More than 200 people met in New York City October 15-16, 2012, at the WIRED Health Conference: Living By Numbers to discuss the premise that better data can lead to better health. The inaugural conference, convened by WIRED and the Robert Wood Johnson Foundation (RWJF), featured speakers from across sectors who talked about how data is helping individuals and physicians manage care on a more personal level than ever before, and also about how that data can be scaled and aggregated to inform interventions that could transform the overall health of communities.
In advance of the event, we caught up with Steve Downs, RWJF Chief Technology and Information Officer, to talk about the latest trends in the capturing and use of health-related data at both the personal and population level, and the promise that these innovations hold for the future of health and health care.
“Big data” is an increasingly hot topic—not just in health IT but across government and the private sector. What does it mean to you, and why do you feel it holds such promise?
For me, “big data” is really an umbrella concept that speaks to the huge growth in the availability of data, and its ability to drive understanding, and ultimately improvements in health. There are whole new sets of data that advances in smart phones and low-cost sensors have made possible, and so there’s an ability to bring this new wealth of data to bear on personal health. And then at the same time, when you aggregate the data across the population level, there is really a new world of potential to discover new relationships and new insights.
As more people collect more data about themselves, we’re going to have data about whole new realms of life that we’ve never had before. I’ve likened this in the past to the idea that you always look for your keys under the light, and all of this new data is in effect expanding that light dramatically. On the one hand, we have created the opportunity for people to gain insights about themselves—what affects my sleep now and how does my sleep affect my pain? How does my diet affect my headaches? That sort of personal insight as an individual can help you take more control over your health. And then, through all of this new light, we’ve created new opportunities to roll data up on an aggregate level and really look at the population, bringing the potential to find really interesting connections among the data.
What are some examples of projects RWJF is funding around data for better health?
There’s a program that I’ve been involved with called Project HealthDesign that looks at what happens when people collect information about their day-to-day lives. We know that health happens not in your doctor’s office, but where you live, work, learn, and play, and so when you actually start recording data about that, it can reveal interesting things about how your condition plays out. The project has been looking at whether those data can inform better understanding of an individual’s conditions and ultimately lead to better treatments. We call the data “observations of daily living,” and they can include data about mood, sleep, diet, pain, physical activity, and the medications you actually take—as opposed to what you’re prescribed. For example, we had a project involving people with asthma whom their nurses discovered were taking their medications incorrectly. There were also patients with Crohn’s disease who could explain what was happening to them over the course of each day, and that led to changes in medications that really improved their quality of life.
What impact has the rise of big data had on RWJF’s vision for better health and health care?
The Foundation is interested in both ends of the spectrum, from the very personal to the broad population view. On the personal side, as much as we like to think of health care as consisting of conditions with clear diagnoses and treatments, I think the truth is a bit messier than that. We don’t really quite know what’s going on and things play out differently for different people. I think there’s a real opportunity for care to be personalized based on an understanding of how a particular treatment is working for a particular individual. One treatment may work for 75 percent of people, but if I’m one of the other 25 percent, we need to know quickly that it's not working for me and try to understand what does work for me in that realm.
One idea we’re exploring is research at a population level. Imagine what we could learn from a company that has 10 million users tracking their fitness, or 1 million women charting their pregnancies, as opposed to more typical trials that involve maybe 75 people. If we can start to tap into more of that data and open it up to research, there are things to be learned, trends that could be spotted, relationships we could understand, and correlations that may surprise us.
What are some specific examples of projects that tap into large datasets to improve health across a community?
Asthmapolis is a great example. This company, started by a former RWJF Health and Society Scholar, basically put GPS sensors on asthma inhalers, so you can gather data every time somebody uses an inhaler, including the time and the location. When you analyze those data, you can find individual patterns and give feedback to individuals that may have to do with identifying triggers. For example, “you typically need to use an inhaler in these kinds of places,” or “it’s these times of the day that affect you most,” or maybe it’s a combination of time and place where events are clustered. But across a population you start to get a new picture. You start to learn about places where lots of people are having trouble with asthma. You can imagine if you scale this over a whole metropolitan area, you could really learn about some significant hotspots for asthma. So that’s another terrific example of individual data that can be aggregated to inform community health decisions.
One of my favorite pieces of recent research, and it was just very simple, was from a company called Massive Health on data from their app called The Eatery, in which people simply rate the healthiness of their food, and then other people who are users also rate the healthiness of other people’s food. They asked some simple questions and discovered, for example, that what we eat gets less healthy with each passing hour of the day. That’s something probably no one really thought to study before, but it’s kind of interesting. They also find that, on balance, people who eat breakfast actually have a much healthier intake at each hour of the day than people who skip breakfast and then overcompensate throughout the day. We’re interested in seeing more of this kind of research. And if the use numbers on these apps get large enough, there could be some really interesting findings in those kinds of data.
What’s the motivation for mapping from an innovation that helps an individual up to data that might help bring about better health at the population level?
There’s a potential shift that seems to be emerging. It used to be that most of the data that would be applicable for medical research were essentially contained within the walls of medical research institutions or health care institutions, and increasingly that’s going to change. You’re going to see companies that are providing really great apps to people are going to get more and more of the data that’s going to be really relevant for research. From the company’s perspective, what’s really important for them is that they deliver awesome products to people, products that work really well and that are convenient and engaging and easy to use. (And if they don’t, there won’t be much data to study.) But as a byproduct, they’re going to generate data that have tremendous social value. That puts them in an interesting position. On the one hand, they have companies to run and products to develop and improve, but at the same time they could make tremendous contributions to society as well. Ideally, they’ll be able to do both. It’s going to be really interesting to see how that plays out.