The world of health care has changed a lot in the past decade, and the Center for Medicare and Medicaid Innovation (CMMI) at CMS has been behind a lot of that change.

Radio Advisory's Rachel Woods sat down with Liz Fowler, director of CMMI, to talk about what innovations CMMI has worked on in the past decade and where it's heading next.

Access our top resources on policy and payment

Read a lightly edited excerpt from the interview below and download the episode for the full conversation.

Rachel Woods: So we know we want to focus on what works and one of the things that you acknowledged is that maybe we just tested way too many models. I think you said 50 models were tested, I want to say almost 30 are actually being used right now. I think we could all agree that isn't actually doing any favors for the sanity of providers. So are you planning on turning anything off if we know we need to make an impact, what are you going to be deprioritizing?

Liz Fowler: Well, you have that right. We're currently operating about 28 models, and we know that they create opposing incentives, they lead model participants to have to really think about how to manage model interactions. So we want to increase participation in the models and we want to avoid the types of complexities that lead providers to have to manage shared savings, for example, who gets the savings if you're in more than one model.

We're really thinking about focusing on a set of criteria for models in the future. Thinking about factors like does the model support or advance one or more of our strategic objectives? Driving accountable care, health equity, supporting innovation, addressing affordability, or this idea of partnership.

We want to think about the potential impact, so really thinking about the potential for savings, for quality improvement, these seem kind of obvious, but also the potential for other payers to be partners in the effort, and then also, once the likelihood of being able to scale this model across the country.

So even by non-participants, and I know this is maybe a little inside baseball for inside CMS, but we want to make sure that we're working more closely with other parts of CMS, like the Center for Medicare and the part of CMS that runs Medicaid, so asking them are we asking the right questions, are we going where innovation is needed in those programs, and this might increase the likelihood that the results of the models can be incorporated back into the program.

Woods: So you've got this system of, I'm going to use the term grading. Looking at what we're offering and figuring out does it actually meet all the aspirations of this administration, of this new leadership team, but are there kind of broad, sweeping takes that we can interpret from that? I'm thinking about episodic payments, bundled payments, total cost of care models. When you look at those broader categories, do you have thoughts on which of those models get turned up or down if you know that 28 is too many?

Fowler: So accountable care, the notion of these total cost of care models are going to be really primary and center to the strategy. I think in terms of how we think about the bundled payments and the episodes of care, they're still important, but we want them to be working in coordination or harmonized with total cost of care.

So we're asking ACOs, we're talking to providers and health systems out there like how those programs can be better integrated. So what do ACOs need in terms of managing episodes of care and some of what they're telling us is can you focus on the sort of very high cost, low volume things that primary care isn't as equipped to manage. So don't go into diabetes, we don't need a diabetes episode, but oncology is an area, some of the very high cost, very I guess complex cases, where having a solution would be helpful and complimentary to what they're doing.

Woods: I saw some comments from one of your colleagues, the COO of CMMI, John Blum, and he had said in a conference something along the lines of we're not going to be promoting models that have more risk just for the sake of having more risk, and in that comment, I actually saw a bunch of different interpretations come out, one of which was even in Modern Healthcare, where some were saying, "This means that CMMI is going to actually turn their focus away from downside risk. They're going to double down on coding, on risk adjustment." Now that's obviously an interpretation. How do you want us to think about John's comments there?

Fowler: I think there's a difference of opinion about how important risk is in this sort of ability to transform the health system. There is a difference of opinion externally and there is a difference of opinion internally as well, and in my view, I think risk is an important element, but not everybody's ready to get there. So I think CMMI's role is making sure that there is a place for those that want to go and bear more risk that are ready, willing and able, that are resourced and have the capacity and the tools to do so, and then how can we bring others along who haven't put a toe in that water and might not be ready.

So I think for us, it's thinking about a continuum of options. So maybe not a one size fits all. On the other hand, we have to reduce the complexity, so not 50 options either. So can we get it down to a manageable number of options? But I do think we need to have those options for those providers and systems that want to bear the risk, and then getting those in at the front end as well.

