Healthcare’s “Top Deck of the Titanic Moment”

Erik Sunset: [00:00:00] Hey folks. And welcome back. This is the DocBuddy journal. I’m your host, Erik Sunset. We are recording this episode today on Thursday, October 19th. This is episode 40 of the DocBuddy journal. So thanks for listening along as the show has grown and evolved, it’s been a lot of fun to get to episode 40 and here’s to many more of them. And this week, we’re back after spending a couple of great days in west Palm beach for the bone society of Florida annual conference. Shout out to Lindsay and Frazier for hosting another fantastic event. And this week is kind of the calm before the storm, before the Becker’s ASC events in Chicago next week. Be sure to visit DocBuddy while we’re all there together to see op notes to see surgery workflow, that way you can revolutionize your facilities, operative report process. Happier surgeons, happier staff, faster revenue cycle. What’s not to like, so we’re in booth number two 10, which is on [00:01:00] the left-hand side of the exhibit hall. We hope to see you there. And for this week. We want to cover the hottest tweet in health. It, I guess, used to be called a tweet. Now the platform is called X. I’m still going to tweet. I don’t know what to call it. If it’s not a tweet. But this particular thread is coming from a guy called will men. Ditis. who’s the CEO of science dot I O. It’s got a really alluring first tweet here. Healthcare is having its top deck of the Titanic moments. Is it let’s see. For full disclosure, we don’t have any affiliation with science.io. Don’t know anybody that works there don’t have a relationship with will personally. But the top line of their website says that. Their AI transforms medical texts into enriched data so that you can build solutions that improve patient care. Turning our [00:02:00] attention to the tweet thread. It’s gotten almost 2 million views at the time of recording the threads on a couple of days old. And for those that aren’t familiar with a tweet thread, you can obviously have just a single tweet that goes out into the internet. Or you can make it more of a long form tweet with a series of them. So when I say tweet thread, I’m just referencing a series of tweets all on the same topic. And we thought this was a really interesting. Take from, well, this is much more through the tech lens. So think Silicon valley think traditional technology. Vendor, as opposed to the traditional health, it. Vendor and through that lens, if that makes sense, this is a lot more of a disruptive point of view. As opposed to something that already exists and either legacy health it or established health it where you’re seeking to improve an existing framework or an existing process. This is much more of a disruptive look like we said. [00:03:00] So in pulling up this thread, we’re going to go through it and then we’ve got some, some thoughts. We’ve got some devil’s advocate points, a couple of critiques, maybe. It’s all in good fun. And this isn’t meant to be an assault on the the essay that we’re going to go through here. But there is a very distinct difference between those on the outside of healthcare seeking to disrupt. Versus those that are in it, either doing the disrupting themselves or simply making too. Looking to improve physician workflow as a part of the system, right. So let’s go through it. Healthcare is having its top deck of the Titanic moment. What happens in the next year will define the next century of American healthcare. And basically everyone is ignoring it. Here’s the real story. It’s a pretty compelling entry there. We’ll define where we think the the thesis is correct and where maybe it’s not so right towards the end here. So here here’s the tweet thread in its entirety.

Healthcare has been defined by [00:04:00] four factors. One extremely limited supply. Doctors are scarce and costly. Principal agent problem with payments. Your insurance pays not you. Inelastic demand and high trust is 0.3. So you need it and have confidence in it. It being healthcare. And number four regulatory capture. This has allowed the industry to behave in incredibly weird ways. Costs can constantly grow. Patient experience can constantly degrade and clinics will still fill up with wait-lists because patients trust it, need it. Aren’t paying for it and have no choice.

In the last years we’ve seen all four of these factors shift. One supplier is expanding. It’s hard to imagine a world where 90 plus percent of care isn’t L L M assisted within the decade. And LLM is large language model. Think chat, GPT. After years of getting brutalized by EMR is we’re close to doctors having technology [00:05:00] leverage. Too, as commercial insurance, degrades patients shift from a payer, PCP centered model to an army of online telemedicine services for each and every condition they might have. This ends with Amazon being your doctor and pharmacy. Cash please. No insurance. Welcome here. Three patients have lost trust in the system. This begins with the medically assisted gutting of America with opioids continues through the pandemic and the horrible experience in clinics today. The answer is not patients demanding better. It’s patients exiting entirely. And that’s exiting the traditional healthcare system entirely. Just for clarity.

