MODEL-BASED SUB-POPULATION ESTIMATES OF MATERNAL MORTALITY RATES AND RATIO FROM SIBLINGS’ SURVIVORSHIP HISTORIES IN NIGERIA (2008 -2018)

Abstract

Despite several interventions, Maternal Mortality (MM) remains high in Nigeria. This is further complicated by lack of reliable estimates of MM for subnational levels such as states and geopolitical regions. A plausible estimate of MM levels is essential to provide evidence based national and state-level planning, resource allocation and monitoring of progress. It will reflect the population diversity in the country and assist in closing MM gaps. This study was designed to adapt sisterhood method and small area estimation techniques to derive plausible estimates of MM rates and ratios for subnational populations in Nigeria. Survivorship history data of 293,769 female siblings provided by 114,154 women in the Nigeria Demographic and Health Surveys conducted in 2008, 2013 and 2018 were analysed. The dataset from each survey was reconstructed into a panel data structure such that each reported sibling was captured as an observation. The MM Rates (maternal deaths per women-years of exposure to childbearing) and Ratios (maternal deaths per 100,000 live births) were estimated using direct and indirect sisterhood methods. Empirical Bayesian technique for small area demographic estimates was used to obtain MM rates and ratios at state-levels. The James-Stein estimator was used to shrink the estimates closer to the population mean values at 95% Confidence Interval (CI). Zero-inflated Poisson regression model was fitted to investigate association between selected covariates and maternal death counts at the community levels. Incident Risk Ratio (IRR) was reported as measures of effect. All analyses were weighted to adjust for the effects of clustering. MM rates in 2008 were high in rural areas and North-West region at 1.21 and 1.65 per 1,000 women-exposure years and lowest in South-West at 0.45 per 1,000 women-exposure years. Levels of MM Ratios were highest in the rural areas and South-South region at 624 and 679 respectively and lowest in South-West (281 per 100,000 live births). In 2013, the levels of MM Ratio were highest in North-Central (712) and lowest in South-West (367 per 100,000 livebirths) and for 2018, it was higher in rural areas (548) compared to urban (523); highest in North-West (901) and lowest in the South-East (268). MM Ratio was consistently lower in the South-West (2008=281; 2013=367; 2018=392) and higher among the Northern regions of the country, particularly the North-East (2008=654; 2013=612; 2018=901). Overall, Kebbi, Adamawa and Taraba states had high MM Ratio across the three surveys. From 2008 to 2018, MM Ratio declined by 18% in the North-West and 54.2% in the South East region. However, there was a 4.8% increase in MM Ratio for South-West from 2008 to 2018. At the community levels, geopolitical zone, knowledge (IRR=1.33, CI=0.98-6.1) and actual use of family planning (IRR=1.92, CI=1.1-9.1) were associated with maternal death counts. This study has derived and shown differentials in subnational estimates of maternal mortality in Nigeria and has identified geopolitical region, the knowledge and use of family planning as major covariates of maternal mortality. This has produced a baseline upon which improvements in maternal mortality in states and geopolitical zones in Nigeria can be based.