Dora Hughes

In this episode, Dr. Liao speaks with Dora Hughes, MD, MPH, the Chief Medical Officer at The Center for Medicare & Medicaid Innovation, part of the Centers for Medicare and Medicaid Services.

Joshua Liao: Could you tell us about your journey and what led you to your current career path and position at Aledade?

Dora Hughes: I am a physician by background, internal medicine physician and I trained in the primary care track at Brigham & Women’s Hospital in Boston. After residency, I participated in a Commonwealth Fund - Harvard University Fellowship in Minority Health Policy. Through that program, I was also able to get a Master’s in Public Health. I think my father would say that’s when I became permanently derailed from my medial career. I don’t think he meant that as a positive. But in any case, I worked in the Commonwealth Fund for a bit in their Quality of Care for Underserved Populations area, spent some time on the Hill working for the late Senator Ted Kennedy, and then the former Senator Barack Obama and had the privilege of joining his administration at the time the Affordable Care Act was passed. It was certainly a very historic and exciting time.

DHughes photo 2.jpg

Dora Hughes, MD, MPH

Chief Medical Officer

The Center for Medicare & Medicaid Innovation

Now I am back, after a stint in consulting and some time at the George Washington University School of Public Health, in the role of Chief Medical Officer and I’m also leading the health equity initiative for the CMS Innovation Center. 

 

JL: I know that [the CMS Innovation Center] launched the Strategic Refresh last October in 2021. For those who may be less familiar, can you briefly summarize the emphasis on equity in that refresh?

 

DH: We have five pillars, if you will. The first one focusing on accountable care, the second focusing on advancing health equity, the third transforming innovation, the fourth affordability, and the fifth area is partnership. And certainly, we’re looking at equity across all the pillars as part of our 2030 vision, however, I’ll tell you a little bit more about the advancing health equity pillar.

 

We have four areas of focus. The first is when we are developing new models, but also modifying models that have at least three years left in the lifespan, how can we better embed equity? That could include, what are some of the care delivery requirements, what’s the payment, financial mechanisms, what’s the target population? All of those types of considerations for new models that could help to advance health equity.

 

Our second area of focus is really thinking about how can we diversify the beneficiaries that are participating or are being touched by our models with particular focus on increasing the number of underserved populations. We think key to that is including more safety net providers, Medicaid providers, FQHCs [Federally Qualified Health Centers]. That’s very much an increasing focus for us to make sure we’re diversifying, capturing the full range of beneficiaries that could benefit from our model.

 

The third area is focusing on evaluation. We have been quite candid that in our first ten years, we didn’t always evaluate our models with respect to understanding the impact on equity. So, we are moving forward trying to make sure that we are thinking about the evaluation design up front so we’re asking the right questions, collecting the data, so at the end of the model we understand the impact that we have had on health equity. We’re also doing a look-back, to see what lessons we have learned from the work and the data that we already have.

 

The fourth area, very critically, is how can we strengthen our data collection for socio-demographic data collection and what type of intersectional analyses that we can conduct because we want to, again, make sure that we understand the quality of care, the outcomes, the beneficiary experience – all of that for populations that may be defined by race, ethnicity, disability, geography, sexual orientation gender identity (SOGI), and other patient characteristics as part of making sure that throughout our models we are improving quality, including equity, across all of our work.

JL: It has been a year since that strategic direction was set. Can you share some of the highlights that you and others in the center have had in this first year?

DH: As you may be aware, we have recently released a strategy update, our one-year look-back to see what we have accomplished, what challenges remain, and looking forward, what will be our priorities over the next year. I have to say we were pleasantly surprised as we looked back to see how much we were able to accomplish across all five pillars.

 

But focusing on equity, we launched a number of new models, announced the models, all of which have a specific focus on equity whether that be through requirements, as an example, to collect demographic data and strong encouragement to collect data on social needs (SDOH). We included benchmark adjustments to help providers, especially safety net providers, who are caring for disproportionate numbers of underserved populations. Certainly, we have focused on evaluation; expect to see some reporting that will be coming out sharing some of our findings. I could go on and on, but we definitely have some specific accomplishments.