Number four, the FDA continues to be slow, to regulate new methods of care that blur the line between medical and nonmedical pill, mills, and other online cash for prescription schemes are allowed to operate for years while patients increasingly asked for nonmedical. Health optimization. [00:06:00] The pieces of the stack that have exists that have existing regulatory entrenchment. UMR and so forth are even stickier than they’ve ever been. When the agency moves slow. It has the effect of allowing captured to persist and entrenched to absurd degrees. There’s no world where we untangle these and I think will is right there, but not necessarily for the reasons he’s giving. We’ll come back to that. Yeah. I just spent the last month bouncing between various healthcare conferences and the party goes on. We have free physician quality, artificial intelligence, inbound, and the most interesting question the industry can ask is how can this automate the fax machine?

The thread continues. No one seems to actually believe that change is coming. The most confident seemed to believe that physicians scribing will get more efficient. If we pour another billion dollars into it. No one is reckoning with the fact that fundamental patterns of care will [00:07:00] change. And every case technology excels at making new firms more efficient than incumbents setting a higher standard for them to catch up to technology struggles, to elevate legacy firms, to a level that exists beyond their industry. Failure becomes the default and expected outcome.

Technology has been a mess in healthcare because one tech firms over promised and under delivered under baked solutions without understanding pre-existing workflows. And to you, health care firms saw their peers trying to adopt and failing with technology. So they safely ignored it. In a world where there is no supply side competition, no net new entrance. This is a perfect strategy. The issue is this is no longer the case. Patients are now building out an alternative medical system. To the side of the traditional one with incredible demand. In the next 24 months, we will see hundreds of novel care schemes pop up that look. Unlike the healthcare system, we are used to. These will be [00:08:00] cash pay often, legally gray and likely mostly LLM assisted. They will provide a better, cheaper and faster option. Even if you don’t believe this libertarian cash pay free market utopia is likely, it is all you have to look at is the integrated care networks for a hint of what the future looks like. Amazon Walmart. CVS are already building this parallel care system. It’s so over. We’re about to see furious competition for every single piece of the stack that has been a monopoly for a generation. From providing care to paying for it. It’s all in play. Healthcare has become a technology industry and a consumer one. And no one is ready. So there’s some pretty thought provoking pieces, too. This tweet thread from will there’s a couple of things though that require at least a devil’s advocates points or at least a counterpoints. And I’m certainly not here to contend that this parallel [00:09:00] system isn’t happening it is, and we’ll, we’ll get into detail later. But there’s a couple of sort of finer points that when you get the outside tech perspective looks so obvious to a, to pure true technology, only shops. Or technology shops that maybe haven’t yet dabbled in healthcare. You can look at some of the massive flops or last decade. Google health being the first one that comes to mind. You would have thought that an organization like Google now alphabet can crack the code. They can just throw developers at it through engineers at it. Let’s think this through and come up with a better way. And didn’t happen and it wasn’t even close. There was never, never a doubt that it wasn’t going to happen. So let’s look at a couple of these tweets. In particular starting with the one that supply is expanding, it’s hard to imagine a world where 90% plus of care is an LLM assisted within the decade. And this is after years of getting brutalized by EMR is we’re [00:10:00] close to doctors having technology leverage.

A little pedantic on my part here, but are we talking about care being assisted by LLMs? Are we talking about documentation being assisted by LLMs? Because there is a. Big difference. There’s a tremendous difference between the two. We don’t want to nitpick the argument because the sentiment is pretty clear that. The thesis from will is that it’s hard to imagine a world where almost all of care is an LLL assisted. There’s no large language model. That’s what LLM stands for one more time. There’s no LLM that can ever perform a physical exam.

