The idea that building better health care systems can improve and save people’s lives may seem obvious, but until now there has been little published with the data and statistical muscle to prove it.
The best evidence yet may have just arrived from a remote, forested district of Madagascar.
Since 2014, a team led by researchers from the Blavatnik Institute at Harvard Medical School, the health care delivery nonprofit Pivot, and the Ministry of Public Health of Madagascar has worked to create a model health system in the district of Ifanadiana that provides universal access to basic health care to the region’s 200,000 residents.
A new study measuring the impact of this data-driven health-strengthening project conclusively demonstrates success. Infant, child, maternal, and overall deaths fell, even in the face of catastrophic damage from one of the strongest cyclones ever recorded, ongoing political turmoil, and deadly outbreaks of COVID-19, plague, malaria, and measles. Families in the district used more health services, especially for mothers and children, and barriers to care such as distance and cost became less severe.
By contrast, the study showed that during the same period, outcomes and access to health care worsened across the rest of the country.
“The number of children under five who died in the district fell by 30 percent in the decade after Pivot launched this project,” said Matthew Bonds, associate professor of global health and social medicine at HMS, cofounder of Pivot, and co-senior author of the study, published in PLOS Medicine. “That’s a great start.”
But Bonds pointed out that the team was not planning on stopping any time soon. “Our goal is zero preventable deaths.”
The data were so encouraging that even before the paper was published, the Ministry of Public Health in Madagascar asked Pivot to expand its project to a region of 1 million people. Madagascar’s minister of public health, Zely Arivelo Randriamanantany, is co-senior author of the study.
Bonds says that the secret to achieving these results—and to collecting convincing evidence that stronger health systems are behind the improvements in care—is lock-step integration of science and implementation since before the program launched.
A new beginning
Bonds first visited Madagascar in 2012 as part of a team asked to assess whether there was an opportunity to improve global health delivery and research efforts in the country. Madagascar was one of the few countries in Africa where health outcomes had been steadily getting worse in recent years.
Bonds was traveling with Michael L. Rich, HMS assistant professor of medicine at Brigham and Women’s Hospital. At the time, the two were colleagues at the international health care delivery nonprofit Partners In Health in Rwanda, where they were doing groundbreaking work leveraging scientific research to improve health in a challenging setting. Rich would later become a cofounder of Pivot and co-author of the new study.
In one clinic they visited in Ifanadiana, they found a nine-year-old girl with malaria. The survival rate for malaria in Madagascar then was 50 percent.
“It’s a disease that is very preventable and treatable,” Bonds said. “Nobody should be dying.”
Things didn’t look good for the patient. The girl was experiencing agonal breathing, the rapid, shallow, gasping breaths that signal hypoxia and too often warn that death is near. The clinic didn’t have any medicine or supplies to treat the girl, but Rich purchased antibiotics, an IV bag, and some quinine from a private source and was able to save her life.
Bonds and Rich saw the great need they witnessed as an opportunity to show beyond the shadow of a doubt what a difference an integrated health system could make in places like Ifanadiana.
Why implementation science is so challenging
Those working in the global health space today can find themselves stuck between knowing how easy it can be to solve a problem and seeing how tenaciously the problem persists, Bonds says. Research can help.
“I’m not a physician, so my goal is to help break this impasse by using science and data to find the best ways to get care to the people who need it,” he said.
In global health, many studies show the impact of single interventions. It has been shown that vaccines and mosquito nets, antibiotics and supportive care, prenatal care, and obstetric care can all, separately, save lives in trial conditions. But it has been trickier to show how to increase life expectancies in the overall population, not just people who signed up for a trial or a specific program. To do that, one must address multiple causes of mortality, which requires an effective health system that can deliver the whole package of basic care that people might need over the long term.
Why is such a simple solution so hard to demonstrate?
The kinds of places that lack strong health systems can also be difficult places to do rigorous scientific studies, Bonds says. It’s especially hard to create both at the same time.
In Rwanda, for example, data collection capacity didn’t ramp up until years after the health care system had begun strengthening, so the opportunity was lost to measure any benefit from those early years.
Teams have to balance the urgent need to deliver care with the challenge of gathering clean data. Sometimes researchers don’t have access to baseline data on demographics, health care usage, or clinical outcomes that predate the intervention they want to study. Sometimes the time frame for the implementation and research is too short. For example, it takes at least five years to measure an impact on under-five mortality rates, and research grants are rarely funded long enough to allow that.
Bonds said the Madagascar team was fortunate in that Pivot had funders, including cofounders Jim and Robin Herrnstein, who understood the importance of long-term research and that collaborators in the government understood the value of building capacity to collect data.
Health and science, inseparably linked
To make sure that they had a reliable baseline against which they could measure progress, Pivot team members began building their research infrastructure at the same time as they prepared to stock clinics with medicine and deploy community health care workers.
To study how the community used the health system and how improvements changed health, they created surveys and built data collection into every level of the system.
They also took advantage of the staggered growth of the health system-strengthening effort to have a solid comparison group. Instead of comparing people from Ifanadiana to other districts with different social, economic, and health characteristics, they compared neighbors who lived in the part of the district covered in an initial set of improvements in 2014 to those who lived in the area covered by a second wave that rolled out in 2021.
Instead of looking at a simple before and after picture, the team regularly collected new data and used it to make the health system stronger.
“Strengthening the health system and the science together fueled a feedback loop that continues to generate better results for the people of Ifanadiana,” said Andres Garchitorena, first author of the PLOS Medicine study, associate scientific director at Pivot, and former postdoctoral fellow at HMS.
For a 2020 study published in International Journal of Health Geographics and a 2021 study in Health Policy and Planning, Felana Ihantamalala, a co-author of the PLOS Medicine study and former postdoctoral fellow at HMS; Garchitorena; Bonds; and colleagues used a combination of remote sensing and researchers traveling on foot throughout the district to map every building, road, and winding path through the forest that people would travel to get to health care. They found that people who lived more than five kilometers from a health center were exponentially less likely to seek care for a child’s illness.
When they saw the results of that study, Pivot focused more effort on the community health arm of the health system, bringing care to the people who were too far from a health center to make the walk with a sick child.
A simple plan
Overall, the intervention was straightforward. Pivot worked to make sure that every level of health care in Ifanadiana has what Partners In Health calls the “Five S’s”—staff, stuff, space, systems, and social support—necessary to ensure that people across the district had access to a fundamental package of care, including prevention and treatment.
Care is available through community health workers who visit people in their homes as well as through widely distributed health centers, ambulances, and a district hospital.
The research generated throughout the implementation allowed the team to analyze copious data from 2014 to 2023 for the new study, including clinical records and household health survey data from a representative sample of the whole district.
Co-author Luc Rakotonirina, Pivot’s deputy national director, who oversees clinical operations and logistics, says that Pivot’s research is an integrated, inseparable part of the operation.
“The community health and clinical care teams help the science team figure out what questions to ask, and the science team brings us the information we need to know in order to improve care,” Rakotonirina said. “We’re all working to save lives.”