The opening day of this year’s Fall Conference of the Washington, D.C.-based NAACOS—the National Association of ACOs—began on Sep. 22 with introductory remarks by NAACOS president and CEO Clif Gaus, Sc.D., and then an appearance by Brad Smith, Director of CMMI—the Center for Medicare & Medicaid Innovation, a division of the federal Centers for Medicare and Medicaid Services (CMS), and a roundtable discussion involving Smith and two NAACOS board members. As with the majority of such events during the COVID-19 pandemic, NAACO’s annual meeting is taking place this week on a virtual basis.
And even though the U.S. healthcare delivery system has been hit hard by the COVID-19 pandemic, NAACOS leaders celebrated the fact that the accountable care organizations operating nationwide have shown major progress in creating value in the healthcare system, despite the challenges of the pandemic. Indeed, Gaus and the two board members participating in the opening session—Emily Brower of the 93-hospital, Livonia, Mich.-based Trinity Health, and Travis Broome of the Bethesda, Md.-based Aledade, a consulting firm that helps physician groups create ACOs—noted that NAACOS-member ACOs have performed exceptionally well in the Medicare Shared Savings Program (MSSP) and other programs. At present, 332 ACOs participating in the MSSP are NAACOS members, while 209 are not; and Gaus, in his opening remarks, noted that NAACOS-member ACOs brought in $2.65 of the total 2019 participation-year savings, versus $1.19 billion brought in by non-NAACOS-member ACOs in that program last year. What’s more, Gaus noted, in terms of contribution to the CMS Trust Fund in participation year 2019, NAACOS-member ACOs brought in $845 million, or 71 percent, of that savings, while non-NAACOS member ACOs contributed $347 million, or 29 percent, of the total, to the CMS Trust Fund. In his brief opening remarks, Gaus told NAACOS member leaders, “I can’t thank you enough” for all their contributions to savings under the MSSP and the Next Generation ACO Programs.
Then Gaus and his team welcomed Brad Smith, CMMI Director and CMS Deputy Administrator and Director, to address the virtual audience. After thanking audience members for participating in making change in U.S. healthcare, Smith said that, “Beginning in January, I began looking at all the [advanced payment] models—we have about 54 currently. In March, I ended up getting pulled into the COVID effort, and ended up being involved in efforts around testing across the country,” he said, which took up considerable time and effort for several months and are ongoing.” Nevertheless, work is proceeding apace on all the current advanced payment models (APMs) and on developing new ones, he assured the audience. Key areas of focus? Creating more “flexibilities” that will encourage providers of all kinds to participate in APMs, and “additional opportunities to participate.”
A summary of the basic facts involved in CMMI’s various Innovation Models can be found here.
Among the programs that Smith highlighted, he stated that “We are very focused in moving forward with the kidney care and radiation oncology models, the direct contracting in kidney care; we’re going to continue to move forward with all the models we’ve announced. Six or eight weeks ago, we announced a new rural care model.” Among the things that he and his colleagues have learned through their comprehensive review of the 54 current APMs, Smith said, is that “We learned, first, we had great participation in CMMI models, and much of that was from your members.” And, referencing the ACO Investment Model (AIM) specifically, he said that, “When it comes to our AIM model, with the upfront payment—we used the learnings from that to help inform the CHART tracks” in the new Community health Access and Rural Transformation (CHART) Model, first publicly announced on Aug. 18.
“We also saw some very important improvements in quality—in the kidney care model, decreases in hospitalizations,” Smith continued. “We’ve seen decreases in emergency department visits.” And, he said, “We’ve seen a number of models with financial success, and a number with really good clinical outcomes, such as kidney care.”
What’s been learned so far? “There have been a number of lessons,” Smith said, “around the importance of setting benchmarks. It’s important to set benchmarks over time. And whether it’s in the oncology models or bundled models, in many cases, the savings came from places different from those we had expected. Early on, we weren’t able to fully evaluate the [value of the] benchmarks. The thing that’s most important to us when we set benchmarks, is that they’re fair and accurate. So with direct contracting, if you’ve had three years of claims to look at—when we started with the MA [Medicare Advantage] rate book, it was reasonably close.” Discussing a number of complexities and nuances, such as calculating the element around the cost of graduate medical education, he said, requires a level of rigor, and the need to make certain adjustments. “We made a few adjustments to the MA Rate Book. And with those adjustments, we were able to accurately predict the future costs. I think that that will mean—we were very optimistic with the way we set benchmarks for direct contracting. I believe they’ll be proven out over time. And if we’re able to accurately predict savings, we’ll be able to extend Direct Contracting over time. We’re also making some changes around risk adjustment.”
Smith also broke a bit of news, stating that “We do have a number of new models coming down the pike; one will be around [Medicare/Medicaid] dual-eligibles. We believe that having coordinated care for dual-eligibles is extremely important. We’re also talking about a geographic version of direct contracting, where providers could take on all of or a portion of Medicare beneficiaries in a particular area.”
Insights gained from MSSP participation
Smith then participated with Gaus and two NAACOS board members in a panel discussion. The two board members were Emily Brower, senior vice president, clinical integration and physician services, at Trinity Health in Livonia, Mich.; and Travis Broome, senior vice president of policy and economics at Aledade, Bethesda, Md.
