Episode Description:
Dr. Robert Pearl, former CEO of the Permanente Medical Group and a professor at Stanford University School of Medicine, discusses his latest book, ChatGPT, MD. He shares how generative AI could transform health care by empowering patients and clinicians through real-time data analysis and insights. These AI tools could significantly improve chronic disease management, helping to reduce the incidence of heart attacks, strokes and kidney failure. In this wide-ranging interview, we explore his vision for the potential of generative AI in reshaping the future of health care.
Guest:
Robert Pearl, MD, author, podcaster, former CEO of the Permanente Medical Group, Clinical Professor of Plastic Surgery, Stanford University School of Medicine
Host/Producer: Carol Vassar
TRANSCRIPT
Announcer:
Welcome to Well Beyond Medicine, the world’s top-ranked children’s health podcast produced by Nemours Children’s Health. Subscribe on any platform at Nemourswellbeyond.org, or find us on YouTube.
Carol Vassar, podcast host/producer:
Each week, we’re joined by innovators and experts from around the world, exploring anything and everything related to the 80% of child health impacts that occur outside the doctor’s office. I’m your host, Carol Vassar. Now that you’re here, let’s go.
Music:
Let’s go. Well Beyond Medicine.
Carol Vassar, podcast host/producer:
Hi, everyone. Welcome to the Nemours Well Beyond Medicine Podcast. With us today is Dr. Robert Pearl. Dr. Pearl is the former CEO of the Permanente Medical Group, AKA Kaiser Permanente, a clinical professor of plastic surgery at Stanford University School of Medicine, and on the faculty at the Stanford Graduate School of Business. He’s also a popular podcast host and a best-selling author. His titles include Mistreated: Why We Think We’re Getting Good Healthcare and Why We’re Usually Wrong, and Uncaring: How the Culture of Medicine Kills Doctors and Patients.
His latest is called ChatGPT, MD: How AI-Empowered Patients and Doctors Can Take Back Control of American Medicine. That’s quite the resume. Dr. Pearl. Welcome to the Nemours Well Beyond Medicine Podcast.
Robert Pearl, MD:
Thank you so much, and for listeners, all profits from my book’s going to Doctors Without Borders, a truly wonderful global charity and relief organization.
Carol Vassar, podcast host/producer:
Hear, hear. Now, your latest book, we’re going to talk mainly about that today, it credits an unusual co-author, and that is ChatGPT itself. How did you come up with that idea, that partnership? How did it work in that practical sense?
Robert Pearl, MD:
Well, in the simplest definition of the practical sense, I approached it as though it was a medical student or resident because I write a lot of papers with medical students and residents, which is that essentially, one or the other of us creates the draft, and then we go back and forth around it. Most of the time, I’m the one creating the draft, and I send it to ChatGPT, it does editing, sends it back to me, and probably went 15, 20 times with each chapter.
ChatGPT provided a lot of additional information. You’ve got to know how to prompt it. Before I even began, I downloaded the 1.2 million words that I had published at the time so the ChatGPT understood my way of thinking. It also had access, obviously, to a lot of the other literature because that’s the way a generative AI tool works. What drove me in the first place was the fact that when I first published Mistreated, I think it was 2015. I saw the system of medicine as the big problem. It was fragmented, paid on a piecemeal basis we call fee-for-service. The more you did, the more you got paid.
The technology was outdated for the last century, although the most common way that doctors communicate, the fax machine, an 1834 invention, and didn’t have really a leader structure. I had hoped that somehow, there’d be some advancement, but nothing much changed. Why did it change? I teach, as you said, at the Stanford Business and Medical School. The Business School, we look at things like evolution of strategy.
Why does it not change across time? There’s always a reason. I came to the second conclusion that it had to be the culture of medicine, the culture that we inherited from our teachers, often 20 years in the past, and that’s how Uncaring came about, but still, no progress happens. Again, what’s the biggest challenge? As I said, we didn’t have a tool. We didn’t have a way to make these advances happen. It wasn’t that anyone didn’t know. We talk all the time about value-based medicine, it’s just that we hadn’t had a tool.
