Download the PDF

On August 8, 2018, the Centers for Medicare and Medicaid Services (CMS) released a sweeping proposed regulation that, if enacted, will significantly change the Medicare Shared Savings Program (MSSP). The proposed regulation, titled “Pathways to Success,” accelerates the path for accountable care organizations (ACOs) to participate in shared risk arrangements while simultaneously softening key provisions, allowing lower revenue ACOs to participate with reduced total financial risk. In addition, CMS has proposed numerous methodological and operational changes. In this paper we provide a summary of the proposed regulation’s key provisions and briefly discuss how they might impact the MSSP.

Introduction of the BASIC and ENHANCED tracks

If the new regulation is adopted, CMS will discontinue all current tracks—1, 2, and 3, as well as Track 1+ under the Center for Medicare and Medicaid Innovation (CMMI)—and replace these options with the BASIC and ENHANCED tracks¹ for contract periods beginning on or after July 1, 2019. Due to the off-cycle contract period, expiring agreements will be granted a six-month extension in order to avoid a six-month gap in participation.

• The BASIC track has five levels of risk designated A through E. The lowest levels of risk (A and B) are upside only and bear many of the same features as the current Track 1. Levels C through E introduce upside and downside risk, culminating in the highest risk level E, which is commensurate with the current Track 1+ and qualifies as an Advanced Alternative Payment Model (APM).²
• The ENHANCED track is for ACOs that can or are required to take on more substantial financial risk. The ENHANCED track has the same financial parameters as the current Track 3, offering greater risk and reward than Level E of the BASIC track as well as qualifying as an Advanced APM.²
• Agreement periods for both tracks will be five years whereas previously agreement periods were three years.

Figure 1 shows both the current and future ACO track options.

The BASIC track will be of special interest to ACOs currently in Track 1 or Track 1+, or ACOs considering entering the program. ACOs in the BASIC track will start at one of the five levels of risk (A through E) based on their experience and Medicare revenue levels and will transition each year³ to the next level of risk, culminating in reaching level E. CMS describes this as the “glide path” approach to taking on more risk, and requires ACOs to transition to higher levels of risk more rapidly (e.g., from level B to D instead of B to C). Furthermore, ACOs reaching level E in the BASIC track will be eligible for qualifying participant (QP) status under the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA). Lastly, unlike current rules, risk track will not drive assignment methodology, with ACOs being able to select retrospective or prospective assignment under both tracks each performance year.

Figure 2 summarizes the financial parameters for each of the five levels in the BASIC track. In general, as ACOs move to higher risk levels, the shared savings rate and the loss sharing limit increase.

How does CMS calculate the revenue vs. the benchmark-based cap?

One of the chief barriers to date for provider-led ACOs moving to shared risk in the current Track 2 or Track 3 is that an ACO’s financial loss exposure is based upon its benchmark, which can be significant; e.g., a bad year could result in ACO participants owing CMS the equivalent of multiple years of their Medicare fee-for-service (FFS) revenue. Understanding this challenge, CMS tested a revenue-based or benchmark-based loss limit in Track 1+, and is now proposing this type of loss limit in the shared-risk levels of the BASIC track. Going forward, the determination of a revenue-based or benchmark-based loss-sharing limit will be formula-driven rather than relying on an ACO to attest to its ownership structure (as is currently the case in Track 1+). This also results in a shared risk model that better adapts to the unique structure of an ACO (e.g., provider-led vs. integrated delivery system-led).

The revenue-based loss sharing limit is calculated by multiplying the ACO participants’ total Part A and B FFS revenue by the applicable revenue-limit percentage for the risk level in the BASIC track. The table in Figure 3 shows an example of this calculation for “Level E” parameters as defined for 2019 and 2020. The loss-sharing limit is the lesser of the benchmark-based cap or the revenue-based cap.

It is important to note that in the calculation of the revenue-based approach, the ACO participants’ total Part A and Part B FFS revenue⁴ is used, including any revenue from services provided to beneficiaries that are not assigned to the ACO.⁵ While this approach may cause a revenue-based loss limit to be greater than a benchmark-based loss limit for some ACOs, we expect the revenue-based loss limit will still be lower than the benchmark-based loss limit for most provider-led ACOs.

