Purva Rawal: We’re a year into the strategy refresh. I think we’re in a really great place where we’ve been able to lay a lot of groundwork for the new strategic direction, and we’ve been doing a lot of thinking about how we want to approach specialty care in our future models. And so, I’ll just take a step back and talk a little bit about our strategic objectives; and I think as I talk about them, it’ll be clear how specialty care is crosscutting, and we really have to have a strong focus on specialty care in order to achieve these objectives. So last year, CMS set out a vision to advance health equity, to expand coverage, and improve health outcomes. And to support this vision, the Innovation Center launched its, as I said, strategic refresh in October of 2021; and really led with five objectives.

The first was to drive accountable care; the second was to advance health equity; the third is to support care innovations; the fourth is to address affordability; and the final objective is to partner to achieve system transformation, where the LAN especially has been a critical partner. And so when you take a step back none of these objectives stand alone. They’re really infused across all of our work and model development. So, I’m going to talk about each one briefly because I think that might help illustrate why and how we’re taking a comprehensive approach to specialty care in our work at the Innovation Center. So, it’s not just an individual model, but how does this cut across all of our work and thinking? So, as I said, I’ll start with driving accountable care. And as part of our strategy refresh, as you said Aparna, we set a bold goal to have our beneficiaries in longitudinal accountable care relationships by 2030, where providers are accountable for the quality and the total cost of their care.

So, when we take a step back, we think about how beneficiaries might experience this. They’re likely to experience accountable care relationships, mostly through advanced primary care, or ACOs, accountable care organizations, and those entities are then expected to coordinate with, or fully integrate, hopefully, with specialty care to deliver that whole person care. And so, if we want beneficiaries to have their full range of needs met, all the way from vaccines and preventive services to specialty care needs for patients or beneficiaries with congestive heart failure or COPD. If we want all of those needs to be met, then primary and specialty care have to be working together to coordinate across that care journey. So really, when we’re talking about accountable care, it’s really bringing primary and specialty care together. Our second objective is to advance health equity. We know that there are significant disparities and healthcare access and outcomes, especially among underserved populations.

And so, to advance equity, we have to test ways to increase access to high quality specialty care that then helps to improve overall health outcomes and to close those disparities. The third objective is, as I said, was to support care innovations. And here, this is actually an area where we’ve done a lot of discussion and speaking with members of the LAN. Here, we’re really considering the role of data, and other supports, that can provide an increased access to high quality specialty care; and to enable that more integrated person-centered care delivery, that I think we’re all driving towards. Access to data that kind of enables communication across primary and specialty care, and that’s available at the point of care, is critical. And that’s really some of the work that sits under that objective. And then our fourth objective is to address affordability and improving access.

We know that improving access to high quality specialty care that patients receive at the right time and in the right place can help improve the over overall affordability of care for individuals and families. And then the last objective is to partner to achieve system transformation. And as I said, the LAN is a really critical partner here, and a piece of how we engage with stakeholders, and also receive feedback on our direction and on our work. We’ve spoken a lot since the refresh was launched last October, about how we have to partner with stakeholders and across CMS and our other federal partners to achieve that system transformation. And specialty care is a core part of our work because we can’t transform the delivery of care, and the care experiences, and outcomes, and health of our beneficiaries without it. So, we’re really excited to continue to partner with specialty providers and facilities as we’ve done with our bundled payment models, which many of you are familiar with our ESRD and other models, but also to build on that; and create new partnerships; and bring more specialty providers onto the value-based payment pathway.

I guess if there’s, you know, as we’re starting out, if there’s one thing, I’d like to leave the audience with today, it’s that we’re taking a comprehensive approach to specialty care in our work at the Innovation Center, and it’s really centered around the patient care journey.

Purva Rawal: There are four main elements or building blocks that are really important to integrating specialty care in a way that drives broader system transformation for patients. So, I would say the first is data, and I talked about this a little bit at the beginning, but we’ve heard from many LAN members and other stakeholders about the critical, and our model participants, about the critical need for data to drive quality and integration. So, for instance, primary care focused ACOs and advanced primary care models could use more information to make high value referrals to specialists.

ACOs could also use data to improve quality and cost for episodic care and some could even benefit from help with subcontracts for certain conditions. So, these are all avenues that we’re looking at and exploring and would welcome feedback on as well. I think the second building block or critical element, is maintaining momentum with our acute care episodes. We know over the last decade that hospitals have transformed the way that they provide care because of episode based payment models, like our BPCIA and CJR models, which many of you all are familiar with. And so, we’re actively exploring policy options for potential successor models for these two initiatives. Keeping in mind the evaluation results that we’ve seen through both of those initiatives, [and] keeping in mind feedback from stakeholders on the importance of episodic models and bundle payments as part of a broader value-based care portfolio.