 

Then another, not directly related to our milestones, but critically important, I think, is in the partnership pillar that I mentioned. We have strengthened our relationship with our colleagues internally, particularly those in Medicaid, with how we want to do more with Medicaid providers in our models. But even outside of the Innovation Center, as an example, we’re working very closely with our colleagues in the Health Resources and Services Administration (HRSA), thinking about how we could include more FQHCs in our models, as an example. I think that both internally we have a number of accomplishments that we are excited about, but also thinking about how are we relating, how are we working with our external colleagues that have just amazing expertise and experience in this area, that will benefit our model work certainly, as well. So, I certainly would encourage all of you, all of our listeners, to check out the report and read for yourself. We of course, always love input and guidance and suggestions for how we can do an even better job over this next year.

 

JL: I think with all this work – and having to drive the work forward, but also partner with others in the community do to this – there must be some challenges you’ve encountered as well. Can you share a little bit about that from the first year?

 

DH: First, to your point, we have taken on a lot of work and so many different areas and all very important. So, struggling to prioritize and think about what do we want to do first and what needs to come later, remains a struggle. But one thing, I don’t know if I would say that it’s a challenge or an issue, but more just “it is what it is,” is that as we move forward with the first round of effort, in some cases these are operational issues – what actually can we do?

 

If we think about our requirements for health equity plans, for example, this is our first foray. So, trying to think about what is reasonable to ask our participants, what data can they collect, what resources/learnings supports will they need, will we get information back, what’s appropriate first to review, are there stringent requirements? All of that. This is our very first step with the health equity plans. There are a lot of operational considerations and basic questions. But we will continue to iterate upon health equity plans and other equity provisions or models for every new model or different phase of the model that comes out. So, we expect that we will continue to improve as we move forward and in a pretty rapid fashion considering the pace of the different developments of different models. We’re excited about that but it is certainly an on-going struggle.

 

JL: I think what I have appreciated about the models that I’ve seen being proposed are those elements that are operationally, need to be addressed, but that really try to orient things around equity. One in particular: there’s a model ACO REACH [Realizing Equity, Access, and Community Health], and there are a number of components to that, but one of those is to use an area level measure for adjusting spending benchmarks. I’m curious if you could describe a little bit about the thinking behind using that measure called the ADI [Area Deprivation Index], or more broadly, other geographic measures that reflect social drivers of health in these models?

 

DH: That is such an important question. For us, I think, as I alluded to earlier, to the extent that we want to pull in more safety net providers or providers who are serving a disproportionate number of underserved beneficiaries, we understand that they may require more resources to provide high level care for these medically and socially complex patients. Thinking about how can we allocate, there are a lot of different ways that you could provide funding, and some of these are being discussed more generally among our partners externally. There are a lot of conversations about what’s financial risk adjustment, social risk adjustment, or providing this up-front grant funding. All of that is certainly part of our considerations for the model. But one area that we have been exploring, and we are testing, is qualifying funding based on the Area Deprivation Index or ADI, as you propose. Is that a way if we peg some of our funding support to the ADI, which is a geographic measure, if you will, is that a way we could steer more, allocate more resources for providers in these areas?

 

We also, for the ACO REACH Model, our benchmark adjustment reflects both a weighted composite measure of the Area Deprivation Index as well as the Dual status of the beneficiaries. So, could a combination of the two steer more resources towards those beneficiary populations who will require more care? Again, this is a test. We don’t know if those are the best equity measures. Are there other measures that should be used to adjust the benchmark? We have mentioned life expectancy, is that a possible measure that would be a good addition to ADI and Dual status? Others have suggested, what about using low-income subsidy or other measures of underserved-ness? We’re still exploring that question internally, we’re still looking for our partners on the outside to help us think through what are the best measures of equity that we can use to adjust our financial support through our models, again, to both attract safety net providers but also sustain their efforts in the long term.