At what level will an ACO enter the BASIC track “glide path”?

CMS has proposed complex rules for determining the level at which an ACO is eligible to enter the BASIC track “glide path,” based on an ACO’s applicant type, experience level, and whether the ACO is considered low- or high-revenue. Figure 4 summarizes the options available for ACOs.

Brief descriptions of the preceding terms follow.

Defining experienced vs. inexperienced ACOs

An ACO is determined to be experienced if either of the following conditions apply:

• The ACO previously participated in a performance-based Medicare ACO initiative⁶ or deferred entry into a second MSSP agreement period under Track 2 or Track 3.
• 40% or more of the ACO’s providers participated in a performance-based Medicare ACO initiative or were part of a deferred renewal arrangement in any of the five most recent performance years prior to the agreement start date.

Defining low-revenue vs. high-revenue ACOs

Low-revenue and high-revenue ACOs are defined based on the ACO participants’ Part A and B FFS revenue for the most recent 12 months of available data, as follows:

• High-revenue ACO: If the ACO participants’ recent total Part A and B FFS revenue (which may include revenue for beneficiaries not assigned to the ACO) is at least 25% of the total Medicare Part A and Part B FFS expenditures for the ACO’s assigned beneficiaries for the 12-month period.
• Low-revenue ACO: If the ACO participants’ recent total Part A and B FFS revenue is less than 25% of the total Medicare Part A and Part B FFS expenditures for the ACO’s assigned beneficiaries for the 12-month period.

Changes to the benchmark calculation

In the proposed regulation, CMS has restructured four key components of the benchmark calculation:

1) Regional adjustment

Regional adjustment to the benchmark is designed to address two issues: incentivizing participation of ACOs that are already efficient compared to their regions (and will therefore have a low benchmark), and creating a more sustainable model for ACOs that have reduced expenditures versus their historical levels. In the current program, regional experience is used to trend and adjust the benchmark starting in the second contract period. CMS has proposed revisions to this approach so that regional blending starts immediately (even in the first contract period) and the effect of the regional adjustment is capped at +/-5% of national Medicare FFS per capita expenditures. Figure 5 summarizes the key changes.

2) Risk adjustment

CMS is proposing to do away with the “newly assigned” and “continuously assigned” designations of the existing program and replace them with an adjustment based on the risk score ratio capped at +/-3% versus benchmark year three (up to a five-year gap) by eligibility category. This is intended to more accurately measure morbidity changes, and limit the effect of changes to coding practices.

3) Trend

Currently, an ACO’s benchmark is trended using national trends for the first agreement period and regional trends (based on the ACO’s beneficiary county mix) for the second agreement and subsequent periods. CMS is proposing to change to a “national-regional blend” approach for all agreement periods. Under this approach an ACO’s benchmark trend is calculated by blending the regional trend and the national trend based on the ACO’s share of total assignment-eligible beneficiaries in each county. This partially addresses a prior issue with regional trends in which large or rural ACOs that dominated a region were essentially competing against their own trends.

4) Agreement periods and rebasing

CMS is proposing to change the length of future agreement periods from the current three years to five years. Consistent with the current approach, the benchmark will be rebased (i.e., recalculated using updated experience data) for each agreement period. Therefore, the benchmark will be rebased every five years instead of every three years. This longer rebasing period will provide ACOs with more stability in their benchmarks, but will also make the risk adjustment and trend (and the changes discussed above) more significant in the benchmark calculations.

Key changes to other program features

CMS has also proposed several key changes to other program features. These changes provide ACOs with appealing options for better effecting high-quality and lower-cost care.

Conclusion

In these proposed regulations, CMS has defined a set of paths for MSSP ACOs. Each path leads to downside risk and has its own set of guardrails. By changing the program to better fit its own goals and stakeholder feedback, CMS has created both challenges and opportunities for ACOs. The effect of these changes, whether positive or negative, will vary significantly across ACOs based on their structure, experience, and region. It is critical for all ACOs to understand the specific implications to their organization to ensure future success.