Determinants of maternal health and family planning service coverage in Nepal: modelling of routine data

Abstract

Executive Summary: The risk of obstetric complications during pregnancy and unwanted pregnancies are linked to higher maternal deaths. Increasing availability and the use of modern contraceptives and providing high-quality pregnancy and childbirth care can reduce the number of maternal deaths. However, service utilization and coverage are influenced by multiple other factors like the poverty gap, literacy rates, overall human development index, and other external factors. Expansion of the coverage of maternal health and family planning (FP) services has been a priority for the health sector in Nepal, and the Government of Nepal is working to ensure that all citizen have access to such services through a series of policy initiatives. It is often critical to evaluate whether these public health initiatives have been effective. In this context, this analysis used statistical models to identify the factors that determine the coverage of maternal and FP services. The analysis used the service coverage data from the Health Management Information System, and district characteristics data from ‘Population Atlas of Nepal 2014’ published by Central Bureau of Statistics. The district is the unit of analysis. Eight statistical models were used to identify district characteristics associated with coverage of maternal health and FP services. Coverage of first antenatal care(ANC) visit, four ANC visits, intake of iron and folic acid for 180 days (IFA coverage), institutional delivery, postnatal care (PNC) visit within 24 hours, three PNC visits, and composite coverage of maternal health services were the dependent variables used to assess determinants of maternal health service coverage. Similarly, the modern contraceptive prevalence rate (mCPR) was used as a dependent variable for FP service coverage. A log-transformed linear regression model was used for statistical analysis. All statistical analyses were performed in R software. The analysis revealed that the coverage of institutional delivery increases with increase in the proportion of female household heads in the districts. Of the 77 districts, 31 districts had higher than average and 30 districts had lower than average composite coverage of maternal health services. Compared to districts with Brahmin as the predominant caste, districts with Magar as predominant caste had 17% higher composite coverage of maternal health services. Similarly, compared to districts in Province 1, Gandaki Province, Province 2 and Bagmati province had 27%, 26% and 15% lower composite coverage of maternal health services. With one unit increase in female mean age at marriage in the district, the composite coverage of maternal health services decreased by approximately 10%. The mCPR of 21 districts was above the average value while the other 37 districts had below the average value. Compared to districts with Brahmin as predominant caste, Districts with Gurung, Magar and Tharu as predominant caste had 32%, 26% and 18% higher mCPR. On the other side, compared to districts with Brahmin as predominant caste, districts with Rai, Chhetri, Limbu and Newar had 29%, 28%, 27% and 21% lower mCPR. Compared to districts in Province 1, Gandaki Province, Lumbini Province and Bagmati Province had almost 37%, 18% and 12% lower prevalence of mCPR. The analysis reveals predominant caste, female mean age at marriage and province as determinants of composite coverage of maternal health services. Similarly, HDI, predominant caste of the district and province were key determinants of the mCPR. Targeted interventions in low coverage districts for both the maternal health services and FP services could be useful.