Episode 29

COVID, AI, and the Future of Medicine

with Eric Topol, M.D.

September 30, 2021

Eric Topol, M.D.
Founder & Director, Scripps Research Translational Institute

Eric Topol is the Founder and Director of the Scripps Research Translational Institute, Professor, Molecular Medicine, and Executive Vice-President of Scripps Research. As a researcher, he has published over 1,200 peer-reviewed articles, with more than 250,000 citations, elected to the National Academy of Medicine, and is one of the top 10 most cited researchers in medicine. His principal scientific focus has been on the genomic and digital tools to individualize medicine.

In 2016, Topol was awarded a $207 million grant from the National Institutes of Health (NIH) to lead a significant part of the Precision Medicine (All of Us) Initiative, a prospective research program enrolling 1 million diverse participants in the US. This is in addition to his role as principal investigator for a flagship $35 million NIH grant to promote innovation in medicine. Prior to coming to Scripps in 2007, he led the Cleveland Clinic to become the #1 center for heart care and was the founder of a new medical school there. He has been voted as the #1 most Influential physician leader in the United States in a national poll conducted by Modern Healthcare. Besides editing several textbooks, he has published 3 bestseller books on the future of medicine: The Creative Destruction of Medicine and The Patient Will See You Now and latest Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again in 2019. Topol was commissioned by the UK in 2018-2019 to lead planning for the National Health Service’s integration of AI and new technologies.


We have a breakthrough disease problem, and I have gotten a lot of backlash for saying that, but we have to tell the truth, even if it's a hard truth.



Gary Bisbee  0:06

Washington, D.C., is my home away from home. I’ve worked here for the better part of three decades as a founder, entrepreneur, policy expert and author.

Don Rucker 0:10

Probably the longest title. Everybody sort of shortened it to ONC for sanity’s sake.

Gary Bisbee 0:15


I’ve learned leadership secrets from many health care executives who understand that Washington is the largest payer and regulator of health care.

Nancy-Ann DeParle 0:25

She said, well, because you’ll never get a husband if you do that.

Gary Bisbee 0:29

I began interviewing health care leaders many years ago because what better way to learn how they think, why they make it to the top and how they remain there?

Think about, what was your most challenging engagement.

Greg Carpenter 0:40

Health care has been the most difficult problem, as you said.

Gary Bisbee 0:43

We’ll talk about that later.

Have any lingering questions about COVID-19? Then you’re in the right place. Today we’ll hear answers from Dr. Eric Topol, an expert who’s been on target about the pandemic from the outset. We dive into the nettlesome issues for which clarity is difficult to come by. We discussed the fact that responses from the FDA, CDC, and the White House have too frequently seemed uncoordinated and not infrequently have lacked in transparency. Dr. Topol discusses COVID testing and asks the question, why isn’t there a greater emphasis on rapid testing? We discussed whether a booster shot is necessary and why, whether we can mix and match vaccines, and how protective natural immunity is. We explore the topic of how we can better prepare for the next infectious disease crisis. Dr. Topol is a professor, scientist, and author. We discussed his classic book, Deep Medicine, which provides a rationale for artificial intelligence in medicine, how AI works, and the overall theme of the book: how deep learning can return humaneness to healthcare. Artificial intelligence provides opportunities for better diagnosis, care, and greater efficiency. While we’re in the relatively early days of the digital health revolution, it’s being led by innovators and consumers who demand and deserve better control over their care.


Good afternoon, Eric, and welcome.


Eric Topol  2:18  

Thanks, Gary. Good to be with you.


Gary Bisbee  2:20  

We’re pleased to have you at the microphone today. This show is about leaders and you’ve been a prominent leader throughout your career hard even to enumerate all your various leadership tasks, but certainly as a cardiologist as a professor, as a scientist, as an author, editor, founder, what do you do in your spare time, Eric?


Eric Topol  2:43  

Thanks, Gary. That’s kind of you. These days my spare time is taken up by our grandchildren. We have three of them here in San Diego and have a blast with them, so that’s all our joy for us.


Gary Bisbee  2:58  

Certainly. What ages if I could ask?