Then, all of a sudden, on December 30th, of 2022, OpenAI releases ChatGPT, and it’s amazing. Even the original one, the free one right now, is very powerful. One of the things that struck me was that this tool is going to become twice as powerful every year. That means that five years from now, it’s going to be 30 times more powerful. It’s the equivalent of a car going as fast as an airplane five years from now. I realized that I had to do two things.
Number one, I had to work with this tool, see its problems, see its faults, see its opportunities, see its advantages. Then number two, I had to move quickly. If I waited a year or two as I did with most books, typical publishing cycle, everything I wrote would be outdated. In fact, even the six months since I’ve published the book, what I see is that the changes are probably doubling of applications into healthcare.
Carol Vassar, podcast host/producer:
As you experience that ChatGPT interface and that partnership, if you will, what kinds of problems did you run up against, aside from the time-lapse that you expressed?
Robert Pearl, MD:
Yeah, so there were two kinds of problems. One kind of problem was just in the evolution of the technology. If I did it today, it would’ve been quarter as hard as it was back then. Downloading data at the time, that’s called GPTs, you call, think of as plugins, connecting your computer with this technology, that was very, very difficult a year ago, and today, it’s much more straightforward.
The corpus of information was two years behind at that time. Now, it’s almost up-to-date in terms of the applications that are being in place one can access, and I’m pretty sure by 5.0, it will be as well. Everything had to be typed. You couldn’t use voice. Now you can use voice. A lot of the things, shortcomings were ones that I knew would get better, I just couldn’t wait to communicate the opportunities that were there. At the time, hallucinations were more common than they are now. I’m predicting that by the next generation,
They’ll be almost gone. At the time, the computer hallucinated an entire expedition to the North Pole. I fact-checked everything the computer gave me. It’s why the book ChatGPT, MD: How AI-Empowered Patients & Doctors Can Take Back Control of American Medicine has a 30-page bibliography because I fact-checked everything. That was the only thing that I found, but it was a big thing, an entire expedition with leadership learning that I don’t think happened. If it did happen, I’d like the listeners to tell me so I can apologize to ChatGPT, but that was the biggest problem of the time.
These, to me, the technology is the small part. It’s going to evolve. As I say, it’s doubling at least every year, maybe more often than that. I’m not worried about that. I do have concerns about whether clinicians will embrace it to the extent that’s necessary to make it as powerful it can be. I believe that done well, it can save hundreds of thousands of lives a year, and make medical care once again affordable for the majority of Americans.
Carol Vassar, podcast host/producer:
Well, let’s dive into that. Outline from the broader perspective, from the high level, how you see generative AI changing healthcare and medicine.
Robert Pearl, MD:
I don’t see any part of healthcare and medicine that won’t change. I can give you a few examples. One of the biggest to me is when I asked myself, either where’s the biggest problem or the biggest opportunity? The same one in American medicine today, why is it that our outcomes are lagging? Life expectancy hasn’t improved since 2010, we have extremely high maternal mortality, we have tremendous healthcare inequality. Why isn’t these things continue to persist?
When I look at this question, I come to the conclusion (that) it’s the epidemic of chronic disease. On the reverse side, if you look at the CDC data, it says that if we could better control heart attacks, and strokes, and the other chronic diseases, that we could reduce the number of heart attacks, strokes, kidney failure, and cancers by 30 to 50%. Now, think about that. If our country had 30 to 50% fewer heart attacks, strokes, kidney failure, and cancer, what would happen to our health? It would soar. What would happen to costs? They’d plummet.
I think one big opportunity has been management of chronic disease. People could ask them logically, “Well, how’s ChatGPT going to do that?” There’s a lot of ways that it could do that. It’s just going to help patients to better prevent and manage their chronic disease. I think the biggest opportunity is that today, we actually have a teacher at the Business School, as I said, and across the campus at the Engineering School, you can find Band-Aids sized monitors, blood pressure, pulse, blood glucose, blood oxygen, almost everything, 99.9% reliable.
Why are they not being used? Because no one really wants [inaudible 00:08:19]. Take hypertension, you’re diagnosed with new hypertension, your doctor gives you a medication to take. Let’s say you had a generative AI tool that could measure your blood pressure every single day, or a monitor, a wearable device that could measure it, and a generative AI tool available around it to collect the data. What are you going to do with it? As a patient, 92 of your rings are normal, and eight are abnormal.
