Andrew T. Chan, MD, MPH, is chief of the Clinical and Translational Epidemiology Unit in the Department of Medicine at Massachusetts General Hospital, where he is also vice chief for clinical research in the Division of Gastroenterology, and director of epidemiology at the MGH Cancer Center. He is the Daniel K. Podolsky Professor of Medicine at Harvard Medical School and professor of immunology and infectious diseases at the Harvard T.H. Chan School of Public Health.

Why is epidemiology and having an epidemiologist on the team important to advances in clinical and translational research?

I view epidemiology as a population laboratory. It’s an opportunity for researchers to test hypotheses they’ve been developing at the bench in a human population and get an early readout about their potential relevance for clinical translation. There are many examples of advances that are made in the laboratory, but their relevance to humans is an open question. It’s critical to have an idea of that relevance before investing in additional studies.

On the flip side, epidemiology can be a source of discovery. We can use these population laboratories to identify patterns in disease or risk factors for disease that haven’t yet been considered, but can be detected in large epidemiological studies for the first time. Then we can take them back to the bench to try to understand the mechanistic basis for the underlying associations with the goal of eventually translating findings to the clinic.

“Over the course of the pandemic, we realized that a big unanswered question was what else people could do to minimize their risk of COVID-19 beyond social distancing, masking, and vaccination.”

Is it new to have a clinical and translational epidemiology unit in an academic hospital? What need is it filling?

I think it is fairly novel. Historically, epidemiology has been somewhat siloed from the clinic, which comes from a tradition where most epidemiologists and departments of epidemiology are housed in schools of public health rather than near patients. That’s been a barrier to translation of epidemiological findings into the clinic. It also leads to less of an opportunity to communicate clinical observations to test hypotheses in large population laboratories.

So we developed the clinical and translational epidemiology unit at MGH to address this gap.  Our goal is to bridge a vibrant clinical service with epidemiological expertise to lead to effective collaborations through methodological tools and population insights.

Our unit is based at MGH, but we have strong connections with many hospitals in the Harvard system, as well as the Harvard T.H. Chan School of Public Health. We view ourselves as a place where we encourage and foster team science and take advantage of the wealth of expertise, knowledge, and scientific interests across the Harvard community.

You were senior author of a study of almost 600,000 people published in Gut that found that those whose diet was healthy and plant-based had lower COVID-19 risk and lower chance of severe illness if infected. What was epidemiology’s role in reaching these findings?

COVID-19 is a good example of where epidemiology filled a critical need. It was a new disease and spreading quickly. As a result, we needed to understand the disease as quickly as possible by collecting large amounts of data. We also needed to apply tools to rapidly analyze that data and return knowledge on what was effective back to the public health community, to clinicians and, ultimately, to individuals.

When COVID hit, we rapidly pivoted to apply our expertise to collect data from a large number of participants using a mobile application, a technology that we deployed and validated in other epidemiological studies. Using digital technology to collect data from people’s homes was vital in doing studies without the ability to see study participants in person.

Over the course of the pandemic, we realized that a big unanswered question was what else people could do to minimize their risk of COVID-19 beyond social distancing, masking, and vaccination. Based on our previous work, we know inflammation plays an important role in the pathogenesis of a lot of chronic conditions. So we hypothesized that diet might be a factor that predicts COVID severity, since the inflammatory response seems to underlie the most severe clinical cases of COVID. We used our epidemiological toolkit to develop a survey of diet, collected data from several hundred thousand people, and got some initial answers about the role of diet and inflammation in predicting COVID outcomes.

“We want to engage with our communities so the work we do in epidemiology will actually have an impact on the people around us.”

What other research in your unit are you excited about?

We’re really excited about applying epidemiology not just in the clinic, but also in the community. I think there’s a high unmet need to develop studies that have a direct impact on the communities that we serve as teaching hospitals and as a university. For example, our group is using epidemiology and epidemiological methods to increase cancer screening in local communities. We have investigators doing studies to improve access to mammography, lung cancer screening, and colorectal cancer screening, using relationships and partnerships that we’ve developed with community organizations. We want to engage with our communities so the work we do in epidemiology will actually have an impact on the people around us.

Looking back at your career, by getting a MPH and adding epidemiological tools to your research toolbox, have you been able to accomplish what you intended?

I’ve been fortunate in my career to have the ability to marry my two passions — taking care of patients and doing research that will have a population-level impact. It’s been so rewarding and fulfilling for me to make a contribution in these ways and to develop a program that I hope will allow others to do the same. We want to create opportunities for people to make scientific advances using all the tools at our disposal to actualize some of the lofty goals we have for medicine in the future.

The growing interest in personalized and precision medicine, for example, has often been informed by using molecular epidemiology, such as genetic information, to address clinical questions. I think we’re at an exciting crossroads where many clinical advances in the future are going to come from the observations that we make in population laboratories through sophisticated molecular tools to better understand mechanisms of disease.