When it comes to child deaths, the world has made great strides in the past 25 years. "In 1990, one in ten children in the world died before age 5," Bill and Melinda Gates write on their blog. But thanks to things like vaccines and better nutrition, "today, it's one in 20."

The death rate for children younger than one month has proven harder to budge. Newborns account for 44 percent of all childhood deaths, and health experts aren't sure why. They know it might have something to do with prematurity, or infections, or complications during delivery. But they often don't know exactly what happened right after a given birth that brought death just a few weeks later. Was the baby not dried off properly? Did the umbilical cord get infected?

In order to better understand the drivers of mortality for all children, on Wednesday, the Bill & Melinda Gates Foundation announced that it's investing $75 million in a series of surveillance sites that will gather data "about how, where and why children are getting sick and dying," according to the release. This Child Health and Mortality Prevention Surveillance Network, or CHAMPS, will be spread initially throughout six locations in Africa and South Asia. It will rely on field workers to take biopsies of children who have perished and on beefed-up laboratories that will perform medical testing.

The foundation hopes that, once put in place, the surveillance system can be repurposed to detect and respond to new outbreaks. According to many global health experts, including Bill Gates, one reason why Ebola wreaked so much damage in West Africa was that the affected countries lacked good healthcare infrastructure.

I spoke with Gates recently about how he hopes to cut child deaths, once again, in half by 2040. A lightly edited transcript of our conversation follows.


Olga Khazan: One thing that surprised me as I was learning about your current initiatives, and the situation on the ground in some of these countries, is that we don’t actually always know when a child dies and why. Why does that happen, and how does it foil our attempts to reduce childhood mortality?

Bill Gates: Our uncertainty about our disease estimates is pretty high. When you look at what are kids dying of in the [Democratic Republic of Congo] ... what they're doing is simply asking mothers what they saw, and that's called a verbal autopsy, and that doesn't really tell you what caused the diarrhea, or what caused the fever. And so we're relying on very, very few cases where we're going out to get samples to figure out what the exact cause was.

In some ways, when we were going after the biggest killers that are everywhere, like pneumococcus for respiratory and rotavirus for diarrhea, we didn't need very precise figures. But now that those are going to be knocked down, depending on the vaccine coverage, to almost zero, the remaining pieces [are things like] where is cholera, where is typhoid? Should you do a vaccine for that or not? Cholera continues to be very episodic over time in different locations.

We've added in these six centers that we're funding, we have two things. One is the latest genetic tools so that any samples we have we can figure out what's going on. The second is, instead of doing verbal autopsies, when we get permission, which we hope will be for five to ten percent of the deaths, we go in and get a minimally invasive autopsy. Then you can go in and using the genetic tools, you can get a sense of what's going on. And so we can take our error bars about these different causes of death and start to resolve those so that the global health community is well-directed. We would make this investment just for the ongoing death-reduction agenda, but this is kind of a dual benefit that we've designed so that these lab tools and expertise could both see an epidemic earlier on and could deploy to nearby areas to make sure that we're a lot smarter than we were in the Ebola epidemic.

Khazan: How will tracking childhood deaths through something like CHAMPS help prevent them? What do you plan to do once you have the information?

Gates: What it will show us is the changes in different locations. We have early work going on for vaccines for respiratory diseases, diarrheal diseases, typhoid, which is both a respiratory and intestinal disease. We need to know how to prioritize the creation and the volume procurement and delivery of these additional vaccines.

Another area to highlight is the first 30 days. As we're making progress on all the things that kill kids from 30 days to 5 years, like the various diarrheas, pneumonias, and malarias, the fraction of deaths that are in those first 30 days, what they call the neonatal mortality rate—that's getting to be a higher and higher percentage. That's a little more mysterious. You'll see something like respiratory failure or sepsis. Actually understanding if we had an antibiotic, would that have solved it? If the child had been kept warm, would that have solved it? If we vaccinated the mother, would that have solved it? We have a lot of uncertainty in those first 30 days and we need CHAMPS data to understand the interventions that will allow us to make to make the same progress on that part of the burden as we had on the older kids.

Khazan: Previously, you called for a kind of “global warning and response system” that’s geared toward outbreaks like Ebola. How will the CHAMPS program fit into that or help with that?

