[Dr. Robert Pearl]: And I saw this sign that to me still sticks in my mind. I think about it all the time. Quality, access, cost. And the bottom is small letters that said pick any two. That was the best that we could do until now because our mindset was always going to be if you provide better access it costs more. If you want a higher quality, it’s going to cost more. If you want a lower cost, you got to limit access. It’s always going to be this tradeoff.

I think we could probably take out somewhere between one and two trillion dollars out of health care spend if we could just do the things that we know are possible that we don't do because the system rewards volume not superior clinical outcomes. So I think the change has to happen in the reimbursement methodology.

I think that generative AI in the United States will lag every other country no matter what the three of us say, no matter how much opportunity there is until we make this evolution. But as soon as you make the evolution now the technology becomes so—"

[Junaid Kalia, MD]: Good morning everybody. Um, thank you for joining Signals and Symptoms today. Really appreciate it. We have an exciting guest today, Dr. Robert Pearl who if I start talking about him, he has been the CEO of Permanent Kaiser Permanente. I mean apparently this year alone he has given 40 keynotes. He's at Stanford. I'll let him introduce himself properly.

[Robert Pearl, MD]: Well uh I'm not sure quite sure where you want me to start in this process about myself. I teach at the Stanford Medical and Business Schools. I uh be I write for Forbes. I have my own podcast called Fixing Healthcare. I have my monthly newsletter that has almost 100,000 subscribers now. And people want more information, they go to my website, robertmd.com. As I said, I teach at the Stanford Medical School as well as the business school. And I can't find a single student who's not using a generative AI tool every single day. I mean, preparing for rounds, learning all the material sitting in textbooks, uh, trying to figure out the tools can be so much better.

So, we're looking at a very fascinating to me cultural clash that's coming up because the people coming into the school are going to be smarter than their professors. That's never really happened. If you think about it, um, up to this particular point, the medical students had no idea what was going on. Uh, they had to learn basic minimal terms. Then they had the always the professor who would come in and give the lectures and teach the pieces and that's still going to be true, of course, for some of the medical information. That's why I think this cultural piece is so crucial but they have a skill that their professors don't and I can't think of a time in the history of medicine and that's going to be incredibly disruptive.

At some particular time I think patients are essentially going to realize that the quality of care they're going to get are going to be is going to be so much better. There's 400,000 people who die annually in the United States from misdiagnosis. Another 400,000 die from not die from but permanent disability from misdiagnosis. That's a huge number of people. And what we know is this technology is going to be as good as or maybe better than it's a second opinion. Imagine for the same price without any added time you could have two doctors. That's really what you're going to be getting in the exam room. I have and I'm sure you do too Harvey. I have hundreds of stories of people whose diagnosis was made possible because the technology was used either by the physician or the patient.

Three years in is not a time for cultural change. It takes a lot longer than that. But it's not going to stop the evolution of this technology. As I said earlier, I think we're going to see these agents who are able to provide care 24by7 continuous monitoring. I mean, think about it. The biggest challenge United States today is chronic disease. 70% of care is 70% of the time that we spend with clinicians is over chronic disease. The CDC says that if we could maximize our care in chronic disease and it's not the impossible level just matching the best of the United States today. What we see is that a 30 to 50% reduction in heart attack, strokes, kidney failures and cancers. I mean that is such a powerful target to go to. And why don't we get there? Because we take care of a chronic problem with an intermittent solution. We see patients in our office every 4 months for a disease that is there and changing every single day.

We could start an a a medication for hypertension and within a month know how well it's working and make an adjustment. Why don't we do it? Because there's no clinician in the United States, in the world probably today, who wants 100 blood pressure readings every single month. But give that data to a generative AI tool first and second derivative calculations. Now we would know very very quickly. The same for blood sugar, the same for heart failure. We could take all these problems and literally monitor them. The hospital at home, what's the limiting factor? We want to know how the patients doing. There's a great opportunity to be able to use a general tool to do it.

Think about in the hospital. You know what do nurses do? They go from patient one to two to three to four. They don't you know maybe it's 4 hours between seeing the patient once and seeing the second time. This tool can tell us who is in trouble. The equivalent of a uh telemetry can be provided for every inpatient not just the small number who are sitting there with a nurse sitting at a screen looking at 12 or 14 different individuals out of might be 200 patients in a hospital at any given time. The opportunities are massive and when you have that many opportunities, you have lots of companies who are developing it. And so the real question to me is going to be will this be a pull or a push? Will be physicians and clinicians who pull this along or will they get pushed by companies that create it and patients who then demand it.