Eric Topol  3:01  

Sure. Seven, three, and eight months.


Gary Bisbee  3:05  

Your hands are full. What we’d like to chat about today is your early years and how you developed as a leader, and then move to your views and wisdom on Coronavirus. You’ve been a prominent thought leader throughout the pandemic, then move to Deep Medicine, your book on AI and deep learning that is already a classic, so congratulations on that. To kick off, what was life like growing up for you, Eric?


Eric Topol  3:39  

I was kind of a precocious kid. I finished my schooling before going to the University of Virginia. I just barely had hit age 16, so I had two grades that I skipped along the way. I landed in Charlottesville and I was way behind all the other kids. I didn’t. I was socially a misfit, so that’s how things got started for me. That’s where I remember finally trying to assimilate, trying to get socialized with the other. For the other kids, I knew it was a challenge. But fortunately, it worked out and I had it had a great time there. Some of my memories at UVA are among the greatest of all, but I started much younger than all the other kitchens by the second and third year I was the Resident Advisor on the halls in the dorms. That brought me up to speed pretty quickly and got me on a really good footing.


Gary Bisbee  4:45  

It certainly did. What did the young Eric think about leadership?


Eric Topol  4:50  

I know I didn’t have enough respect for it. I didn’t realize how vital it was. I was somewhat rebellious. I guess I still am to some extent. It took me a while to understand it’s a sensuality. That’s something older age has reinforced over the years.


Gary Bisbee  5:10  

How about your parents? Did they influence your leadership style at all?


Eric Topol  5:16  

My father actually was diabetic, insulin-dependent. He went blind at age 49 and wasn’t able to work, so there I watched him really struggle, even before he went blind from retinopathy, he was really having a rough time. It was more what both my parents dying young, having lots of chronic medical problems, influencing more, and to that extent, not so much how I could be a leader, but more how healthcare is so precious, and the lack of it is so profound.


Gary Bisbee  5:56  

We’ll get to Deep Medicine later, but you certainly cover that in that book. What about your first leadership role? Do you remember a specific leadership role where you said, “I enjoyed this, I want to be a leader more?”


Eric Topol  6:13  

The first one actually was in college to be the RA for two years. And then after that, throughout med school, there wasn’t really an opportunity for me to take on a kind of a leadership role. As I moved on in medicine and cardiology, I had increasing respect for but really what it was, when I went to my first job at the University of Michigan, there, I was heading up an interventional cardiology program and trying to really spearhead clinical trials brought together a group of friends and people. That’s when I started to get into a groove of how I could be a useful force and leader, how my ideas and innovation could help bring people together and inspire them. It was back in the late 80s when it finally hit me that we could do things that were really potentially impactful work together. And I could provide that kind of glue and the force to help us achieve what we had otherwise not expected we could do. We were the young Turks when we were taking on some of the more established people in the field of cardiology, and it was really interesting.


Gary Bisbee  7:35  

When did you decide to practice medicine or become a physician?


Eric Topol  7:41  

That’s going back to college. I actually was really interested in genetics and I thought I would be a geneticist. I even did a thesis on prospects of genetic therapy in men in ‘75. But I needed to work in college to help defray the costs. And so I wanted to many different jobs. But the one that really influenced me was at the University of Virginia hospital, I got a nightshift job as a respiratory technician. In that job, I was working with the ventilators and equipment and seeing people who were in the intensive care unit who had a Lazarus type effect where they were brought back to life and it really influenced me. I thought maybe I should get into this medical thing because that’s really exciting, so I changed gears along the way. That job really had a notable impact. That’s when I decided to finish the requirements to take the MCAT and apply to medical school.


Gary Bisbee  8:44  

Your career has grown from being a cardiologist to really inclusive of being a scientist, which gets back to your interest in genetics, maybe in college. As I look at you, I think of you as a scientist. Do you think of yourself that way?


