Leveraging routine programme data for understanding under-five mortality patterns in sub-Saharan Africa

Ending preventable deaths of newborns and children under five years of age by 2030 is a central aim of the international community, and codified in the UN Sustainable Development Goal 3.2. Accurate measurement of levels, trends, and age patterns in under-five mortality is essential for tracking and accelerating progress towards its reduction, and providing the evidence necessary to guide effective health policy making. While civil registration and vital statistics systems are either incomplete or absent in most countries in Sub-Saharan Africa, Health and Demographic Surveillance Systems (HDSS) serve as one of the most viable
solutions to the dearth of empirical health and population data. HDSS record demographic and
epidemiologic events of interest at the individual, household, and community level in geographically defined populations on a longitudinal basis. However, it is particularly difficult to obtain accurate information for pregnancies and pregnancy outcomes as they frequently occur between HDSS data collection intervals and are subject to high levels of underreporting. This research will exploit data from HDSS in sub-Saharan Africa that also conduct record linkage with local health facilities and clinics that serve the population. Record linked data will be analysed to ascertain and correct bias in HDSS reports of pregnancies, pregnancy outcomes, early childhood mortality, and their risk factors. Analysis will focus on records of pregnancies or pregnancy outcomes in clinical data that are not associated with a birth outcome in the HDSS data, and how the HDSS toolkit for capturing maternal and newborn events can be improved.