Woods: You mentioned the word scale earlier, and I think that figuring out how to scale these models is actually perhaps even harder than figuring out which kinds of models to focus on, and when it comes to scale, the obvious question is what's going to be mandatory? What are providers going to have to do? Where do you see mandatory models playing into your plan for achieving scale?

Fowler: Yeah, we've been giving this a lot of thought too, and just like the question of risk, I think there's also disagreement about whether we need mandatory models in order to drive success and I think again, in my view, mandatory models do have a place in our portfolio and we should consider them, but they need to be as I mentioned interacting with total cost of care approaches.

So we've got two mandatory models that are moving through the regulatory process, the knee replacement and radiation oncology. Others, I think we're debating and discussing models internally. Some are telling us that if we don't have mandatory models, we're leaving out a whole swath of the health system that will just sit on the sidelines and decide they don't need to make any progress, and I think we don't want that.

On the other hand like I said, we're looking at how we can make sure that these approaches are harmonized and working in conjunction with total cost of care. So maybe that's the long-term and we will need some models that look at mandatory participation in the short run.

This gets to the question of participation too, and making sure that we're incorporating safety net providers, rural providers, others that maybe haven't had the tools and resources and the capacity, and making sure that we're not forcing them into models where they're set up to fail, but we're also bringing them into the fold.

So with the voluntary models, the ones that participate are the ones that are going to be successful, we want to make sure that we're capturing a broader swath but we also want to make sure we can set them up for success.

Woods: So because there's internal debate, and because you have these other goals that you're trying to hit in the backend, is there a sense of the kinds of mandatory models or maybe the kinds of providers that you would want to target for those mandatory models? Is there a thought or a consensus on that?

Fowler: I think those issues are all under debate and it's kind of hard to target certain kinds of models because mandatory tends to look at geographic areas. So within a geographic area, they'll cover providers in that area. And by the way, we also hear from some providers that five years ago, they wouldn't have agreed and they would have opposed mandatory models. Now some of them are even welcoming it.

We've even heard from providers who now that we're moving from voluntary to mandatory are finding themselves outside of the surface area and not part of the geography that's being tested and they're asking us can they come in on a voluntary basis because they liked being part of the model. So I think that's a change. So if you asked if we're making progress, that seems to me to be some progress.

Woods: Well, let me reveal to you what providers tell me. So I've been at Advisory Board for seven years speaking to mostly provider executives and this has been a key topic over those last seven years and if I think back to 2014, there was this kind of certainty that people spoke of when they thought about business model transformation. Risk was coming, value-based care was happening, they were making changes to not just their business model but to kind of the practicalities of operations.

If I'm totally honest with you Liz, seven years later, I almost feel like the folks that I've talked to have slowed down in their thought process. They're still talking about value-based care, but they're talking about it still as this far off ambition. This thing that's going to happen in the future. They're happy to bide their time or hang out at upside only risk, and that's where it comes to how do we push these folks to not just get into alternative payment models but to actually take on meaningful downside risk? Are there other ways maybe beyond the mandatory that you're thinking about to push providers to actually take financial accountability for these models?

Fowler: Yeah Rae, what you're hearing is consistent with what we're hearing and we think that's a problem. Between the pandemic, momentum towards value-based care has really slowed. Part of that is on us. We haven't been really clear about where we're heading—the path we're taking or what's the ultimate destination—and I think that clarity has been missing.

So in the time that I have in this job, really I think my job is twofold—to provide that clarity and lay out that strategy and that future direction and then also help us regain that sense of inevitability. And I think that's really important, and that's why I think this notion of partnership and working with other payers and purchasers and states is really important.

Even if you have a provider that is willing, ready and able and participating in some of our models, they're still missing other payers who are part of that. So they've got one foot in two boats. So even the ready, willing and able partners could use that help from us to help push the other payers in the system to get on board and join this momentum and movement.