Eric Topol  9:02  

Yes. I took a detour for a stretch when I was doing cardiology procedures, balloon angioplasty, and stents for a number of years and headed up that the heart division at Cleveland Clinic for Credit Suisse, but along the way, I never lost my interest in the basic science, particularly genetics and genomics. And as that became more feasible in the mid-90s, I got back to it. And so I always had this hankering for the science of fundamental aspects. And so I was fortunate that I had the opportunity later in my career to really delve into it. And in fact, I made a conscious decision to get out of clinical trials and to do hardcore research in genetics again, so I went from 1975 and it’s almost like a 15, 20-year gap, and got back right into what I really love.


Gary Bisbee  10:01  

You’re superb at it for sure. Where did the interest in wireless technology come from? You were one of the early proponents of that for medicine, and then, of course, AI.


Eric Topol  10:12  

That was really interesting. I kind of fell into that. In the late 90s, when we were starting to see potentially a coalescence of medicine on the internet, there was a company that actually, ironically, was based in San Diego called cardio net. And they said they could monitor their people’s cardiogram and heart rhythm through the internet. I found that intriguing. Basically, I was sent a packet to evaluate it, and it kind of had a said, “Whoa, this is going to be hard, interesting. This can be done.” But what had the biggest impact was when I came to San Diego. So now we’re at end of 2006, beginning of 2007. And while my interest had been kindled in, in this wireless medicine thing, now I went to this conference, we’re talking about February 2007, that it was organized by Qualcomm, which is the number one company here in San Diego, a wireless medicine, wireless company, not at all in wireless medicine at that time. Anyway, at this conference, there was this presenter, of the idea of having a camera in a smartphone, connected to the internet. And I was sitting in the back of the room, I said, I just woke up. And these people were arguing Why would we ever want that? Why? Because we have point and click phones that we have in our pocket that point and click cameras in your pocket that are so high resolution? Why would we ever want to do something stupid like that? Of course, it was November that year when such a thing as an iPhone was created. All of a sudden, I had a eureka moment that, if you could take pictures of stuff like a skin rash or who knows what, you had sensors in your phone, now you’re connected to the internet. So I came to San Diego to start a human genomics genetics Institute. There was no Ingenico, but that was the day I decided no, we were not going to do that. We’re going to add this whole wireless digital medicine. So we have the first academic program in the United States. And what was interesting about that, Gary was that we started to see the bigger picture was you can understand people just through their DNA, me these other layers of information. And sensors, like a camera, or other wearable sensors, would give us this extra dimension to understand people and potentially provide better health care.


Gary Bisbee  12:51  

It sounds like that was the impetus to what now is your interest in AI and individualization, and so on. So that’s very cool. Well, if we could turn for a moment to the coronavirus pandemic, then we’ll come back to Deep Medicine and all that involves, but one of the questions is in regard to testing. It seems like once the vaccines became available, our interest in testing waned. Does that make sense to you? Is that wise, do you think?


Eric Topol  13:24  

No, not at all, testing more than ever, especially rapid testing, the problem we’ve encountered is there was this vaccine-centric strategy. There was never an embracement of a rapid home test, which, of course, could help us navigate schools and kids going into schools and other staff and teachers, and has been used in so many other countries very effectively. But here, we bank too hard on the vaccines getting us out of the woods. And what we didn’t anticipate was this delta strain that was going to be so formidable. And they’re even people who are vaccinated are not fully protected, and don’t know that they may actually be incubating or harboring or actually going to get sick with a Delta virus infection. So we need those rapid tests to keep everyone protected. And that’s something that I hope we’ll see. Because we started the US pandemic, without any testing for two months got way behind the virus spread throughout the country. Now we have a similar problem of not having adequate testing and we’re not measuring up to countries that have really excellent performance.


Gary Bisbee  14:39  

Is there a good home rapid test? Is it a question that the test is actually available, we just don’t use it here?


Eric Topol  14:47  

There are 67 of them approved in Germany. There are two clear in the US. The issue here is it’s not even there are many very good tests there. Dig it down to less than five minutes with higher accuracy than the ones we even have available here, which are too expensive. And in very short supply, they work best when the government of the country gives them out for free, like in the UK and Denmark, Germany and many other countries. And so that’s what we should be doing. Because we want to promote people knowing of their infectious, that’s where they’re that sweet spot. Within minutes, you would know you had to stay home. Don’t go in or interact with other people because unwittingly, you may well be infectious. We don’t have the FDA buy into this principle here. Nor do we have the governmental support of its urgency and utility.