You’re doing great because 92 are normal or terrible because eight are not. As a clinician, you don’t want a hundred blood pressure measurements. We don’t get this data. We don’t actually strive to find it. Now, imagine if the generative AI tool is giving readings back to the patient at the end of the month. It’s a simple calculation. Look at first and second derivatives, see how it’s moving. At the end of the month, it either tells you, “You’re doing great, keep it up,” or at the end of the month, tells you, “You’re not making any progress. In fact, your hypertension is getting worse.”
Clinicians could change it. What are we doing in practice today? “I’ll see you in four months.” What happens at four months? You come in, your blood pressure’s elevated. What do I say? White coat syndrome. We don’t make the changes, and that is why it lags. Hypertension is only controlled 55 to 60% of the time. It accounts for 40% of strokes, major contributing to kidney failure. You look at diabetes. We control it a third of the time. This is a leading cause of heart attacks, and kidney failures, and peripheral amputations.
Think about how much potential is there, and we don’t do it, not because we can’t monitor it, not because of the medications, not because we’re not smart enough, we just don’t know what to do with that data. Generative AI solves that problem. What’s another example? Take a patient. How do we empower patients to understand the problems that they have?
I was on a different podcast a few months ago, and at the end, I was talking to the host, we were off-air, and she said, “Dr. Pearl, you’re a skier, you’re a doctor. My husband fell skiing about three months ago. His arm was over his head. He slid down around a hundred feet. His shoulder still hurts, and he can’t use it like the opposite side. What’s going on?” I said to her, “You said you’ve never used ChatGPT. Why don’t you go home, play with it, put the data in place, see what you find out, and then call me? Tell me what you find. I’ll tell you what I think it is at that time.”
She calls me five days later. ChatGPT said, “Probably a rotator cuff tear. He needs an MRI, and you should consult an orthopedic surgeon because he most likely needs surgery to repair it.” That’s what’s called expertise, not just knowledge. You go to Google, and you click on a link, a bunch of links, (and) you get a lot of information. You get 20 different or 50 different diagnoses of shoulder pain. You get 20 or 50 different reasons why one arm doesn’t work as well as the other arm, but you don’t know specific to yours or your husband’s, in this case, situation.
Then she said the surgeon who operated after he diagnosed the rotator cuff tear and got the MRI said, “If I’d waited three more months, I probably could not have gotten the tendon reattached because the muscle would’ve contracted.” That’s the power. Of course, in the United States today, we have 400,000 people who die every year from misdiagnoses. We know this technology today is actually probably as good as the typical physician. By five years from now, it’s going to be better. How do we use it in that way? How do we fill in the gaps?
You see a mental health provider one week, you don’t see that provider again, let’s say, for two more weeks. In between, what do we know? Nothing. Generative AI can provide that assistance. This is not replacing clinicians. This is augmenting, or hospital [inaudible 00:11:44]. You have hospitals today, patients, a quarter of the people in hospitals today don’t really need to be there. They’re there mainly for monitoring. Imagine if they could go home with a quiet environment where they can sleep, and have the food that they’re used to, and be surrounded by family members.
How do we monitor them? Today, we send nurses out. It’s very expensive and time-consuming. We have a nursing shortage nation. Generative AI hooked up to monitors could accomplish that. Now, if you now link that device into a telehealth center, you can have hundreds of thousands of patients being monitored by a small number of clinicians. Why is that? The technology will tell you who’s in trouble or at a hospital.
I have a bunch of orthopedic issues where I have broken bones that require surgery and hospitalization. The nurse comes by at 8 AM. By noon, she comes back. In between, she has no idea what’s going on, and we know that patients whose clinical status deteriorates, and they go to the ICU, and they’re resuscitated, have a four times higher mortality at three months.
How do we change that so we don’t wait for the monitor to ring to tell us that the patient’s about to have cardiac arrest or the blood pressure is precipitously low? These are just some of the opportunities in my mind. They’re almost endless.
