This episode is brought to you by r one RCM, a leading provider of technology driven solutions that transform the financial performance of hospitals, health systems, and medical groups. R one delivers proven, scalable operating models that power sustainable improvements to net patient revenue while reducing operating costs. To learn how you can transform your revenue cycle performance, visit us at www.r1rcm.com/beckers. Hello, and welcome to the Becker's Healthcare podcast. My name is Will Riley with r one RCM. I'm joined today by doctor Arshad Rahim. Doctor Rahim is the chief medical officer, population health, and clinically integrated network at Mount Sinai Health System. Welcome to the podcast, doctor Raheem. Thank you, Will. Great to be here. Please, can we start with a bit of a personal introduction? Can you tell us about yourself, professionally, and tell us about Mount Sinai? Sure. Sure. Thank you. So my role, is focused on really the, clinical programs, managing about 450,000 lives in value based care contracts, with roughly about 200,000 lives with downside risk, in a, I would say, a quite challenging and interesting market of Greater New York City. You know, very competitive. The other aspect of, big aspect of my role is overseeing the CIN or our clinically integrated network. And, in that, we have, about 6,000 physicians, about 4,500 are employed, and then there's about 1,500 that are community based, what we call our voluntary physicians. So, those are prime my primary roles. I also do keep a small primary care practice and do practice hospital medicine as well to stay active, stay relevant. Excellent. Thank you very much for the context. We're going to talk a little bit today about some of the big challenges, opportunity areas facing health care providers as we, look at 2025. And so I'm very interested to get your take on those things. The first one I'd like to talk about is artificial intelligence. I'm curious about how you're looking at that technology, the the use cases that you think are are most, optimistic. Let's perhaps start there. Where are you where are you focused? Yeah. I would say, just overall, the health system, and there's a a lot of leadership around this and leadership and governance, you know, has been very aggressively looking at applications across all areas of our health system, including, obviously, the research, side as well, using using AI, artificial intelligence. For me, in particular, in my work, the some of the areas that are that are most relevant. One, I would say and I I and I'll probably just caveat by saying, you know, there's always the question of, is it predictive analytics? Is it machine learning? Is it true AI in that realm? So, you know, I think that's that's an ongoing question. But we use a ton of predictive analytics, in managing these populations. Obviously, with 450,000 lives, you can't manage everybody the same way. So that's a huge part, and the models get smarter. We do have clinical algorithms that also get smarter over time, like predicting the risk of kidney disease. That being, you know, a a bit a large untreated chronic condition that has a lot of morbidity, and eventually mortality attached to it. So those, we've actually developed a internal model for that risk prediction and kind of spun it out into a, into something that's, externally recognized called kidney Intel X. And we used to have seven parameters and now have about three or four parameters as the models got smarter. Right. There's also just the idea throughout all of health care of how do you do more with the same or even more with less. And usually, one of your big costs is gonna be your labor cost and hiring more people, and that's very true in population health and value based care. So how do we scale our existing staff using AI solutions? One of the big areas is around risk adjustment. And so using, various NLP and AI algorithms to be able to scale our staff and, review more opportunities, coding opportunities before they get in front of the physician so that we'd have less false positives and less, you know, physician pain, from declining codes that are not really accurate. So the prereview helps, but we need to scale our staff. So those are a few areas, and we'll probably be doing a lot more in, with we've built disease registries for common conditions. And so currently, we're able to say, we have exactly this amount of COPD patients within Mount Sinai health system that are engaged with us. So we're able to identify that denominator. And I think AI will help us a lot in identifying unique insights and in kind of working the model so that we can better target subpopulations. Yeah. Okay. Excellent. Are there any are there any concerns about the technology that you have, particularly? Like, you mentioned, for example, coding and giving advice to the physician like that. Yeah. Do you worry about hallucinations bias, particularly if you're using that technology in that generative way? Sure. Sure. Yeah. No. I think it's it's still very, I would say, exploratory. I think that if you start with a problem where you have no scaling opportunity, if you start with a low bar Uh-huh. You can get better. Bias potentially, is in the things that we're doing, I think that's something we should kind of be concerned with and look at. I would say that right now, there's there's pretty much a greenfield in some of these areas that I'm looking at. Who do you target for care management? Those and but I think being very sensitive and at least at one basic level is measuring your outcomes through race and ethnicity consistently regardless of what technology or lack of technology you're using can kind of keep you honest. Yeah. And those subgroups. So so that we've built in place. So I feel like that's at least one check Yes. On those potential lives. And make models get better and better, right, with better and better data Yeah. That you trust. Yeah. You yeah. Yeah. Yeah. Yeah. Absolutely. I'm interested in the, payer relationship. It's been fractious for many providers, recently, right, with increasing rates of denials, prior authorization requests, and so on. I'm curious about just your take on that in general, but I'm I'm also interested in your in your vantage point from the perspective of, population health value based care leader. And and, like, it feels like, providers are struggling to find the right solution there. Right? Is it is it technology and battling bots? Is it a tougher negotiation stance with payers? Or is there some kind of answer from looking at some of the metrics and the predictive analytics that you can do in value based care to maybe have a different conversation about populations and what they really need. What do you what do you think about those issues? That's a great question. And, and then evolving evolving answers. Okay. Yeah. I'm sure. Yeah. I would say that, you know, still, it's I was reading something this morning about how we likely will have foot in both canoes for a while Uh-huh. Be the fee for service and the value based care. And