That’s just the nature of the beast. The machine will not go hands-on with you. The patient. To work you up. So if we can look past that, Where the tech leverage will really come into play for doctors is if an LLM. Can not only provide exceptional documentation [00:11:00] requiring hopefully ever diminishing amounts of oversight as well, because we know that if a doctor is signing his name on your notes and your charts, whatever documentation in there is the liability of that physician or of that provider. But if an LLM can not only provide this exceptional documentation, but be able to capture discrete data elements in the correct fields within that provider certified EHR for quality reporting. Now we’re talking about real leverage. Now we’re talking about actual leverage for the doctor to be able to spend less time in their software. I am like a broken record here talking about the number one. Driver of physician burnout, being the software that they are forced to use. And the only software they’re forced to use is their certified EHR. But if you can create an LLM that not only captures the documentation needed for for like a basic soap note, but as also dropping those [00:12:00] points for quality reporting into the discrete data fields in the EHR. That’s real leverage. That’s real change. That’s meaningful change in the day in the life of the physician, in the mid-levels and all of the clinical team that supports that effort. So when you look at the tweet thread in its entirety, It is very good. It’s very interesting, but, but it really overlooks the reason where we are today with health. It. That’s meaningful use. You know, we’re going back a decade now, a decade plus. That’s obviously evolved into MACRA MIPS and the many other quality reporting programs, registries that healthcare orgs must submit data to. And by the way, that data needs to be structured a certain way. You’re not dumping PDFs of freeform notes into these registries or to CMS. They’re looking for data points. And to go back to that initial point, if you can craft an LLM that knows which data needs to go, where within a certified EHR. That’s real leverage. [00:13:00]

And. Point of contention here.

Your journey. Mr. It’s not the same thing. Electronic health records are the world we live in today, where data is interoperable to a degree. And I’m certainly not here to be an apologist for the EHR industry, but. The HR is health data that can move. EMR. Sort of the old way electronic medical record, that amounts to a digital silo. So not to pick on our friends. From the outside, looking in here into the healthcare world that are coming from a tech background, but there is a difference and choice of words do matter, especially in a tweet where your. Limited to a certain number of characters. But coming back to the central point with what really matters. There is no healthcare system without providers treating patients and really technology should augment that provider and not be a time suck and the primary factor for the entire occupations burnouts. Even extending down to nurses as well, who are, I’m not a qualified provider in the eyes of MACRA MIPS. It would be [00:14:00] a physician or a mid-level. But burnout knows no bounds between professional distinctions.

Moving on to the next part of this thread that needs our attention is a as commercial insurance degrades patients shift from a payer, PCP centered model to an army of online telemedicine services for each and every condition they might have. The sentence with Amazon being your doctor and pharmacy cash, please. No insurance. Welcome here. For some, this will be true and early for some that already is true. That can afford it. But that’s not really, most people, that’s not making an impact on the bulk of the healthcare consumers in the us. When you look at a little bit of data from the Commonwealth fund, private insurance is the primary health care coverage for two thirds of Americans and the majority of private insurance at 55%. His employer sponsored a smaller share at 11% is purchased by individuals from [00:15:00] for-profit and nonprofit carriers. So this leads about a third of the insured. To CMS based coverage of Medicare and Medicaid. And there are huge numbers of baby boomers that are going to be added to the Medicare roster through 2030. The last number I saw was that expectations are at about 26 million new folks. Joining the Medicare rolls. So that’s a ton of people and. This assumption that we’re going to be moving to a cash pay for anything that you might need. I think misses the Mark A. Little bit. Cause we’re we’re folks either get insurance to their job. They’re paying out of pocket or they’re getting it from CMS. Are we really expecting a massive shifts and to quote. An army of online telemedicine services that ends with Amazon being your doctor and pharmacy. I’m not so sure there is a huge amount of pressure on household budgets. So unless trust within the system is [00:16:00] completely shattered and broken, it’s going to be tough for all of these currently insured Americans to forgo their existing coverage in favor of cash pay services. We’re not talking about the one-off issue. We’re going to cover in a minute. Something that’s a one and done, Hey, I need this script written. I don’t really see my doctor forward or maybe, I don’t know this type of doctor. So it’s easier for me to just go online. I think that’s the exception that proves the rule. I mean, Americans are not well overall, so we’re, we’re talking about moving from physician, patient relationship, treating some type of chronic disease or treating some type of illness to just getting it online. Through telemedicine. And I got to pay out of pocket for it, even though I’m already paying through the nose. For the insurance, I already have. I don’t think so. Not anytime soon.

