4 minute read | January.23.2025
In this Life Sciences & Healthtech edition of The Download, we take on key trends at the intersection of tech and healthcare, continuing the energizing conversation from JPM Healthcare last week.
01 A Private Sector Model for Responsible Use of AI in Healthcare
02 Why Value-Based Healthcare Is Good Business
03 How to Prepare Your Life Sciences / Healthtech Company for a Successful IPO
04 The Download Quiz: Life Sciences & Healthtech Valuations
Jeremy Sherer:
Well, thanks so much for being here today, Brian. We really appreciate it.
Brian Anderson:
Thanks for having me, Jeremy.
Jeremy:
Maybe I should say Dr. Anderson, depending on which hat you want to wear.
Brian:
You can call me Brian.
Jeremy:
Can you tell us a little bit about your background, how you got to where you are today, and then a little bit about CHAI as an organization.
Brian:
I'm a physician data scientist by training. I practiced clinically here in the greater Boston area in primary care and then practiced north of here for about eight years. It was an interesting experience. Part of my practice was working at a federally qualified health center, which, FQHCs if you know them, they don't have as many resources as an MGH might have. And part of my experience there was challenged by the implementation of some of the electronic health records that we had and the time that it takes to document and spend staring at a computer screen. It was really challenging for me as a practicing clinician to get home on time to spend time with my family. And candidly, I got a little bit burned out. And so, pivoting from clinical practice, I pivoted to work at a company called Athenahealth, which at the time was a startup.
I really went into the space of wanting to make the technology better for clinicians. From there, one door opened up to another. I ended up working at a company called Kyruus and then worked at a company called MITRE, which does a lot of work with the U.S. government. In that capacity, I was the chief digital health physician and led a number of strategic efforts in the digital health space working with the U.S. government. It's in that space that I was asked to help lead a lot of the data analytic efforts in the pandemic for the White House COVID Task Force, focusing on vaccines and monoclonal antibodies, some of the therapies that we had. And it was during that time that you had really non-traditional companies coming together, like Microsoft and Google and Amazon and health systems that were inherently competitive. Pharma companies that were inherently competitive. And we were doing a lot of good work.
Jeremy:
It was a unique window when everybody's aligned.
Brian:
Exactly. I think it was in that unique period that we asked, “Wow, we're doing a lot. Can we do something more in a different space?” We looked at AI, and that was the genesis of trying. In March 2021, we committed to bringing our resources together to launch CHAI. The focus of CHAI has always been, how do we bring groups together to build a set of consensus-driven best practices in responsible health AI?
Jeremy:
So there's two things that are important, and I want to touch on both. Let's start with your coalition, because I think a big part of what attracts attention across the industry, and for good reason, is this incredible coalition that you built. It's unusual to have a collection of leading health systems, academic medical centers, and all the rest coming to the table together with all of the tech giants that you just mentioned, as well as startups and smaller health systems. So tell us a little bit about these stakeholders and what you think is bringing them to the table to be involved with CHAI and what they're getting out of it.
Brian:
Yeah, great question. When we started in March of ‘21, it was eight doctors from health systems and two tech companies. So it was fundamentally a private sector-led effort. Very quickly, it grew from there. I didn't appreciate the appetite that so many stakeholders in this space had about wanting to come up with consensus definitions about what good, responsible AI looked like. Yes, we have all of the name-brand recognized big companies like Microsoft, Google, and Amazon, and we have the big AMCs like Mayo and Stanford and Duke and Johns Hopkins, but the coalition quickly grew to include over now, 3,000 organizations that are part of our coalition. 25% of those are startups.
It's very clear that it's not just the entrenched incumbents that are interested in developing these common definitions. It's very much software developers at startups wanting to come together and come to consensus on what the technical specifications look like on, “How do we think about fairness? How do we think about transparency? How do we think about effectiveness and safety in responsible health AI?” Those kinds of best practices can allow a startup to accelerate and grow when you have a common definition of what responsible AI looks like and developers are developing them. That same definition is agreed to by their customers, big health system or small health system, that allows the kind of innovation and growth and the use of AI in safe and effective ways to really scale and go much faster than if you didn't have that kind of common agreement about what the target of responsible AI looks like. So we're really excited.
