Howie and Harlan are joined by Christopher O’Connor, CEO of Yale New Haven Health, to discuss his career path and his experience leading hospitals through Hurricane Katrina and COVID-19. Harlan reports on promising AI tools for taking clinical notes; Howie looks at the financial headwinds facing the companies offering Medicare Advantage plans.
Harlan Krumholz: Welcome to Health & Veritas. I’m Harlan Krumholz.
Howard Forman: And I’m Howie Forman. We’re physicians and professors at Yale University. We’re trying to get closer to the truth about health and healthcare. We’re excited to welcome Yale New Haven Health System CEO Chris O’Connor today. But first, we like to check in on current hot topics in health and healthcare. What’s on your mind today, Harlan?
Harlan Krumholz: Well, Howie, I don’t know. Do you know how hard it is to get capital into startups these days?
Howard Forman: Very hard.
Harlan Krumholz: Very, very hard. So I was interested to see that actually just today, there was a release out today about a company that was able to raise $70 million. I think the place where you can raise money is in generative A.I. these days. This was a generative A.I. company called Ambience Healthcare. But it gets to what I wanted to talk about today. Because what Ambience does, or at least one of the things they do with the AI, is to produce a product that doctors can use to be sort of an auto-scribe. One of the hard things for docs is you want to be looking in the patient’s eyes. You want to be connecting with the person in front of you, but there’s a big documentation burden. By the way, with the electronic health record, this has just caused an immense amount of burnout and dissatisfaction, because people are ending up now documenting far into the night.
Howard Forman: And it’s bad for patients, right? Because you’re staring at a screen and not at the patient, and the patient feels—
Harlan Krumholz: Yeah, that’s the alternative, is that you’re trying to do it while you’re seeing the patient, and then no one’s actually looking in each other’s eyes.
Howard Forman: Yeah.
Harlan Krumholz: One of the solutions in the beginning was have somebody follow you around, and actually sort of be typing while you’re talking. So you have another person in the room. But with the advent of technologic advances and generative A.I., there’s been this introduction of ambient artificial intelligence scribes to alleviate the burden of this sort of clinical documentation. There was recently an article that came out in the New England Journal of Medicine Catalyst, which is not the highfalutin medical journal but is another subsidiary derivative product that they put out, but it’s still academic articles.
Howard Forman: Yeah.
Harlan Krumholz: This one summarize the experience of these ambient artificial intelligence scribes by The Permanente Medical Group. Basically, these scribes were using machine learning and transcribing and summarizing patient encounters in real time. It was really showing a lot of promise. They were doing a lot of things like assessing the quality of the notes, and they were assessing the experience of the individuals in this, the physician-patient interactions, and whether this also reduced after-hours clerical work, and all of these things. It was showing real benefits. Let me just tell you that when they evaluated the transcripts... I’ll say, there was a whole bunch of ways they were doing this, but this just gives you a sense. It got a score of 48 out of a possible 50 points.
They achieved the highest scores on freedom from bias, whether it was internally consistent, whether it was fairly succinct. A little modestly lower on organization, and there were some accuracy issues. For example, there was a case where the physician mentioned scheduling a prostate examination, and the A.I. scribe summarized that a prostate examination had been performed. Another one where the physician mentioned issues with the patient’s hands and feet and mouth. The A.I. scribe said that the patient had been diagnosed with hand, foot, and mouth disease. So neither of those things were true. You know one of the dangers is, when you’ve got a scribe who’s putting things down, of course the doc’s supposed to be proofing it.
Howard Forman: Yeah.
Harlan Krumholz: But you contend there’s this automation bias, where you actually default.
Howard Forman: Yeah. Yeah.
Harlan Krumholz: You sort of read it quickly, and assume that what it’s done is correct. So there’s still work to be done. It’ll be iterative over time. But one of the issues that I’ve raised to you a lot of times is one of privacy now. Because in a sense, you’re now having the encounter recorded and transcribed but also there is now in the room essentially a recording of it. Another friend of mine, by the way, Shiv Rao, who’s a cardiologist, has got another company, Abridge, that’s doing blockbuster, also creating these kind of products. Of course, Microsoft bought, what’s the, Nuance? Yeah, they bought Nuance.