And then I think the third building block is needing to support our primary care models to engage specialists. So, as I said, data is one component of this, but we’re hearing a lot from stakeholders and primary care focused ACOs, and advanced primary care models, how challenging it is to reach past the point of referral for ambulatory conditions. And then on the other hand, we also hear from specialists about how they’ve been left out of models. And so, we need to solve for both problems and bring specialists to the table; and so, we’re looking at different types of incentives and structures that could make that possible. And then fourth, I think the fourth, critical element, or building block, is better quality measurement across primary and specialty care. So, we really want to try to identify new methods for measuring those quality outcomes that are specific to the care that specialists provide, which is often downstream from primary care. And so, we know that assessing specialist performance on primary care measures that they have less influence over is a real barrier to specialists joining or being integrated into ACOs since another area a focus for us right now.

Purva Rawal: We’ve been testing our ACOs bundle payment models for over a decade, and we’ve learned so much. And I think the health system and all the stakeholders have learned a great deal through these model tests about where and how transformation has occurred, and where we can do better. And I think now we’re at a point where we have to bring those learnings together, to test innovations that help evolve primary care, and episodic care, and payment together, again to improve beneficiary access outcomes in their care experience. We just want to keep re-centering on that being the end goal of what we want to achieve. You know, our testing of specialty care and bundle payment models, we’ve generally, they’ve generally shown improvements in lowering expenditures and enhancing quality, episode-based payment models like BPCI, you mentioned, and the bundle payment for care improvement initiative advanced model, and our comprehensive care for joint replacement, or CJR model have, you know they’ve reduced Medicare fee for service payments for the majority of episodes in those models, while at least maintaining quality for beneficiaries. And these model tests, we know, at a high level, that’s what their findings are, but we know that they’ve supported care transformation in facilities across the country; and they’ve really helped lay the foundation for our next challenge to increase access to coordinated and integrated specialty care. So, to your question, I don’t think we can separate accountable care from episodic care. You know, from a beneficiary’s perspective, as Dora said earlier, they want their care delivered in a seamless, coordinated manner that’s consistent with their goals and values.

And so, when we’re thinking about equitable access to high quality episodic care, it’s part of accountable care in that patient journey. So, I can’t really emphasize enough that that’s the lens through which we’re approaching model development in our work. How do we support high quality specialty in primary care? And to illustrate this, I’ll share a patient vignette with you at a beneficiary in one of our REACH ACOs. And I think it always helps to think about the patient journey and the needs from in the health system, like through the perspective of a patient. And here we have an almost 70-year-old African-American male beneficiary with multiple chronic conditions, so, diabetes, heart failure. He’d been experiencing multiple emergency department visits, for instance. So, all of the things that we want to try to weigh, find a way to get better and more coordinated care.

So as part of his, you know, part of being in this ACO, he started to receive monthly in-home visits with the dedicated nurse practitioner and a social worker. He was given a scale to monitor his weight and started to receive education on, sort of, the appropriate use of the emergency department. And the outcomes that we found were better follow-up with the primary care provider, improved management of those chronic conditions, and a decrease in emergency department visits. And so, that’s how we want to start to be able to build models that can help support that kind of seamless and integrated care delivery.

Dora Hughes: About 70% of our Medicare beneficiaries have two or more chronic conditions. Nearly 20% have six or more chronic conditions. 30% of Medicare beneficiaries see five or more physicians annually and struggle to effectively manage the care. We also know that racial and ethnic groups, and other high-risk populations are the most affected with worse health outcomes and reduced life expectancy. We feel very strongly that primary care is part of the solution. Primary care plays a vital role in the healthcare system, or as stated more directly by the National Academies, high quality primary care is the foundation of the healthcare system. They state that without such care, minor health problems can spiral into chronic disease, chronic disease management becomes difficult, and uncoordinated preventable ED visits and hospitalizations increase and healthcare spending soars.

I love that quote. In one sentence, I think it fully captures the challenges that we face with primary care, and really helps to the importance of our investment in these areas. Primary care is associated with improved quality, lower mortality, lower healthcare costs. And because of these reasons, the Innovation Center has tested a number of primary care models. We’ve tested comprehensive primary care, comprehensive Primary Care Plus, and now Primary Care First, or PCF. We’ve learned quite a bit through these models. And perhaps, most importantly, we’ve learned that primary care practices can indeed, meaningfully, change how they organize and deliver care. Our primary care models demonstrate that even practices with limited prior experience, with value-based payment models, can make needed changes to deliver advanced primary care. At the same time, we will acknowledge that the quality and total cost of care impacts have often been modest, not significant, and certainly we’re challenging to detect within a five-year period of our models.

