Eric Topol wears many hats. He is a practicing cardiologist, genomic researcher, chief academic officer of Scripps Health, and the head of the Scripps Translational Science Institute, one of more than 60 biomedical centers supported through an ambitious National Institutes of Health program created nearly 10 years ago to move basic medical research into the clinic.
Before he moved to San Diego-based Scripps in 2006, he helped create a medical school at the Cleveland Clinic, where he was cardiology chairman. He was a vocal critic of the painkiller Vioxx, which was eventually taken off the market. He left the Cleveland Clinic amid controversy after it revoked his academic position.
Topol is a leading proponent of the use of emerging technology to make medicine more transparent and empower patients to make more informed decisions, a topic laid out in his 2012 book The Creative Destruction of Medicine. Xconomy met Topol in San Francisco before a healthcare conference in late May (at which he riled up some of the physicians in attendance). We discussed technological progress—and lack thereof—in medicine and how doctors, patients, drug makers, technologists, and other actors in the world of healthcare are adapting to the rapid changes. What follows is an edited and condensed version of our conversation.
Xconomy: In recent years one of your top initiatives has been to create a medical school and a training situation that’s much more tech-savvy. You were lamenting how little med students were using technology and integrating it. Is that still an initiative you want to pursue?
Eric Topol: It’s been difficult to pursue. We were trying to get a medical school going at Scripps. It didn’t come to pass. We had funding but the president at the time at the last minute didn’t want to go through with it. I had set up a medical school at the Cleveland Clinic, but it wasn’t until the latter part of this first decade that it became apparent that there wasn’t any curriculum anticipating the dramatic changes. A couple med schools are trying to get their students ahead of the curve by giving them high resolution ultrasounds instead of stethoscopes, by having genomic medicine in the program, but overall there’s not much progress. It’s actually misplaced to put the priority on med students. If we wait for them, we wait for a generation of physicians. Turns out almost 55 percent of physicians are age 55 or greater in this country. We want to get to them. Problem is they’re busy and not inclined to do Internet courses. The governing organizations—the AMA and such—are not exactly what I would call tech-supportive or savvy. I do think we have the power of education through Web-based remote learning, but we’d have to make it a certificate, or obligatory. It has to have some teeth in it.
X: So that’s one interface between doctors and technology. But we’re only just now putting together the infrastructure to combine layers of big data, all the “’omics,” clinical data, and so forth, into some sort of actionable data that doctors and clinicians can use. What will it take to get that “soup” of data we’re creating into a form that doctors and clinicians can use?
ET: The New York Times had a story that some basketball players are hiring their own personal data scientists. Doctors aren’t necessarily going to have that. But we’re going to have more data per individual, plus of course what we’ll get when we collate lots of individuals. It spans not only all the different ‘omics, but sensors, imaging, electronic records, social graph stuff. That can all be had now, this panoramic view of a person, but making it useful is not going to be the physician’s charge. What we need are the algorithms, the deep learning from that data, not just at the individual level but at the cohort level. That has to be processed in a simple way to connect the dots.
The problem here in San Francisco, all these great people can do this work, but they’re not medical or health related. They’re working for Twitter or Pandora or whatever, they’re not working in this space because they’re young and this is not their thing right now.
X: How much sensitivity do people coming from the tech side have around regulation and privacy? They’re entering a new world that people in healthcare have inhabited all their working lives.
ET: We’re talking Facebook stuff versus medical information. You download an app and press “I agree” but you haven’t read anything, which no one has. There’s a big disparity between selling your data that you liked something, versus do you have some kind of mental disorder and the whole world’s going to know you’re taking lithium for bipolar disorder or whatever. There’s [a] big issue here of lack of protection in this kind of medical big-data story. HIPAA [the federal health information privacy law signed in 1996] doesn’t touch this. This is about selling your data unwittingly, with the potential—if you have somebody’s genome it’s not unlikely you could identify them. Same holds true with other pieces of each person’s data. De-identification is a concern. The two biggest obstacles of bringing big data into medicine and reaping the benefits of digitizing human beings: number one, we don’t have the analytics and informatics tweaked right, because we don’t have the expertise dedicated to it so far. The other is privacy and security.
X: What are some early examples of healthcare applications and benefits that have come about with this confluence of data, connectivity, and so forth?
ET: There are many examples. You can do your cardiogram on your phone and get a real-time interpretation with an algorithm built into your app. It’ll tell you what your cardiogram shows, and that can pre-empt a visit to the emergency room or an urgent physician appointment. I can get an e-mail now from a patient