The other important stakeholder that we have, of course, is patient community advocates. So, very quickly, as we've grown, certainly with the tech companies and with the health systems, is ensuring that every one of our working groups that includes those stakeholders has patient community advocates. The other really exciting thing about CHAI, yes, it's private sector-led, but we quickly got interest from the public sector. At the federal level, when I think many in HHS saw the work that we're doing in responsible, ethical, safe, and effective AI, they wanted to participate in them.
I think that's really become part of the magic of CHAI is you have these innovators – from startups to big tech, big health systems to small health systems working together in an innovative way to come up with these definitions – informing the policymakers and the regulators at the public sector side. It's very much a synergistic process. You do not want regulations that are uninformed about where the private sector innovators are going, where the new use cases are being developed. And you don't want the innovators developing these use cases being uninformed about where the regulators might be developing these frameworks. So it's that kind of synergistic approach that has come to be our calling card at CHAI.
Jeremy:
What have you done to establish that – first of all, to the extent that works, it is unique and interesting and somewhat novel that we had the idea of an admittedly, partially, but private sector organization setting standards for an industry. Right? That's typically something that we rely on the government regulators to do. But what have you done to set CHAI apart and sort of indicate to federal government and state governments as you're talking to them that they can trust you guys?
Brian:
Yeah, it's a great question. So on the standards part, AI is such a new space, particularly generative AI. Traditional AI, yes, it's been around since the ‘70s. But the use cases and the use of it in very consequential spaces in health is new. Developing technical specificity on what good, responsible AI looks like is something that not many organizations have done. So one of the unique differentiating things about CHAI is the technical depth by which we're taking our working groups. It's not at a 50,000 ft level where we all sort of agree, yes, we want fairness. Yes, we want transparency. It’s software developers, what does that actually look like? And so step one is coming up with that technical level of specificity about what good, transparency, fairness, whatever the responsible AI principle is, taking it from 50,000 ft down to where the rubber hits the road.
The important distinction I would make between what CHAI is doing and what standards development organizations are doing, which do work really closely with the U.S. government, is CHAI right now is focused on developing these frameworks. Now, these frameworks will ultimately become standards at some point. But that standard development process, because it works so closely with organizations within U.S. government, it takes time. It can take between 1 to 3 years to develop a single standard. And so part of what we're doing in CHAI is trying to move at the speed of the innovators, and so focusing more on developing these best practice frameworks and not going through the formal process of the standards development organization. In parallel to the framework development, we are also working with standards development organizations like HL7, like ISR. There are quite a few organizations out there that are very interested in partnering with CHAI, like CTA, to be able to develop formal standards.
Jeremy:
That are known and that are known to folks in Washington and elsewhere.
Brian:
Exactly. I think part of the excitement is, to the second part of your question, how do we do this in a way that's trustworthy? And so part of that, our strategy in that is doing it in as transparent a way as we can. I mentioned CHAI has a lot of members, that's not the entire community in the health, AI or digital health ecosystem. So what we do is we publish all of our things and we make them freely available to the community. And in the draft forms, that initial first draft of the publication that we make, we want to hear feedback from the entire community, from people that are not part of CHAI, from patient advocates all across the nation, from other companies that aren't part of CHAI. We want to offer them an opportunity to give feedback. So we're committed to ensuring that the working groups that develop these technical frameworks publish them and then we have an open comment period. Very similar to how the federal government does their kind of rulemaking. It's 60, 90 day period. And then intentionally wanted to take that feedback in finalizing the draft technical documentation. The other part of building trust is CHAI, as a nonprofit, we want to be seen as a convener and a facilitator to ensure that we're not driving any particular agenda in CHAI. We really want to bring together the innovators to facilitate the consensus process, to build these technical documents. And that's the space that we want to occupy in CHAI.
Jeremy:
So I wanted to ask a little bit about industry. There was a SBV article recently that said that about 30% of investment in healthcare, largely, was in AI last year. There's still a lot of folks, given where we are in the zeitgeist, at CES you couldn’t walk past anything without seeing the term artificial intelligence in it, there are concerns about AI washing. But given your background as a physician, what are you really excited about right now on the clinical side?