And another issue with our primary care models—certainly equity was not always an explicit aim—and Blacks and Latinos have been underrepresented in these earlier primary care models. So, in terms of thinking about how to build upon what we’ve learned, how we can provide better support so that advanced primary care capabilities can be implemented and lead to higher quality, equity, and reduced costs? Now, as obviously we’ve heard from Purva talk at length, building stronger relationships between primary care and specialty care is a major area of focus for our thinking right now. We believe that an enhanced focus on specialty integration will require aligned payment, and payment incentives, as well as quality measurement and data feedback to reduce burden and support practice transformation. So, what specifically could this look like in model design? Increasing specialist focus on ensuring value and appropriateness before scheduling a referral visit and closing the referral loop back to the primary care post visit. Both of those would help to support accountability, shared risk, and team-based care. As another example, and Purva touched on this a bit, facilitating the flow of data, such as improving electronic consultation to efficiently access specialist expertise, to support co-management of patients through collaborative care arrangements or care compacts, and to confirm appropriateness of the referral. These are all very specific examples of how we think that that we can better integrate primary and specialty care. And these are concepts we are exploring for our future work in this space.

Dora Hughes: I acknowledge that in these cases, the specialists may need to play a more involved role for a certain period of time, or even extended time. However, even in these cases, it is so important that the patient continues to see their primary care physician, and that the specialists and primary care physicians stay well connected. This coordination, or co-management, whatever the term you might use, across types of care and settings of care, this is exactly what we’re hoping to foster through our next phase of accountable care models. So, Purva, I will turn to you to see if you disagree with my statement to offer a different perspective, the non-medical perspective.

Purva Rawal: I think we can obviously do more to inform high quality specialty referrals. I think we can do more to provide data to ACOs, to support better episodic care management. And then in our models, I think we can also look at ways to see how our model participants can be using and contributing to interoperable data exchanges that connect providers in the given region. I actually think the work that the LAN is doing here is really important, especially through the state transformation collaboratives, and that’s an important piece of the puzzle here as well. I think the state transformation collaboratives, I think will hopefully help us see how regional efforts can help harness both state and federal government capabilities, as well as private and non-profit organizations, and the role of, kind of, harnessing those capabilities to accelerate transformation.

Purva Rawal: I think most providers, they want to do what’s best for their patients. And the current fee-for-service payment system doesn’t allow them to always coordinate care, to help manage patient and family needs across important transitions from the hospital to home, for instance, or to provide more longitudinal care for beneficiaries with chronic conditions. And value-based care and payment models aim to realign those incentives for all providers, and in today’s theme, and to realign those incentives across primary and specialty care, so that providers are able to do those things for their patients. Over the past year we’ve conducted research on how beneficiaries, but also providers, view value-based care. And on the provider side, while most have heard the term value-based care, some see real downsides because of issues like implementation burden, the potential for financial risk, and honestly, the lack of experience with value-based payment models.

So, I think we have to address these issues and then create long-term incentives for specialists to move from fee-for-service to value-based payment; and to be able to demonstrate how that movement to value-based payment provides the flexibility, and the support, to meet their patient’s needs. And so, in terms of how we can do that in our models, we can examine incentives for specialists to join an accountable entity. We’re also considering the different types of ACO structures, and how we may engage specialists in both hospital-based and physician led ACOs, through different quality and financial incentives as well.

Dora Hughes: If we’re talking about equity, I have to start off by acknowledging the incredible work by the HEAT, the Health Equity Advisory Team, that’s part of the LAN, under the leadership of Marshall Chin and Karen Dale. They have just really helped to influence the thinking and the broader ecosystem, and certainly our thinking at CMMI as well, in terms of understanding some of the priorities and possibilities for achieving health equity. So, taking a big step back, as purpose started out this conversation, she mentioned that advancing health equity was our second objective within our strategy refresh. And I think even within that bucket, there’s four areas of focus that I would briefly mention today. The first is really thinking about, for our new models and for even possibly some of our current models that may have a second cohort or other phase, how do we make sure we are embedding equity in the model design and implementation?

And I’ll talk a little bit more about that in a minute. And the second area is we’ve acknowledged that we have to do a better job serving the full diversity of our beneficiary populations and our models. We want to see racial and ethnic groups. We want to see those that are disabled. We want to see those in rural areas, as well as urban areas. We want to see those defined by sexual orientation [and] gender identity. We want all of that to be fully reflected in our models. And part of our early, part of our first strategy to improve the diversity of beneficiaries in our models, is to increase the number of safety net and medicate providers in our models. And so that is a really important area of work and one that you’ll hear more about as well.