[Junaid Kalia, MD]: So very beautifully said. So we can start with Harvey this side. Why is Qatar and Dubai and Morocco is leading the way in terms of this in terms of Singapore Harley? What I why why there's a cultural slowness, laziness or whatever drag on the US just because of the system is set up or what what are your thoughts on that?

[Harvey Castro, MD MBA]: For me, I think the you know Dr. Parl all three of us are physicians. You know, it's a certain culture, right? You know I always talk about Martec’s law which is basically uh the technology is exponential but our organizations and even ourselves are logarithmic and when the gap increases increases to the point where it's too much a disruption happens and that's why I love uh you know certain hospital systems that are very progressive because they're disruptors they're disrupting the ones that are not that are not doing and so what I love what he said because what's going to happen is our society is going to say “Hey, I need X.”

And I love going to Singapore as an adviser for the Ministry of Health there. I'm seeing things. I'm I'm seeing different technologies and I'm thinking unfortunately in the United States if the culture doesn't change, this is not going to happen here. But I truly think we're going to get FOMO, fear of missing out. We're going to be like, whoa, Dubai is doing this, Singapore is doing this. And going out to our legislators and saying, why is the GDP so expensive when we could do it a different way? And I really think that pressure is going to make change. Dr. Pearl, what do you think?

[Robert Pearl, MD]: Let's go back to look at what happens. You know, the so-called, "Well, I hated doc Dr. Google coming to my office." What was that about? Well, you had patients going and clicking on links through a browser. What could they do with that information? I mean, we have trouble interpreting some of the articles in the literature. They can't even know where to start. So, they're hunting and pecking for answers. Generative AI is a totally different technology. It provides expertise. We have to understand we are now at a moment in history that completely changes the world.

You know the establishment is not going to lead this process. But when you have a tool and I really think that it's not far away where you actually can obtain medical expertise, reliable expertise using a I'm talking a patient at home. We are going to see an explosion and we're going to have to be building it into our curriculum because we have to, medical students need to understand how you take care of a patient, how a disease impacts and that's not a technological question. That's learning how you actually become a clinician.

[Junaid Kalia, MD]: Well said. Very well said. Um again I'm going to keep pushing you back on the cultural change issue and uh both of you have a global view and a generation view but long story short now you're you know uh as I completely agree that there would be tools available and then physicians are going to what how the provider patient relationship is going to change. We always always talked about it should be a joint and partnership and everything but at the end of the day I am writing orders like the word is order MRI brains.

Now let me challenge you guys again. Uh Charlie Wonger once said tell me where the incentives are and I'll tell you what the outcomes will be. Okay. Now you perfectly painted a picture both of you. How do you think that overall reimbursement environment from both governmental perspective and insurance perspective will a evolve over time? What would be the push backs and rate limiting factors and when when that how will we be able to realign hopefully finally with the help of AI the vision of health not sick care?

[Robert Pearl, MD]: This is the most important question the one you just raised. I love the Charlie Munger quote. I I often use it myself. Um in my book ChatGPT, TMD, I say doctors will never do something that either takes them more time or cuts their income. That's the history of of medicine in the United States. It's, you know, every every system of medicine, whether it's private or public, has ups and downs. And that's the uh challenge because something that's going to benefit patients but have one of those two impacts is not likely to be embraced.

And you know I teach a class it's uh 10 weeks. So it takes a while to be able to do it. We will have to evolve the the reimbursement system in the United States if we want generative AI to be used. And so again this is pushpull situation. I think it will happen because we have to use generative AI.

Now if you look at I'm a big proponent of capitation for the reasons you say for the Charlie Munger quote because in a capitated system the reim it starts to emphasize prevention chronic disease control it means rather than using hospitals 5 days a week you start to use them 7 days a week you can go down the whole list of ways that are there I think we could probably take out somewhere between one and two trillion dollars out of healthcare spend if we could just do the things that we know are possible that we don't do because the system rewards volume not superior clinical outcomes.

So I think the change has to happen in the reimbursement methodology. But I'll also point out that this is my third book. My first book was called mistreated where we think we're getting good healthcare where usually wrong was written in 2017 and I talked about some of the same things that needed to happen but we didn't have the tool to make it happen. So, if I tell you guys you've got to take risk, but I don't give you a way to win or have a high probability of winning at the risk game, you're not going to do it. And that's what we tried to do for so long.

