438. Healthcare Deep Dive – How AI and GLP-1 Megatrends Affect the Industry, Why Specialities Need to be Integrated, and Insights on Bundling vs. Unbundling (Ambar Bhattacharyya)

438. Healthcare Deep Dive – How AI and GLP-1 Megatrends Affect the Industry, Why Specialities Need to be Integrated, and Insights on Bundling vs. Unbundling (Ambar Bhattacharyya)

Ambar Bhattacharyya of Maverick Capital joins Nick to discuss Healthcare Deep Dive – How AI and GLP-1 Megatrends Affect the Industry, Why Specialities Need to be Integrated, and Insights on Bundling vs. Unbundling. In this episode we cover:

  • Healthcare Trends, Including High Deductible Health Plans and Cash Pay Models
  • US Halthcare System, Incentives for Preventative Medicine, and Potential for Value-Based Outcome-Oriented Systems
  • AI Adoption in Healthcare, Including Adoption and Risks
  • The Future of Healthcare Technology, New Codes, and Credentialing Process
  • Healthcare, Aging, and End-of-Life Care

Guest Links:

The hosts of The Full Ratchet are Nick Moran and Nate Pierotti of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.

Want to keep up to date with The Full Ratchet? Follow us on social.

You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.

Are you a founder looking for your next investor? Visit our free tool VC-Rank and we’ll send a list of potential investors right to your inbox!

Transcribed with AI:

Ambar Bhattacharyya joins us today from San Francisco. He is Co-Managing Partner at Maverick Capital, an early stage venture fund investing in health and technology. He has partnered with numbers $B+ companies including Hims and Hers, MosaicML, Cityblock Health, Devoted Health, and ConcertAIPrior to Maverick Capital, he worked at Bessemer Venture Partners, Bain Capital Ventures, MinuteClinic (acquired by CVS), and Bain & Company. He has received broad recognition for his leadership, including Fortune Magazine’s “40 under 40”, Business Insider’s “30 leaders under 40 transforming healthcare”, and as a Global Shaper by the World Economic Forum. Ambar, welcome to the show!
Nick. Thanks for having me. Thanks.
It’s a pleasure sir. i You know, I’ve heard of you and I know about your many, many partnerships with billion dollar and DECA corn companies. So you know, congrats on all the success but why don’t you tell us a bit about your backstory in your path to becoming an investor?
Yeah, sure. Thanks. So, you know, my, you know, my background, I grew up, you know, here in America, mostly in rural America. So my father’s a physician, immigrant physician from from Calcutta. And when we, when he came over to the US, he basically went to wherever there were no other, no other physicians around. And that happened to be in rural America. So we, we hopped between rural West Virginia, rural California, for most of my life. And, you know, when we were when we were doing this, it was like our family business. My mom worked as a secretary in the admin, my sister, and I chipped in doing some of the billing the dictations, you know, all sorts of stuff just to help keep the business moving. And so, you know, and there was not even just our dinner table conversations, it was like, all of our conversations were in and around, you know, how does, you know, how does being a doctor work? And how does, how does it mean? What does it mean to run a run a small physician, some physician office, all in a really small town, right? Where we kind of knew everyone, everyone knew each other’s name. So you build these like deep relationships with the community. And you fast forward a bunch of years. And, you know, I want to college unfortunately, I’m like, terrified of blood. And so I, you know, faint, faint, most times that I’m in the proximity of something red, I can never really shake that healthcare, healthcare training, I got it at a young young age, I just became really passionate about the sector. And as I mentioned, like I started my career in consulting, I left there and I’ve joined a small little startup in Minneapolis called Minute Clinic, which was doing some really interesting disruptive activities in the healthcare space. You know, we I was lucky to be relatively early there, and CVS Pharmacy ended up acquiring us. And I was, I’ve been in the venture world ever since. And, you know, when I look back at tying together my upbringing and where I am today, I do think I have this deep appreciation for how the sector works, you know, on a lot of its faults from from seeing it from an early age. And I think that small town community, you know, relationship, like deep relationships that we had, also kind of sticks me and sticks with me to this day, like I’ve, I’m a people person, I love meeting people and bonding with them. And I think when you think about my, my role today, it’s like the two things I’m passionate about. I love people and I love thinking about how to solve problems and in a complicated sector like healthcare, so that’s kind of like the short answer of you know, you know, how, how I got to where I am today.
Wonderful. And can you tell us a bit about your thesis in the investment approach at Maverick?