And the third area focuses evaluation <laugh>, I think Purva it earlier, something that sounds kind of basic, I would say evaluation sounds kind of basic. But we have acknowledged that our first 10 years, we weren’t always asking the question about what is the health equity impact of our models. And if you don’t ask the right questions, you’re not collecting the right data early enough. So, at the end of the model, you’re able to evaluate the impact on health equity. So that is a third area of intense focus. And finally, the fourth area focuses on data. We’ve talked about data quite a bit, but really, we want to make sure we’re collecting sociodemographic data that can inform our model evaluation as we certainly even monitor our models, evaluate our models, thinking about models moving forward. Having the right data is just critically important. So, in the context of specialty care, then how does that align with our focus on equity?

And really, they track along quite well. I mean, first, if we look at our specialty models at a high level, a question where even at the outset we’re asking ourselves, what are our new specialty models that we invest in? Will we have relevance to populations, spacing, and equities, and care, and outcomes? And as example of that, we know kidney models as one example, they disproportionately affect race and ethnic populations. Kidney and stage renal disease requires significant specialty care. And as we improve the care that individuals with end stage renal disease through our models that will certainly disproportionately benefit racial ethnic populations. As another example, our new cancer model, our new enhancing ecology model with cancer as a focus. The question is like, would any of the populations facing inequities, what cancers might disproportionally affect them?

I mean, colorectal exam, colorectal cancer is a prime example. There are other cancers that disproportionately affect racial groups, such as multiple myeloma. So really, even thinking about the outset of a model, does it have relevance? And then within, once we decide to go down on a certain track, how are you making sure that model addresses priorities of concern for underserved groups? And I think, even as you move forward and you start to think about, okay, this is the area we’re going to tackle, what about the model design? What are the model elements that can also help to address equity? I mentioned EOM, our Enhancing Oncology Model already, but I think it’s a prime example of how we’re thinking about model design. This model will provide an increased per-beneficiary per-month payment for enhanced services, for caring for complex patients.

Similarly, if you think about model-design elements, we could point out the ACO REACH model. It will test a health equity benchmark adjustment for providers serving a disproportionate share of underserved populations. So, we’re excited when we think about the new models, -to really start at the very front end, to think about what the potential elements might need to be, what the potential impact may be on the help of certain populations at the outset when we have the opportunity to improve the models, designs, and concepts. And then, if think about specialty models, again, our second area of focus is bringing in the full diversity of beneficiaries. In part, by bringing safety net and Medicaid providers. We are signaling that we want to recruit providers with experience serving underserved patients in our models. And as one example of how we try to do that for our ACO REACH model was that one of the selection criteria was asking applicants about your experience caring for underserved communities, what’s your track record?

What have you done? And so, I think that that approach, certainly whether for primary care or specialty care, will increasingly be one example of how we’re trying to bring in the right providers. A third area of focus: evaluation. We certainly across all of our models, I think Purva touched on episodic, our bundle payments models, really just trying to understand what the overall impact has bee, but also, again, for the specific socio-demographic groups that, what we’ve seen from our early evaluations, the experience is sometimes different, and we need to better understand that. So, we again, we understand how we need to improve the model moving forward. And then the data, it’s always about the data. And just as for our primary care models, certainly for our specialty models, we are requiring new model participants to collect and report demographic data and social needs data when feasible.

This will help with our model monitoring and assessing how well we are addressing gaps in access: in quality of specialty care. And on the data piece, we know that this can be difficult. The CMS Innovation Center, we are absolutely trying to facilitate this data collection reporting. As a couple examples, we are working on FHIR-based questionnaires that can help providers screen their patients. We are also working on both high-tech mechanisms, such as data extraction from EHRs, as well as low-tech mechanisms, such as use of Excel files to obtain this data; because we know that our model participants, depending on their resources, their settings, they will have different levels of capacity, different levels of technological abilities, that are needed to accommodate our data goal. So across these four areas of focus, again, thinking about model design and elements, thinking about how to we recruit more safety net providers to reach our goals and beneficiary diversity, thinking about evaluation, thinking about data, across all of these areas of focus, we think that by the time we look in 2030 and do our next strategy refresh that we will have seen that we’ve been able to move the needle with regards to access and quality of care for underserved populations. So, I’ll stop there. And thank you again for asking the questions and being able to do this in partnership with the HEAT has really just helped us do our work better internally as CMMI.