Oncology Care Model: Looking Ahead to OCM 2.0

Ashley Ridlon, VP of Policy at Evolent Health, shares her insights into ways CMS can build on the benefits and momentum of the OCM and improve the program to help providers succeed in this model.

Episode Transcript

David Schafer: Welcome back to the Specialty Care ValueCast, where we explore the evolving landscape of specialty care management through the lens of value based care, insurance risk, and the patient and provider experience. I'm David Schafer and this is the third and final episode in our series focused on the Oncology Care Model. In this episode we'll be digging into the model design issues, challenges, and possible improvements in the future.

In the first and second podcast we spoke with experts in the model from New Century Health and Cancer Care Specialists of Illinois respectively. Today we'll be digging in more on the model design issues, challenges, and possible improvements in the future. CMMI has indicated that it is working on the next phase of this model and hopes to extend OCM or launch something similar in the future. Which for the sake of this podcast, we'll be calling it OCM 2.0.

My guest today is Ashley Ridlon, Vice President of Health Policy for Evolent Health, a value based care company working with leading healthcare providers and payers to improve healthcare delivery and drive value in Medicare, Medicaid and commercial markets. Ashley can speak not only to OCM 1.0 but the potential future design of OCM 2.0 and what it may or may not look like. Ashley, let's dive right into it. So, in our previous podcast we heard about some of the challenges in the complex OCM program including patient mix and attribution of a particular cancer center to the MEOS, or Monthly Enhanced Oncology Services, structure as well as other challenges in implementing OCM. What are some ways in which the burden can be reduced going forward?

Ashley Ridlon: Sure. Thanks for having me today. I appreciate it. And I'm happy to be here because I think first and foremost New Century Health certainly and Evolent Health, where I work which is the parent company of New Century, and of course the practices that we work with ⁠— and really I would say in the vast majority of the practices in the OCM program ⁠— believe in this program. So we look at this program as absolutely a step forward from where cancer care and payment used to be, which is "do more things regardless of the outcome," to put it really crassly. And that's the problem we're facing in Medicare fee-for-service broadly.

So, you look at this program and it's leaps and bounds beyond where we have been historically and in cancer care in this country. So, you see a lot of care transformation at the practice level. I think you've heard about that maybe in previous podcasts where patients now have more access to their providers. They are getting guidance that helps keep them out of the emergency room, that helps keep them from having an inpatient stay, more care coordination than they had before. So, that is all terrific. In addition, you have at least some practices taking a really hard look at their prescribing behavior and drug spend. That's one of the things I know that New Century has really focused on, and is looking at the comparative value ⁠— both clinical value and cost effectiveness ⁠— of these often very expensive drugs that are prescribed in cancer care. So, those are all really great features of the program.

You mentioned burden, David, and I think this is something that we hear a lot as we talk to practices, that the OCM program is very complex and there are a lot of requirements. So it is very prescriptive, and maybe that was done for a reason. It was, I think, hard to get from where we were in cancer care to where we are now. But could the prescriptiveness be reduced? Could there be a little bit more room for innovation? For example, there's this IOM 13-point care plan which has a lot of really great elements in it but there are a couple of elements that, when you look at it on a case-by-case basis, may not make as much sense.

So, one example is you have to provide cost analysis for the patient. So you're trying to give them a sense of what their costs are going to be in this episode of cancer treatment. Anybody who knows healthcare knows just how hard that is. It depends on your insurance company, it depends on what your out-of-pocket exposure is, it depends on what might change during your course of treatment. So, you have practices that are providing an estimate and this is well intended because financial burden is huge in cancer treatment. But it may be completely removed from what actually the cost of the treatment will be. And who bears that cost? The Medicare program might bear some of that cost. The patient might bear some of that cost. If they have Medigap, that might cover some of their costs.

So, the analysis that they're giving to patients is really not meaningful to them. Maybe there's an alternative to that to really get at the goal of helping patients realize what their financial exposure might be and practices are certainly investing in financial navigators. That is a huge benefit. So, you have a financial navigator that can look at that specific patient's insurance status, Medicare Advantage or Medicare fee-for-service, and help navigate them through the process and help them know what their out-of-pocket cost will actually be.

