So our next speaker is Dr. Michael Volk, who's going to talk about a cirrhosis medical home and an alternate payment model. Thank you, Amina. I just want to make one comment about this earlier discussion about how do you incentivize physicians to do the right thing. I know that some institutions do tie bonus to citizenship, which includes how often you address the BPAs that are firing and things like that. So that is one mechanism, which some people will do. But now I'm going to take this excellent background, and now we're going to dive down into one specific example of value-based care in hepatology, and that's cirrhosis, near my disclosures. So cirrhosis is really an ideal target for value-based care. It's a well-defined disease. It affects about 0.3% of the US population. It kills about 60,000 per year in the US. And many of us make our living on transplant, but of course, cirrhosis kills 60,000. There are only 6,000 transplants per year. So it's a much bigger problem than just transplant. Interestingly, the mortality statistics are equivalent to diabetes or kidney disease. So when you talk to health care administrators, health insurance companies, this often requires some education, because they actually don't realize how morbid and expensive liver disease is. And in fact, the direct cost estimate are about $20 billion per year, which is similar to heart failure. And most people don't really realize this. So as was discussed, patients with cirrhosis have high rate of readmissions. About a third are readmitted within a month, and some of these readmissions are possibly preventable. And you all have seen this frustration. You get a patient readmitted that you see this could have been prevented if they had an outpatient location for their paracentesis, or if they were better educated on how to titrate their lactulose, et cetera. Cirrhosis has clear gaps in care processes. So quality of care is clearly limited. Anywhere from 20% to a half of patients get recommended treatments. And these studies were all done in large academic institutions. So you can imagine that smaller hospitals, it may be even worse. Patient knowledge is lacking. That's one area where that time that we spend to the patients that we don't have in clinic, educating them on their diet and adjustment of lactulose, et cetera, is really critical and can prevent rehospitalization. This is one of the studies we did of patients in a hepatology clinic, showing that half of them thought that sea salt was OK, and that a lot of them thought that they shouldn't be taking Tylenol, but it's OK to take ibuprofen. So clearly, knowledge is lacking. Not only is the cost to the health care system very large, but like most chronic illnesses, most of the cost is focused on a small percentage of the population. And this just shows that, really, the majority of the cost is in the top 20% of the patients. And so that's important, because one intervention for controlling costs is more case management, or more attention, more resources focused on that small group. And so you might have a given intervention that's not cost-effective when it's focused, when it's given to the entire population. But if it's only given to the people that are at increased risk for high expense or high utilization, then it is very cost-effective. This is just one example of a business case. So our largest local payer in Southern California, the Inland Empire, which is San Bernardino and Riverside counties, is the Inland Empire Health Plan, a large managed Medicaid plan in Southern California. And I estimated that there would be about 2,000 members with cirrhosis in their population. And so then when I met with them, that was spot on. They ran their data and looked. And in fact, they did have about 2,000 members. And I also estimated, I told them, look, I think your annual cost is going to be about 20,000 per member. And so they ran their data, and it came back that it was spot on. And so that means they spend $40 million per year in spending on their cirrhosis population. And so that means that if we could reduce their spending, we could save them $4 million per year. And so then that's the business case that I proposed to them, which is if I can save you $4 million, why don't you give me $1 million? And I'll invest that in infrastructure to reduce your costs. So what interventions could we employ? Here's where we still need more data. But just common sense and literature and anecdotal clinical experience tells us that we know that re-hospitalizations are a big cost driver. And we know that across all other illnesses, there's a couple simple things that have been proven over and over that can work to decrease early readmissions. So we need to improve the discharge process. We need to give better patient caregiver education upon discharge. And then we need to get patients back into clinic quickly. It's been shown over and over again that that clinic visit within a week or so is really key. Also, we need to be able to do the paracentesis in outpatient setting. Because otherwise, they end up coming to the ER. The ER is too busy to do the large volume paracentesis. They get admitted. They sit two days. It's silly. So duplicative testing, as was mentioned, that CAT scan that really does not need to be repeated. So we need to figure out other ways. And health exchanges are one way. Educating patients to bring