Yes. So I think that's a perfect segue. Right. Yeah, it is complicated working in a system, but I think what I want to try to share with you working in a big system is there are ways in which I want to impart some, like, empowerment that maybe you could then be negotiating with your system-level leadership on how we bring value in different ways. And so, you know, what I've tried to do is see, well, what can we do in the system to try to say, hey, we can really help ACO performance, and so ACO is a perfect example. Here's my disclosures. This is Dr. Bonsall, by the way. I didn't get to introduce you. Sorry. Yeah. Sorry. I'm Nina Bonsall from Mount Sinai. I just jumped right in. Anyway, so as we already know, we know the U.S. healthcare system is broken. We spend more money, and yet, compared to the other developed countries, we're kind of lowest in terms of quality. We know that as we get older and fatter, we are having an increased demand for utilization, and of course, as Ziad kind of talked about, there's a lot of medical technology that's out there as well, and currently, we don't really have good guardrails around appropriate utilization for a lot of the technology, and of course, all of these things are contributing to increased medical spend. And of course, the funds available to care for this care is declining. The cost is shifting to commercial payers, but that's not really sustainable. We're seeing employers drop coverage. Medicare is running out of money. Ziad alluded to the compressed state budgets on Medicaid, fewer taxpayers in line for aging population, so we all get it. It's a big problem, and we know that there are a number of different alternate payment models. I think with the exception of some pockets across the country that are in that Category 4, maybe California is a little bit ahead of the game in terms of population management. Most of us at our institutions are in kind of a Category 2 and Category 3 type of stage, and I think I just want us to think about we always need to come back to this general construct for success in value-based care. It's actually quite simple, but it's hard to think about how what we do every day could be impacting this very simple equation. And so the way these value-based contracts work, when there's upside or downside potential, it's something called a medical loss ratio or an MLR target, and each payer has their own MLR target, but essentially, you have an allowable spend. You have $15,000 to care for this patient during 2019. If you're able to care for them for less, we have some potential shared savings that's going to be distributed. If you spend more on them, so your actual spend versus your allowable spend is what determines that MLR target. And so I mention that because we always want to come back to that concept. And so let's start with how many of you, I just want to get a sense, are part of an ACO? How many don't know what an ACO is? Okay, so this is really important. So ACO is one of the biggest kind of alternate payment models that we have across the United States right now. At our institution, it's one of our value-based contracts, but we participate in 14 different value-based contracts with varying risk. We have about 400,000 lives in risk at the present time. But the ACO is actually an important thing to understand. So in 2017, when MACRA was passed, CMS basically said you have two choices. You either could be in an alternate payment model where you take downside risk. Okay, what does that mean? That means that if your actual spend is more than your allowable spend, you're writing CMS a check back. So if you're going to take significant downside risk, those are the advanced alternate payment models, then you automatically will get a 5% bump in your Medicare fee-for-service payments for everybody who's in your ACO. And ACOs work at a TIN level, right? So every physician would benefit. So the neurosurgeon, if they're part of an ACO in an advanced care model, if the ACO does well, everybody gets a 5% bump in their Medicare fee-for-service payments. Most, however, are not taking that downside risk. And if you're not taking downside risk, you're in an alternate payment model, but you're only participating in an upside shared savings model. So if you save some money, then that money can be distributed, and how much you get actually is determined by the quality. And so again, the important point that I want to make here is that while the success of the ACO has direct impact on us, right? So if you are in an upside shared savings model, then the MIPS score dictates your Medicare fee-for-service payments, okay? So what makes up MIPS? So if you're on the right side of the graph, that means that your institution is part of an ACO. That's like us at Mount Sinai. We are not in one of the downside risk ACOs. So that means that we have upside only shared savings. And so we have to... Our MIPS score determines all of our physicians' Medicare fee-for-service payments. So how we do in the ACO is actually really important. And historically, the burden has been primarily on the PCPs. So what I want to impart on you is how can we help the ACO, because in turn, actually, our fee-for-service payments are going up as well. So that MIPS score comes from three components. 