And on a little bit lighter note, I got, I got a kick out of these two tweets in the thread. The first one begins, we [00:17:00] have free physician quality, AI, inbound, and most interesting question the industry can ask is how can this automate the fax machine? Welcome to health it in 2023, man. This is much more commentary though, on the shoppers and the buyers in the healthcare space, then maybe the availability or functionality. Of new, powerful technology. This is very much a world where it’s still using paper film. And you’ve heard that. Metaphor from us before. The vast majority of healthcare is using paper film because they don’t know digital film exists. So when we’re talking about improving and automating fax machine workflows that’s how it’s been for the last 15, 20 years.

Welcome to health it in 2023. And the next, no one seems to actually believe that change is coming. The most confident seemed to believe that physicians scribing will get more efficient. If we pour another billion dollars into it. Well, I think the real change that tech is [00:18:00] hoping to affect for healthcare that it desperately wants to make happen is severely limited. By the after mentioned regulatory entrenchment aspects of the system. Once again, this is quality reporting. And then last as a provider, you are already at a cash pay practice. You’re beholden to quality reporting where you’re risking a decrease in your reimbursement. And it’s really not so much risking it. If you’re not participating in any of the quality programs. You are. Gonna have a decrease in reimbursement. And then the final line of this tweet is that no one is reckoning with the fact that the fundamental patterns of care will change. You know, for some the pattern of care can change for some, and it’s those that can’t afford it. The pattern of care has already changed for those that can afford it. And even looking through the COVID pandemic. There were clinics popping up on every corner here in Miami. Can you get a COVID test, bring cash. We’ll test you right here right now. And this is going back to sort of the early middle days of the pandemic where tests [00:19:00] weren’t all that easy to come by. But if you had cash, you can get tested right here right now. I’ll tell you if you have COVID.

The next one is the tweet that technology has been a mess in healthcare because one tech firms over promised and under delivered under baked solutions that understanding pre-existing workflows and then. By extension healthcare firms saw their peers trying to adopt and feeling attack. So they safely ignored it. Point one is true. 0.1 is absolutely true. Over promised under delivered. Under baked. Not so sure. But without an understanding of pre-existing workflows, delivering software to the market. I don’t know that that really falls to the vendor. And, you know, and speaking from my old EHR vantage points, workflows taken really, really seriously by all of the major vendors now. You know, there’s still 700 some odd certified EHR in use. So I don’t know that I can speak for the entire marketplace, but everything in the top half, certainly the top [00:20:00] 25, top 50. Workflows taken incredibly seriously. But that workflow advantage that vendors sought to build into their software was kind of nullified by all of their requirements from CMS around data capture. You have to capture the right data to show us that you’re meaningfully using our software or a software so that you can first get a meaningful use bonus. And then next, just not lose any of our Medicare reimbursement. But all those data capture requirements, turn physicians into their own scribe. This isn’t some universal failure of a lack of understanding of workflows. I mean every EHR worth its salt has a chief. Medical officer to advise on these types of matters. There’s not to my knowledge, a deep state for healthcare that makes so much of the software disaster to buy, to implement and to use. There’s no conspiracy theory that EHR has have to be poor workflow tools. [00:21:00] It’s a product of a R a. R a 2009, which is the American recovery and reinvestment act in 2009, which contained the high tech act bill. That’s the issue. You’re turning physicians into their own scribe. You’ve taken them from treating patients with, they went to medical school to do. And now we’ve got to capture these 25 data points on every encounter. What are you supposed to do about that? There is no. There is no technology firm in the country that could solve for that. And they still have it.

Too bad. Too bad for physicians. That is not that there is some conspiracy. To make health software tough to use. And then point to you that healthcare firms saw their peers trying to adopt and failing with technology. So they safely ignored it. Who safely ignored it. I mean really the only people that can afford to ignore her or it, and that is meaningful use later MACRA, then all of the registries that data has to be submitted to you for your specialty and so on. [00:22:00] The only folks, the only physicians that could ignore it were those on the verge of retiring. Those operating a cash pay practice, which really they don’t need to be on an electronic health record at all. At least not in terms of their reimbursement, holding aside patient safety, holding aside outcomes, management, holding side, all the benefits that the HRS do bring to patients and providers, but there’s no impetus. Nobody said you had to have any HR. It’s all related to Medicare reimbursement. So either you’re operating a cash pay practice or you’re about to retire or your Medicare panel was so small that you could afford that cut in reimbursement. To not adapt. And then meaningfully use that software. Because that’s all we’re really talking about here. Right? Is EHR. , we’re not talking about rev cycle solutions, which are widely adopted. We’re not talking about patient engagement solutions. We’re not talking about appointment reminder software. We’re talking about ignoring the digital transformation that the federal government has tried to make happen in healthcare and has been successful to a degree at huge [00:23:00] expense. By the way, again, not an apologist for the meaningful use. Component of the high tech act. But we would still be in a, in a healthcare ecosystem that was largely on paper without that driver. And that’s the reality. That’s that’s a fact.

