 So our next speaker will be Milda Saunders. Milda is a general internist, and she is an ethicist also. And the patient advocate for organ transplantation here at the University of Chicago. Her early research looked at the role of neighborhood. So beyond individual's own personal characteristics, the role of neighborhood in their outcomes. And more recent work has looked at trying to improve care of patients with chronic kidney disease or people at risk for chronic kidney disease with a special interest in health disparity populations and improving care. Races from looking at equity in the options of renal treatment that are offered, as well as improving the system of care for renal patients. National Institute of Minority Health and Health Disparities is currently writing a textbook on health disparities. And Milda is leading one of the most important chapters. It's the one on improving access to high quality care. So Milda is literally writing the book on health disparities. Warm introduction and for a reminder that I owe your revision. So today, I'm going to talk about to examine the policy context of Medicare funding for regulation for patients with in-stage renal disease. And to do that, I'll describe the in-stage renal disease Quality Improvement Program, which is CMS's paper performance program for dialysis facilities. But then also to go over some brief analyses that help determine the outcomes for in-stage renal disease quality improvement in relation to quality and equity. So background, why in-stage renal disease? One, it affects more than 700,000 patients in the US. There are 700,000 treated in-stage renal disease patients in the US. And Medicare covers virtually all eligible citizens with in-stage renal disease, which is about 90% of patients with in-stage renal disease in the US. And it covers patients regardless of age. In addition, in-stage renal disease represents about 8% of Medicare spending, which may or may not seem like a lot, except they represent only 1% of Medicare patients. In addition, until recently, erythropoen stimulating agents, which are used to treat dialysis-related anemia, represented the single largest Medicare drug expenditure. So Medicare has some stakes in the game. So then why dialysis facilities? So in my work, I think dialysis facilities are important both as a site of care and for potential interventions. Because for most patients with in-stage renal disease, this is their primary interface with the medical system. Of all patients with in-stage renal disease, about 60% receive their care in the dialysis facilities. And for patients who are beginning treatment for in-stage renal disease, that number is more like 80. So interventions here provide a unique opportunity for quality improvement. And we hope disparity reduction, since the federal government, unlike many other diseases, is both the primary funder and regulator of care. In addition to having that motivation, we always wanna have a sort of more noble motivation. So we know that this is a high-stakes disease, both in terms of cost. So patients with treated in-stage renal disease on hemodialysis, it costs about $750 per patient per week. And in addition, it's also high-stakes for the patients. We know that there's a high mortality. So the five-year mortality rate is about 35%, which, as you may know, is about on par with most of the common cancers that we all fear and avoid. So the policy context. So in 2008, that was a busy year for healthcare. So as things were, facilities were required to publicly report their outcomes in a program called Dialysis Facility Compare. So people, both patients and policymakers could look up a particular facility, type in the zip code, and look at the quality outcomes for that facility in the hopes that they would decide whether or not they would do care, and perhaps in the hopes that having to put your numbers out would motivate facilities to provide better care. Many of you probably never heard of the program, and many patients also never heard of it. It wasn't necessarily something that was commonly utilized for patients to compare where they went. In addition, there was a Medicare Improvement for Patients and Providers Act, MIPA, which did two things that were related. One, it bundled prospective payments to dialysis facilities. So instead of this fee-for-service, there was a bundled payment to provide for all outpatient care, including Erythropoietin Stimulating Agents, and the composite payment was set for 2011. In addition, it established the program that I'm gonna talk more about, the End-Stage Renal Disease Quality Improvement Program, which was set to begin in 2012 with a core set of measures, and then to expand the number of measures and the quality threshold in subsequent years. And so the clinical context, as we talk about the measures, so research had previously shown that improved dialysis adequacy and anemia management were associated with greater quality of life and decreased risk for hospitalization and death, which is something that we can all get behind both for our patients and for our healthcare system. And there was increasing evidence, however, which was new that these Erythropoietin Stimulating Agents, in addition to being really expensive, were harmful at high doses, both increasing risk of mortality and of cardiovascular risk, including stroke, at high doses. However, despite this clinical evidence and FDA warnings, there was continued use of Erythropoietin Stimulating Agents, ESAs, above recommended doses. Because prior to this, they were paid for individually. And so there was a multiple motivations that perhaps