their records is another way. Unnecessary testing, we know that there's a lot of, you know, the hep C patient who is not an antiviral therapy candidate, and this is less of an issue than it used to be. But we all saw the hep C PCR checked on a monthly basis. Why? It doesn't really change anything. We see our patients with hepatic encephalopathy that are getting weakly ammonia levels. And then the ammonia level goes up, and they get told to go to the hospital. And the patient's like, well, I'm fine. So we know that that's a lot of areas where there's unnecessary spending. And so we can identify and reduce variation by sharing best practices. We can disseminate the choosing wisely topics. How many of you are familiar with the choosing wisely topics for liver disease? So there's areas where the ASLD, the Practice Guidelines Committee, I was on at the time, and we created these topics and disseminated them. There are certain areas we know are unnecessary. And then, like I mentioned, we have these high utilization patients that are driving the majority of the spending. And so we can do intensive case management. As you'll hear in the next talk, we can employ palliative care, which has been proven, actually, to not only reduce costs, but actually prolong survival in patients with lung cancer. So those are the interventions. What's the structure in which we could deliver these interventions? There's a lot of different possible structures, but we think that the cirrhosis medical home is the best one. This does not mean that we have to be the primary care doctors. It just means that this is for patients for whom cirrhosis is the main driver of all their illnesses and their utilization. And so this is a multidisciplinary team with a patient at the center. We need care navigation, integration of mental health, addiction services, rapid access to urgent care, including paracentesis. So that's the structure that we've proposed. But that's the conceptual structure. What about the actual structure of getting the information we need? So that's where the CQC comes in. The Cirrhosis Quality Collaborative is funded by the ASLD. And it's a multi-center quality improvement collaborative that is launching this fall at 10 sites nationwide with eventual expansion to all the sites. And this provides a structure for quality improvement and cost containment interventions. When I say structure, it's a database at a central data collecting center. And all the different sites are submitting the data. And not submitting manually. This is not increased work for the physicians or the staff. This is an automated EMR interface and also patient-reported outcomes. Patients are entering their own information. So the idea is sustainability, because it should not cost much to enter this data. And so here's just a screenshot of the face page of our registry. And so sites that are participating and will be able to log in. And then each site will be able to see their dashboards. And the background to this is we created a bunch of quality measures. And then we had to specify them and identify how to get the data from the EMR. And so each site will be able to see how they're doing and be able to compare to other sites. Each site will be able to dive down to their individual patients. And so, for example, each site sees that, let's say, only 80% of their patients are getting HCC screening. They'll be able to dive down and see, well, which patients do they need to reach out to get their ultrasounds up to date. But each site will not be able to see individual data from the other sites, just summary statistics. But they will be able to compare summary statistics to other sites. They'll be able to say, gosh, I'm at 80%. These sites are at 95%. And then they can talk to the other sites that are doing well and say, what are the best practices? And they'll learn. So it's a learning network. There's a couple details that we're still trying to work out. And so the prior two talks, when it sounded like a little bit things may still be vague and up in the air, how is this going to look? It's because we're still learning. It's kind of cutting edge. And so one thing that we're still uncertain about is, what are the inclusion-exclusion criteria and the entry and exit triggers? When we talk about population management, we actually, rather than thinking about inclusion-exclusion, we think about entry and exit triggers, because there's this concept of things change over time. But for example, do we include all patients with cirrhosis? What ages? What about the patient who has CHF and cancer and COPD? And oh, by the way, happens to have compensated cirrhosis. We probably don't want that patient in our program, I think, because their drivers of cost are going to be very different. When I talk to health plans, insurance companies, they're very worried about multiple comorbidities. And so that still needs to be ironed out, because they think that a lot of these patients, so their perspective is all these patients have 10 other conditions as well. I actually don't think so. I think Mina's data pointed out that a large portion of patients actually don't even see their PCP. If they have liver disease, I think that was useful. So I think there's a cohort of patients that are more pure cirrhosis. And those are the ones that we need to focus on initially. And then the financial model still needs to