Carol Vassar, podcast host/producer:
You looked at, if I’m not mistaken, from your description of how you worked with ChatGPT, you looked at it as a colleague, as working collaboratively. I’m hearing in the examples that you’ve just provided, looking at generative AI in healthcare as a colleague, and working collaboratively with medical professionals, with patients. Am I on the right track here? Is that the vision that you have?
Robert Pearl, MD:
What you said is totally accurate, but let me add one more piece. It’s the triad: a dedicated clinician with an empowered patient and a generative AI tool. That triad, very soon, if not today, will outperform any of the three alone. Yes, to me, that’s colleagues, collaboration, and cooperation. However, we want to think about it. It requires not just the technology, but the patient empowerment, as well as the dedicated clinician to use it.
I think when we do all of these things, we start to diminish the amount of chronic disease, we start to address another problem, which is the burnout in the medical profession because what people haven’t factored into our thinking is if I see you with right lower quadrant pain, and I diagnose appendicitis and take out your appendix, maybe I’ll take care of you for five days, and then you’re all better. As soon as you walk into my office with severe diabetes, type two diabetes, I’m going to see you four times a year for the rest of your life.
That’s over a hundred visits that I now have to take care of. The workload has become massive, and all we have done is pile things on and make clinicians go faster and faster. We’re down to 17 minutes across this nation on average for a visit. This tool can pick up a lot of the slack. When we lower heart attacks, strokes, kidney failure, and cardiovascular disease, we end up it is far less demand on our clinicians, as well as better health for our patients and lower costs.
Carol Vassar, podcast host/producer:
What are the downsides? This sounds too good to be true.
Robert Pearl, MD:
Every technology that we use, whether it’s the electronic health record, whether it’s going to be a search engine like Google, doesn’t matter what it’s going to be, there’s always issues of privacy, and security, and misinformation. This technology, I think will be actually slightly better, but it’s still at risk. In the short run, we have to make sure the hallucinations are taken care of. What exists today, I would not rely on for some of the applications I just described to you, but they’re coming in two years or three years.
This is not 20 or 30, some futuristic view. Probably these companies could release a version now that did all these things, they’re just making sure that it’s there. Safety is an issue, of course, to some of the existential threats that sit in play. I think the big risk is that we’re going to miss the opportunity. In the book, I write about various eras. You have the late 20th century, the 1900s, when we had massive advances: CT, MRI, ultrasounds being introduced, cardiac surgery, transplantation.
We’ll go down a list of advances that came into play. It was a beautiful and brilliant era. Then we get to the 2000, introduction to the EHR. I can remember people talking about how information will be available with every clinician you see. It doesn’t matter where they’re going to be, and today, it still doesn’t happen. Why is that? The companies have blocked the ability to make information interchangeable between different providers for their own benefit relative to the products that they sell.
Then you get the third era, which was introduction of the internet. Here, we had the opportunity to emulate what happens in travel or finance. We’re able to do so many things virtually, and again, rather than seizing the opportunity, we resisted, and now we’re in this fourth era, the era of generative AI. I think that the tools that we have, the opportunities through a generative AI application, are so much greater than anything we’ve had before.
If clinicians don’t embrace it, if they don’t encourage patients to use it, if they resist it, if they create barriers because they want to hold onto the past, then I think it too could easily become a missed opportunity. I think that’s the biggest challenge and danger that we face.
Carol Vassar, podcast host/producer:
As we look at generative AI, and healthcare inequities, and the social drivers of health, will technology, generative AI specifically, be able to help us address social drivers and make healthcare more equal?
Robert Pearl, MD:
I think it will do a lot to advance it. I don’t want to minimize the problems of the social determinants of health. If you are homeless, if you have food insufficiency, food desert around you, but you can’t obtain clinicians’ help because there’s just not enough of the community where you live, if there’s a problematic job, or what we saw during COVID, where people at lower socioeconomics had to take buses and subways to get to work, whereas other people could stay home and protect themselves against the infection prior to the availability of the vaccine, the inequities are going to be there.