50% is based on quality, promoting on interoperability, which was kind of like the meaningful use step that you might recall, and the improvement activities. And so the most important point is it's rise or fall together. How the ACO does affects everybody. So you should care. And those ACO measures, there is patient caregiver experience measures, there's claim-based utilization measures, and then there's the kind of more quality measures. And as Ziad pointed out, there are a number of measures that, all in pink here, are measures that roll up into that MIPS quality score. The other measures roll into your ACO score, but when we talk about just the MIPS score, you can see those measures that contributed to that. And yes, one of them is colorectal cancer screening. So yes, the gastroenterologist plays an important role there. But I'm going to make the argument that we can play a role in much more than just colorectal cancer screening. So how can we help? So if we said that that revenue side, right, we said that we always want to come back to that equation of how much do we have to care for this patient, and then how much do we actually spend to care for the patient? The two biggest factors that control the revenue side is the STARS rating for a plan, so that's not ACO. That's just your MA plans. If you are partnering with, say, United, and they're a four-star plan in your region, you get paid more to care for those patients than if they were a 3.5-star plan. If a plan achieves four stars, CMS already pays them more to care for them because they're saying you deliver high-quality care, and that trickles down to the amount of revenue we have to play with as we manage those patients. But the most important thing, the biggest driver, is clinical documentation. You always hear, my patients are sicker. So if you're in that conversation where, my gosh, this patient's costing us so much, well, my patients are just sicker. Well, no one knows your patients are sicker unless you actually code appropriately, and I'll show some examples of that in a second. But this is the most important way in which we can increase our allowable spend, and hepatologists do see some sick patients. So coding for those patients correctly can have a huge impact. First in terms of quality, yes, colorectal cancer screening, we can help with that. But we see a lot of patients that may not be seeing their primary care providers as much as you think. You know, they actually see you, and they think, why do I need to see my primary care doctor? I already see you four times a year. We have a lot of patients with NASH. They're on diabetic medications. They're on hypertensive medications. They're on cholesterol medications. Medication adherence in terms of STAR, each of these measures is triple-weighted. There is almost no way for a plan to be four-star unless they're successful in medication adherence. So what can we do? We can make sure our patients are taking their medications when we see them. We can prescribe 90-day instead of 30-day prescriptions. These are all things that have been shown to be very, very effective, and we have the opportunity to contribute. So it's important to highlight that when you're in conversations at the leadership level, that we just don't do colorectal cancer screening. This is how we also can contribute to our collective success. Of course, breast cancer screening, blood pressure control, hemoglobin A1C control, BMI and counseling, I mean, you know, we're seeing all of these NASH patients to be having conversations. These are quality measures that roll up into the STAR ratings, and we need to make sure that everybody knows we can contribute to this. Blood reconciliation post-discharge, statin therapy in cardiovascular disease, and readmissions, and I'll get to that again in a second. But in terms of risk adjustment, this is the single most important way that we can contribute because what drives that allowable spend is the risk adjustment methods in healthcare. And while there are a number of different risk adjustment methodologies, HCC is kind of the most non-black box method. It's very clear, there's a risk adjustment factor for each disease condition, you know, it's a tangible thing that can be calculated, and I'm going to show you an example in a second. Other measures are a little bit more black box. For Medicaid, the CRG risk group methodology is, it's a little bit, it's how things interact with each other, so it's not additive, it's like, do you have depression plus diabetes, that might mean something, but if you have depression and something else, it doesn't mean the same thing. So it's how conditions relate to each other that controls that. So how does the coding part work? So RAF is a risk adjustment factor. Depending on where you live, you get a demographic RAF. So this is our demographic RAF in the New York area. So we're starting with that, and now I see a patient and I say, well, they have viral, they have unspecified viral hep C, yeah, they had some ascites, they've got some CKD, I don't know, stage one, stage two, stage three, whatever, they're obese, and if everything is unspecified, that's your HCC score. So your patient, if however, you got a little bit more specific, you could increase that RAF, the HCC