Yeah, so Maverick you know, my background, we’re, we’re 30 year old Investment Fund, and we have a public arm and a venture arm. And here on the venture side, we focus a lot on on deep thematic research. So we tend to be very, very deep in the areas that we focus on, we try to understand both megatrends, derivative, those mega trends in areas that are of our keen interest. So I spend most of my time thinking about the healthcare sector, given given my background, my interests are and my passion area. And so on a day to day basis, what does that mean? So one is we try to identify these large mega trends. And so, those today, you know, you know, we think about those megatrends as being things like, you know, we call them GPS and GPS, like, you know, things that are just growing exponentially in the world. And there’s a whole bunch of sub trends we’re thinking about that have just been more ongoing value based care, consumerism, and things of that sort. And the way that we identify these themes, we, we look at a lot of macro data, to just understand, you know, how fast are these things actually growing? How fast are these, you know, being predicted to grow, but also do a lot of bottoms up research. So, we talked to a lot of customers, sometimes it’s enterprise customers, sometimes as consumers, we do lots of surveys to try to understand, you know, where the puck is going to go. In a sector like healthcare, it’s particularly challenging, given a lot of that has been historically defined by for lack of better words inertia, where, you know, it’s been a really slow, slow system to change. And so for us, it’s really trying to identify the wedges, that will actually create some change that’ll gain adoption in those areas. And so and then we try to invest as early as we can. We try to partner with entrepreneurs, you know, try to be their first or second institutional partner, to the extent we can.
And what would you say, of all those macro, mega mega trends that you’re following? is maybe most relevant to healthcare right now? Or another question would be, which of those maybe is getting the least airtime that you think about the most, you know, in your investment approach?
It’s interesting, you know, I think the two the two megatrends I alluded to them before, you know, one is, is we call them TPTs, but the intersection of healthcare and AI, and maybe we’ll come back to that in a bit, another one or disease GLP ones, right. So these are the, you know, we’ll go these those Olympics of the world. Now, what’s really interesting about that trend, I think, a lot of people think about it from a from a pharma perspective, right, you know, think what Novo and Lilly are doing in developing these drugs. And, you know, candidly, that’s not an area that we deeply focus on, we focus mostly on health, AI, digital health, technical services, you know, the nuts and bolts of how the healthcare system works. And so in that area, you know, we’ve been tracking, I think people talk about it, and people are very comfortable talking about these GOP ones. But, you know, just a few data points around how powerful that trend is, I think this was, this was lily in their earnings a few weeks ago, you know, said that in the, in the beginning of the year, you know, 2024, January, there was about 5000 scripts, you know, being being prescribed of their GLP, one, in the last week of April. So just four months later, it was 25,000. And so when you see literal exponential growth in the period of four months, you know, that’s something that catches our eye. And we’ve been tracking the space, you know, for 1824 months looking for derivative effects of these GLP ones in particular, you know, what does this mean for medication adherence? You know, what does this mean, for for side effects, you know, from for taking GLP ones, you know, what does this mean, for downstream effects? You know, I think there’s been, you know, other things people have been focused on as people just think about this as a weight loss, you know, diabetes type drug. There’s 27, concurrent phase three clinical trials being run right now, for different indications for GLP ones. Some of them are different formulations, such as oral, some of them are for different things like stroke, heart disease. And so when you start getting deeper and deeper into these into these areas, it becomes really interesting for us of Okay, so what, you know, what actually could happen? Or is there a next generation type of provider that will have to exist, you know, given the existence of, of GLP plans? How will the insurance companies pay for them? Right, you know, this is a huge, huge cost. How will employers have to deal with it, you know, how will consumers access these drugs? Will pharma companies decide to cut out the middleman and go direct, and you’ve, we’ve seen all sorts of innovation in in around each of those areas, as it relates to, you know, you know, that exponential growth we’ve talked about. So that’s like a little bit of how we think about it. So you think about the mega trend, and he starts segmenting into each of those stakeholders, and trying to understand, you know, where are, where those impacts occurring? And then the last thing we try to do, and most importantly, we try to think about the profit pools that might exist or shift because of the impact of those GLP one, so hopefully, that gives you an example of one of the ways we’re going, you know, one or two or three levels deeper on that. megatrend.
Yeah, I mean, you talked about segmentation. And obviously, the stakeholders in healthcare are kind of unique and interesting payers, providers, patients, etc. I’m always interested to discuss with investors, the way they frame sort of the opportunity set that they’re investing into, you know, thinking of market maps and classifications. On Bart, you know, from an investment standpoint, how do you carve up or frame the opportunity set within healthcare sort of at the macro level? Yeah,
so we think about them as five key stakeholders. So one are providers, two are our payers, so insurance companies, three are employers you for our pharma companies. And then the last is the consumer. And if you think about healthcare spend in general it’s about $4 trillion The year and then if you then start segmenting out that 4 trillion, you essentially have government, you know, paying for Medicare and Medicaid. And then you have a self insured employers paying and then you have consumers paying. If you if you cut it, you know, self insured employers that’s, you know, Maverick paying for health insurance or Ford paying for, you know, their employees health insurance, you know, that’s about a trillion dollars of spend a year, Medicare is about 2 trillion Medicaid, it’s about a trillion. That’s how like that fortune, it shakes out. And you can see you can keep you and keep cutting it underneath. So actually within that self insured employer trillion, part of that is is that consumer copay or coinsurance that has to go in there. And within Medicare, it’s Medicare Advantage and Medicare fee for service. And within Medicaid, that is managed Medicaid and unmanaged Medicaid. So you can keep actually segmenting lower and lower. And if you just keep pulling a tread, let’s just pull the thread a little bit more. So in that self insured employer category, that trillion of spend, one of the biggest derivative plays that has been happening for the last 20 years, is the growth of High Deductible Health Plans. So probably many of the folks listening to this pod have have a high deductible health plan where you know, certain amounts of the initial dollars you spend in the first $100 $1,000, etc, are actually out of pocket,