conflicted. So the important program here, the NCDUQIP had a tall order. One goal was to provide cost-effective care for this really high-cost population and to improve the quality of care that patients receive in this high morbidity disease. So we sought to examine how they were doing and whether or not disparities were impacted. So the aims were to determine which dialysis facility characteristics, neighborhood demographics, and were associated with payment reductions under in-stage renal disease quality improvement program in the first year of the program. And to do this, we linked the quality improvement performance file to census data and dichotomized the performance scores which were derived from three outcomes. The percentage of patients with the urea reduction rate greater than 65, which is whether or not dialysis was adequate, a measure of how clean their blood was. So that's good, we want that. A measure of hemoglobin less than 10, which is not good, we don't want that, we want less of that. And a measure of hemoglobin greater than 12, so a blood level greater than 12, which again, is not good that we want less of that. And patients and facilities got any versus no payment reduction. So this is a lot, but I can go through it. So overall, in 2012, only 30% of facilities had any payment reduction, and we'll go through some of the factors that were associated with these payment reductions. So we can see, I actually am not that tall. So we can see that chains, so the group, the corporate ownership of a facility was associated with whether or not a facility got a payment reduction or not. And so the two large national chains were less likely to get a payment reduction than some of the other smaller chains or compared to, and then also compared to non-chain, so independent chain. And the larger number of stations, the larger facilities were also more likely to get payment reductions, as well as those who had a longer length of operation. But what I wanna highlight here, and it didn't quite transfer, is that neighborhood characteristics in which the dialysis facilities were located had an impact on whether or not facilities received a payment reduction. So we can see that compared to neighborhoods with the lowest proportion of African-Americans as the proportion of African-Americans in the neighborhood increased, the likelihood of receiving a payment reduction for low quality also increased. And we'll have you know, so it's poverty, but we control it here, so you're right. When we look at poverty, it is significant. Neighborhoods that had a high poverty rate also were more likely to have a payment reduction. But then when we throw this all in the model, we see poverty is no longer significant, but the thing that remains, including the facility age and the chain, is that the percentage of African-American in the neighborhood is still associated with a payment reduction. So the odds of receiving a payment reduction for low quality if the neighborhood is in the highest quartile of African-American population is 25% greater. And we know that most of the clinical performance due to over-treatment of anemia, which was reduced, and one of the largest reductions in this over-treatment was in facilities in African-American neighborhoods as well as for-profits. Yeah, facilities in African-American neighborhoods still had the greatest payment reduction. Maybe they were just getting warmed up. This was the beginning. They were kind of working the kinks out. So we wanted to examine what happened later in 2016 as facilities had more time to get used to these quality improvement efforts, more time to improve their systems, more time to improve clinical care. So then later we did an analysis which examined which dialysis facility characteristics in neighborhood demographics were associated with payment reductions in 2016. And we linked that data to census data and used the QIP performance scores, which as we said would get more complicated as the years went on. Now there are eight clinical measures and four reporting measures and we again dichotomized any versus no payment reduction. So I'll just talk about the eight performance measures. So it's the same measure, hemoglobin greater than 12, which is seen as bad, vascular access. So the number of patients with a fistula, which is good, the number of patients with a catheter, which is not good, how well the blood is cleaned during dialysis. So dialysis adequacy, how often patients who receive hemodialysis get a bloodstream infection, which we think is bad, and how well hypercalcemia was controlled, which most non-neprologists don't think about much, but we think it's good. And then there were three reporting measures, patient experience, mineral metabolism and anemia management, which just won't go away. And so the facilities who had a total performance score of less than 54 out of 100. So they were given people some wiggle room, had their payments reduced on a sliding scale between 0.5 and 2%, which whether or not that significant was up to debate in the field, but it is enough to make people take notice. And so these results show that one of the most significant factors that was associated with receiving a payment reduction in 2016 was receiving a payment reduction in the prior year. So we know that players that are challenged, I don't wanna say bad players, facilities that had struggles in the prior year continued to have struggles going forward. We see again that chain is important, so some of the two large national chains were less likely to receive payment reductions than non-chains and the independent chains. We don't see the same effect