score, the RAF score tremendously. So now you're saying, okay, chronic hep C, they had an episode of SBP, they've got stage four CKD, they have morbid obesity, they have esophageal varices without bleeding, thrombocytopenia, opioid depression and remission, and you've gone from a score of .390 to a score of 2.875. That translates into a difference in annual payment to care for that patient of 24,000, almost $25,000. So by doing this, you're bringing value, and believe me, system leadership will listen if you speak in this language. And now, how can we reduce avoidable utilization? You know, I think we just need to start to think about this. Many of us have seen a patient and they had a CT someplace a month ago, and you're like, I don't know where that is, there's no way we're gonna get it, let's just repeat it. So you know, that has to change. We have to stop doing that, because that's kind of, it's a pain, but it's stupid, right? I mean, we're talking about utilization and we're driving it by being lazy or not having somebody who's been in the office, now I'm not saying the physician should be doing this, but ancillary staff or put the burden on patients to also bring some of that in. Sites of service is really important for endoscopy. When you do an endoscopy in the hospital, the facility charge is so high that payers are really trying to push back on this. So if you have, and I'm not saying for a variceal bleeder you're gonna do them in an ambulatory surgical center, but for your screening colonoscopies and for your diagnostic endoscopies where the ASA risk is very low, if you have the opportunity to use ambulatory surgical centers, payers will be watching. They are selecting physicians who use ambulatory surgical centers instead of hospitals for their endoscopies, specifically for this, because they're paying so much to the facility fee for the same procedure. Medical ED and admissions, you know, obviously we have a lot of things that we can prevent patients from going to the hospital. So ascites management, screening and prophylaxis, I already mentioned medication adherence. Something that's happening in the readmission world that you guys should be aware of, the payers are putting in no payment policies. So if a patient's discharged with something and they come back within 30 days with a similar diagnosis, same MDC category, they're not paying for it. So the hospital is still on the hook for it, but we're still delivering the care, but we're not getting paid for it. So you can bring value to your institution and therefore to yourself by being aware of this and managing your patients. I think we all know, you know, we don't want our patients to be readmitted. And so getting them in to see you within a week of admission, making sure you're doing med reconciliation, which is really important to prevent that. Prophylaxis spend, unnecessary medications. If you see that a patient's on a PPI because they're all on it, it's time to think about do they really need it or is it just something that got started and never stopped? Therapeutic class substitutions where you can use brand instead of generic, cheaper drugs with equal efficacy. We just have to start thinking this way. And they're watching. I got to, this is actually a dashboard that we have in our hospital for tracking readmissions. Well our readmission rate for the liver service is 25%. So this is 30 day readmissions. So they're watching this. So if you know that you can say, hey, we're going to work on reducing this. This is how we're going to bring value, that will in turn come back to you because leadership's going to be listening to people who kind of are thinking in this way because as I said, they're not even getting paid for that readmission. So if you can work with them to develop strategies to prevent that, you will be value add and be valuable. I saw this paper that was recently published and this is a validation of a risk score in predicting early readmissions in decompensated cirrhotics. So this was just published in Hepatology by Mumtaz et al. So 30 day readmission rates for decompensated cirrhotics ranges from anywhere from 20 to over 50%. So what they did was they did a retrospective cohort study looking in 2013 and 2014 national readmission database. Patients had to have cirrhosis plus one feature of decompensation. 83% had this Elixhauser core morbidity index which uses ICD codes in administrative hospital databases to characterize core morbidities. It's used to predict hospital resource use and in-hospital mortality. Their primary outcome was to develop a predictive risk model. Who are our patients most likely to get readmitted early? And then of course their secondary outcome was a 30 day all-cause readmission. And so the predictors of 30 day readmission, age, Medicare, Medicaid, this comorbidity score, ascites, encephalopathy, HCC. If they had an LVP during the hospitalization, they had a higher risk of readmission or if they had dialysis. Interestingly, being admitted for a variceal bleed had a decreased risk of a 30 day readmission risk. So it remains to be validated but this kind of predictive modeling is going to be important so that you can risk stratify your population where you have limited resources targeting on the patients that are