or health savings accounts included in that number no, that
will help take accounts can actually be a pre tax way of funding healthcare expenses if you have a high deductible health plan. And so I put that like into a parallel a parallel trend. But you know, what’s interesting about these High Deductible Health Plans for you know, is that 20 years ago, these things didn’t exist. And there was a bunch of studies and consultants that came in saying, hey, one of the reasons why health care costs keep growing is that Nick and ombar don’t actually pay for their health care, right? It’s always, you know, the big insurance company that pays for it. So there’s no incentive for Nick or ombar to go shop for our health care. So they said, Hey, why don’t we actually shift some of the costs over to them, such that they’re a little bit more prudent in terms of how they how they think about spending their healthcare, and they would rationalise their health care? So in so that first started with, Oh, it’s $100 copay. $200. If you fast forward now, you know, 2024, the average high deductible for a family is $4,000. It is an enormous amount of cash out of pocket before insurance kicks in. Yeah. So So when when you so we looked at that trend, and we said, hey, that’s really different than 20 years ago. Also, secondly, by the way, it hasn’t really worked. Healthcare costs keep going up year after year. And so I think it’s quite quite arguable as to whether that experiment is right or not. But that led us to like two different discoveries. You know, one is, I think a lot of folks in healthcare had this refrain from 1020 years ago, saying the consumers don’t pay for anything. Ben, in reality, like, when we looked at the data, we said, hey, if consumers are paying $4,000 out of pocket, that’s a lot of money. You know, for a lot of, you know, middle America, lower income Americans, that might be one of the largest, if not the largest out of pocket expense they have on it for the household basis. So we discovered, you know, I think relatively early, that there are many cash pay models that could be very viable, out there, you know, in the healthcare economy. And that can be that could exist in parallel to that trillion dollars of self insured employer trend. And yes, do
you have an example of what that looks like? Yeah,
sure. I mean, you know, companies like good RX are really interesting, right? You know, this publicly traded company, where you basically get a card and you go to CVS or Walgreens. And ordinarily, if you’re picking up a script, your insurance would cover it. A good RX card basically gives you the ability to pay a cash rate for that, so that it never goes to your insurance. But sometimes that cash rate is lower than what your insurance company would have negotiated. You know, that concept, you know, we’re not investors in that in that company. But it’s, it’s it really speaks to, to that trend, at least when it comes to pharma, you know, pharmaceutical spend, I think two is you look at the growth of a number of telemedicine type companies that are that are more cash pay oriented. And those those have also really rapidly grown over the course of time. And so, I think there’s been, you know, a lot of interesting ways that you we’ve seen that consumer behaviour, change and shift because of that in our way we go back to the root cause of, you know, like what has really shifted in the last 1020 years and I think it’s in large part this height, so health plans have kind of forced the hand of the consumer.
Are there areas within healthcare that you you avoid, obviously, of your investment thesis, your tech enabled services and technology company? But even within that set, you know, are there areas that you, you know, don’t engage with the company? For various reasons? You
know, that answer is shifts over the course of time, you know, we tend not to take a lot of regulatory risk, you know, so if things are very binary kind of black and white, we try to avoid, avoid those things. But I say that with a caveat, right, where sometimes a regulatory shift can create a new market. And so, you know, 1020 or 10 years ago, you know, the ACA passed, you know, the Affordable Care Act. And there was the growth of these health insurance exchanges, right. And a lot of interesting companies started in around that time, Oscar being being one of them. And that was a regulatory shift. And there was a chance, you know, many times that, you know, the Affordable Care Act went to the Supreme Court and could have been struck down. But it survived. And, you know, to this day, you know, Oscar is a multi billion dollar company. And so I think, we tend to avoid things that have regulatory risk associated with them that are that is binary. But in the next breath will say that if if we think that it can create a new market and a new profit pool, we will still investigate it, right. And so that’s kind of 111 way we look at it. Another example would be, again, 10 years ago, this concept of accountable care organisations started. And so this is another big mega trend we’ve been following for over a decade where today healthcare has been, it’s been a it’s been a volume, a shop, you know, you go go to the doctor, they send you a bill, you go again, they send you a bill, there’s really been no incentive for preventative medicine, being like, how do I keep Nick, from coming back here, because the doctor is somewhat incentive to have Nick come back being like, once you come back in four weeks, I’ll do a quick checkup. I’ll bill insurance company nickel, pay another copay, but it doesn’t really line up with incentive of, hey, like, the entire goal is, is having Nick be healthy. So these accountable care organisations are at the forefront of this fee for value type shift, you know, where you’re trying to have health systems and physician offices be incentivized financially to keep Nick out of the hospital to keep Nick on his drugs. And so that was a regulatory shift. And that was one we actually legally that we actually leaned into. And I think there’s been a lot of innovation there, you know, in that space, and so, so, we tend to avoid it, but it’s sometimes if it is really dramatic, where it can really shift the profit pool? You know, we’ll take a look at that. So that’s, you know, that that’s, that’s, you know, a nuanced answer to your question.
