3) MNCH interventions, (4) clinical standards and guidelines, (5) commodities and essential medicines, (6) health financing, (7) partnerships and (8) contextual factors (e.g. conflict, political environment, hygiene and sanitation, nutrition and food security, education and human rights). Four SSA countries (Liberia, Zambia, Kenya and Zimbabwe) were chosen based on their U5M ARR between 1990 and 2011 (data available when the study was designed, Figure 1) and their national governments’ willingness to participate. Detailed study methods for each country case study have been published (Kipp et al., 2016; Brault et al., 2017, 2018; Haley et al., 2017). A national document review was conducted for each country to evaluate the MNCH policy framework affecting progress towards MDG#4. Policies and strategies pertaining to overall national health, MNCH and other related determinants were obtained from the WHO African Region office, WHO country focal points and Ministry of Health (MOH) for Liberia, Zambia, Kenya and Zimbabwe. Additional MNCH-related documents referenced in initial sources were subsequently obtained and reviewed (see individual case study supplementary tables in Kipp et al., 2016; Brault et al., 2017, 2018; Haley et al., 2017). An abstraction guide was developed based on the eight study content areas and several cross-cutting questions (Table 1). Each document was reviewed by one author (CAH), who consulted with a second reviewer (MAB) as needed. Information from original documents was recorded verbatim in the abstraction guide to avoid observer bias. Key questions and deductive themes explored during the review of national health policies and strategies and key informant interviews that cut across child survival content areas Utilizing country Demographic and Health Surveys (DHS) closest to 1990 and 2011, one or two provinces were selected from each country that had U5M ARRs comparable with the national ARR and were logistically accessible. Specific rural and urban sites were selected to evaluate differences in MNCH that can exist between urban and rural areas (Table 2). Selected study sites within Kenya, Liberia, Zambia and Zimbabwe Semi-structured interviews were conducted with KIs involved in MNCH from the MOH, donor organizations, community-based organizations (CBO) and health care providers (HCP) (Tables 3 and ​and4).4). CBO participants and HCPs were selected from both urban and rural sites. National level KIs (see below) were recruited from the capital and each local site. In-country research teams collaborated with the MOH and WHO to identify potential KIs representing a range of ages, work experiences and positions/roles balanced between urban and rural sites. Additional inclusion criteria for each key informant group Numbers of key informants interviewed for each country Guides for KI interviews were developed and piloted, mirroring the eight content areas and cross-cutting questions explored in the national document review (Table 1). Interviews were audio recorded, transcribed and translated into English (as needed) by trained research assistants. Transcripts were coded using deductive themes based on study content areas plus additional themes identified upon transcript review. Analyses were conducted using the qualitative software Atlas.ti (Murh, 2004), grouping the on-track countries (Liberia and Zambia) and not on-track countries (Kenya and Zimbabwe) for comparison. Analyses focused on codes related to HGL based on the WHO definition (WHO, 2007). The Institutional Review Boards at the authors’ institutes and both the national and local ethics and research committees for each country approved the qualitative component of the study as follows (see Supplementary file S1 for copies of approval letters): Vanderbilt University Medical Center (Coordinating Center), Kenyatta National Hospital Ethics & Research Committee (Kenya), University of Liberia Office of the Institutional Review Board (Liberia), ERES Converge Institutional Review Board (Zambia), Joint Parirenyatwa Hospital and University of Zimbabwe College of Health Sciences Research Ethics Committee and the Medical Research Council of Zimbabwe.