Downside Risk Is Latest Development in Movement Toward Value-based Care
In January, the Centers for Medicare & Medicaid Services (CMS) released the latest data on the Medicare Shared Savings Program (MSSP), which is the largest of the CMS’ alternative payment models that utilizes accountable care organizations (ACOs) as the mechanism for delivering coordinated, high-quality care to Medicare patients. MSSP grew by 17 percent in 2017, and the program now includes 561 ACOs and 10.5 million expected beneficiaries. Performance-based payouts have exceeded $2 billion since the program started in 2012. But has the program been successful?
“Results for the early stages of federal efforts to encourage ACOs have been underwhelming,” said Alex Azar, secretary of Health and Human Services, in a recent speech to the Federation of American Hospitals. “They were allowed to share in modest cost savings, but not asked to accept responsibility for cost overruns.”
In other words, there were no consequences if costs exceeded target spending—at least not yet.
The Move Toward Downside Risk
MSSP participants could join one of several tracks. Track 1 doesn’t require providers to assume downside risk. ACOs can belong to this track for two three-year periods, meaning they’re only shielded from downside risk for six years. Tracks 2 and 3 have a two-sided financial risk arrangement—ACOs in these tracks share in savings or are forced to repay losses depending on their performance. Track 3 participants take on the greatest amount of risk, though they’re also eligible to share in the greatest portion of savings if successful. New for this year, ACOs can also join Track 1+, which delivers less downside risk than Tracks 2 and 3.
Because it’s less risky, the vast majority of ACOs in the MSSP are in Track 1. But these latest numbers from CMS show that a shift in accountability is happening: In 2017, 91 percent of all MSSP participants were in Track 1; in 2018, that percentage, while still overwhelming, is down to 82 percent. Meanwhile, the number of organizations taking on downside risk in 2018 increased 140 percent over last year. Four ACOs joined Track 3, while 55 joined Track 1+.
“As of yet, there hasn’t been a seminal event or burning platform that tells providers they have to accept downside risk,” says Michael Wolford, senior manager of DHG Healthcare. “Fast forward a few more years, though, and anyone who bills Medicare Part B, including all physicians and extenders are going to start having some of their compensation tied to downside risk.”
HealthTrust member Franciscan Alliance joined MSSP Track 1 in 2015. As one of the original members of the Pioneer ACO pilot project, the 13-hospital chain based in Mishawaka, Indiana, saw significant improvements in post-acute care costs as well as reductions in readmissions. While Franciscan participated in downside risk as part of the Pioneer initiative, the system eventually opted out in favor of MSSP Track 1.
“Despite our success with the program, we weren’t willing to accept that ongoing risk,” says Al Tomchaney, M.D., MMM, FAAFP, senior vice president, chief medical officer, Franciscan Alliance. “There were issues with the mechanics of the program, and there still are.” Tomchaney says there are challenges around patient attribution—CMS assigns patients to ACOs based on where they receive the majority of their primary care—and access to relevant data.
“Medicare is like driving by looking in your rear-view mirror,” he says. “You’re always looking at old data. Getting real-time information is challenging, and that’s what you need to be successful.”
Keeping patients in the ACO network of providers is also an ongoing challenge. While primary care physicians will likely refer patients to specialists within their ACO, it’s a voluntary network and patients are free to go anywhere to receive care.
“Given our proximity to Chicago, a lot of patients may choose to migrate out of our network,” he says. “They may get their primary care here, but when it comes to the next level of care, patients strongly consider going to those academic medical centers outside of our ACO.”
Franciscan Alliance also participates in ACO activity with two commercial payers. “The commercial experience has been a better one,” he says. “We’ve had great success educating those patients on the model. Even though ACO patients still have the ability to go anywhere, we’ve been better able to reduce cost and reduce their utilization while improving quality measures.”
While downside risk isn’t currently a large part of the Franciscan Alliance ACO experience, Tomchaney says they’re open to adding more downside risk in the future as new models emerge from CMS and other payers.
A Better Bundle
Earlier this year, CMS announced the Bundled Payment Care Initiative (BPCI) Advanced model. A replacement for the BPCI program, which is set to expire this fall, BPCI Advanced allows organizations—hospitals or physician group practices—to take responsibility for coordinating up to 32 episodes of care (including three outpatient episodes of care). Applications were due in March, and the program is slated to start Oct. 1, 2018.