We need to focus on housing, and food, and transportation, education, business, et cetera. I think that the people who are suffering right now the most are those who are socioeconomically challenged. They’re working two jobs. Every doctor’s office is closed when they’re off, so they have to go to the ER for care, and that care becomes episodic and not any kind of long-term chronic disease management. They can’t access primary care because there are no primary care doctors in their neighborhood. You can go on and on.
They can’t get the diagnostic tests they need because Medicaid doesn’t cover it. We have a whole lot of reasons that are there, and they’re going to stay there. I don’t want to deceive anyone that we’re going to totally close the gap, but I would say that these are the people who can take the greatest advantage of it. Now, there’s a challenge, and I wrote about this in my Forbes article that I write twice a month for Forbes, and that is that there’s issues of broadband and access to the computers themselves.
It’s why I believe we need to have community health centers. I don’t mean the typical health center staffed by a lot of clinicians. I mean one that’s staffed by the kinds of individuals who often work in the government and the social services. They know the language, they know the neighborhood, they grew up there. They could help sign people up for a variety of programs and they could teach them how to use generative AI. Now, you start to have a leveraged model. That’s going to elevate the bottom more than it’s going to elevate the tops.
The tops are already getting the expertise and care. They may be getting it more expensively, they may be getting it inefficiently, but at least they’re getting that expertise. Now, you could make it available, whereas before this, the only way you could provide that care was to hire more clinicians, and that was the fundamental problem. They’re either going to be too expensive or likely to be in the neighborhood. In terms of telemedicine, the challenge of how do you get the care when you don’t have broadband sitting in your home?
Carol Vassar, podcast host/producer:
You’ve raised an interesting point, Dr. Pearl, in that educating people about generative AI, about how this will benefit them, even physicians, learning that generative AI can be used to the advantage of their work and their patients. Aside from having those community wellness centers, as I’m starting to think of them, what else can be done to really educate the interested players, the necessary players?
Robert Pearl, MD:
It takes about ten hours. People have done some studies to get really good at it. It doesn’t really matter if you’re the patient or the doctor. You’ve got to learn how to do prompts, you’ve got to learn how to ask the right questions. If you ask the question wrong, you’re not going to get the right answer. Medical school, that’s what we learn in medical school, and then we use that with our patients. The same thing, I think, with the technology.
I encourage everyone to, I’ll say, to use it. Not to depend on it totally, but to use it. By that, I mean, think back to a time when you saw your clinician last and you had a new problem, and put in all the information that you would’ve told ChatGPT. Then, figure out ways to put it in better and better. It also hurts if I move my left leg, if I walk stairs. Yeah, it also hurts if I go downstairs because those are different, or it often hurts if I’m churning, or whenever the specific information is going to be, and start to see both the quality of the answers, which is amazing, as well as the ways to improve that by making it more specific.
This experience that has to happen in that way, and as I said, downloading data is going to be problematic. I think, however, that the electronic health record is so problematic right now, and we now have an opportunity to put all that information into place. Most patients have situations where clinicians can get access to the written record, they can get access to the laboratory data, the radiologic findings, medications, obviously, that they’re taking. They can make sure the medications are the ones they’re really taking, at least the frequency, if not the dose, that they are taking daily, on a daily basis.
I think it’s this opportunity to get that experience. It’s like any other expertise. It doesn’t just come on day one. You got to work at it and develop it. Then at some point, you say, “Oh, my gosh, now I really know how to use it.” That’s what I would encourage the listeners who may be clinicians or may be patients, people giving care, receiving care, try it. Just see.
If you’re going to see a doctor tomorrow or today, go into the application, put all the information in place, engage in a dialogue, and ask it the questions you would ask the clinicians ask it, “What is the likely diagnosis, what are the treatments you’d recommend? Which one do you think is best? What are the complications of that particular treatment or that particular medication?” I think that you’re going to be thrilled by how much alignment there is between the technology and the clinician.
If that’s true, just think about that. The difference between a clinician taking care of a patient who comes to you with no knowledge about their problem or with extensive knowledge. Now, doctors hate the patient who comes in with a 100-page printout off of Google with every link being clicked on. Now, why is that? It’s not curated. If you look at GPT, the way that technology is created, by the way, I don’t mean just AI, because the AI of the past, which is called Narrow AI, where large data sets were used to compare it to each other or diagnose and read mammograms or other diagnostic studies, this is a totally different technology.