 

JL: I appreciate you saying that and widening that aperature to the different ways we might measure marginalization or the need to serve different populations. Building on that question, the role of geographic measures, I’m going to ask you to look forward a little bit. In other models, how could you envision geographic measures or other measures incorporating into payment models to address equity?

 

DH: So again, whether it is the Area Deprivation Index, we have certainly talked with our other colleagues in the Centers for Disease Control, as an example, to l earn more about the Social Vulnerability Index. We have also had many conversations with our colleagues in ASPE [Office of The Assistant Secretary for Planning and Evaluation], they have done a review of all the different types of measures, including geographic measures that could be used. I think that this area of thinking through the exact measures will continue to be an active area of exploration and testing, and also working with our colleagues in Medicare. Ideally, we would align if and when possible. So not only with the actual measures that will be used, but the index, How will they be weighted? What’s the algorithm that we would use? I think that that is another area of concern, and we certainly use a composite and use the best weights; but again, this is a test so we’re not sure.

 

But generally, I think the other broader questions that I alluded to earlier. We’re also looking to their recommendations from the Health Equity Advisory Team of the Healthcare Payment Learning and Action Network. We know that they have been thinking about financial risk adjustment and other mechanisms to allocate funding for underserved populations. I feel like we are building as we go. We are learning at the same time as our models are coming out, so this will be something, a three- to five- year, I think at that point we will have a much greater understanding of which of these approaches will be successful.

 

In addition to the types of funding, I do think the timing of the funding, we know for the AIM model, for example, our participants found it very helpful that they receive some of the funding up-front; in some cases, as part of infrastructure funding. That’s another consideration, the timing of the payment, the level of the payment, the structure of the payment. All of that. But it’s just a very exciting area and I think one that will definitely prove to be instrumental to the success of our models and advancing equity in the long term.

 

JL: In the spirit of innovation, which is in the namesake, but in the spirit of innovation, you have to try things and learn from them. If you knew exactly what they would bear out, I suppose it would not be so innovative. So, I look forward to seeing more of that. On that note, are there future areas of exploration or focus areas for CMMI and new payment models that you can share?

 

DH: Yes, one such area was just announced on, I believe, October 15th. The Drug Pricing Executive Order that the President issued which has given the Innovation Center 90 days to think about are there other innovative payment models that we might want to support that would complement and augment the provisions in the Inflation Reduction Act (IRA) that our colleagues in the Center for Medicare are leading implementation of.

 

For that, we’re thinking through possibilities. One of the criteria is equity. If we choose a certain drug class, or a certain population of patients, or a certain area that we want to test, what is the impact on equity? As one such example, we are mindful of the new, novel curative therapies in the pipeline at FDA for patients with sickle cell disease, [therapies] that are likely to come with the high price tags. Given that many of our beneficiaries are on Medicaid, are there ways that we can support states through some type of value-based model that would increase access to these curative therapies for individuals with sickle cell disease?

 

That’s just one example, but we’re also looking across other examples. For many of us, even outside of government, we focused on the Hep C subscription model that the state of Louisiana supported. So we’re canvasing the expert literature, experts externally and internally, to think about what are best ideas and then trying to winnow that down. But again, equity will be one of the criteria that we will use to judge how we will move forward. We’re excited about that.

 

Another area I would mention is that we have focused a lot of attention on racial and ethnic groups for a variety of reasons. One of which is not just the disproportionate health toll that they experience, as important as that is, but also because we have the best data for racial and ethnic groups. Flawed though it may be, still, we have some of the better data. But there are other groups to whom we would like to extend our focus.

 

For example, individuals with disabilities. We do have data from SSA [Social Security Administration]. We know if our beneficiaries are disabled but often it is pretty binary – yes or no. If we want to develop interventions, we really need to know whether they have physical disabilities, or vision, hearing, intellectual [and developmental disabilities]. That type of granular data would be incredibly useful for us so that, again, we could think about what are the best interventions or areas of focus that would be meaningful if we were going to tackle models that could focus on disability, improving the quality of care for that population. Those are two examples of where our thinking may go.