Brian:
To have the opportunity for a physician and a patient to more meaningfully engage, so that the patient and the doctor can understand what the actual problem might be for that patient. As an example, there's a lot of companies in the AI scribe space. Part of the reason I shared earlier that I transitioned out of medicine was because of the challenge of actually being able to connect with the patients that I had because I had this clunky EHR that got in the way. Enabling doctors and nurses to more meaningfully connect with their patients is going to hopefully address some of the significant burnout issues and staffing shortages that we have across the nation, both for nurses and for providers. And create a level of, hopefully, a level of efficiency that will improve the kind of clinical care that patients experience. The knowledge that a lot of these AI models have in terms of, diagnoses, therapeutic, protocols that are potentially the right kind of therapy or the right kind of protocol for a patient, also enables a provider to have at their fingertips a copilot or a companion that would really help them potentially identify the right diagnosis, identify the right kind of treatment.
There are so many stories out there already where you have a provider that's perplexed about what to do with the patient. And the patient takes their history and puts it into one of these frontier models and then brings it back to the provider and really is like, “Wow, that's the solution. Let me help solve it.” You know, one of my close friends, had a kind of cancer, colorectal cancer. And he was seen by one of the most world-renowned oncologists out in California. Didn't know what to do. He took it, put it in I think ChatGPT, and ChatGPT told him the rare form. Here's the mutation, here's a protocol that you should consider. It was just published in the journal, so the doctor didn't know. Took it back to the doctor, who said, “Holy cow. Wow. That's it.” And now he's in remission. And so I think those are the kinds of real-life stories that are going to begin to bring attention to how these kinds of tools can be used by providers and patients to really improve care.
The other thing that really excites me is we have an access problem in healthcare, right? Partly because we don't have enough doctors and nurses. Across the U.S., particularly in rural areas, inner city areas, AI has the power to meet patients where they are and help them connect into the healthcare system and really address their issues. And so I'm really excited about that as well. How can we meet patients where they are if they're in rural environments, inner city environments, and address their clinical needs?
Jeremy:
That's really interesting. So you're a self-described tech geek. You walked into the room and you were playing with cameras here and stuff. You're also a physician by training and, as you're going down this road and on this journey, what are you finding that the tech industry needs to understand about virtual healthcare stakeholders, and conversely, what healthcare folks have to understand about tech?
Brian:
Such a great question. One of the really unique things about health is that it is profoundly intimate and personal. We are all patients. We're all going to be caregivers. And the importance of understanding the unique individual needs and the nuances of an individual from their social determinants of health… the richness of their story… and how personal that is and how it affects their health is really important and a unique attribute in the health space. As opposed to, if you think about fintech or other spaces where it's a little bit more abstracted and it's about numbers. The importance of understanding an individual patient and an individual doctor is going to be really, really important and oftentimes overlooked.
The other part is the consequence out of it. So health obviously has life or death scenarios and uses of AI that could potentially help improve a clinical outcome. And so in those situations where the risks are high and we think about, as a technologist, one of the often used approaches is move fast and break things and develop a minimum viable product and test it. And if it breaks, iterate on that, improve on it. That's an approach that needs to be modulated when we're in the health space, because obviously you don't want things to break when it’s a person's life cycle. And so it's a challenge to innovate and move fast and do that in a way that takes into account all this rich complexity that is unique and personal to an individual. And but yet, delivering a real impact and a real result.
And so oftentimes when I speak to innovators in the private sector space, it's important for them to recognize that partnership and collaboration, like real, meaningful partnership and collaboration with physicians and nurses and their patients, and developing the kind of trust and how we use AI in these consequential spaces, is really important. Innovators, particularly in AI, if you're in it every day, don't appreciate that we have a really bad or really profound trust problem with AI. Survey after survey out there shows the majority of Americans don't trust AI. If you add a consequential space like health into that, they trust it even less and so if we want to move fast and innovate, we have to develop the kinds of real collaborative partnerships with the stakeholders that are part of. And that includes nurses, doctors and patients.
Now, what does the healthcare system need to learn from the technologists? Part of the brokenness in the healthcare ecosystem as it relates to technology is the siloing of so much of this and the misalignment of incentives. Now, those obviously need to be addressed at a policy level. But from a technologist’s perspective, it's really important when you're partnering with the healthcare ecosystem to understand one, that the incentives aren't aligned, but two, there's a real challenge with healthcare systems always thinking about, well, we're unique. “There's no one like us, right?” The clinical workflows, yes, they're complex. Yes, they're challenging. A level of willingness to collaborate with the technology side of the stakeholder is really important. And oftentimes I see that kind of collaboration break down when the healthcare providers or nurses don't understand how to do that in a way that allows for that kind of innovation and iteration. There's oftentimes an approach that it has to be perfect before I use it. And that's not how you innovate. Right. And so it's that balance between how do you innovate and move fast and do it in a safe way so you're not breaking things, but do it in a way that you're actually going to potentially disrupt something that might make a doctor or nurse uncomfortable. And I think that's a real challenge for some of us in this space.
