Health Policy Vitals

Eight key insights on this week's ACO participation announcement

By Piper Su, Eric Cragun, and Adam Lustig

Earlier this week, the Centers for Medicare and Medicaid Services (CMS) announced the 2016 ACO cohorts, including new and renewing ACOs for the Medicare Shared Savings Program (MSSP) and ACOs joining the Next Generation ACO Model. Overall, the announcement signals continued progress towards Secretary Burwell’s goal of shifting 30% of traditional Medicare payments to alternative payment models by the end of this year.

Here are our eight quick takeaways on what the information reveals about the evolution of ACO participation.

1. Providers have sustained and growing interest in the ACO model and are moving towards two-sided risk

With 147 renewing ACOs and 100 new ACOs in the 2016 MSSP cohort, the number of MSSP ACOs grew to 434. In addition, 21 ACOs joined the new Next Generation program. The 100 new MSSP ACOs this year is up from last year’s 89 new ACOs and likely reflects providers' expectation that the transition to value-based payments is becoming more inevitable in both public and private markets. Moreover, CMS reported that the number of beneficiaries served by ACOs in Medicare’s programs has grown from about 5.9 million in 2014 to 8.9 million at the beginning of 2016. The growth of participation in two-sided risk models (22 ACOs in Track 2 or 3 of MSSP and 21 ACOs in Next Generation) further evidences the growing move toward risk-based payments.

2. ACO participation is growing in organizational diversity as well as numbers

The 2016 cohort includes some large, high-profile organizations joining for the first time (e.g., Henry Ford in Michigan and North Shore-LIJ in New York), but also includes 39 ACOs in rural and underserved areas joining under the ACO Investment Model (AIM). This group is larger than the groups that joined the similar Advanced Payment Model in 2012 (20 ACOs) and 2013 (15 ACOs). While the larger size of the 2016 AIM cohort may be due partly to the lack of a similar option in 2014 or 2015, it is notable that there is an increase in small and rural providers joining the model. In addition, MSSP added its first ACO led by a post-acute care provider, with Genesis Healthcare jumping in to the program.

3. Participation decisions seem only loosely correlated with past ACO performance

As might be expected, nearly all of the top performing ACOs renewed participation for 2016. But even among ACOs that struggled the most relative to their benchmark in 2014 (the latest year for which we have performance data), renewal rates were surprisingly healthy. Similarly, ACOs joining Tracks 2 or 3 and Next Generation had a wide range of previous performance in ACO programs, indicating that they may see greater opportunity from different program attributes in terms of benchmarks, attribution, waivers and financial risk.

4. Opening up Track 1 to a second performance period was likely key to retaining participants

One of the biggest elements of last year’s final rule for the MSSP program was the decision to allow MSSP participants to stay in Track 1 for another agreement period, rather than requiring them to move to two-sided risk models. While this may slow the immediate progression along the risk spectrum for some organizations, it also likely had the desired effect of retaining a large number of program participants. It’s not clear how many organizations would renew absent that change but it seems unlikely that all 147 renewing ACOs would have shifted to two-sided risk.

5. ACOs opting to sit out in 2016 will have much to consider in the coming year

While the program changes that CMS made last year were enough to retain a majority of ACOs, there were a number of organizations that will not continue participation in 2016. This drop could be attributed to any number of factors, including administrative issues, frustration with the program or uncertainty around forthcoming regulatory guidance. There are a number of decisions that will be made in 2016 that may impact provider choices around future participation in a Medicare ACO, including updates to the financial benchmarking methodology in MSSP and regulatory guidance for implementation of the new MACRA-mandated alternative payment model (APM) track.

Rules around which MSSP tracks qualify as APMs for bonus purposes and determination of the performance year for APM participation will have implications for ACOs’ participation decisions. If CMS decides to only count participation in two-sided ACO models toward the APM requirement, ACOs that did join in the 2016 cohort might end up clamoring later for the option to switch tracks so that they can earn APM bonuses in 2019. Likewise, organizations that decide to seek greater risk in their commercial contracts may also alter their Medicare ACO strategy in future years.

6. Broader strategic considerations loom large in participation choices

CMS created Track 3—with elements like prospective attribution and a higher sharing rate—and adjusted elements of Track 2—for example, introducing prospective attribution—in the hopes of swaying more ACOs to take on downside financial risk. While the number in these tracks did jump significantly (up to 22 from just 3), more than 90 percent of ACOs still opted for Track 1. Moreover, the fact that historic Medicare performance did not strictly correlate with participation decisions in 2016 signals that ACOs are making participation decisions based more on broader delivery strategy and risk tolerance than on individual, programmatic elements. Ultimately, Medicare’s success in persuading more ACOs to shift to two-sided risk may depend more on multi-payer alignment of payment incentives than Medicare program adjustments.

7. Attrition may be inherent in meaningful transformation efforts

More than 60 ACOs from the first three MSSP cohorts chose not to renew participation for 2016. And based on the figures CMS released, another 23 from more recent cohorts seem to have quietly left the program before the end of their performance periods. In addition, just nine ACOs remain in the Pioneer ACO program, which started with 32. But program attrition may not, in and of itself, represent a programmatic failing.  Rather, we should expect some attrition in programs designed to drive dramatic changes in payment and delivery. As CMS’ Patrick Conway noted in a recent interview with us, testing the limits of innovation means that there are some programs that may not fare well in the long term.  It is debatable, however, how much attrition is natural and how much is related to legitimate program shortcomings.

8. The information provided this week is helpful but more is needed to fully understand program trends

CMS' announcement gives us useful visibility into the health of the ACO program. Still, the lack of specific information, such as the specific number or names of providers in Track 2 or Track 3, limits our ability to do a full analysis. Similarly, information about the number of beneficiaries served by each ACO would be helpful to better understand the geographic spread of ACOs and how ACO size is factoring into participation options.

We will be watching in future months for additional data on ACO program participation and performance.