Um, Barwin, the majority of Payers are activated and, and engaged through an employer. Yeah. Is it reasonable, you know, to have sort of a value based outcome oriented system, when people are changing jobs frequently. And, you know, the incentives with that would suggest that short term, try and eliminate as much, you know, cost to the payer as possible, and just pass the buck on to the next payer, when somebody switches employers?
You know, it’s a, it’s an, it’s an awesome question. It’s a really fundamental question. I mean, when you compare, you know, the US healthcare system to a lot of other countries. That is, that is one, one big difference is that a lot of other health systems are nationalised. And so there is a long term incentive to make sure that people are on the right drugs early on, they are staying healthy early on, and you every, you know, it’s that phrase, like, you know, a penny of prevention is worth $1 of $1 of cure, you know, you know, that that phrase, it’s, you know, it’s a, it’s a really, it’s really true in a lot of other lot of other economic settings. I do think you’ve touched upon a really fundamental issue here in the US, which is, people do switch jobs, and with that, they switch health insurance. And that does create, potentially an incentive for less preventative medicine in order, you know, and, and, you know, further, you know, pushing off the buck, you know, to the next employer to next thing to avoid it. But there’s a counter trend to that, too. So the counter trend stat is self insured. Employers are also also incentivized to keep talent, right, and they’re also incentivized to recruit phenomenal people. And I think here in the US, culturally, so far, there has been a linking between looking at a for every individual employee, their compensation package and their benefits package. And so there is this incentive for HR and the CFO to not overly, you know, skimp on something and saying, Hey, this is a $10,000 high deductible health plan, right, which would be maybe the extreme or there’s no health insurance offered here. So because then, you know, I think there’s a fear that talent would just go elsewhere. So that keeps, that keeps the bar relatively relatively high over the course of time, we are seeing really is taken off in the last few years, there are new programmes called one of them, healthcare, a lot of acronyms already throughout Accountable Care organisation is one of them. And the Affordable Care Act is another one that has caught the eye of a lot of a lot of investors and a lot of customers. This is called the Accra. And this is actually how do you get self insured employers to actually move their employees on to health insurance plans that are on the exchange, right, so they’re actually able to shift it more akin to the to the Affordable Care Act. And so we’re seeing, you know, a few cracks in the armour there and neck where maybe that that that shifting will occur, and people can actually keep to sell the same exact plan, even if they ship from one job to another because of things like egress. So that hopefully that that gives you at least our take on it.
You know, we had Adrian goon, on the show of forward, and he made a pretty good argument that consumers will pay for premium. If it’s, you know, 10x better than what they’re receiving, currently. So I think his analogy was, you know, do you watch satellite TV? Or do you pay for Netflix? You know, do you listen to the radio? Or do you pay for Spotify? Obviously, these are much more much lower purchase price than a lot of things in healthcare. But you know, he’s trying to innovate in a way to say we can make sort of the the PCP and primary visit, you know, 10x better and more efficient and easier and more tech forward than, you know, what you have currently. Do you believe that this sort of philosophy will work in healthcare that consumers will just pay out of pocket? You know, for services or for a plan? That’s 10x Better than maybe what they’re receiving from their self insured employer?
I think so. I know, and, you know, I listened I listened to Adrian on on your pod as well, to refresh myself on his on his views. Listen, I’ll say a few things, some, some where I agree, and some where I disagree. So we’re I agree, I think there, I think consumers have now been, their expectations have increased dramatically over the last 10 years in terms of what they expect from from the healthcare system. Part of that is through, you know, companies that we’ve been fortunate enough to partner with, at early stages, companies like one medical companies like HIMS, and hers, that have really pushed the envelope for what it means to define having the consumer at the middle of that experience. This could be as simple as you know, same day appointments or next day appointment, as simple as being able to book on an app, as simple as having access to telemedicine at your fingertips. As simple as knowing exactly how much you’re paying for prescription drug at a certain time. You know, as simple as having it come in discreet packaging, like whatever it might be, you know, straight to your door versus having to leave. I think there’s a lot of things that have happened the last 10 years that of, you know, entrepreneurs have innovated in. I do think COVID was an acceleration of that anything, a lot of folks hadn’t tried simple things like telemedicine, they hadn’t done an at home test before. And now that people are familiar with that, it’s really, it’s really easy for people being like, Oh, I’ve actually done that. And, you know, why can’t I do that all the time? So I think in that respect, I think it is very much consumer sentiment has changed, I think the bar has increased for the rest of the world, health systems, insurance companies, etc. I think the place where I think there’s still a lot of opportunity is, even with those, even with those innovations, it’s the system is still way too fragmented. And what I mean by that is, if there’s a significant portion of Americans today that have some level of polychronic disease, that means they have multiple diseases that are chronic in nature. So think diabetes, and high blood pressure is maybe the most common example. And in situations like that, you end up going to multiple physicians, you know, you go to a cardiologist for your heart, and you end up going to endocrinologist for your diabetes. And by the way, there’s a primary care physician who’s kind of shepherding your experience throughout and dealing with other things. I think there remains a very large opportunity to integrate a lot of that care. And I think when when you look back at the last 1020 years, I think each of those specialty experiences actually improved a lot. The primary care experience has improved a lot cardiologists experience has improved a lot. But the integration between all Those points is not. And so when I look forward on a 10 year basis, and I think about how can technology, how can AI actually transform healthcare? This, to me is one of the lowest hanging fruits is where you can actually have real time communication between different specialties. You can actually have one integrated technology technological system, looking at the records and the visits from specialist to specialist to understand, hey, ombar needs this, Nick needs that in a more integrated holistic way. And I don’t think that was even imaginable 18 months ago, but today, I think it is. And I think when you when you fast forward 10 years, it’ll it’ll become just as obvious as telemedicine is today. And just as obvious of as preventative medicine is today. So that’s kind of where I think where I think the buck is going. And I think consumers will soon begin to expect this of being like, Wait, why do I have a cardiology appointment in February, and my endocrinologist appointment in May, there’s like, oh, that’s when their schedules were open. Like, that doesn’t make any sense. Like that, that has to change, like real time. And I think like we’re finally at a point where I think, you know, some obvious things like that will
umber does that suggest or education’s system for new healthcare professionals needs to change as well, in order to provide a more integrated health solution instead of these sort of siloed? disciplines?