In many respects, BPCI Advanced looks a lot like previous CMS bundled payment models. It covers 90 days post-discharge (or post-outpatient procedure) and is reconciled retrospectively.
Target prices will be provided in advance, and they will be risk-adjusted based on factors like patient case mix, regional benchmarks and historical data. But unlike previous CMS bundled payment models, including the now-scaled back Comprehensive Care for Joint Replacement (CJR) initiative, participating providers will be on the hook for downside risk from day one. Reconciliation payments, both positive and negative, will be limited in years one and two to 10 percent of the target price, before rising to 20 percent in years three through five.
The program qualifies as an Advanced Alternative Payment Model (APM) under the Medicare Access & CHIP Reauthorization Act (MACRA) Quality Payment Program. It shifts Medicare Part B spending to one of two performance-based payment tracks: the Merit-Based Incentive Payment System (MIPS) or the Advanced APM.
For MIPS, performance is measured through data that clinicians report in four areas—quality, improvement activities, advancing care information and cost. A final score in these four areas determines the provider’s adjusted payment.
Providers participating in qualified alternative payment models can select the Advanced APM track, allowing them to bypass MIPS and earn even higher value-based incentive payments. In order to qualify for the Advanced APM track, however, there’s a minimum threshold for revenue at risk—25 percent of payments and 20 percent of patients must be tied to a qualified Advanced APM in 2019. Over time, those qualifying thresholds will increase substantially—to 50 percent of payments and 35 percent of patients in 2021, and 75 percent of payments and 50 percent of patients in 2023.
If providers can’t meet those thresholds, there are lower ones for partial qualification (20 percent of payments and 10 percent of patients in an APM in 2019, for example). In the current performance year, there are nine qualified APMs. (Read more here.)
Test models from the CMS Innovation Center are just that—tests. Take CJR, for example: It was scaled back from a mandatory bundle in 67 markets to a voluntary bundle in half of those. Also in 2017, CMS canceled mandatory hip fracture and cardiac bundled payment models before they even launched.
Broad, bipartisan support of MACRA is really the only signal healthcare organizations need in order to have confidence that a growing portion of their reimbursements will be tied to value, says John Young, M.D., MBA, HealthTrust chief medical officer.
“The quality-based payment structure is not going away,” he adds. “The specifics of what that looks like may change and continue to evolve, but it’s definitely here to stay.”
The Evolution of CINs
In addition to ACOs, physicians can also create and join a clinically integrated network (CIN). Similar to ACOs in their value-based care objectives, CINs are groups of physicians, usually sponsored by hospitals, while ACOs are generally made up of physicians, hospitals and other healthcare providers. CINs often serve as the physician network of an ACO. CINs coordinate care across specialties, while ACOs coordinate care for specific populations across an entire episode of care.
Wolford says it’s hard to know how many CINs exist in the United States—there’s no publicly available data—but the actual number might not be all that relevant. “About a decade ago, CINs were all the rage, and everyone seemingly had to have one,” he says. “Hospitals pushed really hard to stand these networks up but didn’t push with the same vigor to get any contracts in them. There were CINs in form and structure, but there wasn’t enough for participants in these networks to do, which was suboptimal.”
But that doesn’t mean they’re still useless. If the past five years saw a rush to create CINs, the next five years will witness their evolution. “Providers believe there is now value in being in one and even more value in being a part of a good one,” Wolford says. “Over the next five years, the rush will be to consolidate CINs, make them perform and demonstrate their value.”
Preparing for an Unclear Future
The specifics of value-based care payment programs may still be in flux, but there are many steps providers can take today to prepare for future models. Those steps include convening physician leaders, establishing dedicated administrative positions and understanding technology requirements and limitations, including how well the organization can currently track key quality metrics.
A data strategy that addresses interoperability as well as the ability to marry claims data with clinical data is also paramount, Tomchaney says. That’s a work in progress at Franciscan Alliance, which will install an accountable care and population management system module in its electronic health record system this fall.
“Claims data doesn’t paint the complete picture because not all clinical data shows up on a claim,” he says. “This will help us match the two and give us a better understanding of what’s happening across the care continuum.”
The first deadline for BPCI Advanced was March 2018, and the next application period is expected in January 2020. However, Young expects other options from CMS to open up and other payers to test out alternative payment models—even ones involving downside risk.
“Dipping your toe into a voluntary model in a small way isn’t going to make or break your organization,” Young says. “But it most certainly will help prepare you for the inevitability that is value-based care and two-sided risk.”