most likely to get a readmission. This is actually data from commercial payers. And so for Medicare, and for Medicare, inpatient costs are probably, they account for about 40% of healthcare spend. In commercial populations, it's specialty care. And that Ziad kind of alluded to that, that we're in the crosshairs here and people are looking at that. And so I throw out there for you to consider, when you're in an RVU model, there's a lot of times that you're seeing patients for things that you don't necessarily need to physically see them for. So think about, obviously, if you're in an RVU model, I understand the challenge, but I'm just throwing out there that we need to start thinking about this. We're now leveraging e-consults. So we have e-consults, including for hepatology, across 20 subspecialties. And now we have the payers at least paying for it. With the exception of Medicaid, it's $40 per e-consult. And there's an RVU value to it as well. And so if you think about, so it's 0.7 RVU. If you think about a level three visit is probably like 1.2, 1.3 RVU. So two of these would equal a level three visit. The idea is that you can do more of these and then have higher acuity patients where you're doing that coding and that management to prevent the unnecessary utilization. E-consults work particularly well when it's really just lab-based stuff, right? There's really no physical exam that's gonna change what you do. Endocrinology was the first one that started this with TSH and kind of thyroid abnormalities. They could just make recommendations via the EMR. But you know, for hep C treatment, for risk stratification of NASH, do they need a fiber scan or not? These are things that could be very amenable for e-consults. And then one thing that I've seen in our ACO is that a lot of times patients see the specialist so much that they really don't ever see their PCP. And I don't even know if the specialists realize that sometimes. And so it's important for us to just think about this. So when we looked at our ACO data and we grouped our members into three different groups, and I just draw your attention to group two. These are members who are attributed to a specialist because of plurality of care. And they have not seen a PCP in the past 18 months. This is our ACO population. So 10,000 out of, we have 40,000 lives in our ACO. So about a quarter of them have not seen their PCP in 18 months. So here we're saying that the ACO success depends on seeing the PCP and all of those quality measures and they're not even seeing them. So how can we help? So what we've developed, and this is about to go live in Epic, is we've teed this up. If we cold call patients, if we just say, hey Mrs. Smith, you haven't seen your PCP in 18 months, they're gonna say, why do I need to see a PCP because I already see my cardiologist six times a year. And so unless they hear it from the specialist, that's us, unless they hear it from us, they may not even see the value in seeing their primary care provider. So it's our job to have that conversation. And in some cases, you might wanna be the PCP for that patient. And that's fine, many people practice that way and that's totally fine. But what we've developed here is the ability to either make yourself the PCP, in which case you're responsible for all of those quality measures or link them to primary care. And so the way this works is we've taken a combination of Epic data and counter data plus claims data. The value of that is by claims, you can see if they saw PCP outside your system. But claims have an inherent lag of about three months. So by having the Epic sweep as well, you can see if they had an encounter in the past month or so, in the past couple of months with a PCP. So it allows you to have that conversation and then decide whether or not you're gonna become the PCP, in which case, there's a lot of things that come with that. And these are things we provide to primary care providers, or you allow them to be linked back to the PCP. So we provide great value in helping our patients even link back to the PCP, because they actually sometimes don't understand the true value. So our patients have significant health risk, and therefore, they have high potential total cost of care. That's the only, actually, it's this patient population that value-based care works, because if you have a very high potential total cost of care, but you have effective strategies to try to minimize that, the delta can get pretty big pretty quick. We have effective risk adjustment strategies, and our patients definitely are sicker than other people's patients. And we can address that risk through population health management programs. We are positioned to impact overall quality of care outside of colorectal cancer screening, and I think we play a big role in trying to link our patients back to the PCP. So I think this is just to give you ideas of how, depending on where you are in your institution and your level of interest in that, engaging leadership around these conversations and making yourself kind of visible and a leader in this space will allow hepatology to get more and more in front of system leadership and see how we can bring value. Thanks, and I'll take any questions. Thank you. Again, back to the horse out of