Gary Bisbee  15:44  

Unfortunate to say at least. Kind of relates back to our whole strategy and resources available for public health, which, which really needs to pick up. But in terms of the vaccines, we’ve seen a lot of heard a lot about breakthroughs. How do you think about these breakthroughs? How should we interpret that?


Eric Topol  16:06  

The term sounds kind of alarming, “breakthrough,” but actually, what we’re really concerned about is people getting ill, to the point that they may even have to be in the hospital, or actually in the hospital, or in the ICU or dying. It’s one thing if you just happen to have a test that is positive once you’ve been vaccinated and you don’t really have symptoms or they’re very mild, but then so many of these people that have these breakthrough infections are getting quite sick. I’ve had many colleagues—because we got vaccinated very early in December, January—who have gotten quite ill right on the verge of going to the hospital or even being hospitalized, so this has become a real problem because it wasn’t envisioned. We didn’t have it at all in the first six months, essentially. Only when July started and the Delta variant took over in the US. It turns out, it’s probably more time and it is delta, but it’s some type of interaction of the two because delta is so much more infectious. So now I think this is a problem. We need to get a booster third shot program moving we have lots of infighting among the leadership of the US at the FDA and NIH and CDC and the White House. So this isn’t good. This is not a good recipe for moving forward. But I hope we’ll get the resolution and get the follow the science. Obviously, there’s this competition of global vaccine equity. But there’s no reason we can’t do both because we have lots of vaccines that have been distributed in this country that could get us on solid ground to prevent these breakthroughs that are much more worrisome in people over age 60, or in health care workers.


Gary Bisbee  17:58  

Are you comfortable with the science we’ve seen out of Israel? I think they were testifying in front of the FDA today, and what we’ve seen reported from Pfizer and Moderna and J&J. Are you comfortable with the science there? Are we really positioned to think that we all have this third dose?


Eric Topol  18:22  

I have to say, I think the Israeli scientists are first-rate. I know several of them having visited there, and what they’ve done, because they are a laboratory for the world, they’re the only place that has millions, over 3 million people who’ve had a third shot. And they’re following them. They have, of course, electronic health records for everyone so, even though it’s a small country of fewer than 9 million people, we’re learning a lot from them. What bothers me is when people just discount their data and say it is not a value. Now, it isn’t the same representative of the US data. There’s a different lack of diversity relative to our country. But it provides unique insight. That data is vital, especially. We don’t have the public health commitment to have the right data in our country, so we’re very dependent on outside countries that collect and analyze their data, and Israel is doing a great job of that.


Gary Bisbee  19:25  

Question is arisen about mixing and matching. In other words, if you had a first shot that was J&J, Pfizer is now available or Moderna may be available as a booster shot. Should you take that when you can get it if you started with J&J, do you have any thoughts? Is there any science on that, Eric?


Eric Topol  19:45  

Yeah, there’s quite a bit on that. It isn’t so much on the J&J. But we can impute that from the ad know, viral vector of AstraZeneca because they share a lot of features. What’s remarkable—if you take an antiviral vector first and then an mRNA, whether that’s Madonna or Pfizer as the second goes—you get the best vaccination immune response that’s ever been seen more so than two mRNA shots are more than asked to AstraZeneca. So that would be the best if you’re looking forward to the most potent immune response. That would be it, especially if you separate eight to 12 weeks between them. Because that’s another thing is that Pfizer, three weeks dose spacing was probably too fast. And it didn’t allow for as much of a memory and the immune response as we’d like. Now interestingly, we don’t really have any data to support the opposite direction. If you had an mRNA vaccine first, will you benefit from J&J? No, we don’t have any data to support that it doesn’t look nearly as promising.


Gary Bisbee  20:57  

What about the inner nasal vaccine? I’ve heard you talk about that for various points and times and it sounds like something that we should be really digging into. But I don’t know that that’s happening. What are your thoughts about that?


