The MedPAC Blog

Comment letter on the Medicare Shared Savings Program

by MedPAC Staff | Feb 02, 2015

 

Today MedPAC submitted a comment letter on CMS’s proposed rule regarding the Medicare Shared Savings Program (MSSP). The MSSP and the Pioneer Program are two options for Accountable Care Organizations to have the opportunity to share in savings with the Medicare program if they achieve spending below a target and meet certain quality metrics. The complete letter is available here.

This post highlights a key issue discussed in the letter: how the benchmarks for ACOs are determined. A benchmark is an ACO’s spending target for its population of attributed beneficiaries. If an ACO achieves this target, it is eligible for shared savings. Currently, benchmarks are set based on an ACO’s historical spending for the population of patients attributed to it. MedPAC has two major concerns about this policy design: (1) it may not be sustainable, particularly for relatively efficient ACOs, and (2) it does not promote equity among ACOs in the same market.  MedPAC believes that the first issue could be addressed in the short run by maintaining benchmarks for particularly efficient ACOs instead of lowering them in the next contract cycle when the ACO reduces spending. The second issue could be addressed in the longer run by transitioning ACOs to a common benchmark, for example based on local fee-for-service (FFS) spending in their markets.

Short Run

In the short run, ACO benchmarks are now “rebased” if ACOs’ spending during the first contracting cycle was below their initial target.  In other words, if an ACO reduces its spending for attributed beneficiaries in the current ACO cycle, its benchmark would be lower in the next three-year ACO contract cycle. This implies the ACO should be able to continuously improve over its past performance and manage the cost of its aligned population to grow no more than the average growth in FFS.

This formula is not likely to be sustainable, and in particular, punishes historically efficient ACOs, who may have begun the program with a lower benchmark. To address this issue, MedPAC proposes CMS not rebase (or not fully rebase) the benchmarks for efficient ACOs, e.g., those that have per capita adjusted spending below the national average AND per capita adjusted spending below average FFS spending in their market. ACOs that do not meet the two-part test would have their benchmarks rebased. 

Long Run

Basing ACO benchmarks on historical experience (per current law) was a reasonable starting point for the program. It set what should be an achievable target for the ACO; its historical experience trended forward by the growth in FFS at the national level. The logic is that if the ACO can beat the national trend, it can share the savings with the Medicare program because spending for its beneficiaries will be less than it would have been if the ACO did not exist.

However, the concerns stemming from basing ACO benchmarks on past spending are becoming apparent. In addition to the aforementioned challenge of sustainability, historically-based benchmarks mean that ACOs are often in the same market with other ACOs with very different benchmarks.

To illustrate why this may be inconsistent with the ACO program’s goal of encouraging efficient (high quality, low cost) care, imagine a market with two ACOs. One ACO has been historically low spending, and has a benchmark below local FFS spending in the market. The other ACO has been historically high spending, and has a benchmark above local FFS spending in the market. If the first ACO struggles to keep spending below a target that is already lower than local FFS spending, it may not be able to share in savings and may have a harder time remaining in the program. On the other hand, the second ACO, with a benchmark above local FFS, could have an easier time meeting its spending target and sharing in savings, even though its spending is still higher than local FFS. In this example, historical benchmarks are effectively punishing the more efficient ACO, which is competing with the less efficient ACO on more difficult terms.    

One way to ensure that all ACOs in a given market are competing on equal terms is to move ACO to a common benchmark in a market, for example, one based on local FFS spending. The proposed rule explores several options and captures much of the argument for and against transitioning to common benchmarks. In our letter, MedPAC urges CMS to keep the following principles in mind when considering alternatives:

  • There should eventually be equity among benchmarks for ACOs in the same area. This is important because more efficient (higher-quality and lower-spending) ACOs should be rewarded and providers and beneficiaries encouraged to join them. Equity means starting with equal benchmarks in an area, properly adjusting for risk, and protecting against differences in coding.
  • The goal is not to create as many ACOs as possible: the goal is to create the conditions that will allow efficient ACOs to be successful. In the long-term, equitable benchmarks across Medicare’s three payment models (FFS, ACOs, and MA) should help to determine if ACOs are a more efficient organization for the delivery system than MA plans or the other FFS providers in an area. We do not envision ACOs as the most successful option in all markets; there will be some markets where MA plans dominate and others where FFS dominates. Ideally, equitable benchmarks would encourage providers to organize into the most efficient delivery model for their market and encourage beneficiaries to choose a plan or provider that best meets their needs.

Said differently, the MSSP should transition from historically-based benchmarks unique to each ACO to a common benchmark for each market. This would provide equity and encourage efficient ACOs to join and stay in the program. 

Next Public Meeting

PRESS ROOM

For the latest news and press releases from MedPAC


Click Here

Join Our Mailing List

Sign up to recieve e-mail notices of Commission meetings and new posts to our website.

Sign Up Now

Filter posts by subject