It’s pre-loaded with everything on the internet. It has all the medical textbooks, all the journal articles, it’s then transformed. It provides 2 billion ways that data can be looked relative to each other. Data from more recent journals is more significant than old ones. Why? The new journals have the old information in it. Peer-reviewed journals are more significant and more valued than ones that are not. You go down a list of these different ways. That’s where I think that we need to be. I would encourage people today to start experimenting and then see what the doctor says and compare them.
When you walk into that office and the patient, like the woman I spoke about, whose husband had the shoulder problem, when the doctor said, “I think there’s a torn rotator cuff,” she didn’t have to ask, “What’s a rotator cuff? What does it mean? What’s a tear?” She knew that already. When [inaudible 00:24:02] said, “I think he needs an MRI,” she said, “Thank you.” She didn’t have to question that. When he recommended surgery, if you go almost anywhere, people say, “See a doctor.”
Why do they say that? It’s not to help you. It’s not that you need to see a doctor. I’d tell you that to protect me, the designer of the program, because if you don’t, if something goes wrong, you’re going to blame me. In this case, the technology was so specific. See the orthopedic surgeon because you probably need surgery. Understand how different, what is knowledge, and what is expertise. What generative AI does is elevate expertise. Patients are not going to be doctors. Doctors train for a decade to do it.
Doctors using it will be better than doctors who don’t, but patients can become very knowledgeable, maybe to the level of a medical student, maybe to the level of a resident. When with that information, they can leap far higher in the diagnostic and treatment trajectory.
Carol Vassar, podcast host/producer:
The way you describe it, it sounds like we’re not going to lose the human element of the practice of medicine. Am I interpreting that correctly? I think there’s a worry on some people’s parts that all this technology, all this data, all of this generative AI is going to make it all very rote. The human element of medicine is really the art part of the medicine. Talk about how humanness stays in.
Robert Pearl, MD:
I would say it’s almost the opposite, and let me explain why. There was a great study at the University of Arizona where they took the messages that doctors and patients exchanged with each other. The patient had a question and sent a message to the clinician. The clinician responded back with a message. Then, they gave the same questions to a generative AI tool and had its answers.
Then, they brought in both patients and clinicians who were given all of these answers, not knowing whether they came from a clinician or from the technology. Lo and behold, what’d they find? Doctors rated the information given as four times better when it was given by the technology than the human. Patients rated it nine times more empathetic. Think about that. The technology is more empathetic than the human. Now, it’s not that the human didn’t have the emotion. The human didn’t have the time. The human couldn’t prioritize it.
That is the challenge that sits in place. I think it’s going to be the opposite. What it’s going to do is to liberate the clinician. When you’re seeing patient after patient every 17 minutes across the day, it’s very hard to maintain that traditional doctor-patient relationship. I’d say it’s almost impossible. You can barely get through the acute problem, though this is all the chronic disease. I saw an interesting study that said that for the primary care physician to address everything that was recommended in the literature, there’d be 57 steps, and it would take 27 hours a day.
It’s impossible. To me, if this technology frees clinicians up, if you’re doing well after a month on that blood pressure medication, you don’t have to come in. That frees up a visit. If a patient has a new symptom and a problem, and it’s the kind of thing that could be taken care of with medical advice, you can do a quick telemedicine visit. You can empty your workday, whereas today, you just race your tongue, it’s hanging out of your mouth, you can barely get to the finish line. At the end of the day, you go home not feeling like you did the best that you knew you could do, not because you didn’t want to. It just was physically not possible.
Today, we’re asking people to lift like a 200-pound weight. Maybe there’s a few people who could lift it fine. I can’t lift 200 pounds. I don’t know if you can lift 200 pounds, but now we’re going to lighten that weight, and whether we can lighten it to 100 pounds or 50 pounds, I don’t know the number that we can get to. We have to learn and try, but if we can take a third of the patients who today are coming to the office and find a way to solve their problem without bringing it to the office, we’ve now created the added time so that we can invest in the patients who can benefit the most.