I I think that that's, you know, that's true, and and that's part of it. So then you have not you have two negotiations that are actively going on, but they do feed each other, which is the fee for service rates and the value based care arrangements. I kind of live in those. I have more more of a role in the latter. Yeah. But I'm, you know, I'm I'm a part and part of the team that does the overall. I would say that, it's, you're kind of finding who really wants to partner and sometimes maybe who can partner and, kind of who is it. So I do think there's there will be the haves and have nots in this in this. I think that, would you go with just a few payers where you have a better relationship and go deeper versus a lot of payers where it may have been historically? I could see that happening definitely. Yeah. And, because it doesn't really appear that there's any lack of patients to care for, but, you know, the arrangements have to be fair so we can at least cover our cost. And that's true on both sides. Yeah. You know? Value based care takes a good amount of resourcing, So we need to have, you know, either the some of the upfront coordination payments or the, or essentially what are the shared savings or the various incentive kickers. They have to kind of at least exceed your costs for providing, all those cares, all that care. It's a it's kind of a payer by payer process, right now, and it depends. And we've had, you know, some successes. I do think the other thing in this context is for value based care, just having an arrangement, I think I kind of say we're past the honeymoon of value based care. Mhmm. We're about, you know, twelve years post Affordable Care Act, and, people have had enough experience. And there's some positivity. There's also some jadedness over the expense that it takes. So I think that the payers that recognize that and that truly partner with you And then also listen and really listen on the metrics that truly matter and kind of limiting to that. For example, we know there's about 17 total quality and efficiency metrics that we really focus on. Right. As much of our contracts that can be focused on that and not other things that are essentially almost distractions Yeah. That or that we would have to organize newly around, you know, that's that's very important for us moving forward. Yeah. So perhaps you're saying then with more data, more experience, you're learning what's truly predictive of the a future state, and you're able to optimize and organize around that then Exactly. Including your real commercial relationships. Yeah. Exactly. Yeah. Exactly. Yeah. Yeah. And you can kind of see I've done a lot in the, recently in the specialty space. As, you know, CMS really wants to bring specialists under the tent of value based care, it's at least half of the $4,400,000,000,000 or more. And, and we kind of know what works and what doesn't. And so it it's important that the payers support us on that. And so that we can also just expand this well beyond, primary care and frontline care and into that specialty care as well that's, you know, a core part of how we deliver care, especially at teaching hospitals and academic medical centers. Yeah. Excellent. Okay. Thank you. That's great. Great. Let's let's move to the conclusion and and talk a little bit more about 2025. What's on your mind for 2025? Priorities, outcomes. Can you flesh that out a little? Yeah. Sure. You know, the one I think one core area is, even though I mentioned the importance of specialty care, and that is very important in frontline, primary care access and engagement. And I would also say frontline specialty access engagement kind of and when I look at that, I think of cardiology, some of the medical subspecialties, gastroenterology there. I think that will continue to be key to really generate, stickiness for the health system from, fee for service and, top and, you know, top line revenues. Yeah. But also, in order to really manage populations, that's gonna be key as well. That combination of specialty. And Yeah. Yeah. Okay. And that rate of, essentially, how often are people coming in for, for preventative services? How often they're coming in for chronic disease management early? Yep. And that, I think, is is kind of a win win. Patients want that. Sure. And I think it the care feels good for providers to be delivering it, because we're we're change you know, changing the trajectory as opposed to just being reactive to a disease state. Yes. And so and then building the infrastructure around that proactive model and executing is not a core competency of No. I can see that that's very difficult, right, in many ways. Right? Because we're used to access being through a certain generalist. Yeah. You know? Yeah. Yeah. We build it. You come. Yeah. You know? And this is a different model. How how does it how do you do that then? How do what are some can you talk a little bit more about that? Yeah. So I think foundationally data matters a little bit. And it's interesting. I'm actually thinking about this problem for a a country out, let's say, far to the east of us who's asked us to work on this problem with them. Okay. And so I've been thinking about it quarterly. But foundationally, you do need some baseline data on your population. Right. To be honest, even things such as, you know, a few disease states, agent sex, and how do you reach them? What's their address? What's their phone number? And just accuracy around that. And then operationalizing very a a bunch of different patient engagement strategies and approaches. Yeah. And, you know, it it really kind of working those throughout the year, ultimately targeting metrics. We look at our Medicare population, and we'd love to have it a hundred percent, have at least one comprehensive visit per year. But we, if we could at least get eighty five percent plus. And now we're doing the same thing in our commercial population, especially anyone who's 40 and older and really kind of segmenting that group. So and then you you have to kind of build databases. You have your EMR has to work for you. A lot of patient engagement. You have to have, I think incentivizing. You know, most systems will do it and incentivize to do. If you're just a pure RVU model, it really doesn't work in this for this. What you do have to do is incentivize on outcomes. And we've had good, some good success with that in primary care and specialty care. And some of the outcomes, you know, it's it's the team. It's not one or the other. It's it's the overall system driving forward and Yeah. Takes into account home based care, but all those modalities. But, yeah, having those kind of be, you know, big dot metrics and regularly tracking towards them, and that will help inform your operations too or see whether your operations are working or not. Yeah. Excellent. That brings us to a a lovely conclusion, I think, doctor Eam. Thank you so much for sharing your perspectives on, on all of these issues. Well, it's been lovely talking to you. Thank you. Likewise. My pleasure. Thanks.