Caroline Buckee is mad as hell—and that’s good for public health
Chris Sweeney | Winter 2020
Sitting in her office on a brilliant autumn afternoon, Caroline Buckee rattles off a brief list of all that’s wrong in the world. “The Amazon is burning. U.K. and U.S. politics are imploding. The humanitarian crisis in Syria. Rapacious environmental destruction. Yemen is home to the largest cholera outbreak in the world. Seventy-three thousand clinical cases of dengue recorded in Bangladesh since the beginning of the year,” she says, before taking a breath. “There are so many things to keep us up at night that it’s hard for the average person to take existential threats seriously.”
But Buckee isn’t your average person. Her career across a range of public health disciplines has been fueled by outrage at the human toll of health inequality. And as associate professor of epidemiology and associate director of the Center for Communicable Disease Dynamics at the Harvard T.H. Chan School of Public Health, she takes one existential threat especially seriously: a global pandemic. It could be a novel strain of influenza that tears around the world or an emerging pathogen for which we are woefully unprepared.
Causative agent aside, such a pandemic is “a nonzero probability,” Buckee says, lapsing into public-health-speak. In other words, it most likely won’t happen—but there’s a chance it could. And if it does, it will be devastating in every sense of the word. “It has the possibility of killing many, many millions of people,” she says, her soft British accent making the warning seem all the more ominous. “It’s a low-probability, high-risk kind of event.”
Over the past decade, Buckee, who pins a visitor with her direct gaze and gestures frequently with her hands when explaining heady concepts, has been building systems to track and forecast the spread of lethal infections, including malaria, dengue, Ebola, and cholera. Being able to follow the path of a parasite, bacterium, or virus is crucial to getting ahead of it and stopping a localized outbreak from blossoming into an epidemic, and an epidemic from ballooning into a pandemic. Buckee’s work blends tried-and-true epidemiological methods that go back as far as John Snow with bleeding-edge technology. Among her more high-profile pursuits is using location data from mobile phones to study how and where diseases spread. “In terms of epidemic forecasting, being able to know where people are going is hugely important,” she says. “And mobile phone data answers very specific questions about how people travel around.”
Finding Meaningful Patterns in Everyday Movement
Drawing on massive aggregated, anonymized data sets provided by phone companies, Buckee looks for what she calls “coarse-scale mobility flows.” It’s the type of information that can show, for instance, that 40,000 people moved from County A to County B one day, and that 10,000 people moved from County B to County C another day. In numerous peer-reviewed studies, Buckee and colleagues have shown that these large-scale trends can be turned into epidemiologically relevant patterns of movement. These patterns could then be translated into risk maps so that health officials can better allocate resources to prevent and fight outbreaks.
Until this point, the studies and computer models have all been retrospective. Now, Buckee is eager to move these methods from the pages of scholarly journals into the offices of health officials in countries where outbreaks are catching fire in real time and the resources to fight back are limited.
“Caroline has been very attuned to making sure she’s building tools that are not only useful for academics but are useful on the ground,” says Marc Lipsitch, professor of epidemiology and director of the Center for Communicable Disease Dynamics. “It’s going to take some time to get them up and running, but as she and her team continue to build more user-friendly systems and help train people in the necessary computation, these tools will make an impact.”
Buckee is now working with colleagues in Bangladesh—home to both one of the world’s largest refugee camps and some of its most populous cities—in hopes of building out this analytical pipeline. As with any translational research, putting it into practice is a complex, iterative undertaking. It involves bringing together a mobile phone company that houses and anonymizes raw data with epidemiologists in the country’s ministry of health who can run the mathematical models needed to transform the data into accurate disease forecasts.
To be sure, privacy concerns abound whenever officials consider leveraging mobile phone data. The list of Big Brother scenarios is seemingly endless and includes oppressive governments using such data to stifle dissent and track opposition. “Given the politics and debates around the use of cellphone data, it could be seen as a bad time to be working in this space,” Buckee concedes. “But at the same time, the data are immensely valuable for public health, the risk of epidemics is growing, and mass migration is on the rise. We need all the tools available to try to respond, and this is an important data source to regulate properly and to use safely and effectively.”
Buckee’s work using mobile phone data has made her one of the scientific whiz kids changing the face of global health. It has been heralded by journalists, academics, and public health advocates and even earned her a spot on MIT’s Technology Review’s “35 Innovators Under 35” list in 2014. Flattering as that may be, what Buckee wants more than anything is for her work to help remedy the gross health inequities that persist around the world.
“All of this innovative research at institutions like Harvard—work that has potential application in the field—needs to reach the most vulnerable populations, whether it’s in the context of helping eradicate malaria in Kenya or improving pandemic preparedness in Dhaka,” Buckee says. “That’s my motivation.”
