Defining conurbation (extended urban area) boundaries for maternal healthcare applications in sub-Saharan Africa

Background: While urban areas in sub-Saharan Africa (SSA) have historically enjoyed a wide range of advantages in maternal health (urban health advantage), this advantage is diminishing or reversing. Concurrently, SSA is currently witnessing the fastest urbanisation rates in the world. Therefore, there is need for an improved understanding of maternal healthcare utilization in conurbations (an extended urban area, consisting of several towns merging with the suburbs of a central city) that captures the lived experiences of women. However, current definitions of conurbation boundaries vary between countries and international entities and may not fully reflect the reality of women seeking healthcare. Here, we define boundaries of 23 conurbations in 22 different countries across SSA by incorporating the lived experiences of women. Further, we compare the coverage of three maternal health indicators based on the existing and enhanced conurbation boundary. Methods: To define the boundaries of conurbation we assembled secondary data on local administrative and health boundaries, data from city-specific sources and satellite based data from 5 different sources. We spatially overlaid these sources to create a preliminary conurbation boundary. To reflect the lived reality for maternal health, we then adjusted the resulting boundary following feedback from researchers and/or health professionals with expert or lived-experience knowledge of each conurbation. We then computed and compared the coverage of three maternal and child health indicators based on 3 original definitions of conurbation boundary (health boundary, satellite-based boundary, and the enhanced boundary). We used the most recent demographic and health survey (DHS) and accounted for its complex, multilevel cluster survey design. Results: Preliminary results show that there were 5 sources of global and regional satellite based sources of data that were then used to define a conurbation. However, these sources provided varying spatial extents of conurbations. The coverage of the maternal and child health indicators were variable across the different definitions of conurbations. The coverage based on the modified boundary reflected more realistically the lived experience of women when seeking healthcare. Conclusions: The results shed light on the impact of conurbation definitions on maternal and child health indicators and on the role of urbanisation in the evolving landscape of maternal healthcare in SSA.