 

We have also signaled publicly already that we are doing more work thinking about primary care. We know that that is particularly important for all populations, and of course, I am biased as a primary care physician, but especially for underserved [populations]. We have also signaled that we are interested in exploring dementia and that maternity will always be a priority, so [thinking] across a number of areas, and also our diabetes strategy. It is just a lot of exciting work that’s in various stages of our model lifecycle approach. And, I think you’ll continue to hear more and more. Every month there will be new announcements that will be rolling out and new opportunities, frankly, for our partners and advocates externally to engage. We started having a number of round tables and other events. We want to continue them and we have found them quite valuable, and our partners externally have as well. So, continue that drum beat as well.

 

JL: One of the ideas behind my work is that activity doesn’t necessarily itself equal progress. That being able to go in the same direction and coordinate is very important, otherwise you just get a lot of well-intentioned activity in many different directions. One of the things that I think comes up often in these conversations is about multipayer alignment. There are definitely challenges around characteristics of populations, etc. How are you thinking about multipayer alignment and how does that play into what comes next in payment models?

 

DH: It’s such an important question and area of very active discussion. I would say active, animated, agitated – I’m not sure of the right descriptor, but we are laser focused on multipayer alignment.

 

And just to take a step back for the listeners who may not understand what that means. What we found in our models, if we’re targeting safety net providers, for example, if 50% of their patients are on Medicaid and only 10% are on Medicare, then they don’t have an incentive to really participate in our Medicare-focused models. So, if we are able to align our Medicare, quality measures, payment, etc, align and harmonize across Medicare and Medicaid, and to some extent our commercial partners, then as a provider I will have a greater incentive to participate in the model.

 

If I know my Medicaid payments and my Medicare payments and hopefully commercial payments, if they all have very similar requirements, or at least are directionally aligned (I think is how we sometimes describe it), provides much more incentive, reduces the burden, provides better data across all of the different payer sources. That is an area we have announced that we are quite focused on, multipayer alignment. Certainly, it’s hard work but we are working very closely with a number of our colleagues in Medicare and Medicaid. We are also trying to align working very closely with our colleagues at HRSA with focus on the FQHCs.

 

Again, these programs have all very different quality requirements, reporting requirements, financial structures and mechanisms, streams of funding, but we think that if we’re going to be sustainable and successful in the long term, we have to figure out how we can all move forward in the same direction. So, this is an area that you’ll continue to hear more and more about. We’re just trying to be very deliberate and systematic, both in our Federal programs and policies, but also in our conversations with states and commercial partners about how we can do more in this area.

 

JL: Before we end here, is there anything we haven’t talked about that our listeners, you think, would really benefit from learning about payment models and equity, and what you and others at CMMI are working on?

 

DH: One thing that we haven’t talked about and I think is very important, as you alluded to – what is the progress, what have you accomplished in the first year. One thing that we have included in the strategy update report is our attempt to lay out some measures so that we will be able to track and monitor our progress over the years. It has also been very tricky because, of course, we don’t want 10,000 measures, so what’s the core set of critical measures for which baseline data is available. So, there are a lot of different factors that have gone into thinking about that. But I would encourage you and the listeners to check out our strategy update report to see the proposed quality measures.

 

Again, these might not be the right measures. We may need to expand. We may need to reduce. But this is our first step to really put forth measures that we will seek to measure our progress, accountability to ourselves, and to the beneficiaries that we serve moving forward as we remain focused on our 2030 vision and achieving those objectives. That would be one other development that I certainly would call your attention to.

 

JL: Dr. Dora Hughes thank you again for joining today. I certainly benefited from hearing your perspectives and I trust our audience will as well. Thank you again for joining us.

 

DH: Thank you, thank you so much.