Eric Topol  21:12  

Yeah, it’s frustrating because we basically double down and put everything on a shot, but shots are activating our immune system in our bloodstream, not in our upper airway, the lining of our nose, and mouth and throat. And so this is a problem because that’s where transmission is in Sweden, we haven’t had a transmission-centric plan. And so while we got behind all these manufacturers of the shots, we haven’t gotten behind the nasal and oral vaccines. In fact, just last week, an oral vaccine that had can be stored without any refrigeration, and guess they basically this new so-called mucosal immunity, which is that barrier established of your airway. That’s what we need now. So Well, actually, we needed that at the beginning of the pandemic, I hope we can put our pedal on the metal and get this done because there are 20 of these out there. But we haven’t supported it. And these are often startup companies, and some have considerable promise. So I hope we will do this. And an oral vaccine is very attractive, particularly if it doesn’t require any refrigeration.


Gary Bisbee  22:29  

What you’re thinking about natural immunity, I’ve heard a lot of discussion about that. I’m not sure what the science is, but there seems to be pretty solid immunity if you’ve actually contracted COVID. What’s you’re thinking about that?


Eric Topol  22:45  

Over the months now, we’ve seen a lot of new data to tell us that prior COVID natural infection is a very important type of immunity. And the reason why is when we give a vaccine, we’re just giving basically the spike protein, that’s just one part of the virus. Whereas when you get infected, you don’t want to be infected. But if you do, you’re basically making antibodies to many other key parts of the virus that come into play. And it looks like you develop very durable antibodies and cellular T and B cell responses. The problem is that that’s not good enough. That is if you add a vaccine, one dose on top of the natural immunity, you get superhuman, and it’s called hybrid immunity. So we have these people out there you’ve had COVID and they think that they’re bulletproof, but they’re not. But they would be pretty much if they just would get the darn one-dose vaccine. One of the things that is really unfortunate is we have no plasticity at the CDC. We have these cards and you have to have two shots or else you don’t get your card. I don’t know why we’re using cards in 2021. Anyway, what they already have a line for is if you had prior COVID and you have a test to back it up that counts that’s as good or better actually than one of those have any vaccine so we just can’t figure that out. And then another thing is so many countries do that. Now we just can’t have we’re so bureau pathak about our practices. It’s just silly stuff.


Gary Bisbee  24:31  

Are we going to reach a point of what they call herd immunity or population immunity?


Eric Topol  24:37  

Population really would be nice. We would have had it if we didn’t have delta. We were looking really good at alpha by May, this country, early June. It was as good as it ever had been and had Delta never come and arrived in this country, which of course was unavoidable but we were getting there. But now it’s a reset because this is so hyper contagious. Now we need 85% of the people in the country 85% or more with either vaccination-induced immunity or will take prior COVID ideally with the one dose. But that’s 85%. We’re at 54%. So yeah, and if you add in some private COVID, we might be at 65%. We’ve got 20% of more Americans that need to get vaccinated in order for us to be inside because we have such a ferocious strain of the virus now.


Gary Bisbee  25:33  

It sounds like the plan for employers to mandate which mandated vaccine is causing a lot of political problems, but that would probably hasten our movement toward getting at 20%. You’re talking about vaccinated? Thinking about transparency, starting with the snafu at the CDC on testing seems like 100 years ago, but as I talked to leaders and health systems and health insurance, there’s been a “lack of trust,” developing that space, perhaps on just seemingly a lack of transparency, without trying to be political about it. What’s your thought about that? How can we develop the trust again, and our agencies here?