100%? You know, it’s interesting, if you look at I’m not a physician, given given my fear of lead, but I’ve I’ve played one enough, you know, in different settings, where I’ve been able to follow a lot of what’s what’s been happening in medical education and things of that sort. I think there’s a lot that is there’s a lot of change that needs to happen in that system to one understand the rapid adoption of, of technology in that system. I think most most folks are now ingrained with electronic medical records. But I think population health management tools, certainly how AI will have an interface inside of healthcare, how to really focus on on Social Determinants of Health, all of those things have historically not really been taught in a traditional medical education. I think that has to change, I think to as you bring up this point of tying my last comment to multi specialty, I think the way that rotations work today is that you know, you get six weeks, eight weeks, 10 weeks, you know, following an OB following a cardiologist doing an ER stent, etc. You know, how do you how do you go up one level and say, Hey, from a systems engineering perspective, this is how we should think about not just the hospital workflow, but the patient workflow, both pre visit, Intro visit and post visit. I think that systems systems engineering type type thinking is really, you know, that’s like, in some ways, historically has been the role of the government. But I think more and more that can be the role of health systems, insurance companies, etc. And I think consumers will demand it, because technology will actually show them that it is possible to do some of the some of these things in real time and at a higher quality than has been done before. So that’s that’s kind of where, where, and how I view where I think the metadata cases have been had just has to change actually, for that to be adopted. Know,
as we think about systematic approaches to healthcare, and leveraging best practices, you know, my mind does go to private equity owned hospitals and health organisations. I’m curious for your take, I’ll share a quick anecdote with you. But I have a buddy who used to work for a non private equity owned hospital, notable hospital spine surgeon and had a good experience. And then he moved to a private equity owned hospital and he noticed the rate at which surgery was diagnosed for conditions was much higher, as opposed to physical therapy and and, you know, what he believed was the appropriate diagnosis. So I guess my question for you is, you know, is private equity owned healthcare, a positive or a negative with regard to you know, real health outcomes for for patients?
It’s an interesting question, acknowledging I’m not a private equity investor, I think I can answer it without any self interest. So in our in our venture fund is not buying any hospitals anytime soon. So you know, maybe I’ll say two things. I think there has been numerous studies comparing the outcomes between nonprofit hospitals for profit hospitals, and religious associated hospitals in the US over the course of time. And, somewhat surprisingly, there has been no difference seen from a patient outcomes perspective across different segments, which is really fascinating, actually. So you know, we hear about anecdotes like this, you know, where, you know, someone switches from x hospital to white hospital and there’s a higher prevalence, but when you look at it on an overall basis, you know, that hasn’t been seen now. At the same time, there have been many stories about where, you know, then in healthcare, one of the monitors is, you know, you know, no margin no mission, right you know, even if your nonprofit hospital, you know, or nonprofit entity you like you need to have some level of margin to be sustainable such that you could do your mission. And of course, there is a fine line between maximising margin and maximising mission. And I think in health care, one of the things that matters the most is that the people at the top and whether it’s the board level C suite, all the way down to the, to the nurses, social workers and doctors, you know, don’t get it twisted, right. Like, it’s at the end of the day, it is the patient’s life, it’s the baby’s life, it’s a senior’s life that you’re trying to optimise. And so, you know, cutting corners is not appropriate, right, you know, in some shape or form, driving efficiencies is appropriate, right? And so how do you find that fine balance, and that’s where a lot of human judgement comes in. And so there are some of those horror stories, you know, where you do hear about thing, things shifting, unfortunately, to that way, but I, but on that, but the data would show that, whether it’s a nonprofit, religious, religiously oriented or for profit, it can happen in any of those settings. Unfortunately,
I see. You’ve touched on AI a couple times, I’d like to kind of dive into that. And we’ll start you off with a softball here. So the adverse impacts of AI generated hallucinations and healthcare can be more severe than in other sectors, you know, how does one help large orgs embrace the advantages that AI can bring while sort of abating or dealing with the risks of hallucination? Yeah,
it’s, it’s a, it’s a really, it’s a really good question. So the way that that we’ve thought about AI intersecting in with healthcare, let’s just focus on providers for a second. So hospitals and physician offices, where those adverse events could happen more often, right, from an from an from an elucidation perspective, I think one spectrum we look at it is there’s there’s the admin back office functions. And then there’s the clinical facing functions. And each of them, an AI can have an impact on either of them. So today, when you look at Where’s AI adoption, in healthcare, most of it is on the admin side. And what that means is, so another another stat of every dollar spent in the US on healthcare, about 25% of it is spent on the back office. And so it’s a very significant percentage. And so, that is things like billing, or something called revenue cycle management, which is the concept of how do providers put together a bill sell the seller, you know, send that bill to an insurance company, the insurance company may deny that bill, or you know, accepted, they may send a payment back, that payment may match what the charge was or not, that’s called payment integrity, and you kind of have this and then the provider can argue it again, saying You promised to pay me $100 I actually got 98. Pay me the last $2 It’s this never ending cat