Health eHeart: A Framingham Study for the Social Media Era?

Robert M. Califf, MD; Mark J. Pletcher, MD, MPH; Jeffrey E. Olgin, MD; Gregory M. Marcus, MD, MAS


May 16, 2014

This feature requires the newest version of Flash. You can download it here.

Editor's Note: Variously described as a modern day Framingham and a virtual study, the ongoing Health eHeart™ study from the University of California, San Francisco (UCSF) enrolls participants online and leverages data from mobile apps (some provided) and social media. Dr. Califf interviews the investigators and discusses novel strategies for conducting cost-effective, scalable clinical research trials.

The Health eHeart Study

Robert M. Califf, MD: Hello. This is Rob Califf from Duke University, and I am delighted to be interviewing the Health eHeart leaders here on the UCSF campus in preparation for the American College of Cardiology (ACC) Scientific Sessions. It turns out some West Coasters cannot make it to the East Coast. They are on call and cannot make the long flight, so it is great to be out here in San Francisco, where I was an intern and resident. Maybe we can start by having you introduce yourselves. Mark, you can go first.

Mark J. Pletcher, MD, MPH: I am Mark Pletcher in the Department of Epidemiology and Biostatistics and the Department of Medicine here at UCSF. I am a cardiovascular epidemiologist and also a general internist.

Jeffrey E. Olgin, MD: I'm Jeff Olgin, Chief of Cardiology here at UCSF and co-lead with Mark and Greg on the Health eHeart study.

Gregory M. Marcus, MD, MAS: I am Greg Marcus, cardiac electrophysiologist and Director of Clinical Research for the Division of Cardiology at UCSF.

Dr. Califf: Mark, tell us a little bit about the background of Health eHeart. How did you get into this line of thinking about epidemiology?

Dr. Pletcher: We came together around a common vision for harnessing new technology that is transforming our society and helping it transform the way we do clinical research. There is a real opportunity in cardiovascular disease, and in epidemiology in particular, for taking advantage of common basic measurements that we can get both through traditional clinical research methods and also through new methods of collecting information, using sensors and social networks, etc., to help us do clinical research in a different way.

Dr. Califf: Jeff, you are an electrophysiologist, and I tend to think of electrophysiology (EP) guys as hanging out in the lab and doing procedures and things like that. This seems like a distant step from that. What is the marriage here between hardcore EP and epidemiology?

Dr. Olgin: We EP types always gravitate towards technology. Part of this came out of a clinical trial that Mark and I are leading currently and the sheer cost of doing that type of research, with the heavy personnel requirements and the slow slog to enroll patients. It just seemed that there should be better, more efficient ways to do that, so we wanted to try to leverage and build technology to enable that.

Cutting the Cost of Research

Dr. Califf: How does this relate to traditional and clinical research?

Dr. Marcus: We are faced with a challenge of increasing costs in a time when funding is dwindling. We recognize that there is a pervasive use of technology in our daily lives that has largely remained untapped in doing research. My research had been very patient-oriented in the EP realm, but trying to enroll a large number of patients, we had the idea of creating a cohort, recognizing that the Internet and technology gave us a new opportunity.

Dr. Califf: Give us a picture of the nuts and bolts. What is the magic that is going to dramatically reduce the cost and increase the efficiency?

Dr. Olgin: It is still early, so a lot is yet to be determined, but our sense is that we have built technology to try to recruit more broadly so that we do not need more bricks and mortar study centers to recruit people. We are leveraging social media and the Internet to replace much of what occurs during an in-person visit, and that will hopefully reduce the number of personnel needed to deal with that. Everything from the consent process (which is all done online in the study) to the every-6-months visits (which are also done online) to much of the data collection and even some novel outcome ascertainment and integration with electronic medical records, in theory, should make it less costly to do the research. The way that we think about it is that the more patients we enroll in the study, the cost per patient is driven down, which is quite the opposite from a traditional study, where you price it on a per-patient basis.

Dr. Pletcher: We think very carefully about our costs scale up. In a traditional study, you pay for each measurement in terms of coordinator time, bricks and mortar, rent, equipment, and all that stuff. We are trying to design this so that we pay for the central infrastructure, but then the marginal cost of each additional measurement is essentially zero, so that we can scale up without it becoming expensive.

Leveraging Fitness Trackers

Dr. Califf: What about gadgets? I am sitting here right now comparing the Fitbit® and the Nike devices. I noticed that there is a 20% difference in the number of steps I take every day, and you can probably guess which one I like better. Are you giving people gadgets? How are you getting the biological measurements that come in through the Internet?