A Winding Career Path to Public Health
For as long as Buckee can remember, she’s been keenly aware that the world is a tragically unequal place. As she was growing up, her family moved frequently because her father, who has a PhD in theoretical astrophysics, worked in the oil industry. They bounced around every few years, with stops in Alaska, Norway, Canada, the Middle East, and the United Kingdom. Each location offered up its own set of peculiarities for a young Buckee to chew over.
For a while, she liked the idea of becoming a veterinarian. But she soon realized that she was too softhearted to spend her days diagnosing beloved golden retrievers with cancer and putting down house cats. Wanting to keep a foot in the animal kingdom, she opted to study zoology as an undergraduate at the University of Edinburgh. She signed up for summer course work that took her to the Usambara Mountains in Tanzania to study plants, insects, and birds. It so happened that one of the areas in which she was doing fieldwork sat across from a malaria clinic. “I remember there being a bunch of mums and babies lined up at the clinic,” says Buckee, “and all I could think was, ‘Well, hmm, I’m catching butterflies.’”
Even as field biology lost its allure, she was becoming increasingly intrigued by evolutionary theory and the hidden stories that data can tell. Pursuing a master’s in bioinformatics at the University of York in England, she worked on the first DNA sequences of the macaque malaria parasite Plasmodium knowlesi, which sparked a lasting interest in pathogen genomics. After that, she earned a DPhil in mathematical epidemiology at Oxford University, where she worked under Sunetra Gupta, an acclaimed novelist and theoretical epidemiologist, and homed in on how ecological factors influence the population dynamics of Neisseria meningitidis, the bacterium that causes meningitis.
Upon completing her doctorate, however, Buckee found herself staring down a personal existential crisis. She began waffling over her career decisions, or lack thereof. “I didn’t know if I wanted to be a scientist or not. I never did. I still don’t,” she says, followed by a self-deprecating chuckle. “I get bored too easily.” She took a year off, traveled, went surfing, and briefly kicked around the idea of becoming an avalanche search-and-rescue pilot. Eventually, she came around to a life of science and signed on for a postdoctoral fellowship in Kilifi, a coastal city in Kenya.
It was there that Buckee started to explore human malaria parasites and the geography of infection, which led to the realization that mobile phone location data was a public health treasure trove. Starting with data provided by a major mobile phone operator in Kenya, she focused on malaria. “Human migration is a huge problem for malaria because when you get infected with malaria, you don’t get symptoms for two weeks,” Buckee notes. “This means that you can travel pretty far before you develop symptoms, and the clinic where you show up could be far from the place where you had become infected. This distance makes it hard for control programs to know where transmission is actually occurring.”
Blending tried-and-true epidemiological methods with bleeding-edge technology such as analyzing mobile phone location data, Caroline Buckee has built systems to track and forecast the spread of deadly infections—including malaria, dengue, Ebola, and cholera.
One of her first research papers on the subject, published in 2012 in Science, showed that cell phone data was a potent tool for mapping malaria risk because it allowed researchers to look beyond the limits of mosquito dispersal and identify high-risk areas where the infections had originated. From there, the projects snowballed. She linked with Norwegian mobile phone operator Telenor, which has networks in the Middle East and Southeast Asia, and conducted a study in Pakistan showing that mobile phone data could predict dengue outbreaks more effectively than could traditional surveillance methods. “One of the surprising things from that study was the sheer volume of long-distance travel, which traditional models had been significantly underestimating,” Buckee says.
Low-Tech Tools, High Human Impact
It’s easy to pigeonhole Buckee as the “phone lady.” Yet, her work with mobile phone data represents only a sliver of her young career as an epidemiologist. She has developed models to examine the efficacy of quarantine policies to contain various emerging infectious diseases, including Ebola, influenza A, and severe acute respiratory syndrome, and she’s creating new ways to study parasitic evolution as it unfolds during the spread of disease. But she understands that not every situation calls for high-tech solutions and millions of data points. In fact, sometimes the best methods are gratifyingly low-tech.
Not long after Hurricane Maria crashed into Puerto Rico in September 2017, Buckee began working with colleagues around Harvard and at Carlos Albizu University in Puerto Rico to figure out how many people died as a result of the storm. Initial figures from the Puerto Rican government had put the death toll at 64, which to people on the ground seemed like a shocking underestimate, given the intensity of the storm and the catastrophic damage it inflicted on the entire island’s infrastructure. Power outages in some areas stretched four months. Supply chains collapsed. Roads were washed out, and hospitals and clinics were inaccessible. For elderly people who relied on ventilators or needed daily medications to manage chronic diseases, the circumstances were life-threatening. “We were hearing stories of people burying loved ones in their backyards, so we knew the numbers were inaccurate,” Buckee says.
But exactly how off was the official death toll? In the wake of the storm, a restless Buckee mulled over ways to provide a more accurate estimate. “I was so anxious about it. It kept me up at night,” she recalls. She and her colleagues knew that an accurate estimate would account for lives lost not just during the initial storm blitz but also in the days and weeks that followed. So they went knocking on the doors of 3,299 randomly chosen households across Puerto Rico. They asked residents about infrastructure damage, displacement, and deaths caused by the storm between September 20 and December 31, 2017.