We run our practices as though everyone is basically the same. No, there are some patients from whom we need an hour. There are other patients who’d rather not see us in person at all; they just want an answer, and we don’t have any ability to sort that out. We offer the same kind of care to everyone. This technology accomplishes that. It tells us who’s at the greatest risk, who needs the most attention, who’s doing just great, and who needs just a pat on the back.
Carol Vassar, podcast host/producer:
Dr. Pearl, what keeps you up at night when it comes to the practice of health and medicine?
Robert Pearl, MD:
As you said, everything I’ve talked about sounds great, except this one problem with it. If we lower heart attacks, strokes, kidney failure, and cancer by 30 to 50%, what’s it going to do to physician income? In the book, I talk about anything that either slows the clinician down, like the EHR, or anything that undermines the income, and that was the risk of some of the internet tools that are there, won’t be embraced.
I believe that we’re not going to be able to use generative AI to solve the challenges of medicine today, save 100,000 lives a year, and lower the cost of care until we can move from a fee-for-service pay-for-volume. The more you do, the more you get paid. If there’s any good or not, that doesn’t matter. You get paid twice as much if you see a patient two times to solve a problem you could have solved in one visit. We can go down the whole list of problems that are there.
We’ve got to move to a capitated model, and it’s got to be capitated not like today at the insurance level, because the insurance companies get paid a single fee, but then they pay doctors a fee-for-service, and all the problems come up again, but at the clinician level. We need therefore groups of doctors working collaboratively, cooperating with each other, use this technology together, sharing information to be able to get the best results.
What keeps me up at night is will our nation be able, in time, to make this shift from fee-for-service to capitation? If not, I think what we’re all going to just do is we end up having an increase in the disparities, the inequities that you and I talked about earlier. We’re going to see our life’s expectancies continue to stagnate. We’re going to consider all the problems of today to continue because why would they be any different? I also know the opportunity is out there to be able to make the change, but we have to shift that reimbursement model.
What I write about a lot right now, so I think about it a lot at night, I know how it can happen. The only question on my mind is, will it happen? If not, then I’m going to be very, very sad because it’s going to really erode the practice of medicine. You talk to clinicians, they were having problems in their own office, and it includes private practice. They joined hospital systems, so they brought in private equity, and now they’re suffering twice as much. They’re being driven to do things that they don’t want to do that not is not in the best interest of patients.
We can go down the whole list of ways in which their life has, it’s a different set of problems, but it’s certainly no better and probably worse. I see the profession at some point deteriorating, people deciding they don’t want to be doctors because the problems are so great. We’re lucky because today, we still have a lot of superb applicants into medical school, but that’s my fear, and that American medicine, by the way, like medicine in some other countries, will become a second-rate job, and it will not be paid the way that it should. It will not attract the talent that it should. It will not have the value, the esteem that I believe that it should.
To the clinicians listening in, this is the time that we need to make the change, and I encourage you to lead the way now that you have the tool. Maybe you couldn’t do capitation before because you were so afraid. How am I going to lower the cost? If the cost goes up, I’m at risk. Now, what we have is the tool to accomplish it. Sure, you got to have a good negotiation. You can’t change it in one year. We can change diabetes and hypertension in one year, but we can’t change the complications. They take longer, five years.
How do we get into a collaborative type methodology? This is what keeps me awake. It’s actually going to be the next Forbes article that I write, and if people want information about it, they can go to my website, RobertPearlMD.com.
Music:
Well Beyond Medicine.
Carol Vassar, podcast host/producer:
Fast-paced and engaging conversations like that with Dr. Robert Pearl are what we serve up on a regular basis here on the Well Beyond Medicine Podcast. What and who would you like to hear from next? Leave us a voicemail on our website, NemoursWellBeyond.org. While you’re there, subscribe to the podcast, check out any episodes you may have missed, and leave a review. That’s NemoursWellBeyond.org.
Our production team for this episode includes Lauren Teta, Susan Masucci, and Cheryl Munn. Next time, we’ll talk the magic of collaboration between healthcare partners and the benefits for children’s health. I’m Carol Vassar. Until then, remember, we can change children’s health for good, well beyond medicine.
Music:
Let’s go, Well Beyond Medicine.