Howard Forman: Nuance, which has the DAX product. I wanted to mention a couple of quick things on this. Number one, I looked at the Nuance DAX product probably six months ago, because somebody inquired of me to look at it. It’s fascinating to watch. It’s exactly what you describe. It’s important to know there are two models that you can approach. You could either do what you described, where the doctor reviews the note themselves. Or there can actually be a relatively contemporaneous other professional who’s sitting there listening to the tape and reading the report to make sure it’s capturing it accurately.
So there are ways to deal with the accuracy issue beyond that which you described. The product you’re talking about, though, which is Nabla Copilot. What’s amazing about that is this is a very early-stage startup. They immediately go and do this pilot at Kaiser over the summer and then do a larger pilot in the fall, which is the one you’re reporting on. Then they get another, I don’t know, $30 or $40 million.
Harlan Krumholz: $70 million, $70 million they were—
Howard Forman: But is that Nabla or is that the Ambience one? Because there’s several—
Harlan Krumholz: Oh, no. That’s the Ambience Healthcare.
Howard Forman: No.
Harlan Krumholz: Oh, you’re talking about Nabla. So you’re talking about another one.
Howard Forman: Nabla. I mean this whole area is... what’s great about it is you have many firms competing at once. That gives me hope—
Harlan Krumholz: By the way, the Abridge one, which has also raised... the one I talk about, my friend Shiv started, has raised a lot of money, and is embedded within Epic. He’s got a partnership with Epic now.
Howard Forman: There you go. So we’re going to see a lot of competition in this space, and it’s happening so fast. In a matter of like nine months, we’re seeing major competitors both from the enormous size of Microsoft down to these really small startups. I think this can transform clinical outpatient and even inpatient practice.
Harlan Krumholz: Hey, let’s get on to Chris O’Connor. I’m really excited about the interview coming up.
Howard Forman: Chris O’Connor is the chief executive officer and president at Yale New Haven Health. In these roles, he manages the operations of the health system and its five hospitals as well as its physician network, Northeast Medical Group. Prior to his appointment in March 2022, he held multiple leadership positions at YNHH, Yale New Haven Hospital, and Yale New Haven Health. Most notably as president of the health system through the early months of the COVID-19 pandemic, during which time he was crucial in adapting operations and ensuring PPE for the system.
Widely recognized for his leadership and strategic acumen, O’Connor has held leadership positions at the Hospital of Saint Raphael, St. Elizabeth’s Medical Center in Boston, and the Ochsner Clinic Foundation in New Orleans through the devastation of Hurricane Katrina. O’Connor received his bachelor’s degree in economics and his master’s degree in hospital administration from George Washington University. We both welcome you to the podcast. It’s really nice to have you. We’ve been talking about having you for a while. I wanted to start off, because I didn’t even realize that you’re a Connecticut native.
Christopher O’Connor: Born right here at Yale New Haven.
Howard Forman: Yeah. I’d love to hear the origin story about where you grew up, and when did you first get motivated towards economics and then hospital administration?
Christopher O’Connor: Oh, boy. This is going back, and you suggested short answers. My father was a corporation counsel for the city of New Haven. My mom and dad moved here just before my birth in 1969. I was born in 1970. My mom was a nurse, and so she began, actually worked at Yale New Haven Hospital initially and then quickly transitioned to the newborn nursery at the Hospital of Saint Raphael. Was there for 30-plus years, until the acquisition by Yale New Haven Hospital, and then ultimately retiring. So my familiarity with New Haven has been really through my entire life, and quite a change that has occurred within the city in my duration.
It’s been great to watch, because I think it is thriving right now as a result of some incredible investments that began, I think, from the university’s perspective. Then the hospital, with the acquisition of Saint Raphael, has quickly, I think, joined forces to really foster a different economic environment in the city. It’s something that, while it’s one component of the system, it is a major one given the disproportionate impact that Yale New Haven Hospital has on Yale New Haven Health.