“The 2019 success is awesome and to be celebrated,” Brower said. “The other insight that many of us had is that the muscle that we built to be successful in the new models helped us to be successful in COVID work. Such as population health risk assessment, identification of patients at risk, and care management. And I think that the experience we’re having this year is the recognition that FFS models aren’t super-resilient, not that we’d expect to have a once-in-a-century pandemic happen regularly. But what people are yearning for is not just the benchmark and its accuracy, but the fact that it’s prospective in nature and population-based. Could you tell us a little bit more, now that you know who will be participating this year, tell us what that landscape looks like?” she asked.
“We are thrilled about the level of applications and participation in Direct Contracting,” Smith replied. “Over 100 folks applied to participate. Half are standard DC intermediaries with experience. About 25 percent are new entrants, folks who have served folks primarily in MA [Medicare Advantage]. And that was one of the key goals, to bring them into Direct Contracting. And about 25 percent are high-needs direct contractors. About 60 folks are in the implementation period, during which you can align beneficiaries with your program. And a lot of folks will launch next year. With COVID, we’re giving everyone more flexibility; we’ll have a second opportunity for folks to join in 2022. But we’re thrilled with the participation we’ve seen for 2021.”
“You mentioned the geographic model,” Brower followed up. “When you look over that mix of folks who are going to be participating in Direct Contracting, I know the benchmarking papers are pretty fresh, but what are you thinking about that?”
“We did an RFI [a request for information] on the geographic model about a year, year and a half ago, and got some feedback,” Smith replied. “But whether it’s MSSP or other models, our core belief is that the more risk we can allow people to take on, with the more flexibility they can have, the more successful they’ll be. We’re creating a lot of flexibilities around care management and benefits, and our hope is that through that kind of mechanism, we’ll see lower cost and better quality. What will be unique about this is that today, many organizations are working based on prior relationships with beneficiaries, but that will change with the geographic model. The folks who are going to most benefit are not necessarily those who are already seeing their doctor two or three times a year, so that will be very important.”
“The pandemic certainly exposed the fragile nature of fee-for-service” operations, Broome said. “What does everything look like five years into the future?”
“I totally agree with what you and Emily have said about [revelations around the creation of] value in the pandemic,” Smith replied. “Not only in CMS, but with private payers, this move to value will only go faster. And at some point it becomes really important that we hit a tipping point and move it from a kind of a side project to the core way that we reimburse. The geographic model will hopefully move large numbers of beneficiaries into value-based care. So in the next five to ten years at CMMI, if we’re successful, we’ll move value-based care from something like 8-10 percent of an organization’s revenue, to 80 percent or more of their revenue” for a large percentage of patient care organizations. “So we’ll be focusing on things that we have a high likelihood of scaling up. And the place where your members in particular are going to have an advantage is that, because they’ve been in MSSP and Next Gen, they’ll be ready and prepared to work on that type of risk level.”
“The way that ACOs were able to super-fast deploy virtual care” was remarkable this spring, Brower testified. “I know Travis did, we did. So there’s this whole new set of capabilities, and it feels very exciting and also very responsive, in terms of convenience, access. How are you thinking of incorporating those experiences into the shorter-term or longer-term future?”
“Obviously, we increased the flexibilities around telehealth tremendously during the pandemic,” Smith said. “So what parts will become permanent parts of reimbursement? A number of flexibilities will stay in place. And in Next Gen ACO and DC, we already have a number of those flexibilities, including originating site and home-based; and our hope, esp. as we get more folks engaged, is to add even more flexibility around site of care. We’re looking to become even more flexible around that over time.”
Working with dual-eligible populations in value-based care models
“You mentioned duals, and duals have been an area of success in MSSP in particular,” Broome said. “What are your thoughts on what’s next for the dual-eligible population?”
“We think duals are a huge opportunity for value-based care,” Smith said. “They’re some of the most expensive beneficiaries in the country, and their care isn’t always coordinated. We can tremendously improve their quality and cost of care. In one of our earlier versions of Direct Contracting was the idea of duals integration, and we’re reviving that. So for either plans and MCOs [managed care organizations] or providers serving them, we could create for them the option of managing them on the Medicare side.” There is great opportunity, he said, for dual-eligible members “to stay at home and improve their quality of life and lower cost.”
What’s working
“Based on your review of the 54 models,” Brower said, “we’d love to hear a bit more about the most promising practices. What did you see in there that we should know and grab? That could help us as an audience focus and drive change.”
“Among successful ones,” Smith said of APMs that have shown greater promise, “we saw a really clear thesis about where the quality and/or cost improvements might come from. The Maryland total cost model is essential a global budget. Our belief is that organizations could work withing a global budget, could decrease hospitalizations and outpatient usage. We thought that if we capped the budget, hospitals could really move to control costs. And the home-based model—we have a number of quality metrics, and we see those quality metrics around keeping people in the home as reducing hospitalization. We saw a $650 million reduction in costs in the model from that.”
Meanwhile, Smith said, “When I think about the middle bucket of models, where we’ve seen a lot of transformation, but not the level of quality or cost improvement, we have seen decreases in hospitalizations and other areas. The challenge is how we structure the models. Both in the case of bundles and the oncology model, instead of our being able to scale those models nationally, we’re having to tweak those models and do more tests on them. We want to make sure that cost savings are sustainable, and that we don’t have to keep revising the models all the time—we want something sustainable that people can count on. So getting the structure of these models, getting the benchmarks right, is an imperative. I believe it’s going to be very hard to move the majority of payments into value-based, if everything has to change every three to five years. We’re really focused on scaling models so they can be used in the regular program”—meaning the Medicare fee-for-service program.