Jeremy:
Absolutely. So, it’s mid-January 2025 right now. And everyone is thinking about the upcoming change of administration. So, at the risk of being the 1,200 person to ask you about this this month, how is CHAI thinking about that upcoming change? And, in terms of continuity and the efforts that you had underway in Washington, issues in the states, what’s front of mind?
Brian:
Yeah. So some really exciting opportunities are on the horizon. As I said, CHAI when we started, fundamentally led by the private sector. Yes, the public sector, yes, the current administration joined. We have representatives from the FDA, ONC, NIH, VA all participating in our working groups. But this is fundamentally a private sector led effort where it's bipartisan in nature. Like nobody wants AI that's going to hurt someone, right? We all want AI that's fair, safe and effective for all of us. And in the conversations that I've had, that my team has had, from the Hill day that we had back in November to ongoing conversations we've had with leading senators and representatives from Congress, as well as at the state level, one of the things I've appreciated is their desire to enable or allow the private sector to have time to really develop these guidelines and guardrails.
I think one of the more consistent things I've heard recently is we don't want a top down government approach where the government says, here's the regulation, right? Without it being informed by what is the private sector coming to consensus on. As an example, how do you measure bias generative AI? I can tell you right now, we do not have consensus on that. And so if there were a regulation that came out specifically with like, if you have bias, this is what's going to happen. One of the challenges, we would in the private sector immediately say like, we don't even know how to measure it. Like, how do you do that? And so, the conversations and the feedback that I've gotten has been great. Private sector, go do your thing, like build this consensus, build these best practice frameworks. And we want to work alongside you. And yeah, as clarity comes, certainly we'll develop the appropriate regulations. I've also heard you know, we certainly don't want to build regulations that are going to stifle innovation.
The other important part is how does how does this ecosystem self-regulate? How do you create those internal guardrails that ensure that kind of safety and efficacy in this space? It's moving so fast. And part of what we've done has been launching things like the model cart, this AI nutrition label, or developing private sector specific uses of a quality assurance lab. And so these are fundamentally grounded in a collaboration between the AI developer, the vendor, and the customer. And so what we are hearing in CHAI, very loudly from the customer side of this is, we want greater transparency into how these models are built. We want to be able to have a vendor take their model and have it tested by a quality assurance lab. You know, we love the PowerPoint slides. PowerPoint slides are great. But to have the kind of trust that I was talking about earlier, having that independent entity that can validate a model's actual performance, is a really powerful way of building trust and advancing a procurement process through to a final deal. Right? Like you don't want a vendor to be stuck in a pilot or in a bake off. If the customer can get the kind of transparency through things like a nutrition label or the use of an external validation from a quality assurance lab, those are the things that we're identifying in CHAI as ways that we can, internal to the private sector, develop these guidelines and guardrails. And the public sector, at a federal level and at state level, I think that's a great idea, right? Like, how can we have the private sector begin to develop these kinds of approaches? And then, again, I never want to speak for a policy or a regulator, but they might be able to make use of those at the more appropriate time. And so what I've been hearing from incoming administration officials is a strong willingness to partner with us in that, to support us, to give us the time and space to be able to do that. I'm really excited about that.
Jeremy:
It's incredibly exciting. Not to give you a crystal ball, but if we're looking ahead to 2025, thinking about what you’re excited about, what CHAI is working on, what the possibilities are. What do you hope is going to be CHAI’s biggest win in 2025?
Brian:
Two things. The first one is an effort that we'll be announcing in greater detail. I'm happy to share with you. Now, there's been a lot of talk about agentic AI, right? We were at CES. Everyone's talking about AI. AI agents essentially are agents. They're operating in semi-autonomous or autonomous way. In whatever the context is, in the health context, one of the exciting things that we're going to be focusing on is the ability for an individual, you and me, to have a multiple number of AI agents that could be operating on our behalf, helping us, advocating for us, assisting us in navigating a complex healthcare system. But the challenge is, coming back to what I said earlier, because health is so personal and individual, your values, your priorities might be different than mine, different than someone else's. And so how we build these AI agents in such a way that allow them to be aligned to your definition of human flourishing or your values as it relates to health is both critically important but also really, really exciting. And so one of the things we're certainly seeing in 2025 is this movement from a kind of single monolithic model that you have to go to the web to use to the potential of having agents that are on your phone, or on a personal device that you can take with you or have working in the background for you and in the health domain. That has really profound implications.
