Abstract
BACKGROUND: Uganda hosts the largest refugee population in Africa, which exerts much pressure on the district health systems. While refugee-hosting districts (RH) receive targeted investments, the extent to which these influence maternal and child health (MCH) service coverage remains unclear. Using routine facility data, we examined differences in MCH coverage and trends between RH and non-refugee-hosting (non-RH) districts and also explored the effects of government health financing and health system performance on MCH coverage. METHODS: We conducted a retrospective analysis utilizing routine health facility MCH data from the Uganda District Health Information System and district-level government Primary Healthcare (PHC) expenditure data from 2020 to 2023. MCH indicators were ANC1st trimester, ANC4, Institutional deliveries, mothers' Post-natal care (PNC), Measles1 and DPT3 vaccination. We computed a composite coverage index (CCI), health systems performance z-score and compared trends across RH and non-RH districts. Mixed Effects Models assessed the association between government expenditure, RH-status, health system performance over the years. RESULTS: RH districts consistently had modestly higher coverage of ANC1st trimester, ANC4, Institutional deliveries, PNC, Measles vaccination and CCI trends. Government expenditure was significantly higher in RH districts and refugee-dominant (RD) districts (p < 0.001 vs. p = 0.007). Refugee-dominant districts had higher but non-significant MCH coverage. Unadjusted models of MCH indicators and CCI were positively influenced by government financing and health systems performance z-score except for DPT3 and Measles, respectively. Adjusted models revealed that ANC4 coverage was 7.4% points higher in RH districts (7.42; 95% CI:0.753, 14.090; p = 0.029) and increased by 3.6% points for every unit increase in z-score (3.60; 95% CI: 0.729, 6.462; p = 0.014). CCI increased by 1.6% points and 2.3% points for every unit increased in the government expenditure and z-score respectively (1.55; 95% CI: 0.310, 2.788; p = 0.014) vs. (2.31; 95% CI: 0.642, 3.975; p = 0.007). CONCLUSION: Novel approach - leveraging routine facility data, revealed MCH coverage was modestly consistently higher in RH districts over the years and RH status influenced ANC4 coverage. Overall district-CCI depended on Government investment and health systems performance implying increase in PHC financing could be a key driver to universal district-level improvement.