Data-Driven Done Right

Much of Madagascar’s Ifanadiana District is passable only by foot, electricity is nearly nonexistent, and basic health and human services are almost entirely absent—posing a variety of complex and persistent health challenges. In low-resource settings like this, a data-driven approach to strengthening health systems can prove especially valuable, as PIVOT, a medical NGO, demonstrated in this region.

More data is not necessarily better, but given the growing emphasis on data within global health and development, finding ways to use it more effectively is essential—and PIVOT’s data-driven approach could help inform and improve other efforts. We came to know the work of PIVOT as part of a collaborative effort, including researchers from Johns Hopkins, Stanford, UCLA, and Virginia Tech, all working together to improve tuberculosis diagnostics for low resource settings. As we quickly learned, PIVOT’s data-driven approach proved very conducive to conducting research with an NGO partner, particularly in an austere setting.

Prior to PIVOT’s engagement in the district, studies reported a 1 in 6 mortality rate in children under-5 years, and a lifetime maternal mortality rate of 1 in 14, both of which are among the highest mortality rates in the world. PIVOT began working in Ifanadiana in 2014 with the goal of delivering health care to the rural poor in Madagascar, and the hope of serving as a model for the rest of the country.

Interestingly, some of PIVOT's founders come from Partners in Health (PIH), and like PIH, PIVOT emphasizes integrating its operations within the local, regional and national health systems. More than 95% of its staff is native to Madagascar and virtually all of their activities are conducted hand-in-hand with local and national health officials. Although a health systems approach to global health is not unique to PIVOT, unlike most medical NGOs, PIVOT is obsessively data driven. This makes sense once you learn that its founders include 2 astrophysicists (Jim and Robin Herrnstein), a Harvard professor with dual PhDs in economics and ecology (Matt Bonds), and a physician scientist with decades of research combatting tuberculosis (Michael Rich).

The organization has evolved since its inception, but a data-driven, health systems approach has remained central from the start. Before beginning any major programmatic activities, PIVOT had the foresight to initiate an expansive, longitudinal study of the population it aimed to serve, including a sample of more than 1,600 households both within and outside of their catchment area. Balancing the need for data with respect for the beneficiaries, PIVOT conducts follow-up surveys on its study population once every 2 years and guards these households from unnecessary sampling or intrusion even from its own collaborators, partners, and senior staff. In addition to collecting population-based longitudinal data, PIVOT has also focused on improving health management information systems (HMIS). PIVOT leverages each of these 2 primary data sources to modulate its own programs and interventions as well as regional and national policies with an emphasis on scalability.

The first follow-up data became available in 2016 and has proved invaluable, highlighting where efforts have and have not born results. The quality of data has allowed PIVOT to not only manage limited resources, but also expectations, which has refined its programmatic approach and improved relationships with their Malagasy collaborators—a fact we witnessed firsthand in Ifanadiana this past month. In just 2 years, PIVOT’s efforts have nearly doubled the number of people treated for fever or diarrhea, decreased neonatal mortality by 36%, and decreased under-5 mortality by 19%. Most notably, PIVOT’s data-driven approach to health systems strengthening has demonstrated substantial improvements in 2 critical areas: malnutrition and universal health coverage (UHC). After a baseline study revealed stunting in more than 50% of children, PIVOT decided to layer its malnutrition efforts across all 3 levels of the district health system: community health workers, health centers, and the district hospital. The push achieved a near-100% success rate in over 3,000 children screened since the program began and solidified the malnutrition program as one of PIVOT’s key battle horse programs.

Regarding UHC, PIVOT’s baseline study showed 1 in 6 children died before the age of 5, and only 20% of kids received treatment for fever and diarrhea—and identified treatment cost as the primary impediment to seeking care. To address this concern, PIVOT removed all user fees for children by paying the Ministry of Health (MoH) facilities directly for over 40 drugs and 20 consumables, resulting in health utilization quadrupling. The MoH later invited PIVOT to join its UHC committee, which sent a team to Ifanadiana this past March to review PIVOT’s data, understand the costs and potential impacts of removing point-of-service fees, and explore the possibility of piloting its national program in the district.

PIVOT data has also shown where their efforts have had limited, or in the case of maternal health, unequally distributed success. Through its infrastructure program, PIVOT created maternity wards in its health centers, developed an ambulance program that provides rides to the hospital for emergencies (including delivery complications), and funded a surgeon skilled in C-sections at the district hospital. These investments doubled the number of women delivering at health centers, but only among women living within 5 km of a health center. Armed with this knowledge, PIVOT is now working to devise a comprehensive maternal and reproductive health program for 2018 that will extend its reach to women who live further away.

Unfortunately, not all organizations are equipped technically or materially for such a data-driven approach to the delivery of basic health services, but the importance of this kind of data should not be understated. Used correctly, it can help foster accountability and encourage organizations to confront weaknesses and failures—an approach that engenders trust. Such values are often rare in the global health community where exposing failures can mean less funding. However, a commitment to data collection, analysis and dissemination is critical to ensuring an organization’s long-term success, and to the iterative development of a sustainable approach to strengthening health systems in the world's most limited and complex environments.

Jeffrey D. Freeman, PhD, MPH is a research associate at the Center for Humanitarian Health with the Johns Hopkins Bloomberg School of Public Health.

Cassidy Rist, DVM, MPH is an assistant professor at the Center for Public and Corporate Veterinary Medicine, Department of Population Health Sciences, at the Virginia-Maryland College of Veterinary Medicine.

Join the thousands of subscribers who rely on Global Health NOW summaries and exclusive articles for the latest public health news. Sign up for our free weekday enewsletter, and please share the link with friends and colleagues: Subscribe to GHN