the barn and our value. One question is words like, you need to think more about doing these things, not being lazy and doing a CT scan in clinics, and you should have a conversation. I have conversations with my patients about going back to their PCP because I do not want to manage necessarily or give the information about weight loss, what I should eat, you should see a dietician because it's a lot of time that I don't have, as well as doing medical reconciliation. Out of my control as a physician is very simply all the infrastructure needed to order those tests, look up that CT scan, talk to the patient, and many times say, you need to go back to your primary care. They don't really care about me because I spend 10 minutes with them, they never listen to my heart and my lungs, and they're always on the computer typing in things. They don't care. So I think culturally, everything has already changed in medicine to sort of add on more things to document so we can get up for billing and reimbursement, up the reimbursement. We're not individually responsible, plus we don't individually get the benefit. So the person that does all this is slammed, earns less pay probably because they're still not generating the RVUs because they actually sit and talk with the patient, and then they're doing their charts till midnight. And so they're retiring. Yeah. Yeah. Well, I think that there's a lot of stuff that the providers don't need to do. There's a lot of things that can be teed up by ancillary staff, if you have them. If you don't have them. Yeah. Well, I think that ... Where are you at? Yeah, I think that when you go to them with the data and you say, look, we're going to work on this and we're going to cut this and work out a model, but each system is different. So each system is at a different level of maturity to deal with this. So some have invested at Mount Sinai, we've invested a lot of money on a pop health platform. There's now incentive models for primary care providers for quality, for decreasing preventable ED. They get bonuses based on that. So again, each system's at different levels of maturity. It's not a one size fits all. Some will never get to this and that's okay too. But if you're in a system that values this, then you want to understand it so that you can be able to have the conversation to negotiate how you may also benefit from this. But if you don't know and you just say there's nothing, yeah. So each system's at different levels of maturity. Thank you. That was an excellent presentation. I'm Andrew Muir from Duke. Ziad guides me on all this stuff. But I thought your slide around the documentation was the best example I've seen about how you could really impact it with your documentation. But to piggyback a little bit on her earlier question, if you are that person who then does the right thing, does the good documentation, or if you decide to become the PCP, like either in your system or are you aware of any examples where that then does translate to a model that incentivize the person for that behavior? Like can you confidently reassure that junior faculty member that that will be- Come back to them somehow. Yes. And not just make the system fatter. That's the challenge. I totally understand that. I mean, I think that if your department values kind of your billings, right? So each system, each department reimburses their physicians in various different ways. The ACO hook or the MIPS hook is interesting because you can engage specialists whose Medicare fee for service payments are actually benefited by success. So it's kind of like, it's your Medicare fee for service payments, right? So in the MIPS model, if you're successful in the ACO, everybody's Medicare fee for service payments go up, right? So if your reimbursement is percent of your billings, then obviously you're already going to see the benefit of that by contributing here. It all depends on the structure of the departments and how they are paid and their compensation levels. I think a lot of people are trying more novel ways. We're going to be piloting a practice that it's no RVU model at all, right? So it's just, you have your, this is PCP of course, not so much specialists, but what we're going to try is a model in which it will be for Medicare patients because that's where you can have the kind of greatest potential of moving that MLR target. And so you'll get paid on managing a panel. How you manage that panel, whether it be by phone or by face, doesn't matter. You're guaranteed a salary. So if you are able to decrease the cost of care, there's obviously, there's benefit there. So there, you know, there's going to be these different types of models. I just use the fee for service one because people can feel it most tangibly. And I feel like that's the hook to at least start to get specialists engaged in this work and see how they might be able to benefit. So Lynn Taylor from Rhode Island. Thank you for a very illuminating talk. I have a question about one of your more minor points though. You mentioned or you highlighted this, that payers, for example, will not pay for readmissions within 30 days. If it's related to the same, it's related. So we all live under that. So the payers have the reins, the payers who didn't do your transplant