Evaluation of Large-Scale Maternal and Child Health Programs in India & Kenya

Abstract

Major reductions in morbidity and mortality could be achieved through high-quality, high-coverage implementation of proven maternal and child health interventions, including facility-based delivery by skilled birth attendants and prevention of mother-to-child HIV transmission through the provision of antiretroviral drugs. Methodologic advances in implementation science could identify opportunities to improve the effectiveness of these programs. This dissertation illustrates how epidemiologic methods can be applied to implementation science research using data from two large-scale maternal and child health programs; namely the BetterBirth Program, which implemented the World Health Organization’s Safe Childbirth Checklist in Uttar Pradesh, India, and the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) funding of prevention of mother-to-child HIV transmission (PMTCT) activities in Kenya. This dissertation A) reviews the development of the BetterBirth Program’s implementation package to identify methodologic lessons learned that can assist future researchers seeking to develop and evaluate complex interventions; B) performs regression analyses to investigate the relationship between coaching intensity, adherence to essential birth practices (EBPs), and maternal and perinatal health outcomes in the BetterBirth study in order to provide insights into the optimal coaching regimen for future interventions; and C) estimates the impact of per capita PEPFAR funding for PMTCT on infant and neonatal mortality and HIV counseling, testing, and receipt of test results during antenatal care in Kenya using a quasi-experimental dose-response analysis of publicly available data. These evaluations illustrate how the collection and analysis of quantitative data can advance our understanding of the overall effectiveness of large-scale maternal and child health programs and provide insights into how the implementation of these programs should be improved.