Presentation at the Global Maternal Newborn Health Conference, October 20, 2015
Background: Advancing the health and wellbeing of women and children remains an important focus in the post-UN Millennium Development Goal era, with many of the proposed Sustainable Development Goals concentrating on maternal and newborn health (MNH). To accelerate achievement of these upcoming goals, researchers have begun calling for the disaggregation of national-level MNH data, elucidating pockets of regionalized disparities and allowing for focused advocacy and equitable improvement for all women and children.
Methodology: Here we utilize spatial statistical methodologies and Bayesian regression techniques to map disaggregated skilled birth attendance in Burundi, Kenya, Rwanda, Tanzania and Uganda. Most recent household survey data, UN statistics, and other data sources representing population distribution, subnational age structure, fertility and growth rates were utilized to estimate births and pregnancies. Skilled birth attendance was modelled utilizing the Integrated Nested Laplace Approximation technique in R software and national household survey data. Additional data layers such as travel time to facilities were quantified using geo-referenced health facility data via EmONC.
Results: Distributions of women of childbearing age, births and pregnancies were mapped at 100m resolution for each country. Maps of disaggregated skilled birth attendance were also built for the region, including antenatal and perinatal attendance. These spatial layers were combined to illustrate regional unmet skilled birth attendance need, emphasizing local heterogeneity and highlighting areas of highest need for MNH services and advocacy.
Conclusion: The results of this research produce novel, high resolution surfaces of skilled birth attendance, including antenatal and perinatal attendance, highlighting subnational heterogeneity throughout East Africa. These findings begin to fill a critical knowledge gap regarding where and to whom subnational disparities are occurring. This has important policy relevant implications and may be used to inform targeted resource allocation, as well as predict related MNH outcomes such as neonatal and maternal mortality.