and mouse game. Yep. When you take a step back, and you say, and that’s all very manual, by the way, you know, you know, tonnes of people on an insurance company side tonnes of people on the on the provider side, that a lot of that can be automated using AI, because they can go through an electronic medical record, parse out, what were the appropriate codes, what are the inappropriate codes, put them into the billing system, someone automatically send it over to an insurance company, they can actually check it, send it back, that whole thing, that whole process, the revenue cycle management, the prior authorization process. Ai automation is at the forefront of that. So we’ve seen a lot of adoption there. And to your point around hallucinations, the risk of a hallucination there is relatively low, because there’s tends to be a human in the loop. It’s tends not to be a medical diagnosis. It’s more about a building, you know, a building type situation. And there’s a lot of reinforcement learning happening every single time it goes back and forth. Right. So you know, provider sends a bill accepted or not thumbs up, reinforced, learning, rejected, okay, thumbs down and not reinforced learning, right? So it’s kind of built that feedback loop has built into it. We’ve seen a lot of adoption in their second place. We’ve seen a lot of adoption, is in the medical transcribing space. And so right now, anyone who’s been to physician office in the last 10 years, has this experience where the physician is not looking at them in the eye. They’re staring at a computer screen. You know, patients don’t like it. physicians don’t like it, nurses don’t like it. And so how do you actually use technology to auto scribe a visit? This improves physician productivity? It is just using ambient listening to do this, no human can actually go to a doctor will go through and check to make sure it’s right. That’s like a way of preventing hallucination hallucinations for happening. And also real time reinforcement learning, right, as soon as a doctor corrects a note, it’s there to train the system better. We think things like that have health systems have started to adopt that insurance companies have started to adopt that. And I think the train has left the station there. And there’s a lot there’s enough reinforcement learning and, and humans in the loop, that the risks so far have been relatively minimal. Now, on the flip side, your tie, you know, when you talk about hallucinations, potentially creating real harm, it is around more of the clinical side and the diagnostic side. And so, that is, you know, if a physician is using a chat GPT to actually do a diagnosis, how do you actually know if the diagnosis is accurate or not? How do you actually make sure you’re not violating HIPAA? Right? So patient patient privacy, you know, because it is illegal to have me kind of broadcast at a world Nick’s medical record, you know, there’s Nick, as a federal rights, where is that information being private, that being leaked is a crime and it’s a fine for the health system? And then how do you actually make sure that even researchers you know, who are doing like the latest in clinical trials or drug discovery or whatnot, are using AI in ways that is auditable, that’s trackable, and are actually aware of the hallucinations that are happening. So on that basis, we think there’s a whole new layer of the stack that has to be created. And that’s in the responsibility AI layer or trustworthy AI layer. And I think health systems are conservative by nature. What they’re doing today to solve this is they’re forming committees, just saying, Okay, we have an AI committee now that is looking at looking at hallucinations, or looking at AI usage by clinical staff, admin staff and research staff. And they’re trying to understand where and how they’re using AI. But as the usage increases, the next thing that they that they want to make sure that is happening is to make sure it’s being done in a safe way, both for compliance reasons to make sure that Nick’s medical information is not being shared with the entire world inappropriately. And also to track hallucinations, right, and to make sure that if I type in, you know, you know, headache, you know, and stuffy nose, it’s not saying heart attack, right, or something like that. And it’s being able to track that over the course of time. So we’re seeing health systems adopt more and more, both technology as well as candidly consulting services to understand what’s happening there to track that. It is one of the areas where I think we will see a lot of innovation happen. And candidly, I think that innovation may look a little bit more like healthcare innovation, or it might might look more like cyber security innovation, right, we’re in the cybersecurity, there’s a similar type of threat from both internal and external actors, where you want to prevent information leakage, you know, and that’s whether it’s through phishing attacks, or people inadvertently, you know, sharing internal information, you know, and putting it into non safe places. It that also may be through to making sure that folks are putting information in the right way there’s access control, like who actually has access to the latest AI algorithms versus having it blocked in the system. So we’re actually thinking about this in a much broader way of like, how do you prevent your clustering around and hallucinations? But our question is a little bit more broader, like, how do you prevent the inappropriate use of AI in healthcare? And how do you actually track how AI is being used in healthcare? How do you audit it? And if there’s a lawsuit or something like that, how do you follow that trail back to like, what exactly happened? Where Did something go wrong? For a bunch of those purposes? So that’s kind of how we see both sides of the spectrum? Yeah,
I mean, part of your answer, it leads me to think about more comprehensive solutions in general. Where do you stand on sort of the bundling versus unbundling argument? As we think you gave an anecdote before like transcription, revenue cycle management, maybe you got prior off, you’ve got, you know, a whole chain of workflows involved, from the initial patient provider discussion all the way to diagnosis and coding and everything else. Do you think that most of these workflow elements will be handled on a point by point solution like best best of breed, you know, point solution, or do you think they will be bundled comprehensive solutions like a bridge that, you know, end up taking most of the market?