Eric Topol  26:19  

They all have great intentions and some really sharp, extraordinary people, but the communication has been dreadful. The best example recently, as you can appreciate, is when over a month ago, the president says we’re going to have a booster program. And we’re going to give everybody who is eight months, and it’s going to start on September 20. And that lead rankled the people are ultimately two key players resigned from the FDA and the CDC and some of the people FDA revolted. And then they changed from eight months to six months to five months back to six months, and who knows what will be next Monday. This is not communication that is acceptable. And it has to be much more careful and consistent and driven by the science and evidence. And so I‘m kind of stunned. I knew how bad it was last year. But this year, I actually thought we get in the groove. But unfortunately, the miscues have continued. And I hope they will improve and get to where they ought to be they need to be because that’s what engenders trust. I started putting out on Twitter several weeks ago that we have a breakthrough infection breakthrough disease problem, which hadn’t been acknowledged by CDC. When I did that, I got a lot of backlashes: “Oh, no, you’re gonna help feed the anti-vaxxers.” What if you tell the truth and you get the public so that those who’ve been vaccinated gear up and wear masks? And know that there are some liabilities out there that, if they got vaccinated early, they may be at risk. That was instructive for me that people don’t want the truth. They want “happy talk.” If you just give “happy talk,” people just like that. That’s not the way to communicate either. We have to tell the truth, even if it’s a hard truth.


Gary Bisbee  28:27  

If we were talking two, three years from now, what do you think we’ll look back on as lessons learned from the pandemic? Communication is the obvious one, you’ve been talking about that. What other lessons do you think we’ll learn about our approach to public health?


Eric Topol  28:45  

We got it our public health in many respects, well, before the pandemic, we were printing trillions of dollars to the Afghanistan war, while our public health commitment and resources were eroding. It was never as good as it should be. Even though CDC commanded global respect. We need to have pandemic preparedness like never before and we should be leading the world in this. There’s no excuse that we don’t have stockpiled vaccine readiness for any virus. They’d been on the list of the top three or four, whether it’s influenza, Coronavirus, NEPA, whatever we should be ready so that the day that hits we’re squashing it. Obviously, there’s an interaction with climate change that’s forcing a lot of this, we should be all over that as well because it’s the interaction between the climate effects and the pathogens, that’s also putting us at increased risk globally. So we have to work with all of the players in the world, we have to have a much more deep commitment. And I can’t just be money, you just throw billions or trillions out. That’s not enough. We have to be innovative and we have to have a commitment to and cooperate with all the forces on the planet.


Gary Bisbee  30:04  

Let’s turn to a happier note, which is Deep Medicine and how AI can help return humaneness to healthcare. I love the book, it’s just really an action-oriented book. I hate to ask this, but how can you consolidate 300 pages into a very short interview? But could you just give us a kind of a top-level view of Deep Medicine and what you were accomplishing with that book?


Eric Topol  30:31  

It’s counterintuitive. Most people would say, “How could technology enhance humanity? How could that make us more unique?” I think that’s what grabbed me, that everybody’s talking about AI to do this or that and in medicine and health, but the overarching potential that’s so exciting, is how we can use it to bring back the patient-doctor relationship, the way it used to be, where it was precious, there was this deep empathy and exquisite bond, human to human bond. And so we can use AI to do that if we go after it, and that there are so many other things that the book gets into the AI could do it still, you’re in the early phases of that. If we don’t lose sight of where we can go with this, we actually could get a quality of medicine that you and I would want, all of us would want, that your doctor has your back and really cares for you. The flip side is there’s a global crisis of burnout and depression among clinicians. And why is that a lot of it is because people feel they’ve lost their ability to care for patients, they’ve lost their way. So we could even get that back, which is the morale and the excitement of what it is this noble privilege that we have of being trusted to care for patients.


Gary Bisbee  31:46  

The last decade was important to this with the HITECH Act, which basically digitize medicine, and also an explosion of technology. Where would you rate us today? Eric, in terms of just the technology and data available to personalize medicine and pursue some of these optimistic sorts of things that are in the book, where do we stand? Are we in the first inning, second, third inning?


Eric Topol  32:32  

It’s hard to put it in an inning. We’re certainly not at the seventh. We’re probably in that first or second inning. We’re long on data. We have whole genome sequences, sensors galore that capture data continuously, almost any physiologic system of a human being, we have no shortage of data, but our ability to process that data in a meaningful way. We’re serious, we’re severely impaired. Where will we be eventually, in the years ahead, is we’ll take all of a person’s data, their biologic layers, their anatomy, their physiology, their environment, their electronic records, all of those data, and all the pertinent medical literature up to the moment for that person to prevent illness. That’s where we can go, that’s what’s so exciting. But we are not even we haven’t even taken the first baby step. We haven’t even stood up as a baby yet sat up, but we will over the times ahead.