So when we go a little bit deeper, this parallel system that we’ll references throughout the tweet thread seems to revolve more around disruptors, like hymns. And if you. Watch football on the weekends, you’ll have seen the hymns ad. If you listened to any podcasts that have advertising on them, you’ll have heard of HIMS ad. But if you’re not familiar, hymns provides telemedicine services for men, with ed, with hair loss and so on. So with a service like that, this disruptive factor for what would have been your primary care physician, or maybe your, your urologist. As you get access to a script writer for these non-essential drugs that you want. You pay cash for it and you move [00:24:00] on. And sure they’d love to sell you a subscription to whatever it is purchased periodic check-ins as required with a physician that’s writing the script and that’s it. So that, that is disruptive. That’s taking patients out of the, their primary care office or their urologist office in the case of hymns and so forth. And it’s moving it online. So that change is already happening. It’s already taking effect. But there isn’t going to be a parallel cash pay system for folks with actual medical issues. So how can you diagnose and treat something like type two diabetes remotely? Y you can’t. Not in a way that will be a better experience than what we’ve already got. That’s not to say it’s a good or great experience in any way. It’s just that you’re still doing a console. You’re still going to the lab, still having blood drawn. You’re still waiting for results. Then you’re gonna talk to a physician again, you’re going to get on a plan. You need drugs, if it’s drugs you know, and, and so on. The other half of this, [00:25:00] is that, how far would you have to drive listener? To get to a CVS minute clinic. I’ve got like five. That are a Stone’s throw from my house and even more here in my office. This alternative delivery model. Isn’t, it’s no longer revolutionary that revolution has happened and you’ll see continued vertical integration there in the space. Absolutely. But the availability of walk-in clinics and urgent cares, like that’s. That’s already there. So, and in some facets of this, this thread, I’m not totally sure what we’re talking about. I mean, we’re talking about how LLMs are going to grease the wheel for this sea change in healthcare. But a lot of these revolutions are already occurring. And at what scale do we expect them to deepen, especially to peel people away from sort of traditional medicine into this alternative parallel model.

But all that to say. Again, this is not a critique of the of the thesis. It’s, it’s interesting. It’s given us something [00:26:00] really different to talk about here. And personally, I love it. When tech takes an interest in healthcare. I love that in America, we exist in a free market where if you have an idea, You can try to monetize it. That’s awesome. But the thesis here that LLMs can underpin an alternative to traditional healthcare though. It shows some of the, perhaps some of the naivete of tech when addressing health care has many issues. So sure there’s potential with Elma LLMs to solve for scale on a P and L really scale in the sense of healthcare is patient throughput and even secondary markers, throughput like messages from patients. But it overlooks the physician oversight aspect of their relationship. It also overlooks the legal and compliance aspects of physician oversight for the care and outcomes of their patients. Scale or patient throughput. Isn’t what improves patient outcomes. It just isn’t. [00:27:00]

All that to say, I think will is right about the parallel cash only care delivery mechanism. But it’s already happening. It’s already happening with sites like hymns. With a CVS minute clinics. With Walmart’s clinics and so forth, but a sea change from the fee for service world, we’re living in. Is going to take a lot more than the advent of physician grade generative text software.

And this might be more. B roll type material than anything else, but the The tweets on X or the Zeitz, if you will, some are calling them some of the replies to this thread are. Are interesting. They’re thought provoking because you see the advent of LLMs across health care, some are being integrated into broadly, widely adapted EHR is. We’ve discussed healthcare systems, leveraging LLM [00:28:00] technology for their providers to send replies to their patients with, because it’s such a drain on their time. So this, this isn’t going away and, you know, I’m grateful for a town square, like Twitter that can. Bring a lot of different points of view. To lights. So reply here to the thread. Who do you see holding liability in this new model and system? When it comes to malpractice and abusive care? How has that mitigated? I’m a physician, you as PAs and NPS under me, you make more mistakes than an hour, then you can imagine. And the reply from the author is that for telemedicine today, when you get a prescription, after going through a structured form, That’s 99% machine and 1% human who is responsible. How does that change? When we 1000 acts of the volume and crank the acuity NPI holders will act like human liability shields. Increasingly inhuman.

