Measuring the impact of development assistance on vaccination coverage: evidence from low- and middle-income recipient countries

Thesis (Ph.D.)--University of Washington, 2019Over the last decade, substantial amounts of development assistance have been invested among low and middle income countries to strengthen the performance and outcomes of the immunization program. Multiple development partners have collectively instituted different financing obligations targeting both new vaccine introductions and scale up of traditional vaccines. This dissertation explores the development assistance landscape and the impact on vaccination outcomes, with the goal of making a timely contribution to guide subsequent investment decisions as the end of the decade of vaccines draws nearer. In the first chapter, Tracking donor funding towards achieving the Global Vaccine Action Plan (GVAP) goals: A landscape analysis (1990-2016), development assistance for immunization is characterized by purpose, recipient and time. The study utilized data from existing project databases, annual reports, and audited financial statements of multi-sectoral agencies supporting immunization where funding was categorized by objective, estimating allocations to different vaccine types versus health systems strengthening. Using generalized linear models, funding projections were made to inform the progress of current global targets. While development assistance has remained resilient over time, findings from this analysis suggest that resource targets stipulated in the Global Vaccine Action Plan may not be met by 2020. This as a result calls for renewed financial assessments while strengthening existing resource efficiency at recipient level in order to achieve the desired child health outcomes. The second chapter, Effect of donor funding for immunization from Gavi and other development assistance channels on vaccine coverage: evidence from 120 low and middle income recipient countries, builds on the first chapter further evaluating any impacts of aid on vaccine coverage. Following the launch of Gavi in 2000, recipient countries have not only continued to scale up underused vaccines but have concurrently expanded the scope of their routine immunization products which to date include thirteen new vaccine introductions. In this study we apply a novel approach where disbursements are disaggregated into funding for specific vaccines versus that for health systems strengthening. For these aid categories, we evaluate the impact on coverage for a wide spectrum of routinely administered vaccines namely; DPT3, pneumococcal vaccine (PCV3), Pentavalent3, Measles2 and Rotavirus2 vaccines. Findings from this analysis suggest varying improvements in coverage as a result of aid with stronger effects occurring among the newer vaccines. From a policy perspective, these findings offer promise for continued investments in immunization, but perhaps even more importantly, highlight the need for improved strategies for fiscal sustainability and efficiency in order to achieve universal immunization coverage. The last chapter, Determinants of inequality in vaccination coverage for DPT3 among sub-Saharan countries, investigates predictors of inequality for DPT3 vaccine coverage. Different measures of inequality are computed using 5km by 5km level vaccine coverage estimates to quantify existing geographical disparities in coverage in sub-Sahara. Using vaccine coverage estimates from 2000 to 2016, we quantify inequity using three measures. First, we assess the shortfall inequality which is the average deviation across subnational units from that with the highest coverage for each country. Secondly we estimate the threshold index which is the proportion of children below a globally set subnational coverage target, and lastly, a Gini coefficient which represents the within country distribution of coverage. We use time series analyses to quantify associations with immunization expenditures controlling for country socio-economic and population characteristics. Development assistance, maternal education and governance were associated with reductions in inequality, with governance augmenting the observed relationship between development assistance and inequality. Results from this analysis also indicate that countries with the lowest coverage suffer the highest inequalities. We also demonstrate growing inequalities among countries which have since met national coverage targets such as South Africa and Kenya. Burundi, Comoros, Lesotho, Namibia, Rwanda, Sao Tome and Principe and Swaziland had the least shortfall inequality (