Dr. Olgin: It depends. We allow people who own devices already to connect their devices, but we have also partnered with some companies to give away free devices. For example, in January, we gave away 1000 AliveCor™ ECG devices, and to speak to Mark's comment about marginal costs, we have collected 18,000 ECGs in less than a month without any marginal or personnel costs. We have some additional studies that we are proposing in which we will give out devices to look at activity and so forth.

Dr. Califf: You have an alliance with the American Heart Association (AHA). How does that work?

Dr. Marcus: There are several facets to it. We have several common goals because ultimately we are interested in the cardiovascular health of the public. The AHA recognizes that. One thing that they can provide is access to a large number of individuals, which is important for us to take full advantage of this scalability and to enroll not just a large number of participants but a large number of engaged participants. We are working with them to gain access through newsletters and email campaigns and also to reach out to underrepresented groups that may not have easy access to technology.

Dr. Califf: I was visiting the University of California, Davis, a couple of weeks ago, and they told me that access to healthcare there is lower than in Appalachia, which was hard to believe so close to such affluence, and there are so many doctors there. Even in the state of California, it is an issue. It would be great if the AHA's help and the low cost would allow people to participate. Where do you see this expanding in terms of the AHA? Will all affiliates be involved? How do you see it working with them?

Dr. Olgin: Hopefully all of the affiliates will become involved. We have been working mostly with the national office, so it is pretty broad coverage. And this is not limited to the United States. Of our 7000 members in the study so far, 140 different countries are represented.

Regulatory and Ethical Issues: What About Privacy?

Dr. Califf: One thing that comes to mind with this kind of research is regulatory concerns -- missing data, missing people, and you are not seeing them physically. How do you know it is a real person? Is the US Food and Drug Administration on board with this?

Dr. Pletcher: We have not done any drug studies, although we hope to be able to do those in the future. There are some interesting issues about identification and assurance of that identity over time that we will have to think about. We worked quite closely with our institutional review board (IRB) here at UCSF, and they have been forward-thinking about how we can design this in a way that is going to meet the ethical and regulatory requirements of clinical research.

Dr. Califf: For example, with respect to the consent process, there was a Secretary's Advisory Committee on Human Research Protections (SACHRP) meeting a couple of weeks ago. There was an interesting discussion because there has been resistance to online consent in a lot of venues. You have the opportunity to do tests and retest and verify that the person actually learned something. I assume you are doing that with your technology?

Dr. Marcus: One of the exciting things about this study is that although it is a mechanism for conducting research in a more efficient way, the study itself is an experiment in methodology, and so we are very interested in studying the process itself as we move forward. We have learned a lot in working with our IRB to see how reliable these things are. We do have some validation schemes built in as well.

Dr. Califf: What about privacy? You cannot have a discussion about something like this without someone raising the privacy issue. Have there been concerns about privacy?

Dr. Olgin: Everybody brings this issue up, and we thought about it from the very beginning. The approach that we have taken is very straightforward, which is that we let all of our participants know exactly what our privacy policy is. It is very transparent. We treat the data exactly the same as in any other clinical research trial. We do not sell the data. We do not give the data to anybody who is not already part of the study. All of the data are housed within firewalls and are compliant with standard encryption and data safety. We have a certificate of confidentiality from the National Institutes of Health, for example. We think the data are as safe as any clinical research.

Dr. Marcus: The way that the consenting process is built is modular. When people consent to the study, they do not necessarily consent to have their Fitbit connected or an app downloaded. Before we do that, we obtain their consent specifically for that purpose.

Dr. Olgin: The IRB has allowed us to not repeat all the pages of legalese, so our modular consents are very targeted, very easy to understand, and participants are very clear about what is going to happen.

Dr. Pletcher: Our intent is that people will actually read it. If it looks like another legal document is on the way, they are not going to read it.

Dr. Califf: Are these forms publicly available?

Dr. Olgin: They are available to people who join the study. We are happy to share them. There is nothing proprietary about them.

The New Era in Patient-Driven Research

Dr. Califf: This is a good segue into something we are both involved in -- the Patient-Centered Outcomes Research Institute (PCORI). This is a new institute that was formed by Congress. It is not a government agency; it is funded by a combination of the Medicare Trust Fund and a tax on the major insurers. The board consists of members of academia, industry, and government. They decided that with $6 million a year, they could pay for patient outcomes research, and they put their biggest bet on a national network. They put out research funding applications for 2 kinds of networks: clinical data research networks (CDRNs), which have to have at least a million people in integrated health systems, and patient-powered research networks (PPRNs), which is a new concept that did not exist before. You started as a PPRN. Can you explain what that is so a regular person would understand?

Dr. Pletcher: We are very excited about the opportunity to create a PPRN from the Health eHeart study. We have created what we are calling the "Health eHeart Alliance," which is an alliance of organizations and patients who will use the Health eHeart study to do research. It allows us to engage our study participants in the process of thinking about science and coming up with ideas that are going to be useful and relevant and to do research so that the results will actually matter to the participants. It is a very exciting opportunity to do things that we have never done before and really partner with our participants.