be finalized, worked out. Probably in the beginning, it's going to be more like a level one model. So we get paid X amount of money per month to support the infrastructure. We don't really share any savings, or a little bit of upside potential. But eventually, either we're going to have to prove to the insurance companies that we really can save them money or we're going to have to take a risk. And that's kind of the future of healthcare, unfortunately, is that health systems and practices that are not willing to take financial risks are probably not going to be as financially successful. So that's it. I know I breezed through some of the stuff that you had already covered. But I just want to see if anyone has any specific questions about the CQC, what's the future for that, or a cirrhosis medical home model. Thank you. Michael, do you want to share what some of the quality measures are going to be that you're tracking? Yeah, thanks. We created, as many of you know, quality measures can be divided into structure, process, and outcome. And an outcome, they can be divided into clinical and patient-reported outcome. Structure measures are a little bit hard to measure, and they're pretty far distant from what we care about, which are patient outcomes. Process measures are really easy to measure. But the problem is, with process measures, is that individual centers or physicians can game the system by including and excluding patients, and more easily improve their scores that way when you look at process measures. So ultimately, we tried to focus mainly on outcome measures. So we have a few process measures, such as HCC screening. Did the patient get a HCC screening? That's kind of one of the easiest ones. I was, Fasiha Conwell and I were on the PQRS advisory panel for CMS, and that's really the only one that panned out out of all the measures that CMS was considering was HCC screenings, because it's just the easiest, the simplest, the best evidence. But we have a lot of outcome measures, so obviously survival, re-hospitalization. As you mentioned, other process measures, whether the patient got in to clinic visit within two weeks after hospital discharge. We have patient-reported outcomes. So we're using the ASI-7 for ascites, because ascites control is one of our key things. We tried to structure them around the domains of liver disease, so encephalopathy ascites, HCC viruses, and other patient-reported outcomes. We're having the patients tell us if they get re-hospitalized because, as you all know, sometimes they get re-hospitalized at a different center. And we're also focusing on substance abuse, because that's a main driver, a big driver of how well our patients do. So we're including data, patient-reported outcomes on their drinking habits, also whether they got referred to substance abuse counseling, whether they attended substance abuse counseling. So that's an example of some of the measures. Yeah, and I, oh, good, sorry. Hi, I'm Simona Jacob from Yale. I have a question and a follow-up after that. So do I understand correctly that CMS is interested to apply cost-related, reimbursement-related to HCC screening rate? And if that's correct, would you add to your cirrhosis home a walk-in ultrasound? Because we all have patients, you know, you put the order in, you want them to be screened, and then they don't show up, so you have to do it again. So, you know, if you get penalized by not achieving a target HCC screening, how are you gonna overcome that? And any thoughts on that? Yeah, good question. So the HCC screen is a PQRS measure, and so your institution or your practice can use that as one of your MIPS measures that will affect your reimbursement. As far as the interventions for how to improve HCC screening, so we did a study, you know, years ago, where we looked at that, and, you know, one of the simplest ways was just creating a tracking system. Then we also just tried to align their ultrasounds with the clinic visits, so we scheduled them for the clinic visit, and they'll ultrasound right before or right after. Now, that only works if your patient can come to your institution for the ultrasounds. I live in a very heavily managed care environment where patients can come see me in clinic, but their imaging has to be at some ambulatory imaging center, so that makes it harder. But, yeah, I mean, there are novel ideas, such as, you know, more mobile ultrasound units that if, in the CQC, if we identify there are some centers that are really, you know, blowing it out of the water with their rates, and we find out what are they doing, and how can we learn from them. So then the cirrhosis medical home, you are not anticipating in the future to include an ultrasound, so I think in Italy they had that day hospital model when patients would go in, they'd get ultrasound, they'd get EGDs if they want to, so. Yeah, yeah, no, that's a great possible intervention. That would be a great idea, yeah. Yeah, so Paolo Angelli Pubstad, and that actually showed improvement in mortality at the cirrhosis day hospital, and so I totally agree, because then they could get their paracentesis, they could get observation, they could get IV fluids, things like that, so that's one possible intervention that may work. All right.

value-based care model

cirrhosis patients

Cirrhosis Quality Collaborative (CQC)

patient-centered care

quality improvement