On August 8, 2018, the Centers for Medicare and Medicaid Services (CMS) released a proposed rule that will significantly change the Medicare Shared Savings Program (MSSP) if enacted. This paper is the sixth in a series of Milliman white papers on the proposed rule.
The MSSP proposed rule has been met with a wide range of reactions. Some stakeholders have suggested that this will be the beginning of the end of the MSSP.1 Others have praised CMS for making changes they believe move the program in the right direction.2 One thing that is clear is that all accountable care organizations (ACOs) will be affected. There will be some winners and some losers, and the impact of the rule change will vary by ACO depending on its current situation and unique characteristics. Given this variability, we have examined the rule from the perspective of different ACO situations.
Other papers in this white paper series identified the following key characteristics that will determine an ACO’s risk-sharing parameters and financial benchmarks under the proposed rule:
- Participant revenue
- Costs relative to its region
- Prior experience within the MSSP3
In this paper, we explore these characteristics plus a few others. While this is not an exhaustive compilation of all ACO characteristics relevant to the proposed rule, this report highlights the primary considerations that affect most ACOs. We used these characteristics to identify what might be considered “winners” and “losers.” We are defining ACOs as “winners” if the proposed rule provides some type of benefit to the ACO compared to the current rule and vice versa for “losers.” The winners’ benefit could be a more favorable benchmark, lower risk exposure, or additional options that were not previously available. Our categorization of ACOs as winners or losers is meant to be generally applicable—there will certainly be individual ACOs within some of our “winner” cohorts that do not benefit, and vice versa for the “loser” cohorts.
Winners: Low-revenue ACOs
ACOs that are comprised primarily of physicians (i.e., those without hospital participants) are likely to be classified as “low-revenue” ACOs under the proposed rule. In general, low-revenue ACOs will have a smaller (or similar) amount of risk exposure as they have under the current rule, while gaining more flexibility and choices. Compared to the current rule, these ACOs will have the ability to maintain relatively low levels of risk for a longer periods of time (up to an additional 10½ years) before moving to the ENHANCED Track.4 Additionally, they will have a lower loss sharing limit due to the revenue-based loss-sharing in Levels C, D, and E of the BASIC Track, relative to the current Track 2 or Track 3.
Figure 1 illustrates the loss-sharing limits for a hypothetical ACO under various levels of the BASIC Track. In levels C, D, and E (the three levels with downside risk), the loss-sharing limits for low-revenue ACOs are based on a percentage of participants’ total Part A and Part B revenue during the year. Conversely, under the current rule the loss sharing limit is a function of the benchmark. We expect the revenue-based limits to be considerably lower than the current benchmark-based limits.
Figure 1: Comparison of Loss-Sharing Limits for Low-Revenue ACO under Proposed Rule and Current Rule
|Total Part A and Part B Revenue for ACO Participants
|Total Benchmark Expenditures
|Loss sharing limit
1Under the current rule, Track 3 uses prospective assignment and Track 2 uses retrospective assignment. Therefore, the total benchmark expenditures for a given ACO would not necessarily be the same under Track 2 and Track 3 but we have made this simplifying assumption for the example.
2Loss sharing limits under Track 2 increase in each performance year.
Level E of the BASIC track is nearly identical to Track 1+, and therefore low-revenue ACOs with previous experience in Track 1+ will not see any impact to their loss-sharing limits, although they can remain at this level for longer under the proposed rule. However, low-revenue ACOs that had been planning to move to downside risk soon have the option of a more gradual transition to risk (in Levels C and D) under the proposed rule than they have under the existing rule.
Losers: ACOs beginning Track 1 in 2016, 2017, 2018, or hoping to start Track 1 in 2019
ACOs that entered their first agreement periods under MSSP Track 1 in 2016 or more recently, as well as those planning to enter the MSSP in 2019, will have less time and lower gain-sharing in upside-only risk tracks. Under the current rule, ACOs can be in Track 1 up to six years and an ACO’s participants can be in Track 1 even longer if the participants switch ACOs. Furthermore, the proposed upside-only risk track will share in 25% of gross savings rather than 50% under the current Track 1. Figure 2 illustrates the shared savings from 2019 to 2021 for a new 10,000-life ACO under the current and proposed rules. For this illustration, we have assumed the ACO achieved 5% gross savings each year and progressed as slowly as possible through the BASIC track glide path. Under this illustrative example, total shared savings in these three years would be nearly 60% lower under the proposed rule (50% due to lower gain-sharing rates each year and 10% due to the delay in the 2019 program start date, from January 1 to July 1).
Figure 2: Comparison of savings for an illustrative ACO joining the MSSP in 2019
||Average assigned beneficiaries
||Benchmark expenditures PBPY
||Gross savings percentage
||Gross savings (current rule)
||Gross savings (proposed rule)
||Shared savings (current rule)
||Shared savings (proposed rule)
||Impact on Shared Savings (proposed rule - current rule)
There are some mitigating factors for this cohort. ACOs currently in Track 1 that were already planning to move to downside risk will not see a major negative impact, and ACOs starting an agreement period in 2017 or 2018 can finish their current agreement period before starting the BASIC or ENHANCED track. One change that may be beneficial for some ACOs starting in 2019 is the regional benchmark adjustment begins in their first agreement period rather than the second agreement period under the current rule. As mentioned by CMS in the proposed rule, 80% of ACOs had a positive regional benchmark adjustment in 2017.
