Economic evaluation and equity impact analysis of interventions for maternal and child health in Tanzania. Evidence for fair and efficient priority setting

Abstract

Introduction. Tanzania has seen a progressive decline in maternal and child mortality over the years. The last two decades have been a landmark with about 50% reduction in maternal and child mortality. However, the recorded improvements in the health status of mothers and young children in Tanzania is masked by geographical variation in the reduction of maternal and child mortality. In 2010, the under-five mortality in the Lake zone was reported to be 109 deaths per 1000 live births compared to the Northern zone where it was 58 deaths per 1000 live births. Key interventions addressing maternal and child health problems are inequitably distributed. There is a 57% difference in maternal mortality between poor and rich pregnant women. Similar trends are observed in interventions which address health problems in children under- five, though to a lesser magnitude with a gap of 10% to 15% between poor and rich populations. Economic evaluations of interventions for maternal and child health are imperative in generating evidence and informing context-specific allocation decisions to achieve rapid reductions in maternal and child mortality. The aim of the study is to generate evidence on a selection of maternal and child health interventions so this can inform priority-setting decisions in the direction of increased coverage for effective interventions that improve health outcomes and redress inequity. Methods. The health system implementation costs, including programme costs, were quantified to calculate the cost-effectiveness of adding rotavirus and pneumococcal vaccines to the Expanded Programme on Immunisation. The costs for the provision of diarrhoea and pneumonia treatment to children were quantified. We employed the ingredient and step-down costing approaches for the analysis of costing data. The cost and coverage data were collected from one urban and one rural district hospital and a health centre in Tanzania in 2012. Secondary data on disease epidemiology, national level intervention coverage and effects were retrieved from published literature and government reports. We used DALYs, QALYs and LY as the outcome measures and estimated incremental costs and health outcomes using a Markov model. For the equity impact analysis we used the Lives Saved Tool (LiST) to estimate potential reductions in maternal and child mortality and the number of lives saved across wealth quintiles and between rural and urban settings. Results. The introduction of rotavirus vaccine alongside the current diarrhoea treatment is highly cost-effective compared to diarrhoea treatment given alone, with incremental cost-effectiveness ratio (ICER) of USD 112 per DALY averted. The 13-valent pneumococcal vaccine is cost-effective, with ICERs of 258 per QALY gained and USD 245 per LY gained for Tanzanian settings, compared to no vaccine and 10-valent pneumococcal vaccine. However, the differences between pneumococcal vaccines were not robust with scenario analyses. Varying key model parameters may switch the results in favour of either of the pneumococcal vaccines. The probability of being cost-effective for both vaccines was at a much lower level than willingness-to-pay for health of USD 609 per capita Tanzania gross domestic product (GDP). It is probable that using both vaccines is highly cost-effective at a price far below a willingness to pay for health of USD 609 per capita Tanzania’s gross domestic product. The scale up of key, highly cost-effective interventions is likely to save more than twice as many mothers and children under five in the poorest population quintiles compared to the richest quintile in Tanzania. Increasing intervention coverage to equal levels across quintiles would also reduce inequalities in maternal and child mortality. Conclusion. This study has shown that it is possible to use currently available methods and tools to generate evidence for policy decisions in low-income settings. Combining available information on the burden of disease, economic evaluation and equity analysis to develop evidence-based health policies and plans to ensure fair and efficient resource allocation is possible, but remains a challenge. The use of scientific evidence is an important element in informing both policy and prioritisation decisions about health interventions. Health policy developed on the basis of systematically generated evidence is likely to be acceptable and achieve the goals of universal access to health services regardless of need.