I think that’s the that’s the multibillion dollar question. And so, here, here’s what we’ve seen. We’ve seen today I mean, it’s uh, you know, it’s like the third inning right now, it’s like we’re past the first inning, but like most of the game has yet to be played. What we’ve seen is two things. One is, we have seen the best of breed players rapidly try to launch products number two, and three. So I’ll give you an example in transcription. The 800 pound gorilla in transcription is is not a startup. It’s a company called nuance, which is owned by Microsoft, and it was acquired for $16 billion a few years ago. They have the preponderance of market share in the in the transcription market and healthcare today. But it is a point solution, it is you know, they were product called Dragon or product called DAX, and they go through and they sell the health systems and, and do it and that has been kind of a standalone product. plethora of new startups have come their britches, the Sookie used INNOPOLIS, the ambience of the world, just to name a few that have come up with, I’ll call them better, faster, cheaper type ways of doing nuance, right? They’re there, they’re faster at transcription, they’re more accurate. They’re using, you know, interesting combinations of models. They’re they’re cheaper than nuance, lots of lots of innovation there. And so, so two things have happened, you know, like in this chess game. So thing number one is nuanced, owned by Microsoft, Microsoft, obviously, the partners with open AI, they have, you know, played the Microsoft playbook of how do you how do you destroy slack, you give away teams as part of the bundle, right. And so you, you are seeing the beginnings of, you know, that nuanced Microsoft bundle selling of the health systems. And we’ll see how that goes. nuanced, I would say today is not the best of breed player, it is not better, faster, cheaper, but it is bundled. And the flip side, you have, you know, all of these startups that have started as finished, as transcription companies, they’re all very rapidly launching products, two, three, and four. Some of those products are in the coding space, some of them in this clinical summarization space, but they’re quickly increasing the surface area in which they interact in the healthcare system. And so, it is a really interesting way that you’ve asked this question of why like, and I’ll reframe it, which is, you know, which bundle will wit is it the bundle of the Microsoft product suite, you know, which has done a really good job of, you know, in the Excel teams, you know, Outlook type type world, just maintaining to be, you know, to multi trillion dollar companies, it’s a real, it’s a real, it’s a real beast, versus this emerging, but bundle that you’re seeing startups play, which is everything on the admin side, you know, it says transcription is decoding it’s the clinical summarization, and so on and so forth. And how, you know, how and when health systems decide that between those bundles, I think that’s that that is the battle of the next decade on the admin side, it’s gonna be fascinating to watch.
Maybe maybe one more question on this topic related to AI? Do you think that the market opportunity for AI technology startups will be larger with existing codes, or with new codes that are derived based on you know, what can be done going forward with with AI?
And you mean, like reimbursement codes? Or you mean, like, as right? That’s, it’s, it’s really, it’s really interesting. I’ll, I’ll go back to, you know, comment before that we talked about which was, how do we think about what areas we don’t invest in? Right? And like, well, if it’s, you know, if it’s really regulatory, driven, you know, we really think twice or trice, you know, to make sure we understand that risk? Well, it’s interesting about what you just said about these new building codes, new billing codes don’t arise that often. But there’s been a few in the last few years where AI can be very applicable. And so there’s codes in and around remote patient monitoring. So how do you monitor a patient once they leave the hospital? There are codes around chronic care management. So how do you make sure if someone has diabetes or whatever, even if they’re not in a physician visits, they’re in that, you know, January to May gap that there’s some continuity of care. So what’s really interesting about what you’re alluding to is a lot of remote patient monitoring, especially with the growth of multimodal can potentially be done using AI. So what is today a social worker visit or a nurse visit or a physician visit? In 18 to 24 months, there’ll be some sort of device that is able to do that today, it’s an iPad. But tomorrow, it’ll maybe something maybe that makes sense. Something else that we’ll be able to do that that physical assessment, do the audio transcript, maybe with a human in the loop just to make sure things are right, but very much driven. And you look at the chronic care management side, those are phone calls. You look at you look at all the innovation in AI, around call centres around being able to have a voice respond in real time with low latency. That technology is improving. literally by the week. And so when you fast forward, I don’t know how my five years went by fast forwarding a few quarters, the government will have a really interesting decision on their hand about an inch private insurance companies to have like, do you reimburse AI driven phone calls? Do you reimburse AI driven, remote patient monitoring type systems. And the ying and yang will be this, health insurance companies want to keep their costs down. And so they’ll be worried about the number of bills being generated by AI agents. The flip side, if these AI agents are actually doing a, a, a high quality job, are able to reach these patients or their family members in lower cost ways you actually may be preventing the most expensive thing, which is a hospital visit. And that will be the health economic analysis that needs to happen. So whether it’s a new code that enables that or whether it is the existing code allowing for for a technology to do that, that is a question. There is a precedent, right. And the precedent is 40 years ago, Only doctors could build for codes. Today, you could have nurse practitioners built. Right. And so you have seen different different practitioners or different licensures being able to do more and more of at a crucial time. Yeah, exactly. How does that intersect with AI? There’s there’s those there needs to be a credentialing process. Does there need to be a licensure process? Which sounds like a strange concept. But again, the FDA already has processes for these, you know, like they have approved algorithms, you know, that they you know, for AI today, not generative AI, but predictive AI? Is there a similar type of bar that has to be met here? I think these are at the system’s level of like some of the biggest questions we’re gonna have to wrestle with in the next few years. Awesome. Or if
we can feature anyone here on the show, who do you think we should interview? And what topic? Would you like to hear them speak about?