Winners: ACOs with high market share
Under the current rule, ACOs that comprise a large share of their markets (including many rural ACOs) have a very difficult time generating savings after the first agreement period because the regional trends for these ACOs are largely driven by the ACOs’ own experience.
The proposed rule addresses this issue by introducing a blended national-regional trend. For ACOs with high market share, the blended national-regional trend will be heavily weighted toward the national trend. This mitigates some, but not all, of the risk of an ACO lowering its own future financial benchmark through significant cost reductions in its current period. While the proposed rule addresses a portion of the high market share trend issue, the regional benchmark adjustment will continue to have a limited impact on these ACOs because they make up a large portion of their regional benchmark.
Losers: ACOs with costs much lower than their regional benchmark
Because some efficient ACOs would be able to generate shared savings without achieving further cost reductions, CMS limited the impact of the regional benchmark adjustment in the proposed rule in two ways:
1. The weight given to the regional benchmark adjustment will not exceed 50% in any agreement period (the maximum is 70% under the current rule).
2. The total impact of the regional benchmark adjustment (after blending) in each beneficiary entitlement category cannot exceed 5% of the national fee-for-service (FFS) expenditures (there is no limitation under the current rule).
As previously mentioned, CMS noted in the proposed rule that 80% of ACOs that renewed for a second agreement period in 2017 had costs below their risk-adjusted regional benchmark (and therefore benefited from the regional benchmark adjustment). Many of these ACOs may have anticipated driving further cost reductions, widening the cost difference compared to their regions’ costs. This provision may limit the ability of some ACOs to benefit from these significant cost differences.
Winners: ACO’s interested in rebasing “off-cycle” in 2019 (i.e., within their current agreement period)
Under the proposed rule, an ACO with an existing agreement period ending in 2020 or 2021 can choose to enter into a new 5½-year agreement period commencing in 2019, at which time the ACO's benchmark would be rebased. This may be beneficial to an ACO if:
- Its financial benchmark under its existing agreement period is inadequate based on current cost and utilization levels within its region.
- Or the ACO is “efficient” compared to its region and the ACO is in its first agreement period (and thus not subject to the regional benchmark adjustment).
Under the current rule, an ACO does not have the option to rebase off-cycle in 2019.
Low-revenue ACOs appear to be the most significant winners under the proposed rule because they will enjoy material reductions in risk exposure for up to 10½ years if they stay in the BASIC track. CMS made it clear in the proposed rule discussion that it wants to build on the early successes of physician-led ACOs. It will be interesting to see if this proposed rule affects how ACOs structure their participation lists in the future.
Conversely, high-revenue ACOs that were not planning on transitioning to downside risk appear to be the most significant losers because they will be required to take downside risk sooner. These ACOs will need to decide whether they are willing to be at risk for the total cost of care of their beneficiaries. Given that ACOs have been a large catalyst for the movement to value-based care, their decisions may have a significant impact on population health efforts within their communities.
This paper highlights the importance for individual ACOs to consider their unique situations when assessing the impact this proposed rule will have on their organizations. While some ACOs inevitably will be inclined to leave the program, other ACOs may find a viable path, or even be better positioned, for the future.
The chart below summarizes the high-level impact of the proposed rule on ACOs based on various characteristics (some of which are not discussed in detail in this paper). Although the list is not exhaustive and many items are interrelated, the list highlights some of the key themes of the proposed rule. The high-level impacts on ACOs are separated into three categories: financial benchmark, upside potential (“shared savings”), and risk exposure (“shared losses”).
The impacts shown are relative to the current rule, not relative to other ACOs. For example, ACOs would always prefer to have lower costs relative to their regions because the regional benchmark adjustment will increase their financial benchmarks. But ACOs might receive less benefit under the proposed rule than they would have under the current rule.
In the chart, a green circle with a checkmark indicates a clear favorable impact and a red circle with an X indicates a clear unfavorable impact. When the effect is moderate or uncertain, we removed the checkmark or X. For instance, new ACOs might benefit from the regional benchmark adjustment that is applied immediately, but it is also possible they will see a negative effect from this.
|Moderate or uncertain unfavorable impact
||Moderate or uncertain favorable impact
|Clear unfavorable impact
||Clear favorable impact
1National Associations of ACOs. Proposed rule likely to drive exodus of Medicare accountable care organizations (ACOs). Retrieved on September 17, 2018, from https://www.naacos.com/press-release--mssp-nprm.
2 Meltzer, R. (August 13, 2018). CMS's ACO proposal resurfaces discord over pace of risk-based models. FierceHealthcare. Retrieved on September 17, 2018, from https://www.fiercehealthcare.com/payer/cms-s-proposed-rule-acos-draws-praise-from-obama-s-national-coordinator.
3The proposed rule does also consider experience in other CMS and CMMI programs, such as the Pioneer ACO Model, Next Generation ACO Model, and Comprehensive ESRD Care Model.
4Under the current rule, it is possible for an ACO to have a total of nine years at the Track 1 or Track 1+ level. Under the new rule, low-revenue ACOs that are still in Track 1 can have an additional 10½ years before moving to the ENHANCED track. Therefore, some ACOs that started in 2013 could end up taking a total of 17 years before moving to the ENHANCED track.