fellowship, hepatology, all this have the reins. Should we, when we think about health care costs, should we as physicians do something more actually to put the guardrails on the payers so that they're the ones that are controlling health finances? So for example, in Rhode Island, we have among the lowest Medicaid reimbursements in the nation. So subspecialists increasingly will not see Medicaid recipients. And meanwhile, the largest provider of Medicaid in our state is a for-profit. It's the second fastest growing for-profit in the state. Revenues over $3 billion a year, the CEO makes millions and millions into millions, while more and more doctors will not take patients that get their Medicaid through neighborhood health plans. So why are we so worried? What else should we do to be, so to get us out from under the whole control of payers? Isn't that part of the bigger picture here? Those repayment, I mean, I think there's a lot of discussion happening about that right now. When you partner with payers, I mean, we have to push back, right? Because nobody, we can't just say, oh, okay. There's actually like a lawsuit going on about that nonpayment policy for the payers. So I'm not at all saying that that's the right thing. And we do need to put, we have to make payers more accountable. I'm simply stating that that's what it is right now. And so that's what we're up against. But I agree with you completely. The payers need to, I mean, their bottom line also has to change. And I think that when you engage with them, we're certainly pushing back when we negotiate contracts. And that's why, you know, you might not want to go to downside risk, but upside only shared savings is a good model because you're still getting paid for the most part for your care. But if you're delivering higher value and lowering the total cost of care, you're actually getting paid more than you would just with your baseline rates. The Medicaid issue is obviously a state to state issue and hard to, you know, that's a lot of policy and advocacy that we need to do. And so it's a little bit harder to, you know, think about a solution for that. That's, you know, obviously a big problem. Thank you. Hi there. Thanks for the talks. Heyman Shaw from Toronto, Canada. Just so everyone here knows, we share a lot of the same problems that you guys have in Canada. So you're not alone. Interestingly, there is a movement in Canada happening to sort of have more of these accountable care organizations and bundle payments. And we're at the beginning of it, whereas you guys are now several years into it. And, you know, one of the things that I haven't heard discussed here yet is the idea that the highest performing, most cost efficient health care sort of systems in the world, the main reason for that, the main driver is probably that those countries spend the most on social services. And that as a result, their health care expenditures are lower. So I'm wondering what your advice would be to a Canadian who's sort of starting out in this, to speak to the system administrators and the policymakers around how to make sure that as we move towards bundle payments and ACO type models, we actually can increase social services spending so that we're not facing the same issue here, you know, 10 years from now that you guys are facing today, which is that there's this, you know, we're talking about value, but we're actually not talking about the things that are truly driving our health care costs. Yeah. I mean, the whole idea of social determinants of health is obviously a very, very big topic. And certainly without addressing that in certain populations, right? So it's less relevant for your commercial populations, but certainly for your Medicare and Medicaid and depending on where you live, right? That's going to be bigger. I mean, our care management team have a number of predictive risk scores and models that take into account social determinants of health. So I think when we get into all of these predictive risk scores for unplanned admissions, for readmissions, there are a lot of models that include the social determinants of health that help stratify those at most risk. And we're partnering with community-based organizations because they're already out in the field working, but we need to kind of be an umbrella around that so that we know who our partners are. So if the biggest issue for the patient is food insecurity, okay, let's get them involved with this community-based organization. If they're having an issue around transportation or housing, we're partnering. So we can't, the health system, the problem is the health system sees the downstream impact of what's wrong with our entire society, right? Because we're the ones that are getting the burden because it's translating down into poor health outcomes. So we can't do everything. And so I think when you think about developing your models, partnering with appropriate community-based organizations that can help address some of the social determinants of health, and including those factors in your models when you risk stratify your population.

healthcare system challenges

empowering individuals

U.S. healthcare system

alternate payment models

value-based care

risk adjustment coding

patient care coordination