Maternal mortality in Zimbabwe: a geospatial analysis of socioeconomic factors

Abstract

Background: Over the years, Zimbabwe has struggled to reduce the incidence of maternal mortality making it one of the most dangerous places to give birth in, globally. The lack of understanding of the geospatial mechanisms and socioeconomic circumstances within which maternal mortality occurs in Zimbabwe is a hindrance to its sustainable management as a public health problem. Methods: Four rounds of the nationally-representative Zimbabwe Demographic and Health Survey (ZDHS) data collected between 1999 and 2015, were employed to geospatially analyse the levels and changes in the socioeconomic factors associated with maternal mortality in Zimbabwe. Computation of maternal mortality rates and ratios across provinces, mapping of hot spots and cold spots of maternal mortality, and logistic regression models of selected socioeconomic factors with maternal mortality across provinces comprised the analyses employed in the study. Results: Of the total maternal deaths recorded in Zimbabwe between 1999 and 2015, the majority were recorded in provinces such as Manicaland (20%), Mashonaland Central (13%) and Harare (13%). Except for predominantly urban provinces such as Harare (580 to 501 death per 100 000) and Bulawayo (490 to 460 deaths per 100 000), MMRatios across provinces increased in the 2010-2015 period. Provinces with notable MMRatios increase include Masvingo (630 to 790 deaths per 100 000) and Matebeleland North (490 to 728 deaths per 100 000). Compared to the 1999-2005 period, the 2010-2015 period also had a higher concentration and spread of maternal mortality hot spots across the country. Harare and Bulawayo were cold spots. Selected places in the immediate peripheries of major cities, borderline areas and the stretch of areas in the predominantly rural provinces such as Masvingo, Mashonaland East and Central were maternal mortality high spots. Overall, parity, age, education, cost of healthcare and religion were observed as significantly associated with maternal mortality. Statistically significant (p < 0.05) regression analysis outcomes with the highest odds include those involving the cost of healthcare in Matebeleland South (aOR = 25.50) and Harare (aOR =14. 12). In terms of wealth index, the highest odds were noted among those identified as “poor” in Harare (aOR=2.45) and Mashonaland Central (aOR=1.69. On religion, Protestantism in Bulawayo (aOR=2.25) and Pentecostalism in Matebeleland South had relatively high odds of association with maternal death. In terms of statistically significant odds by education, primary education in Masvingo (aOR=13.34) and Harare (aOR=4.04) were the most notable. The strength of association between maternal death and education categories such as primary education overall improved between the 1999-2005 and 2010-2015 period (aOR = 2.95 to 2.35), while with those noting distance to the facility as a big challenge the odds had worsened (aOR = 0.80 to 1.29) X Conclusions: maternal mortality in Zimbabwe is a national crisis requiring protracted investment to nullify it. Equally important is recognising that the experience of maternal mortality across time and space is not homogeneous, thus in the short term, it is imperative to identify places that require more attention than others. The occurrence of maternal deaths involves the intersection of multiple geographic demographic and socio-economic factors operating at different levels and strengths The formulation of maternal health policies will thus need to be cognisant of the dynamism and relativism of the socioeconomic and geographic contexts within which maternal mortality is instigated.