Howard Forman: I want to come back to that, but I was curious to start off... Now it makes more sense to me about why healthcare might’ve been an interest to you early on, but when did you know you wanted to go down that specific path? I ask this because a lot of my healthcare management students in the School of Public Health look at you as a role model, for good reason, and they come from a very similar sort of economics hospital administrative background.
Christopher O’Connor: Yeah. You know, it’s interesting. I took an EMT class when I was a sophomore as an undergrad at GW. Not many universities offer you to take a credit delivering class. It was two nights a week, and I loved it. I grew up as a kid watching the show Emergency! and said, “Oh, I would love to take a EMT class.” I had no idea I’d use it. It just so happened that in parallel to that, my father’s law practice, he was in private practice at the time, hit the bubble bursting in the real estate market. His practice had some significant economic shortcomings, which required me to step out of school between my fall semester and spring semester of sophomore year.
My mom, who I mentioned was a nurse at Saint Raphael, she called up a colleague in the emergency department, and said, “Hey! My son just got certified EMT. Could he work in the emergency department while he figures out what he’s going to do?” I did just that. I became an ER tech, and I fell in love with it. It was, from the first day I put on scrubs and was doing CPR in the resuscitation room, watching the team work together in incredible ways, without talking, and just for the sole purpose of keeping that patient alive or helping somebody in need. I fell in absolute love with it. The people there were spectacular.
So it was then, and talking to a number of physicians and the administrative staff at Saint Raphael, that I said, “Maybe there’s something here between my business interests and my love of economics and the principles that drive business and healthcare.” That’s when it really shifted. I did my senior thesis work on the designation of trauma systems across the country. At the time, the trauma alert systems were just coming in. Maryland and New Jersey were two states that were early adopters, that I did that work. It actually turns out in the study that I did that it’s both beneficial for patients and cost, which was still the optimal outcome of the work that we do. But that’s how it really began, and I haven’t looked back since.
Harlan Krumholz: Chris, we’re so fortunate to have you in this position. These are trying times. You’re a feel-good story. A local kid, you come back, you know the area so well. But you’ve got one of the hardest jobs in America, not necessarily the Yale job, but I mean leading healthcare systems right now turns out to be one of the most challenging. You’ve got various constituencies, various different interests. You’re being battered by healthcare policies that you have no control over, at least trying to influence. And the demands keep increasing every day, every day.
I want to ask kind of on a personal level, how do you manage that? Because the incoming’s got to be so intense. You’ve got some view of the long term, so you’re trying to implement a longer-term strategy, but the number of crises that end up on your desk every day, just because you’re running this enormous healthcare system. You’ve got a lot of capable people around you. But still, I mean, in the end of the day, the buck stops with you. What’s your strategy about managing that kind of responsibility?
Christopher O’Connor: Well you said it, Harlan. It’s really the team. You rely on your team, because only a fool would sit in this seat and think they can manage it all. I’m incredibly dependent on the colleagues that I work with and have worked with over my career. I think it is certainly true that when I moved into the CEO role, it is different. While I had it at Saint Raphael, the scope and magnitude here is just, it’s uncomparable. So it’s forced me to rely even more heavily on that team, and that’s creating new skill sets for me. I will tell you, I’m still developing those, and there’s still plenty of opportunity to do it more effectively. But that’s what it comes down to. It’s about building those relationships, because the times will get hard, and you’re going to need to lean on those even more so when you have those difficult issues that you have to overcome.
Howard Forman: You spoke at the beginning really eloquently about the transformation of New Haven. Harlan and I haven’t been here during those years, but we’ve been here since then.
Harlan Krumholz: I’ve been here a long time, Howie, I’ve been here a long time.
Howard Forman: I know you have. I know, but still not—
Harlan Krumholz: I came here in ’76. I first came here in—
Howard Forman: But not ’70, not ’70.
Harlan Krumholz: ’76, ’76.
