Introduction: Although health expenditure in sub-Saharan African countries is the lowest compared with other regions in the world, most African countries have improved their budget allocations to health care over the past 15 years. The majority of health care sources in sub-Saharan Africa are private and largely involve out-of-pocket expenditure, which may prevent healthcare access. Access to healthcare is a known predictor of infant mortality. Therefore the objective of this study is to determine the impact of health care expenditure on infant mortality in sub-Saharan Africa. Methods: The study used panel data from World Bank Development Indictors (WDI) from 2000 to 2015 covering 46 countries in sub-Saharan Africa. The random effects model was selected over the fixed effects model based on the Hausman test to assess the effect of health care expenditure on infant and neonatal mortality. Results: Both public and external health care spending showed a significant negative association with infant and neonatal mortality. However, private health expenditure was not significantly associated with either infant or neonatal mortality. In this study, private expenditure includes funds from households, corporations and non-profit organizations. Public expenditure include domestic revenue as internal transfers and grants, transfers, subsidies to voluntary health insurance beneficiaries, non-profit institutions serving households or enterprise financing schemes as well as compulsory prepayment and social health insurance contributions. External health expenditure is composed of direct foreign transfers and foreign transfers distributed by government encompassing all financial inflows into the national health system from outside the country. Conclusion: Health care expenditure remains a crucial component of reducing infant and neonatal mortality in sub-Saharan African countries. In the region, where health infrastructure is largely underdeveloped, increasing health expenditure will contribute to progress towards reducing infant and neonatal mortality during the Sustainable Development Goals (SDGs) era. Therefore, governments in the region need to increase amounts allocated to health care service delivery in order to reduce infant mortality.
The study used pooled panel data from 2000 to 2015 for 46 countries in SSA. The source of data for this study was the World Bank Development Indictors (WDI) [31]. We used infant mortality rate and neonatal mortality rate as outcome variables. The infant mortality rate is measured as the death of a child less than 1 year old per 1000 live births and the neonatal mortality rate is measured as the death of a child less than 28 days per 1000 live births. Predictor variables included total health expenditure measured as percentage of GDP and income per head as measured by GDP per capita (Additional file 1). Higher health care expenditure is expected to be associated with lower infant and neonatal mortality. Different population age groups, namely those under 14 years and above 65 years, were measured as a percentage of the total population. These were included to control for different country demographic structures. Relative to the younger population, the population age group above 65 years is expected to increase infant mortality outcomes by increasing death rates. To control for the varying levels of infant and neonatal mortality in SSA, HIV prevalence rate, maternal mortality ratio, fertility rate, access to improved water and sanitation, measles vaccination coverage, and school enrolment were included in the model. In addition, we used immunization rate as a proxy to measure the effect of the use of preventive health care services on health outcomes such as infant mortality. To model the two health status outcomes (infant mortality and neonatal mortality), we used random effects models on the pooled panel data from 2000 to 2015 for 46 countries in SSA [21]. To predict health status using these models, we added the covariates total health expenditure as a percentage of real national income, gross domestic product per capita real income, which acts as a control variable for the demand for health services and other economic factors. The total health expenditure is further grouped in to public health expenditure, private health expenditure and external health expenditure. The demographic variables represent population age groups of under 14 and over 65 years age, respectively, and expressed as a percentage of total population. In this model, a random effect was added for country to control for unobserved heterogeneity and the outcome measure and predictors were transformed using a logarithmic function where appropriate. In addition we used demographic variables to control variation across the countries The modelling approach for the panel data from previous studies is as follows [21]. Where Yit is the outcome variable in country i at time t, X is the matrix of predictor variables, including the intercept, and β is the matrix of fixed regression coefficients. The total variation in the model is broken up into two parts. Between country error represented by the random effect term υi and within country error denoted by εit. In most panel data analysis, there was the need to test for random effects or panel effects in the model. The Breuch-Pagan Lagrange Multiplier (LM) test was used to make a decision between random effects regression and simple OLS regression. The null hypothesis in the LM test is that variances across the countries is zero. This is, no significant difference across units (i.e. no panel effect). Here we rejected the null hypothesis and conclude that random effects are appropriate at the 5% level of confidence (p-value < 0.001; 95% CI) [32]. Secondly, the Hausman’s specification test was employed to compare estimates from the random effects and the fixed effects models. The null hypothesis in the Hausman’s test is that the error term for country is not correlated with the predictors. Here we failed to reject the null hypothesis and conclude that random effects is appropriate at the 5% level of confidence (P-value = 0.077) [32].