It's things like that where there is a very low-risk cancer, stage zero or one, that can be treated and cured very efficiently. And yet the patient has to focus on advanced care planning because that's a requirement. Or on the flip side of that, you might have someone in very late stages of cancer that should be looking at advanced care planning and yet you have to look at a survivorship plan. So, could there be a little bit more flexibility where the practice gets to cater these requirements? Maybe there's some options or some flexibility to the patient's exact needs. I think that's definitely a piece of this.

I would say another really big part of this is the structure of the pricing and in how that's calculated in the formula. And that's kind of a whole different topic, but it is difficult to understand. I mean, practices are hiring consultants just to figure out how do I know how I'm going to do in advance, how do I make projections, and then how do I know what works? And that's all very complex as well.

David Schafer: Thank you Ashley. And I think you left off on a really interesting point discussing the pricing model and I'd love to talk a little bit about that now. So, you mentioned that given the high cost of cancer treatments, especially around chemotherapy drugs, correctly establishing the benchmarks and targets that practices need to try to beat is one of the biggest challenges of the program. It's very hard obviously for practices to reduce their costs below the CMS established targets and as a result succeed in the program. This means that CMS needs to correctly determine what the costs should be absent intervention and then practices must abide by that and try to get below that target price. And so, can you talk a little bit about how this is done today and some of the subsequent challenges with the pricing model?

Ashley Ridlon: Sure. Great question. This is one of the biggest challenges in the OCM model. So, it's imperative in any of these CMMI models ⁠— and I used to work at CMMI, I've been on that side of the table ⁠— you have to be able to establish and assess how you did against a benchmark. What would the costs have been absent what your intervention is? And ultimately for CMMI to decide whether or not this model worked and thus whether it can be expanded, it has to have a benchmark. That makes complete sense. Benchmarks in this program are based on your historical spending, they're trended forward, there's a geographic adjustment and they are intended to get at what the cost of cancer treatment would have been.

The challenge is if you look at different cancer types and especially even within a cancer type ⁠— subtypes of cancer ⁠— so for instance a breast cancer HR2 positive. That might be a very different cost, and not just the cancer subtype but the progression of cancer: whether it is localized cancer or whether it is metastatic cancer, it spread to other parts of the body. The costs for metastatic cancer are going to be higher. The costs for certain subtypes are going to be higher. So why aren't we looking at pricing in a little bit more of a fine-tuned way? So, that's one question we've been asking. I know my colleagues over at New Century Health have been doing some analysis on that front.

I think there are probably multiple ways to get at this challenge. And certainly I think the CMMI program team acknowledge that it's challenging. It's also challenging because you have expensive drugs and you have sometimes new drugs that come onto the market. There's a novel drug adjustment when a new drug does come onto the market. But you know, if you think about different practices and what their case mix is, you might have very different outcomes in the OCM program if you have a very high percentage, say, of a very expensive cancer that might be underpriced when the benchmarks are established.

So again and just as a review, the benchmarks are based on historical spending trended forward. You have some adjustments that are made and then CMS takes a discount. So, that is 4% below the benchmark for the one sided risk model and it's 2.5% below the benchmark for the two sided model. What that means is if you get below the benchmark you've saved Medicare money in effect, but CMS keeps that discount. So, they keep the 4% below the benchmark or they keep the 2.5%. That's your target price. That's what you really have to beat in order to get your population based payment.

So the bar could be very high if the pricing of the benchmark and the targets are actually under pricing what the actual cost of cancer would be in absence of your intervention. I think that's the biggest concern. If benchmarks and targets are underpriced then no wonder only half, or maybe a little over half, of the OCM practices have ever received their population based payment. It should be higher than that. Another thing that we hear across dozens of practices we've talked to is maybe they got a performance based payment and they feel like they're doing really well in the program but they have absolutely no idea why they got the payment, the reconciliation.

So, CMS looks back at the episode and they determine whether or not you achieved your population based payment. But then they also reconcile that multiple times to account for claims that come in later. So, sometimes that even reduces your payment later on. So one of the challenges is it takes a long time for you to actually learn how you did in the program. Here you are, you're treating a patient in the six month episode of care and then you don't know how you actually did with that patient in terms of your cost and quality metrics until months down the road. I mean it could be nine, 12, 18 months down the road that you actually learn how you did. And so, that is one of the really big challenges in the program.