Gates: We're hoping to get partners to come in so that instead of the six centers, we can have 20, and that would add dramatically to disease surveillance in poor countries. We need ongoing capability that's sampling places like Democratic Republic of Congo, where studies are done so irregularly that the uncertainty of what health is like there is very high. In any case, what we would see is you'd either hear symptomatically about something unusual going on and you'd go over there and gather samples that would go to these labs. We'd need to pre-clear that the samples could move across country boundaries to the regional center. We'd need epidemiologists who could go into the affected areas and understand what that burden is. You ought to be able to respond in a week or two, whereas with Ebola ... at best, that took months to do.

Khazan: What do you mean when you say "surveillance?"

Gates: You'll have labs and they'll have a big radius where they're gathering samples from childhood and other deaths, but if they hear there's an increase in death rates or some unusual symptoms, they'll go out further and gather samples from outside their normal radius. They have to have political permission and pre-agreed understanding, and language skills. Then they put them into their advanced genetic PCR-type capabilities, and then making those publicly available to the world. Saying, "hey, the signature we're seeing here is quite unusual." Or there's been a shift in the burden.

If you're lucky, the disease would show up in their normal catchment area. But more likely, you'd hear reports of excess deaths or unusual symptoms, and you'd take this map where you cover everywhere, and these are just the focal points to get that capacity up very quickly. Not having to think, "how do you send contaminated blood samples back to the U.S.? What does U.S. customs think when they get Ebola blood vials?" There were huge problems in Ebola in dealing with samples. There were a few months were there was almost no processing going on.

Khazan: Do you ever worry whether initiatives like this will be enough without fundamental change at the government level in some of these countries? How far can NGOs go when they’re dealing with authoritarian or dysfunctional leaders?

Gates: Some authoritarian regimes run very good health systems. [Laughs] In terms of the epidemic, we all have to worry that there are parts of the world that don't have good health systems. There are exemplars, even in very poor countries. Rwanda's an exemplar. Ethiopia, 10 years ago was terrible and has now gotten very good and continues to improve. The ones that really stand out as a big problem are parts of India, Pakistan, Nigeria, the DRC ... Those are all areas where we have a particular focus to make sure that well-running systems and trained people are there. If you have epidemic outbreaks in the Central African Republic, we will catch it sooner if we have good primary healthcare there. There's no substitute over time for having great functioning primary healthcare. There's a lot of money from the U.S. and others going into the three Ebola-affected countries. Hopefully it should be enough money for them to get a good primary healthcare system going. There is progress over time, but you won't get everywhere, and those are the places where you'd have to be more worried about an epidemic starting.

Khazan: How is CHAMPS different from other attempts to help these areas or to improve health in low-income countries?

Gates: Most of the data [that exists]—those are good, those are super important because they give us a broad statistical base, but [the surveys] only happen every five years or so. In terms of determining cause of death, they're not that useful. Then you'll have things where people will go in and do a medical study in one location. They gather data and then leave. What we haven't had is something that's an all-disease, semi-permanent using all the latest technology in poor countries to really ascribe cause of death. Questions like, Where is there cholera in Africa? We actually have a cholera vaccine, but where should it be used?

We're building up local capacity with these people. If these countries get richer, this is the kind of thing they'll fund for themselves. But the uncertainty of the data we were operating on was fine when it was just pneumococcus and rotavirus, but for this next phase, we need a lot more data. We're going to build capacity and be ready for epidemics.

Khazan: Anything else you wanted to mention?

Gates: I'll just mention briefly that there's a company called CureVac in Germany—they and some others are injecting RNA into people and it surprisingly gets into their cells and surprisingly changes the protein manufacture. In the early trials, this stuff looks very good. It's being used for a variety of things, including cancer and infectious disease. We made an investment in them. Part of the beauty of this technology is that you can change what you're making to go after some new Ebola-like thing very rapidly. It's a vaccine platform that will take a number of years to prove out. But it's the kind of thing that ... you might have a tool that would be much more responsive. It's the kind of R&D that the world needs to be investing in; it has a profile that might make it work for epidemics.

In the case of Ebola, the epidemic was stopped without any tool. ZMapp [an Ebola antibody drug] will be in full production two to three months from now. If the next epidemic is Ebola, ZMapp will come and make a huge difference, as might the vaccines. But unfortunately the next epidemic might be something we haven't seen before. It might be flu, which we are outrageously still not ready for a big flu outbreak.