for the size. And we also surprisingly see the persistent effect of the percent African-American in the neighborhood. So we see here both in the bi-variable and the multivariable model where we control for all of these things that having a large percent of African-Americans in the neighborhood is independently associated with receiving a large payment reduction for low quality. And that's after you control for poverty. So this leads one to conclude, facilities in African-American neighborhoods and obviously with a higher proportion of African-American patients are still more likely to see payment reductions. And so here we see, just to give you a sense of what we're talking about, so the facilities in the highest quartile were neighborhoods that had greater than 25% African-American, but we see that when we look at the patient population within those dialysis facilities, the patient population was more than 60% African-American and so the quality received in those centers in those neighborhoods has a greater disparity impact on African-American patients. We do see, this is a mixed ad, good news, bad news, so overall fewer facilities were receiving a payment reduction. So we saw in 2012 it was 30% to receive any payment reduction. In 2016, the number had decreased to 6%. So this may mean if you debate whether the quality measures or what we wanna measure for sort of quality outcomes which are still debated, but say that that is true and that we got it right, that may mean that more patients are receiving high quality dialysis care. However, the same people who have disadvantaged time and time again are still people who are left out of that rising, rising tide, lifting all boats. Still the people who are receiving lower quality care and are in facilities that may continue to provide lower quality care because they got less money to provide that care going forward. So conclusions and implications. We know that healthcare disparities are in part a quality issue. They're complicated reasons, but we know that healthcare disparities are people not getting the care that they need at the place that they needed in the way that they needed to provide good healthcare outcomes. General quality improvement programs and policies can increase or decrease disparities and to avoid unintentionally increasing disparities in stage renal disease and other quality improvement programs should measure the change in the disparity as the quality improves and provide incentives and or resources both for targeted quality improvement or for disparity reduction. And I wanna thank Marsha Chen, who's a mentoring collaborator and another collaborator on the paper and then some funders of this work. And I'll open it for questions. Hi, John Lantos from Kansas City. Thanks, it was very challenging paper. I mean, it raises the general question of how to tailor incentives to achieve social goals. I mean, the idea of incentivizing programs that are higher quality or punishing those that are lower quality seems to make sense. Although your last slide seemed to suggest an opposite approach. That is to target funds towards the centers with lower quality to help them improve their quality. So if you had to choose between these two approaches, which would you recommend? So I'm not gonna choose. I think it's a both and. So we know that as we saw from the numbers, there's some saying that we're gonna measure this and we're gonna incentivize helped improve quality for the vast majority of dialysis facilities. And so that's a good approach. But what we then can do is say that there's some facilities that serve patients that no one else will because in order to be accept, you have to be accepted at a dialysis facility. And so those facilities, some of those facilities may just be bad players. And so they will get penalized. They have no interest in improving and maybe those will close down. But let's assume that that's not the case for the majority of facilities that get penalized year after year. So then we can target resources, technical assistance, risk adjusts based on social disadvantage or clinical status to provide additional funds and training to help those facilities provide high quality of care to the patients that they serve who are most vulnerable and most in need of the care that the facilities provide. Excellent presentation. I thought anecdotally that the lead expensive drug in Medicare is Viagra. And I know there are many more people who are having ED than ESRD. And can you comment on that? And then I have a truly serious question. Okay. So they're probably sub-overlap in the patient population. However, it's the patient population and the per-cost of the drug. And so arithmetic and stimulating agents are expensive per dose and dozed frequently, either three times a week or weekly. And so potentially, and we've sound in research to be overdosed. And so potentially that was responsible for the high cost. Right. There are many debates in Medicare case management for farm benefits. Is the average elder, should they get three Viagra tablets, seven Viagra tablets, or 21 Viagra tablets a month, if not more. So that just another anecdote. The real thing that I wanna let people know is Rithropotin is a very important drug. And in high risk neonates, those with neonatal encephalopathy, it's neuroprotective. And there are currently NIH RCTs funded for that. We could get no buy-in from the ebomakers to sponsor any of the phase one or phase two trials of those agents because this is so lucrative. This is another example of some of the health disparities that inadvertently we see affecting vulnerable populations. Thank you.