The idea that you can capitate an insurance company and then pay doctors and hospitals on fee for service and expect you're going to get a different outcome is absurd. It's not going to happen . particularly if you don't have 100% of someone's practice. You go back and look in the political arena of the United States when the ACA was passed. How did they create accountable care organizations? You couldn't it didn't require they did a compromise. You couldn't require the patient to tell the doctor they were part of the ACA. ACO, how do we know what's going on? It's all a broken system.

And I think that generative AI in the United States will lag every other country, no matter what the three of us say, no matter how much opportunity there is until we make this evolution. But as soon as you make the evolution, now the technology becomes so powerful . for this opportunity to say I don't need to see everyone because I know how they're doing. I'm monitoring them every single day. It's not that it's a second rate care. It is so much better. It's like having your own private nurse and private doctor in your home every day evaluating your health and giving you information along with a nutritionist as well as a lifestyle coach.

You can go on all the pieces that are there. You know, every Sunday night, this is the evolution. And I think it happens naturally as soon as you move from pay for volume to pay for superior outcomes. Higher quality, easier access, lower cost.

[Junaid Kalia, MD]: I 100% agree. We are actually going to move towards a capitated system which actually does go from outcomes. We've already seen DRGs being implemented in the hospital system. I still get a lot of pitch texts and everything but a lot of uh pharmaceutical innovation is now happening because of generating AI. What are your thoughts on improving how AI will have an impact on that process?

[Harvey Castro, MD MBA]: Yeah, I'll jump in. Uh big picture as you know we you mentioned the pharmaceutical space. Uh there's several companies saying that they can now decrease the cost of a drug from 2 billion to 100 to 200 million. So that's a 5 10% um cost now. And if that's the case and that brings us that much closer to personalized care leveraging AI. Um, I literally just posted uh today about digital twins and my personal journey and I basically created a rag model digital twin and I've been using it for everything and as a result I've been able to lose weight. Um, my numbers are looking great, everything. And I'm pushing my biological clock back, but I'm leveraging AI and using my physician degree, putting it together to do this. And the cost is nothing because of this technology. And I wouldn't be able to do this, you know, 5 10 years ago, but I can today. So I my point is this. This technology is getting cheaper. Us, the humans, our creativity, we're putting it together. And how can we take it to the next level? And that's how every single person here on this call is showing the future.

[Robert Pearl, MD]: Well, I'm not as optimistic that the generative AI tools, as opposed to the narrow AI tools are going to be that much that helpful in terms of the creation of new drugs. But where they will have I think major impact is going to be between the creation of the drug and getting the final FDA approval for all the reasons that Harvey just talked about because it is going to be able to demonstrate efficacy and it's going to be able to uh be able to run a lot of clinical trials at a very very low cost and that as Harvey talks about is uh you know is often a limiting step for companies but I want to piggyback a little bit on what Harvey was just talking about. He was talking about keeping himself healthy and the longevity sitting in place.

You know, after I was appointed the CEO in Kaiser Permanente, I was invited to give a talk at the university at the Oregon Health Science building and at the end of my talk, I had 30 minutes till the car was going to take me to the airport. I walked around and I saw this sign that to me still sticks in my mind. I think about it all the time. It was, you know, the typical student-made handmade sign. I'm, you know, sitting on the wall. I'm not sure who created it. Across the top in big letters that said quality, access, cost and the bottom in small letters that said pick any two.

That was the best that we could do until now because our mindset was always going to be if you provide better access it costs more. If you want higher quality it's going to cost more. If you want a lower cost you got to limit access. I mean this is the model that sits right now. You look at prioritization. What's it designed to do? Lower cost by diminishing utilization. It's always going to be this tradeoff. And now I want to say that the future of medicine is the realization that by increasing quality, by increasing access, you lower cost. It's a complete flipping of the world on top.

But to accomplish it, you accomplish it by using the Generative AI tools to be able to do the things that Harvey said to be able to keep people healthier, to prevent disease, to better control chronic disease. Because it's not that doctors didn't know what to do, they just don't have the time to do it. And that is what this tool is going to do. You know, it's going to expand in non the non-healthcare world. It's going to make things bigger. It's not going to take away from people.

But let me add one last piece. When I talk about capitation is at the delivery system level. We're talking about to clinicians and to hospitals. Because if you don't get it at that level, you don't change the doctor patient interaction. You don't shift from asking how do I do more things that are more complex that I can build more. You have EHRs not designed around building but around clinical care. You don't do the your mindset just goes in the wrong direction.