Howard Forman: No, but so you’ve seen the transformation, and you described it aptly. Harlan and I were also here in the early aughts, as they say, when Yale New Haven Hospital was covered in a negative way, appropriately so, on the front page of The Wall Street Journal for its collection practices. They really led the way, I think, in revising those collection practices. Harlan has a tremendous scholarship in the area of financial toxicity. Yale New Haven has tried to address that at the root. In addition to that, Yale New Haven has won awards now for community benefit, and is reasonably highly rated by the Lown Institute for community benefit. I was just wondering if you could speak to, how do you balance, when a hospital is losing money or struggling, how it continues to make a commitment to the community around it, beyond just the healthcare provision?
Christopher O’Connor: Well unfortunately, as I walked into the office, it’s really the first time that Yale New Haven has experienced those losses. So we’re embarking upon new territory. What I will say is that our commitments to the communities go beyond our profit and loss statements. They have to be enduring. These are communities that we are going to be in for hundreds of years, we believe. Therefore, it can’t be on an episodic, year-by-year basis. We have to be in this for the long term. That’s what’s so important about, when times were good, we had the fortune to build some balance sheet depth to enable us to support programs, even when we were having year-to-year losses, as we’re in right now.
I think that’s so critical. These communities, we are the largest employer in most of the communities that we are in. Therefore, our moves make a difference in the sustainability of New Haven, Bridgeport, New London, and Greenwich. So it is essential that we are committed over the long term with these communities and will continue to be. I think it’s a priority. It’s an institutional priority. I’ve been incredibly supported by the board, as well as Yale University, in those efforts. I think that commitment remains incredibly strong.
Harlan Krumholz: One of the interesting things that I learned about you was, I didn’t realize that you were actually down in New Orleans during Katrina and were in a position where you had to help manage the response. The floods there are still extraordinary. I went down there afterwards and saw some of that damage. I mean, it was catastrophic. I wonder if you could just reflect a little bit on some of the lessons learned. What did you experience in that, and what did you take away from that? I hope you’re unlikely to confront anything quite like that again, but it must have been a turning point.
Christopher O’Connor: Well I’ll tell you, you contrast the experience during COVID, which was a very long-duration event and you didn’t have a window into when it would end, to Katrina, where it was a very short-duration event. We thought we knew when it would end, after the hurricane passed. Unfortunately, for the levee breaches, it went on long beyond that. Those are very different experiences, but I think there were lessons that ran true for both. One is, communication is extraordinary. And it’s always imperfect, and we’ll always learn from what we could do better. But I will tell you, I think in both instances, at Yale New Haven during COVID and down at Ochsner during Katrina, we were out in front, and particularly with our employees, we were communicating regularly. I think that’s essential.
The other thing I mentioned earlier in a different context, and that’s the importance of team. I had no idea who our boiler plant workers were down there, who our plumbers were that were dealing with the steam and some of our HVAC issues. But boy, when you’re in the middle of that type of a response, where—even though Ochsner for the most part, certainly at our main campus, we remained relatively dry—water came up to literally our curbside. We had enormous impact from the storm—roofs blown off, I mentioned our steam was impacted, because we had to vent. The chimney was kind of blown over, and therefore it didn’t allow the steam to operate effectively. We lost water service.
I mean that experience of relying on the entirety of the staff. And obviously our clinical staff, it goes without saying they were operating in incredible conditions. It was impressive how the team with a capital T was so essential to getting through that. It was an amazing experience, as difficult as it was. There’s a psychology of those of us who were there during the storm itself, versus those who came in to relieve. It felt like they missed something. That’s not what you want either, because we wouldn’t have been able to do what we did if we didn’t have the backup team there—more for the clinical staff. It was such an invigorating... I mean, if I didn’t think that I went into the right industry, after living through that experience…. I think most of us come into healthcare, no matter what our roles are, to try and make a difference.
Boy, did we feel like we were making a difference during that time. But it also came with some frustrating times. I mean, I remember when there was a false report that there were mobs overtaking pharmacies, because of the insecurity that was felt across the entire city, before the 82nd Airborne came and hit the ground and really took things back. And watching people cry and break down. We had a physician, actually a physician leader who broke down and had some serious mental illness. These events create opportunities where you can’t mask some of that vulnerability. So it was an immense experience. I could probably talk about it for days, but probably some of the most rewarding and difficult time in my career that I have experienced.