As I mentioned, you might have gotten a PBP and you're feeling great. You have no idea how you did it. On the flip side, you might have not achieved your performance based payment and then unfortunately you don't know what you should change to try and do better the next time. So, it makes improvement and rapid cycle improvement in particular really challenging. CMS has tried to address this by providing really detailed quarterly feedback reports. That's certainly helpful. But I think the time to financial reconciliation could certainly be shortened. Maybe that could be an improvement project of CMS. I think that that would really go a long way.

David Schafer: Thanks Ashley. And let's follow up with one of the concepts you talked about and that's the performance based payment. So, practices that achieve a savings below the target price so below that 4% discount in the one sided, upside only model or the 2.5% discount in the two sided both upside and downside model will have the ability to recoup savings. And you mentioned that not all practices are achieving a PBP. And so, at the end of this year OCM practices that have not received a performance based payment or that PBP will have a choice to make. They can either move into two sided risk or exit the program altogether. Both Evolent and New Century Health believe in the power of two sided or full risk models to truly drive value in our healthcare sector. Do you think the OCM practices are there yet? What is needed to ensure more practices can and will enter the two sided model and also succeed in it?

Ashley Ridlon: I think fall is when practices really have to make a decision. Some of them, of course, they've achieved their population based payment. They have a choice. They can stay on one side or two sided. The ones who haven't, which is a little under half, they've never gotten PBP. They will have to make a choice, either exit the program or move into two sided risk. Are they ready? I don't think we really know yet. I can just tell you anecdotally having had conversations with a number of practices, most of them say, "We're not sure yet. We're still assessing. We're still looking into this." They're looking at their volatility over the course of the program. They're looking at, "Are we willing to take some downside risk even though the potential upside might even be more beneficial?"

Of course the discount is lower in the upside model. But we want to have some level of confidence to go into the two sided model that we can really put some skin in the game because we have confidence that we're going to be able to save money. That is just a high bar to to overcome.

I've talked to one practice that was very sure that they wanted to go into two sided risk. They're doing well, they're getting PBPs. They are looking at maybe going in in January of 2020 and that's great. I know New Century Health is working with practices to conduct a volatility assessment to determine what is their outlook for two sided risk. And New Century has a risk-sharing model so they can come in and share some of the upside and downside risk and that might make more practices feel comfortable joining the two sided model.

But really — and this is true of ACOs, bundled payment, whatever the model is — you need to have some ability to predict your future performance based on your prior performance and your patient mix and just your unique circumstances. How many physicians are in the practice? What's our geography? What is our patient mix? And be able to look out and project that maybe we'll win some, we'll lose some, but on net we think we'll come out ahead. And that's going to be really important. You know you're doing all of the care transformation activities that are driving down unnecessary readmissions to the hospital or unnecessary inpatient admissions. That in many cases can be low hanging fruit. You can do a lot of those activities and drive down those really high cost care outcomes that you want to avoid when you can.

And then what? So I think the question is, and sometimes it's sort of the elephant in the room, what do we do about drug spending? And I've heard a number of practices say, "Well it's out of our hands. We can't really control drug spend." But we've also seen some success among practices that have taken a hard look at that. They're using really robust clinical pathways, clinical decision support to help physicians know at the point of care, "How do we treat this patient? What's really going to be the highest quality and the most cost effective care in this unique circumstance?" Data is really, really powerful. And so, I think the more the practices are equipped with data to be able to make those assessments the more comfortable they're going to be in moving into a two sided risk model.

Ultimately, I hope practices will look seriously at two sided risks because this is the direction the Innovation Center is moving. You have the CMMI Director Adam Boehler out there saying, "I want to blow up fee-for-service", and really moving aggressively away from fee-for-service into some amount of risk: two sided risk and even up to full capitation. So this is why I think you don't want to experiment with it if you're going to completely lose your shirt. And I think you should do your due diligence to make sure that's not the case. But once you do have some level of comfort in two sided risk, I think it is worth considering because I really do believe this is the way of the future in Medicare that you're going to see more and more providers pushed into risk whether that's on a mandatory basis or whether that's on a very broad voluntary basis. I think it's going to be more the norm in the future.

David Schafer: Thank you, Ashley. This is all really comprehensive and informative to our listeners. Just if you could add one final thing or if there are any other improvements you think CMS can make as it is contemplating the next phase for OCM, now is your time. What are you thinking?