Who is responsible. Man. [00:29:00] Right now there’s only one person that’s responsible and that’s the physician. That’s the provider. There’s no anyone else to hold the liability there. So that NPI holders will act like human liability shields.

Who could you imagine being a provider? So doc, we’re going to attach your name. To this LLM populated form. That a prospective patient will fill out. You don’t really need to read anything. All we need is your signature. In the E-prescribed module.

No chance. There is nobody. No physician will be signing their name. To a process like that. Dr. Aryan Nanda replies sounds like a pretty terrible deal for the physician. How would they be able to accurately assess the differential? Consider co-morbidities think through the logistics of the recommendation or the recommended diagnosis and treatment plan. When they’re given 30 seconds per patient, [00:30:00] why should their license be on the line? And right now, it is, there are no laws around consumer or patient protection with the use of LLMs in their care. There’s only one party. That’s responsible. It’s the physician. So who’s responsible.

The replies continue. We’re facing a critical shortage of MD’s largely driven by MDs, demanding funded residency calves, I guess, adapt or society will route around the shortages.

The next reply. If by routing around you mean patients getting suboptimal care, which puts their lives at risk, then I guess that’s what the market wants. Residency caps have been an issue for 30 years. It’s not new. It’s more due to the caps, tied up to politically charged Medicaid. And Medicare.

This brings me back to my thesis around LLMs that you’ve heard here on the DocBuddy journal. In some instances. The [00:31:00] extreme power and leverage that an LLM can provide for you. Is a solution. In search of a problem. Not all the time. Not every time. But enough of the time. The you kind of scratch your head and go. What are we really talking about here? If we’re looking at moving the. Responsibility of care from a physician to a machine. And to machine, which is not all that difficult to trick or not all that difficult to spit out. An inaccurate response or a response with no basis in anything. Then we’re in big trouble. That is not what the market wants. That’s not what the market wants at all. I just can’t imagine.

Okay. So we’re getting a little long-winded, like I said, there might’ve been more B roll type material than actual a podcast episode material. All of course include a link [00:32:00] to this tweet thread in the show notes and the episode description. You’ll be able to take a look at it. If you’re active on Twitter, you can put your own 2 cents into this thread. That’s gotten just about 2 million views at the time of recording. If you do have any thoughts, or if you think that I’m off base, be sure to tag at DocBuddy app. On X, if you do make a reply again, this is not a critique of will or any of his ideas about sitting here, sort of both on the entrenched health, it vendor side of things for a lot of my career. And that providing our own disruption with doc, buddy, I can see it from both sides. I think they. I think that the thesis that LLMs will underpin this revolution in care delivery. I don’t see it. Not unless something something drastic changes with the liability of patient outcomes. Not unless something changes drastically with the availability of physicians, which is of course a problem. We want to [00:33:00] keep them happy. You want to keep them practicing as long as they can. Just the bent for the overall benefit of the system. Obviously individuals will be. Free to make their own choices and their own careers. And when they decided to hang it up but very interesting food for thought. And we’ll open invited. This is something that comes across your desk. We’d love to host you on the episode. We can get some clarity on some of these points. We can talk it through. Really am grateful that there’s a lot of smart folks thinking about healthcare, about health it and ways that technology can improve the experience for both physicians, patients, their organizations, and potentially even lower costs for payers as well. So much as it fits into this parallel carrier delivery mechanism that is being defined as largely cash based in Will’s thesis. And on behalf of the entire doc, buddy team. Thank you for listening. Be sure you’re subscribed on apple podcasts or Spotify. If you like the audio only if you like the video component, be sure to subscribe [00:34:00] on YouTube as well. We hope to see you at Becker’s ASC event next week in Chicago, you are welcome to reach out to me, Erik at doc, buddy.com E R I K at doc, buddy.com. If you’d like to set up a one-on-one meeting while we’re all together in Chicago, for everybody in ambulatory surgery center, land. And until next time, I’m your host, Erik. Talk to you again soon.