Harlan Krumholz: Chris, I just want to thank you so much for taking the time to be with us today, and to give folks a little bit of a view into who you are and what you’re doing. Anyway, for us, it’s been a terrific, a terrific—
Howard Forman: It’s fantastic.
Christopher O’Connor: Well, you guys are a pleasure, and I always enjoy time with both of you. Anytime, and it was a thrill for me as well.
Howard Forman: Thank you. Thank you very much.
Harlan Krumholz: Wow, that was terrific. What I really love about that interview is Chris was willing to share a little bit about who he is in addition to his perspective—
Howard Forman: Oh my God, yeah.
Harlan Krumholz: ... on the health system.
Howard Forman: I knew too little about him, and it was really fun to hear him tell it in his words.
Harlan Krumholz: Yeah, it’s really terrific. But now we’re at another favorite part of the podcast for me, which is to hear what you’re thinking this week.
Howard Forman: Well, what I’m thinking are things that you were thinking, because you brought this up to me a few weeks ago. There’s been a flurry of new stories over the last few weeks, and I’m starting to see common threads to all of them. As I mentioned, you mentioned one of them to me about a week or two ago, but then two more financially consequential events have occurred that is bringing this into focus. As our listeners recall, Medicare Advantage, which is the private delivery of Medicare through private firms, has been an incredibly lucrative business for many players in healthcare, but most notably the companies that run the plans themselves.
Humana is at the top of the list for that. They’ve been the greatest beneficiary. The stock remains up 1800% since the passage of the law that created the modern Medicare Advantage plans 20 years ago. This compares with a 350% increase for the overall market, so this is vastly more than just your average stock. But Humana reported a large loss two weeks ago, and you were the first to point it out to me. It’s now down 25% as a stock since September. This week, another firm called Cano Health filed for bankruptcy. This was a fast-growing primary care practice and practice manager that has been actively sought for Medicare Advantage delivery, so in other words, primary care for the Medicare Advantage patients.
They have been circling the drain, no exaggeration here, for a while now. They are down more than 99%. In fact, it may be 99.9% from their all-time high set just three years ago this week. Their problems were truly multifactorial, but one contributing factor was their reliance on Medicare Advantage at a time when scrutiny of Medicare risk adjustment was growing. And this is before the Biden administration instituted yet greater restraints on risk adjustment, and other changes that can cost Medicare Advantage plans money.
Then third, beyond Humana and Cano Health is Devoted Health. It’s a so-called unicorn, because it’s worth $13 billion. It’s a startup that only deals in Medicare Advantage plans, and is heavily weighted to two states, Texas and Florida. And yet in the five years of its existence, it has not turned a profit, and it’s now having to raise even more funds to just continue operating. I have no doubt that Medicare Advantage has been a hugely lucrative space for a long time, but the good times may be coming to an end. We’d want pricing to be fair. It’s not like I want them to lose money.
But we also want to make sure risk adjustment has to be accurate, and I’m going to ask you about that in a second, Harlan. We should want quality of care to be high. What I think right now is this may be a pivotal moment in the history of Medicare Advantage plans. The firms that run them are pushing back on the continued cuts in risk adjustment payments, and time will tell. I want to ask you, Harlan, whether you think we are sufficiently good at risk adjusting at the administrative data level. So that this should not be a problem going forward, that we can accurately price these plans.
Harlan Krumholz: Well, first of all, let’s just say when we see losses in these companies, it’s not just whether or not the risk adjustment works, because they’re spending money on a lot of different things. I’m always a little skeptical when I see a company like Humana seeing that loss and then pegging it on Medicare Advantage. It’s to their—
Howard Forman: Yeah, that’s fair.
Harlan Krumholz: It favors them to say, “Hey. This is why we need to be paid more, because this is not viable.” The company, of course, has got lots of different things that they’re spending money on. But I do think that we’re at a moment in time where, if we wanted to, we could produce highly accurate risk adjustment. Meaning, for people who are listening, that we could predict what should these people cost under ideal conditions? You’ve taken on risk.