Ashley Ridlon: Well, I don't know if I can narrow it down to one but let me see what I can do. So, we talked a little bit about being a little less prescriptive in the care model that could be a benefit. We talked a little bit about fine-tuning the pricing model, just really making sure it's accurately pricing episodes. Our proposed solution for that would be to cater the pricing model to cancer subtype and progression, metastatic or localized. We think that's a very simple way to do it and it could actually simplify this very complex formula, math that I can't even begin to understand. I'm really glad we have people on our team who can understand it and translate it but it's become such a complex model that I think there's some ways we can really simplify the math and make it understandable clinically to practices. I think that's really important. Just really making sure the pricing is right so that the targets are really accurate, precise, and achievable.

And then, I think optimizing the two sided risk model. The two sided risk model should really look attractive to practices. Yes, you've got skin in the game. And I've got to really hand it to CMMI; they made some changes to the two sided model so that it is more attractive to practices in the sense that the downside risk is pretty manageable. It's an asymmetric model where the upside is higher than the downside. So, you have some skin in the game, but again you are protected to some extent. There are caps on the upside and the downside. But that is something I think CMMI was really smart to build into the two sided model. But I would look at, "Are there ways to optimize that even further?" I think that's another area.

And then lastly, something we didn't really get into but it's something that we're exploring, is the current OCM model has an option for practices to pool together. So multiple practices come together, they pool for the purposes of establishing their benchmark and targets and, presumably, actually pooling their risk across practices. But nobody's taken this option up and I have been curious about that. Could multiple practices effectively pool together? And I look at a similar program in CMMI, it's the Bundled Payment for Care Improvements Model Advanced (BPCI-A). They have a model where individual provider groups can apply directly and do bundled payment. They've got multiple types of bundles. They also have an approach where a convener entity is the applicant to CMS and the convener applies and then they go out and they aggregate multiple provider groups together under one convener umbrella.

I can see a lot of value in that model for the OCM program. So, why not just translate from BPCI Advanced the convener approach into the OCM model? It would be completely optional, but you allow a convener to apply and to pool practices together, aggregate their risks. If it makes sense you could allow the funds to flow through the convener such that the convener can have risk share arrangements with the practices that might look somewhat different than the PBP structure. So, it could almost look a little bit more like full risk. You could do the upside and the downside in different ways, and leave it open to the convener in the practices to work out the detail in their program agreements. But that can be a really interesting model. It could help streamline some of the operations of the model so you're providing a little bit more administrative simplification and perhaps even cost savings on the administrative side.

We see practices out there that are hiring multiple vendors and that can be costly. So, a convener could be a one-stop shop for the practices and help them save on the administrative side, as well having access to the data across the practices. There could be a lot of shared learning, shared data. And then that convener would have the relationship with CMMI as a program participant. So, they're actually in the room — or more likely on the webinar, on the phone — during all of the feedback calls and the program calls, so they're really embedded in the program.

I think that would be a very easy lift for CMMI. It's the same group that runs BPCIA and OCM. So, that could be really interesting. And there are probably lots of other issues and ideas out there on how to improve the OCM model. So, what I would say is I hope that the practices that are participating, the practices interested in maybe participating in the future will come together with their national organizations and work with CMMI to really make this a great model. I think there's a lot of really wonderful features in the OCM model today. We don't have to start from scratch. I think it's really building on what works, simplifying it, getting the pricing right, really optimizing and pushing practices over time into two sided risk. This could be a really effective model and could really save Medicare a lot of money and do it while helping improve cancer care and doing it in a way that practices can survive and thrive. So, we're excited about the future. There are challenges to be sure but we think they can be worked out.

David Schafer: Thank you Ashley. We're very lucky here at Evolent and NCH to have a policy guru such as yourself. And to our listeners thank you so much for tuning in to learn more about the Oncology Care Model. We hope you have many takeaways from this podcast, and for more information you can visit NewCenturyHealth.com to learn a little bit more about the Oncology Care Model. Thank you so much.

The Specialty Care ValueCast is produced by New Century Health, a specialty care management company focused on oncology and cardiology with nearly two decades of experience improving health outcomes and reducing costs for provider and payer organizations.

Previous Video
Positioning Providers to Take on Risk
Positioning Providers to Take on Risk

As the market moves toward value-based care, one physician describes how New Century Health gave his practi...

Next Video
Taking the Pain Out of Prior Authorization
Taking the Pain Out of Prior Authorization

Prior authorization is a burden for most practices. Two partner physicians share how New Century Health's a...