In Precision Medicine, ‘We Need Dialogue Across Disciplines’.

Since the first human genome was sequenced in 2003, the precision medicine field has made strides in advancing individualized treatments for patients, based on information from the patients’ own genomes, environments, and lifestyles. While the majority of precision medicine research has centered around individual diagnoses and treatments thus far, less attention has focused on the role of data science in advancing population health.

On Wednesday, November 20, the School of Public Health will hold the Public Health Forum, “Harnessing ‘Precision’ to Advance Population Health and Health Equity” featuring Kirsten Bibbins-Domingo, Lee Goldman Endowed Chair in Medicine, chair and professor of the Department of Epidemiology and Biostatistics, and professor of medicine at the University of California, San Francisco School of Medicine.

“Precision medicine is often reduced to being all about genetics, but I think that’s a misread on how it was originally conceived,” says Bibbins-Domingo, who is also a general internist and cardiovascular epidemiologist. “When I think about the promise of precision, it’s the ability to harness tools in a wide variety of research activities that hopefully will improve the way we deliver care in the clinical setting, as well as the way we think about population health.”

Ahead of the seminar, Bibbins-Domingo discussed the current status and challenges of precision medicine research, and how the healthcare and public health fields can collaborate to advance population health.

How would you define precision medicine?

What precision medicine refers to is the wonderful array of new tools that we have at our disposal. We have vast amounts of curated and real-world data, and the ability to make more precise measurements on individuals and groups that characterize their health states and other aspects of health. We also have the computational power to integrate measurements across a variety of data sets to do this in a faster way. It is the combination of these tools that I think of when I think of precision.

In which clinical areas has precision medicine been most effective, and what are the field’s biggest challenges moving forward?

If we take the very narrow view of precision medicine—that it is targeting a particular therapy to a particular mechanism of disease action—we’ve probably made the most strides in cancer, because we’re talking about the genetics of the tumor, and we have had success in the therapies that target the tumors.

In other arenas, it’s been more challenging, especially in the more narrowly conceived ideas of genetics, genomics, and ability to target therapies. One of the biggest barriers in the clinical setting is that we collect data on certain types of people but not other types, so our genetic databases are vastly underrepresented for any group that is not of European descent. We also have gaps in the ability of frontline providers to talk about new advances and help patients to understand where they might be useful. Third, we are in an environment where there is a lot of hype about these things, and we’ve had less of an ability to separate hype from what is medically useful.

As a scientific community, we have to think very critically about how we evaluate the scholarship that is being produced, as well as understand the real utility of it in clinical practice. We tend to have these conversations in silos, among people who use traditional approaches to evaluate evidence and people who are producing evidence with newer techniques. But the ability to integrate evidence to help patients make better decisions in the clinical setting is lacking, and we need dialogue across disciplines to have those discussions.

How can the healthcare and public health fields increase this dialogue and work more collaboratively to advance and benefit from precision medicine?

We need to get away from the hype and rhetoric. Is genetics going to solve our major population health problems? Absolutely not. That doesn’t mean we need to throw out our understanding of biological variability, and our understanding of why we get individual differences in disease or distribution of disease.

This is also an environment where we are increasingly realizing that we have not reached the outcomes we hoped to achieve in the clinical setting, and that we do have to understand social context and other ways in which health is shaped. Ironically, it is the same types of data tools that will allow the clinical setting to understand the biology of disease in patients, and the social context in which patients’ health develops. My hope is that the broader forces that shape health will not only allow us to improve health in the clinical settings, but require us to think across disciplines.

But when we talk about the cross-disciplinary discussions between medicine and public health, the challenge is to not over-medicalize the social context in which health emerges. It’s great that healthcare settings are embracing the social determinants of health and understanding the social context of health, but the risk is that it tends to over-medicalize the things that take place out of the doctor’s office, but are just as important for health.

You’ve done a lot of work in cardiovascular disease prevention and care—how do you hope precision medicine will ultimately help reduce the prevalence and incidence of cardiovascular disease?

All of us have some risk of cardiovascular disease and we’ve had decades of scholarship that help us understand what we need to do to prevent it. What has been lacking is prevention for people who are at very high risk of cardiovascular disease. We need to integrate population health approaches that allow us to understand how we reduce the risk of people with average or increasing risk of this disease as they age, and how to target therapies specifically to these people. We’ve pushed clinical prevention for population health cardiovascular disease to its limits—we have lower blood pressure targets, we’re increasingly trying to prescribe cholesterol medication to everyone, but we have less of an ability to find the highest-risk people and make sure that we target them most intensely for therapies to reduce their residual risk. That’s how some of these newer computational approaches can allow us in the clinical setting to not just understand the average risk and the effects we know they have, but more individual-level risks to target interventions more effectively for those higher-risk individuals.

I’m particularly interested in this for people of younger ages, because that is where we have a gap in knowledge. We don’t have the ability to delineate very well which young people are at high risk over the next 10 years to have an event. In prevention, you want everyone to exercise and eat right, but you really want to find the few people who have a high 10-year risk when they’re young, and that’s where we need more discovery and an ability to assess sets of information that will help us find and understand those risks.

—Jillian McKoy