You’ve taken on a large number of people and said, “We’re going to provide care. Pay us for this in a lump sum, and we will manage the care for these people.” We are at a point in time where we could estimate that with a high degree of accuracy. It was previously known that it was probably the risk model was favoring the Medicare Advantage firms and overpaying them, because there was a sense that we should be moving people into Medicare Advantage, and we wanted to incentivize businesses to grow in this area.
Howard Forman: Right.
Harlan Krumholz: That was never said out loud, but I think in the hallways people knew that this was true. The question is, is that changing now? What’s going on? There are things like introduction of GLP-1s, where these anti-obesity medications could have widespread use. The sickle cell treatment that we talked about, if people start using that, millions of dollars. Of course that’s going to be constrained by venues that can actually provide the service. But there’s a whole range of advances that are coming down the pike that are expensive—chemotherapy for cancer, a range of things.
So the question is, do these models need to be updated? Not because of how sick the people are, which is how these models have normally been done. But because of what are the availability of products, and how is that affecting the range of things that people might use given a degree of medical need? So it’s a complicated area, Howie, because it’s not just about people. It’s about the pricing and the ecosystem.
Howard Forman: Related to what you talked about in the intro segment, A.I. and using A.I. might help to better analyze patient charts, to know exactly how they should be coded. Right now, we very often have professional coders sort of trying to figure out how to code.
Harlan Krumholz: I’ll take you a different way. The question to me is whether or not this A.I., if properly regulated and overseen, can actually cause widespread greater efficiencies within healthcare, because of the ways in which it can enable more people to be seen more efficiently with—
Howard Forman: Using appropriate guidelines.
Harlan Krumholz: And for people at appropriate levels, like take everyone up.... For any level of training, you can actually see more complex patients because you’re being assisted by AI. You’ve been listening to Health & Veritas with Harlan Krumholz and Howie Forman.
Harlan Krumholz: Okay, look. I’ve got to figure this social media thing out. It was so easy for me when I was just doing Twitter, and I kind of—
Howard Forman: Twitter doesn’t exist anymore, Harlan. It does not exist.
Harlan Krumholz: I signed off of Facebook kind of, although I have an account, and I wasn’t really using LinkedIn. But now you’re pulling me into LinkedIn.
Howard Forman: LinkedIn.
Harlan Krumholz: We’re both agreeing that X has got issues, and at least shouldn’t be the only place we are.
Howard Forman: Right.
Harlan Krumholz: So I’m only going to tell people, “Hold on tight. I’ve got to figure this out.”
Howard Forman: Yeah.
Harlan Krumholz: But yeah, I know you’ve got it figured out. So tell us about you and Threads.
Howard Forman: Well, I don’t have it figured out, but I am on Threads, at T-H-E, the number 4, M-A-N, @the4man. But you can also email us at firstname.lastname@example.org. And besides Threads and Twitter and LinkedIn and our podcast, I am fortunate to be the faculty director of the healthcare track and founder of the MBA for Executives program at the Yale School of Management. Feel free to reach out via email for more information on our innovative programs, or you can check out our website at som.yale.edu/emba.
Harlan Krumholz: If you like the podcast, please rate us and review us on your podcast app. We always read your reviews. I guess I should say, if you don’t, yeah, also rate us. We’re eager to learn.
Howard Forman: Yeah, no. And by the way, it helps other people find our podcast when you do that. You’re not doing it just to give us a rating. You’re doing it to get people to know that our podcast is out there.
Harlan Krumholz: Yeah, you’re helping us. You’re helping us. We appreciate all the listeners, and we appreciate any help you give us. Health & Veritas is produced with the Yale School of Management, and the Yale School of Public Health. Thanks to our researchers—Ines Gilles, Sophia Stumpf—and to our producer, Miranda Shafer. Amazing people, we so appreciate you.
Howard Forman: So thankful. Yep.
Harlan Krumholz: Talk to you soon, Howie.
Howard Forman: Thanks, Harlan. Talk to you soon.