And so one of the things we're going to be launching in CHAI is an effort to how do we think about enabling developers to develop these kinds of models with that understanding of how do I adjust the model or tune the model in a way that allows it to be, sensitive to your values, my values? In a way that, serves all of us in a way that is cognizant of how your values might be different than mine. And when we do that, we have the opportunity to unlock a lot of specific use cases in the health domain that allow the kind of human flourishing that we all want, in a pretty consequential space. So it's going to take a lot of people and a lot of consensus to develop what that framework is. So we're excited. We're going to be pulling in ethicists, spiritual leaders, philosophers, certainly a lot of technologists, doctors and nurses and patients to develop that kind of framework for evaluation.
The other really big thing is the model part. We've seen amazing adoption of this already. A number of startups have already begun partnering with us and building out the applied model card quickly.
Jeremy:
Could you recap, just for folks who aren't in this day in, day out, what the model card is.
Brian:
Yeah. So the CHAI model card is essentially a nutrition label. When you go to the store, you look at a box of cereal or a can of soup, you turn it around and there's the nutrition label. It has a high-level description of carbohydrates, proteins, fat, that sort of thing. And you know what's in it. A model card for AI essentially helps you know what's in it. Like, how was the model trained? What kind of data was the model trained on? What are the indications? What are the limitations? What are the ingredients? Those are basic things that you as an end user, a patient or a doctor, would want to know to help inform, should I use this tool on this patient or should I, as a patient, have this used on me? That kind of base level minimum threshold of transparency is one of the big efforts that we launched at the end of last year, bringing consensus from the vendor side and the customer side was really important because then that sets the stage for this year for the adoption and utilization of that. And that kind of level setting, an agreement on what transparency looks like, I hope, and I'm really excited to see this, will be a process of how we build trust in the AI space for health. And in particular, as we're building that kind of trust out as the customers, as doctors and nurses and patients trust these tools more, that's going to unlock the kinds of more consequential use cases for AI that we want to truly help a person in a life or death situation. Right. Because if you don't have that kind of transparency as a doctor, I wouldn't want to use a model that I didn't know how effective is it on a person, the person I have in front of me on death’s bed. So we're really excited about the adoption and utilization of the model card and nutrition label and this project on human flourishing that will be launching.
Jeremy:
We're excited about it. And for everything that CHAI will accomplish in 2025. So thanks so much for making the time. Really appreciate it.
Brian:
Thanks, Jeremy. Thanks for having me here. I appreciate it.
Brian Anderson, the CEO of the Coalition for Health AI (CHAI), joins Orrick’s Jeremy Sherer to discuss how industry is coming together to develop responsible AI standards in healthcare. Some of the topics they cover include public sector collaboration and CHAI’s model card that, like a nutrition label, tells a customer, patient or doctor exactly what the AI model is made of.
Learn more about this dialogue among 3,000 tech and life sciences companies who have joined CHAI, including tech giants such as Microsoft, leading health systems and hundreds of startups. Brian shares insights based on his experience as a digital health startup operator and a physician.
“Part of the magic of CHAI is you have these innovators – from startups to big tech, big health systems to small health systems, working together in an innovative way to come up with these definitions – informing the policymakers and the regulators,” Brian said.
Albert Vanderlaan:
My name is Albert Vanderlaan. I'm the head of global capital markets at Orrick, Herrington and Sutcliffe, based in Boston. I'm joined today by Dr. Farzad Mostashari, the CEO of Aledade and the co-founder, as well, of Aledade. Very excited to have you here today. Do you want to give a quick background about yourself?
Farzad Mostashari:
Sure. I'm a public health doctor. Originally, I trained in internal medicine, and I worked at the New York City Health Department for ten years. Then I went into the federal government. I helped lead the efforts to roll out electronic health records to American healthcare during the Obama administration. Then I founded Aledade, over ten years ago now.
Albert:
That's amazing. We're going to be talking a lot today about value-based health and value-based healthcare. It would be great to get a little bit of background on what that really means, so our audience can have a frame of reference for where it stands now.