Gary Bisbee  33:36  

Yeah, we will. I’m confident of that. If we were here at the end of this decade, Eric, I think Deep Medicine would look like an absolute roadmap for how we were going to proceed. What clinical applications of AI exist today. Can you point to one of them that is a clinical application that uses AI and it’s useful today?


Eric Topol  34:02  

One for physicians that’s most widely adopted and still early is the interpretation of scams. So already it’s getting into many health systems for radiology, all different types of scans. One that is pretty well validated prospectively is diabetic retinopathy, which going back to the earliest part of our conversation is an important area that I see for 50% of diabetics never get screened, and it’s a preventable cause of blindness. So that can be done in grocery stores with AI now you were a person sits down and they have a retina picture taken in AI deep learning algorithm gets your immediate answer a whether you have any retinopathy and what should be done about it. That’s something I feel ophthalmology is zooming ahead of and has probably made the most momentous controversy so far in AI. A lot of people don’t realize it. And in Europe, skin lesions are going to be anyone be able to get a differential diagnosis with probabilities by taking a picture from their smartphone. So things are moving for sure.


Gary Bisbee  35:17  

How are physicians accepting this?


Eric Topol  35:21  

Not well. They feel that somehow or other, if this change is not acceptable, there are some good reasons for it, like lack of randomized trials and the kind of compelling data we all would want. But too bad reasons are feeling threatened. Not understanding that this could be the greatest thing that ever happened in terms of making their lives easier, better, restoring the ability to really care for patients. So I hope that over time, we’ll see physicians and clinicians in general embracing AI, we need that. Because if they do that, not in a blinded way without understanding the nuances and liabilities, but if they are more open-minded, we will make progress. Medicine changes very slowly and is less associated in this country with more reimbursement. It’s a sclerotic, ossified, hard-to-move profession.


Gary Bisbee  36:21  

What can we do to increase adoption and kind of speak to and perhaps eliminate some of these barriers to use clinically?


Eric Topol  36:32  

I’m hoping if we gear up with doing the right kind of prospective, rigorous work, that is transformative, we will see that it will be irrefutable evidence to move ahead. There’s a commitment now. We have new standards for the leading journals of what they’re going to accept. And so the, we’re fortunate that just in the first five years of having an AI in medicine, we’re seeing a commitment to high-quality efforts, so we’re going to start to go at a pretty high clip and making progress.


Gary Bisbee  37:11  

Chapter nine in your book, Deep Medicine was entitled “AI and health systems.” You brought up the idea of predictive modeling. in that chapter, what’s your thinking about where we need to go with predictive modeling, it strikes me that that’s when we’ll have to work some magic with some of the physicians to get them to embrace that. What are your thoughts about that?


Eric Topol  37:38  

There are two levels area of predictive modeling. One is when a patient comes in the hospital, that you would have the AI give you a heads up about their risks that you might not be able to simulate, you might be able to see it from your clinical acumen. The other one that is even more intriguing, from a standpoint of how big it could be, is the ability to do remote monitoring and keep almost the vast majority of people in their home rather than in a hospital when I was involved in the National Health Service UK review because it’s it has universal health care, the commitment to that was easy compared to this country where hospitals are a major lobbying force, but for modeling who can be kept at home, and we’re not decompensate anticipating when there are the earliest signs of potential decompensation, we will go there, we will not use hospitals as we do now in a decade or more. So that’s exciting. Health systems in this country are not aware of that opportunity largely because they rely on the hospital’s $1.3 trillion, the number one line item of American healthcare, so it’s gonna be hard to change the mindset on that.


Gary Bisbee  38:58  

It’ll be hard, but I’m a subscriber to the fact—and your book points us out—the individual is going to lead the way. They want what you’re espousing in the book and pretty soon they’re going to not accept anything but that. Do you have that feeling as well?