Oh, gosh, that that’d be so if I if I had a magic wand, then you can interview anyone, I’m going to stick with healthcare here. I think it’d be really, really interesting to interview folks who’ve had a systems level type perspective on health care system on the healthcare system and how it’s changed. So when I think back the last decade, I think about the folks who have either run, you know, CMS Centre for Medicare services, either under Republican or Democratic administrations, or folks who run the innovation arm for those for CMS. So, folks like Seema Verma, you know, who was under Trump, or Andy Slavitt, you know, prior to that, I always find to be really fascinating type folks where they’ve been able to really zoom out and think about, you know, how does changing, you know, you know, or tweaking one law here, actually have a downstream effect and different different derivative areas there. I always loved those conversations, because they can go in a multitude of different areas and, and always helps me Zoom, zoom back out. And so I would pick someone, someone from that systems level type thing.
Perfect ombar What book article or video would you recommend to listeners? Yeah,
I read an article a couple of months ago, published by a professor at Penn named Zeke Emanuel, and he talked the article, I’m going to butcher the name, but it essentially is titled, Why lie, I don’t want to live beyond 75. It’s a very provocative title, you know, for an article, especially for a medical professional and a professor at Penn. But in that article, it’s relatively short, it’s 10 full pages. It talks about a lot about ageing, and around a lot of what happens to the human body as it ages, both from a mental perspective around dementia and Alzheimer’s, but also from a physical perspective about the onset of diseases, cancer, heart disease, etc. And it’s written in a very provocative way, where he’s like, Hey, if I’m sick at 75, like, you know, pull the plug, you know, that is almost as provocative as that. But it was one of those rare ones that made me really think deeply around, not just extending life, but how do you maximise the quality of life? How do you actually think about what do we what should we do from a healthcare perspective such that 1015 20 years from now, we’re able to change the title of that article from, you know, why I don’t want to live to you know, be on 75 to beyond 85. Like, are there ways we can we can actually prolong those, you know, the onset of disease, their prolonged dementia, or the onset of dementia? But if we can’t, really thinking about how do we think about end of life and the flu upside is how do we enjoy, you know, the rest of our life up until that moment what you know, you can pick your number 175 or 80, or whatnot. But it really was one of the one of the most humbling articles and provocative articles I’ve read in the sector. I’d highly encourage it for folks love
it. Well, I noticed your you were a psychology undergrad. So maybe we can have a good discussion about what life really means without death. And then finally, your ombar Do you have any habits, tactics or techniques that are a secret weapon.
And this is a strange one. But one of the things that I try to not do every day, I try not to read industry news every day, I still get all the newsletters in my inbox, but they’re auto forwarded to a folder. And I read them all on Saturday mornings or Sunday mornings. But I tend to find that once I you know, if you get those in your inbox every day, your mind immediately begins to be reactive. And you’re just reacting to the news, you’re reacting to someone else’s thoughts that are in your in your head, versus having the bandwidth and the ability to start your day thinking more proactively about the world and about the date ingesting those data points. And so I started this habit a few years ago, where I might be like four or five days late to some piece of funding news, or some exit or something that has happened. And I realised like the timeliness of knowing such and such a piece of information. It’s just not that meaningful. And so that’s like a little hack I have in my life. And I think it’s not necessarily a time hack that saved me, you know, a few minutes a day, but it’s more of like a mental hack of how do I always make sure that you know, in those prime moments of working hours, the thoughts in my head are as much as possible my own proactively versus kind of reactive to inbound news.
Perfect. He is ombar but the ceria in the firm is Maverick ventures. Barra thanks so much for all your your insights and wisdom today. I feel like I have a much better understanding of healthcare and what we have in store in the coming years. So thank you.
This is awesome. Thanks a lot.
All right, that’ll wrap up today’s interview. If you enjoyed the episode or a previous one, let the guests know about it. Share your thoughts on social or shoot them an email, let them know what particularly resonated with you. I can’t tell you how much I appreciate that some of the smartest folks in venture are willing to take the time and share their insights with us. If you feel the same, a compliment goes a long way. Okay, that’s a wrap for today. Until next time, remember to over prepare, choose carefully and invest confidently thanks so much for listening