Farzad:
There are many flavors of it – accountable care, value-based care, risk-taking providers. A lot of different flavors for it. But the shorthand that I like to think about is, do you make money preventing strokes or do you make money treating strokes? Value-based care is basically the one pocket in American healthcare where you make more money if you prevent strokes.
Albert:
Got it. So, it's really that alignment of interests that you're trying to do more preventative care or do that just from a baseline perspective to ensure on the back end you're not running up the cost?
Farzad:
Yeah. When I was in the New York City Health Department, Mike Bloomberg asked us, like, how do we save the most lives? It was such a startlingly simple question. Because then you start working backwards and you're like, oh, well, if we really cared about saving lives, what we would all be focused on are things like controlling blood pressure, which we do about 65% of the time in American healthcare.
Albert:
Interesting. Wow.
Farzad:
Right. And so then why don't we focus on preventing heart attacks and strokes and controlling blood pressure, which we know how to do? We don't need to invent new technology or medicines for that. We just need to pay attention to it. Why don't we? And that fundamentally, what I realized was, it wasn't a knowledge problem. It wasn't a technology problem. It was a healthcare incentives problem. Bringing that alignment between what's good for the patient, clearly being what's also good for the provider, and what's good for society. Bringing all three of those into alignment is what value-based care really is.
Albert:
Okay. That makes a lot of sense. I wish we were doing more of it. But in that frame, what is the future really of value-based care from your perspective? And really what is Aledade doing to forward that progress?
Farzad:
Winston Churchill once, it's attributed to him, said that Americans always do the right thing after they've exhausted all other possibilities.
Albert:
Sounds accurate.
Farzad:
So that's why I feel with value-based care, that's where we are right now, is we have exhausted all other possibilities. If you think about how do you control healthcare costs growth in a world where there's no limits on how many MRIs you can get or what fancy test that may or may not be good for you… How do you control that? You get into the fee-for-service mechanisms, right. Like let's shift all the cost to that patient. So they have skin in the game, right? And we've seen the failures of that because people then stop taking their medicines that are lifesaving just as much as they stop going to the useless test. Then you get into denials, prior authorization. How do you like that? We have exhausted all other possibilities. Value-based care is really showing that it works, that if you give primary care, in particular primary care providers, the incentives, the data, the regulatory opportunity to take total cost of care, total quality, be the quarterback of care for their patients, that they can reduce healthcare cost trends while giving patients more, not less. I think that's the unlock for this whole thing is better, more informed and more engaged primary care.
Albert:
That makes a lot of sense, especially given the overall construct of how interfacing that primary care needs to be with insurance and other providers and giving them that incentive to really quarterback it and overall help patient care and outcomes. How is Aledade working with them and orchestrating your construct within that ecosystem?
Farzad:
Yeah, so Aledade has now become, over the past ten years, the largest network of independent primary care practices engaged in value-based care. We allowed the practices to keep their autonomy, while we bring to them the technology, the data, the contracts, the coaching, and the playbook in order to succeed in the value-based care world. We have about 2,500 practices across the country now, over 2 million patients, which translates to something like $25 billion a year of medical costs that we're managing. We're getting about 2% lower healthcare cost trends year after year after year. So, you know, by five years in, you're 10% lower total cost of care than you otherwise would have been.
Albert:
Wow. That's very impressive. How can you accelerate that or what's your growth strategy to acquire more primary practices to come into that ecosystem?
Farzad:
A lot of this right now is organic growth. So, we succeed, in some of our markets where we've been there the longest, is where we continue to have the most robust growth. In the state of Delaware, this year, our practices got a check from us for shared savings, for actually reducing hospitalizations and reducing cost trends. That check paid them more for those patients than they had made billing fee for service visits all year long for those patients. So when those practices talk to their neighbors and friends, that's when you really get the organic growth spinning. We did do a tuck-in acquisition this year or last year in ‘24 in Michigan. That may be something we do more of. It was a great group, a management services organization that had great relationships with hundreds of practices that they were supporting. Culturally, a good fit. So there may be more of those in the future.
Albert:
Gotcha. So we're sitting here on the eve or a couple of days away from a presidential inauguration. We're not sure fully what some of that package is going to look like from a regulatory perspective. But any insights around where you guys are thinking the future of the regulations may impact your business, or what that may look like from a regulatory environment to accelerate this type of value-based care in the future?