Eric Topol  39:18  

Yes, the people that are following this, particularly, they’re savvy, that we should not at all, underestimate what you’re bringing up, which is democratization of medicine. That there’s a yearning for having more charge more control, being able to capture their own data, get interpretation. We’re going to see almost all the common diagnoses, whether it’s, I mentioned skin, ear infections in children, urinary tract infections, heart arrhythmias, and on and on and on. They’re going to be able to get a lot of those screened by themselves. Someday they’re going to say I’m not going into hospital, I’m having the sensors in my home, and you can send help if I need any kind of thing. We’re not there yet. We’re not at the point where people have smartphones, and they’re doing their own smartphone ultrasound imaging throughout their bodies. But we’ll be going there someday too in some respects.


Gary Bisbee  40:18  

No question about it. You mentioned the health systems. The health insurers probably in the same boat aren’t perhaps aware of how fast this might come when you meet with health system leaders or health plan leaders. What advice do you give them for being aware of what’s coming down the pike here?


Eric Topol  40:41  

There are a myriad of opportunities with AI, and we haven’t scratched the surface. There are a lot of back-office operations that can be replaced with AI, human scribes, we have something like 60 70,000, human scribes that go into a clinic, visit with the doctor so that the doctor can actually talk to the patient. We don’t need that we should have synthetic notes and that is much better than you write the notes we have. So there are all these different ways that we can implement different types of AI in our daily practice of medicine, to not just save costs, but deliver better care, but also the idea that we get to have more time with patients. That is the gift of time. I look forward to the day when health systems in each city are competing with each other, giving time for patients and doctors to come together. Not seven, not 12. But much more than that time. The gathering is a critical metric that we don’t have right now.


Gary Bisbee  41:49  

I agree with that for sure. I see more and more founders starting companies that have to do with AI. I would imagine they kind of make a pilgrimage to your Scripps Research. Do you see an increasing number of founders with good AI applications for healthcare?


Eric Topol  42:11  

Oh, absolutely. There are 100s, perhaps well over 1,000 in every aspect, whether it’s medicine in life science, drug discovery, it is incredible because this is the biggest potential shakeup in the history of medicine, just because there are so many different nodes of entry to disrupt when you can automate things when you can be more accurate. That’s not something we have we haven’t really discussed yet. But we have a serious problem, we medical errors in this country. And if we can start to get those down to whenever we get down to zero, but 20 million a year of serious diagnostic errors that you and I all of us are going to have one, at least in our lifetime. We can’t afford to have that continue, so there’s a lot in the diagnosis space. But also we’re starting to see things more in terms of predictive modeling, treatment, remote monitoring, you name it. There are AI startups and the tech Titans, Amazon and Google and Microsoft, and Salesforce and all these others. They’re making a big play on this as well.


Gary Bisbee  43:24  

Eric, unfortunately, we’re running out of time. This has been an awesome interview. We’re very grateful for your time. If I could ask one last question, going back to the beginning about leadership, we have a number of early-stage leaders in this audience, what advice would you have for an early-stage healthcare leader?


Eric Topol  43:46  

The idea is to not accept any dogma challenge, and work with your team to kind of be that inspirational force of questioning, not accepting things that are widely accepted. Oftentimes that leads to better ways to do things or new ways to think about things. That’s what I’ve been doing for decades, and that goes along with telling it like it is, which is something we discussed the pandemic. Reading that cultivating that the constant vetting and reassessment of things are really healthy, it’s important. That’s what we should be promoting among young people. So that they are not just passive, that we all become activists and question things continuously, and come up with better ways. Because whatever we’re doing now, there’s a better way of doing it in the future.


Gary Bisbee  44:43  

Dr. Eric Topol, author of Deep Medicine, our audience needs to buy it, needs to read this because it’s a good one and it’s important for the next decade or two and healthcare. Eric, thank you so much. Terrific interview.


Eric Topol  44:57  

Thanks so much for having me. I really enjoyed our conversation.

Subscribe for Updates​

For exclusive access to Think Medium content and program updates, subscribe here.