Dr. Olgin: And we can engage the participants in all aspects of the research. They are no longer just passive participants in providing data. They are potentially an integral part of everything from conception and design to actually recruiting and carrying it out.

Dr. Califf: It has been a tremendous learning experience for me as part of the coordinating center. We have 11 of these CDRNs, such as the entire city of New York or Chicago. You are involved with 5 of the University of California systems and a huge network together with the Department of Veteran's Affairs, so altogether, up to 100 million people are involved in this. It is amazing, particularly to see the rare disease advocates. They take no prisoners, and they are not interested in our excuses for why we are so slow and not getting answers. As a cardiologist, it has always bothered me that heart disease patients have seemed to be less strident about getting answers. What is your formula to get people activated?

Dr. Olgin: It is a good question. Heart disease is different from many of the diseases that are in the PPRNs and CDRNs in that much of it is risk of developing disease rather than actually having disease. The approach that we have taken so far is driven by the prevalence of risk factors, and we can do something about them if we get enough people involved and learn and develop new prediction science.

Dr. Marcus: We are also fortunate that a lot of the sensors are naturally related to cardiovascular health. They have never been rigorously assessed with hard cardiovascular outcomes. Laypeople understand Fitbits and checking their pulse and their weight, which is a nice synergy for us.

Dr. Califf: That is great. The other thing that strikes me, as I have looked at this, is that every network I have talked to would like to know the heart rates, blood pressures, and oxygen saturation of participants. Those data are relevant to any new treatment or even to a comparison of old treatments. It is a real opportunity for cardiologists to help other people. Let's fast forward 5 years. Where do you hope to be with Health eHeart?

The 5-Year/1 Million Goal

Dr. Olgin: Our ultimate goal is enroll a million people. Hopefully within 5 years we will be close to that. Our main goals are to try to collect enough big data to get better risk-prediction algorithms for disease progression and incident disease and to test new paradigms for chronic disease management. We are at the cusp of a revolution in cheap scalable sensor technology that will enable us to measure such things as beat-to-beat output from a wearable device and beat-to-beat blood pressure in the not-too-distant future. Being able to see the value that these bring to health outcomes will be very important.

Dr. Marcus: In addition to enrolling a million people, we want, from a layperson perspective, for the Health eHeart study to be a well-known household name such that everybody knows what it is and asks: Why are you not participating in it? From a researcher or physician standpoint, it is considered an available resource to answer important research questions. We have already developed an ancillary study proposal. Hopefully it will be a way to conduct research not just at UCSF but throughout the nation.

Dr. Pletcher: We have thought about that from the beginning and tried to build it as a tool that can be used both for observational research and for clinical trials. If you ask about our vision in 5 years, I would love for us to be doing multiple simultaneous clinical trials and observational studies using our Health eHeart study infrastructure.

Dr. Califf: I am going to be out at Google this afternoon. I have been amazed to learn that every time you do a Google search, you are participating in multiple randomized trials, and that is publicly known. I have no understanding of how they got there. Basically they concluded that they could not draw causal inferences from observational studies about how a computer screen affects future advertising revenues, so they developed multiple factorial randomizations.

For my last question, I am going to give you a proposition about where the future is going, and then I will ask you what the average cardiologist should be doing to get ready for this new era. We share a goal of reducing the cost of learning by at least a log order (and hopefully 2 log orders), which means that you can answer twice as many questions. The ignorance that we currently have about the right thing to do will rapidly change if we can scale up the learning system or the research system (whatever you want to call it). Five years from now, let's assume that most people will have wearable sensors, and their medical information will be analyzed in a big data fashion to give them and their providers the best decision support they can get. How should a cardiologist think differently to be ready for that? In the clinic now, we have to figure out what to do without a lot of help. We try to remember or we use UpToDate®. I learned that 80% of doctors just do a straightforward Google search now. What are you advising your trainees, Jeff?

Dr. Olgin: One thing is that patients are going to have much more control over their data, and they will be providing new kinds of data that we have never even thought about in medicine before. We need to figure out how to embrace and leverage that, both for research as well as for clinical management. The second thing is that the line between research and clinical care that has been so black and white until now is going to be blurred. As providers, we should be participating in these kinds of randomized trials that Google does all the time for what is accepted as standard care therapy. We should figure out which of those standard care therapies are really better by doing randomized trials by provider.

Dr. Califf: They randomize millions of people in an hour, and if Health eHeart reached its expectations, you could at least randomize hundreds of thousands in an hour, which would be a major breakthrough. Thank you. It has been a great discussion.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.