Farzad:
I'm a federal policymaker. And people are, I think, oftentimes feel like the regulatory risk or pen stroke risk, that it's a little more complicated than it actually is. That it's more capricious and more unpredictable. I actually feel like it's not that… like you listen to what people are telling you, right? What do the policymakers want? If your business is aligned with what policymakers want, you will face regulatory opportunity. If your business is built on an arbitrage, if your business is built on a loophole, if your business is fundamentally misaligned with what's good for America, maybe you'll be able to delay it. But you won't sleep easy. I feel like we are completely aligned with what policymakers want, which is better care and lower cost. We're giving patients more, not less. Through Republican, Democratic, Republican, Democratic administrations, value-based care over the past 15 years has been something that's been a consistent thread. I anticipate that there will be some different flavors. But there will be continued support for better care, lower cost and really incentivizing, and I think this is a very conservative idea, is creating incentives that align private profit with public good.
Albert:
Makes a lot of sense and also in the context of a lot of healthcare companies over the past few years have really been struggling a little bit. Is part of what you just described why you think it's been successful over that time period when others have been struggling?
Farzad:
Well, it is an interesting time in Medicare Advantage, particularly. I think is a lot of what you're referring to is health plans have had lower margins and risk-taking providers and Medicare Advantage have had a hard time. Aledade has as our base, as our core engine, a government program, the Medicare shared savings program, with traditional Medicare patients. That business has been remarkably consistent in its rules, in the regulations and in our results. We last week or, just a few days ago at the J.P. Morgan conference, we announced our expected ‘24 results in that program, which was to go from $538 million of revenue in ‘23 to 750 million of revenue in ‘24. So continued very strong growth off a big base there. That is part of it. But I also think in Medicare Advantage we are very actually bullish on it. I think that's going to be the next leg of the business for Aledade.
Albert:
Okay. Fascinating. I think that that makes a lot of sense. Again, just in the context of what you're trying to accomplish. So, you mentioned J.P. Morgan Healthcare that we're just coming off the back end of as well. Any other insights from others that were at the conference that you found to be really resonating in this space or other insights showing in terms of maybe big pharma and where they're leaning into the value-based care proposition or the opposite?
Farzad:
There was a lot of talk about where the future of Medicare Advantage in terms of policy is going to be. Clearly the Republican administration is going to be a little more, I think, inclined to be pro Medicare Advantage. Versus, I think some of the concerns that folks in the industry had about the Biden administration feeling like they were being tough on Medicare Advantage plans. For me, one of the insights was that a lot of what people ascribe to politics is actually just actuarial math. The actuarial math seems to indicate that the actuaries with good intentions misestimated what the growth rate was going to be for utilization post-COVID. We're still living through the waves of COVID rebound. There was a big rebound of utilization into ‘23 and into ‘24. You heard United Healthcare reporting yesterday continued increase in utilization in Q4 of ‘24. I think it was just a mispricing of how much the feds allowed for cost growth, and how much actual cost growth took place. That is going to be mathematically, by law, corrected in ‘26 and ‘27. That sort of depression down in the funding level inexorably leads to increased funding levels in ‘26, ‘27, which everyone is going to say, “Oh, that's all politics. Trump is going to pay the MA plans more.” It's like, no, it's not. Actually, it's just if you read the regs, if you read the law…
Albert:
It's just good old fashioned read the law and put actuarial math to it.
Farzad:
There you go. There you go.
Albert:
And in that same vein, you know, a question that I've always had around that value-based care proposition and doing so much on the preventative side, as we have an aging population in the U.S, is there a shift over time to how that value-based proposition goes from the preventative to what is needed towards end of life and all of that? Where is there a balance? Is there a balance in terms of how that is going to look as we go forward with an aging U.S. population?
Farzad:
Do you mind if I give a personal anecdote?
Albert:
Absolutely.