With that firsthand data, they estimated that there were 14.3 deaths per 1,000 people during the time frame. They then compared that rate against government death records for the same period a year prior and found that the mortality rate in 2017 was 62 percent higher. Buckee’s team—which also included Puerto Rican collaborator Domingo Marques, the François-Xavier Bagnoud Center for Health and Human Rights’ Satchit Balsari, and Harvard Chan School biostatistician Rafael Irizarry (who is from Puerto Rico)—estimated that the actual death toll of the storm probably exceeded 4,600 people.
When government officials estimated the death toll from Hurricane Maria in Puerto Rico at 64 victims, Buckee’s team of researchers carefully collected and analyzed data and showed that the actual death toll likely exceeded 4,600 people.
Published in the New England Journal of Medicine, the study served as a flash point for activists who had been challenging the official tally since day one. Media around the world latched on to the findings, and cries for further investigation rang out from San Juan to Washington, D.C. Before long, the Puerto Rican government revised the death toll to just under 3,000 people. The experience was extremely stressful, Buckee says. “But we made a huge impact.” Most important, she adds, intense media pressure forced the Puerto Rican government to release death registry data just five days after the study came out. The study also started a conversation among federal agencies about how to prepare for and quantify the impact of natural disasters—not just in terms of mortality but also in terms of long-range impacts on medically vulnerable populations.
A Game-Changing Asset for Bench Scientists
Oftentimes, the impact of Buckee’s work is far less obvious. Flaminia Catteruccia, professor of immunology and infectious diseases at the Harvard Chan School, was once assessing whether using a sterilizing compound to stop female mosquitoes from mating could reduce malaria transmission rates. She asked Buckee to help model the impact that approach would have on malaria spread. Catteruccia had hypothesized that interfering with mating behavior would sharply drive down transmission, but Buckee’s models showed that the effect would actually be minimal. “It was a huge paradigm shift for us,” Catteruccia says. “We actually moved away from trying to develop inhibitors to mating behavior—that’s how important Caroline’s work was to our lab.”
More recently, Catteruccia has been studying applying antimalarial compounds to bed nets to stop mosquitoes from becoming infected with Plasmodium falciparum, the most common malaria-causing parasite in humans. Catteruccia’s team had shown that the approach worked: Mosquitoes that landed on surfaces coated with the antimalarial compound atovaquone were completely blocked from developing P. falciparum infection. Once again, she asked Buckee and her team to develop a model to determine whether this approach would actually have an impact on malaria transmission. This time, Catteruccia’s hunch was right. The models developed by Buckee’s team found that adding an antimalarial compound to conventional insecticide-treated bed nets would lead to a steep drop in malaria transmission.
It’s game changing for a bench scientist to know the potential impact of an intervention so early in the research phase, Catteruccia says. Not only did it make for a stronger paper, she says, it also helped funders understand that this is an approach worthy of investment and further exploration.
Promoting Gender Equality in Science
Catteruccia admires Buckee’s scientific acumen and willingness to reach across disciplines to push boundaries. But she equally admires Buckee’s work outside the lab to promote gender equality in science, higher education, and beyond. “She’s not afraid of voicing her concerns,” Catteruccia says. “She’s a very strong woman, and that’s who we need to be leading the field.”
Last April, Buckee received the Harvard Chan School’s Alice Hamilton Award, which honors a pioneer in the fields of toxicology and occupational health and who, in 1919, was the first woman appointed to the Harvard faculty. Near the beginning of her lecture at the award ceremony, Buckee brought up a slide that featured a graph. It showed that the number of tenure-track women across Harvard University has essentially been stagnant over the past decade. Buckee went on to present a slew of data points demonstrating that women at Harvard—and in the sciences in general—face professional, societal, and personal barriers that combine to effectively exclude them from opportunities and hinder their success.
“If you look at the history of women at Harvard, it is a relentless and persistent push against grudging acceptance,” she told the audience. “I think each of us need to ask ourselves about what we’re doing and if we’re actively trying to change this.”
Depending on whom you ask, using an award acceptance speech to call out the institution giving you the award—which also happens to be your employer—could be seen as controversial or even self-sabotaging. To Buckee, it was the perfect platform. “This is a major issue. There are important conversations happening, but they’re often happening among people who can’t make change,” she says. “The people that need to be part of this cultural shift are senior men.”
Such candor is a reminder of how fiercely determined Buckee is to rooting out inequity in whatever form it takes, whether trying to close the gender gap at one of the wealthiest institutions in the world or trying to close the health gap between the world’s most affluent and most resource-poor nations. “A lot of what I do is fueled by rage,” she says. “I find anger to be quite motivating.”
—Chris Sweeney is a senior science writer at the Harvard Chan School Office of Communications.
Portrait by Kent Dayton for Harvard Chan.
