Cost-effectiveness of Strategies to Reduce Maternal Morbidity and Mortality

There is little guidance, and few evidence based analyses, on how to cost effectively scale-up local efforts to reduce maternal mortality in the poorest countries
Investigators:
Delphine Hu
Zachary Nieder
Kelly Blanchard
Dan Grossman
Country Contextualization of a Global Policy Model

In some of the poorest areas in the world, ranging from rural Uttar Pradesh, India to Badakshan, Afghanistan to Northwest Nigeria, one third to one half of the deaths in women of reproductive age relates to childbirth and unsafe abortion. Nearly all are preventable. Global progress has been suboptimal in meeting MDG5 which calls for a 75% reduction in the maternal mortality ratio by 2015 as a key step in reducing poverty and supporting social and economic growth. Our focus on India, Nigeria, and Afghanistan reflects the magnitude of the burden of maternal deaths in these countries; India accounts for one-quarter of all pregnancy- and delivery-related deaths, and Afghanistan and Nigeria have the highest rates of maternal mortality in the world. When coupled with our inclusion of Tanzania, the work relates to four countries that account for more than 40% of the maternal deaths worldwide. We have conducted similar work in Mexico.

Reducing maternal mortality has become a publicly stated priority by governments, international agencies, and national health systems but major challenges remain.  While there is general agreement that interventions such as universal access to high-quality intrapartum and emergency obstetrical care are the ideal, there is little consensus on which sets of interventions - and in what order to introduce them - are the most affordable and cost effective. There is little guidance, and few evidence-based analyses, on how to cost-effectively scale-up local programmatic efforts (including those that involve new technologies) in the poorest countries, and especially in those without an accountable government or functional health system.

Given the diseases and global burdens for which we do not yet have the technology, vaccine, or means to cure, this project addresses a public health failure that we know how to fix and know what needs to be done technically. Rather, the critical challenges include complex issues like the performance of the health system, delivery of health services, health worker shortages, linkages (or lack thereof) between health and nonhealth sectors (e.g., transport). These are the very same challenges that need to be solved for delivery of antiretroviral therapy (ART), community interventions for malaria, for management of the rising burden of chronic disease, for emergency care for road traffic accidents. This project provides an example of how the formal inclusion of these factors in a model-based analysis can more comprehensively capture the influential factors that determine "real world impact". In turn, analyses generated using this modeling framework allow for countries to design realistic strategies that consider the particular resource and infrastructure constraints they are facing, and that evolve in phases such that both the pace and scope is realistic, affordable, and cost-effective. Finally, the approach we take allows us to identify the most important uncertainties (i.e., information of high value) that should be priorities for empiric studies, including operational research.

We use decision analytic tools, including computer-based models and methods of evidence synthesis, to improve the ways in which we translate meaningful measures and metrics into actionable realistic strategies, to identify focused investments informed by rigorous analysis, that can be made now; to illuminate in a transparent manner, the critical knowledge gaps and heterogeneous levels of evidence for different interventions and programmatic approaches. Our general analytic approach is that we integrate the best available country-level data (including system capacity) and quantify the expected benefits (e.g., averted deaths, averted morbidity, reduction in the MMR, proportional mortality ratio, and lifetime risk), and the economic consequences (cost resources needed and cost savings) associated with a wide range of integrated interventions. We identify the most effective and cost-effective of these in the context of community-level, district -level and national-level capacities, and model their uptake incrementally.

We purposefully choose a modeling and analytic approach that is well suited to reflect the most important determinants of maternal mortality and the most important details about the setting and capacity. Selected features of our core model and our analyses follow: First, our calibration methods allow us to reflect a wide range of inter-country situations, differentiating rural and urban regions, and even specific states; second, the flexible structure allows for differentiation of the level of the health system, ranging from services delivered at the household and community level, to those at the primary-level, to those at the secondary and tertiary-level (e.g., district-level and above); third, the model permits integration of assumptions about the availability of transport, facilities, and quality of care, such that we are able to not only reflect the status quo - and take into account existing strengths and weaknesses - but we are able to superimpose strategies that adopt an incremental approach to building capacity; fourth, our analytic framework considers several stages and phases of investments to improve maternal health, and can provide insight into how the design of a strategic approach can influence the pace at which financial resources are needed, as well as the magnitude of these costs; sixth, the model generates a wide range of outcomes so that key messages based on the analytic insights can be constructed to frame the most relevant points in ways that will be most appropriate for different decision makers.