Farzad:
People ask me like, how is value-based primary care different? How do you guys get, you know, better care, lower cost? It seems so hard to do. It's not hard to do because the current system is so bad. I know this because I'm taking care of my mom, who's 86 and has congestive heart failure. This summer, let me just go through the story of how someone who's getting care from a wonderful, smart, amazingly caring primary care doctor in a fee-for-service, academic medical center practice, right. Her doc wrote her for a prescription for SGLT2 inhibitor, Jardiance. My parents went to the pharmacy and they said it's going to be $600. They didn't fill it. They didn't tell me. They didn't tell the doctor. Then she has trouble breathing. We call the practice and they say next appointment is 3 or 4 months from now, go to urgent care. We go to urgent care, they say go to the emergency room. 40 hours in the emergency room, 40 hours sitting next to her bed, and she comes home. She doesn't get a phone call from the primary care practice. Like, “We heard you were in the ER. How are you doing?” Right. To your question, when we were in the emergency room, they said before we transport her, you got to sign this piece of paper, which is like, do you want to be intubated? Do you want to have chest compressions done? Do you want a feeding tube – in the emergency room? Right. That's state-of-the-art fee-for-service medicine right now. Right? When you don't fill the prescription, the doctor doesn't know about it. When you need urgent primary care, they send you to the ER. The ER is 40 hours, and they don't know you, and they slide over that form for end-of-life care. When you go home, your primary care practice doesn't know you were there. Basically doesn't have a mechanism to have that workflow.
In Aledade primary care practices, none of those four things would have happened. Her doc would have known she didn't fill the prescription. They would have said, call us first before you go to the emergency room. They would have followed up with her after an ER visit. And we have a program, a counseling program, that has a Net Promoter score of 92 that talks to patients for six hours about their end of life wishes. So that's the difference between value-based primary care and wonderfully well-intentioned, intelligent, smart doctors being burned out because the system doesn't support them in the way they want to be.
Albert:
Very clear. I appreciate you sharing that story, too. I think it brings to light a lot of the transition that, I mean, when I go back to talking to my grandmother, who, you know, died years ago, unfortunately. But she always was like, where's the family doctor? You had the family doctor, you had that ecosystem. They knew those people, and they went into that. Then we shifted over the years to the paradigm you just described, which is very unfortunate. But I think what you're working towards makes so much more sense just from a personal perspective, too, which I think put aside value-based care versus other. It's just personal care too, right? I think that's a huge aspect of it. I really appreciate that. Well and really appreciate you coming in today. Thank you so much. This was a pleasure to have you.
Farzad: Thank you.
Farzad Mostashari, the CEO and Co-Founder of Aledade, makes the business case for the value-based healthcare model and talks about the policy landscape with Orrick’s Albert Vanderlaan.
What is value-based healthcare? “Bringing that alignment between what’s good for the patient, clearly being good for the provider, and what’s good for society,” Farzad said. “Bringing all three of those into alignment is what value-based care really is.”
Farzad brings a valuable perspective as a physician, a policymaker in Washington and now the CEO of Aledade, the largest network of primary care practices engaged in value-based care – with 2,500 practices, over 2 million patients and $25 billion in medical cost under management.
Market conditions are improving as we enter 2025. The life sciences and healthtech IPO landscape is showing signs of life following renewed issuances in late 2024. Companies with compelling clinical data are finding interested investors, particularly for companies with clinical proof-of-concept data in key therapeutic areas like oncology, CNS, immunology and cardiovascular diseases.
Here are five key learnings to help you prepare your company for a successful IPO, based on a series of curated events hosted by Orrick and BofA Securities:
Early preparation leads to better outcomes. Companies should begin preparing at least 6-9 months before their target IPO date by:
Insider support remains essential. A successful IPO requires strong support from existing investors – and we expect this trend to continue. Insider support demonstrates conviction in the company. However, companies must ensure that they balance insider support with new investors to grow their investor base in preparation for post-public trading activity.
Financial readiness requires significant time. The audit process is usually the longest lead-time item. Companies with private company audits can undertake a public company “uplift” when the IPO nears, but others without the benefit of any prior audit will need to go straight to a full PCAOB audit – a substantial lift in either event for small teams at many biotech companies. Working with auditors to reserve resources and establish timelines is critical to ensure expectations are met on time.
Internal infrastructure is a must-have (even for companies that delay an IPO). Successful IPOs require well-built infrastructure that is put in place in advance of an IPO (the earlier, the better). Companies looking to an IPO must:
Most companies should undertake a dual-track process to ensure the best outcome for stakeholders. The current financing environment appears to be opening more than in the past two years, but most companies should still explore a dual-track IPO and M&A process to maximize value for stakeholders. The dual-track strategy requires coordinating advisory teams across both tracks and efficient management of parallel workstreams.
A) 25%
B) 34.7%
C) 42.2%
D) 50%