Officials at the Centers for Medicare & Medicaid Services have said they are re-evaluating their portfolio of payment models. In its June 2022 report to Congress, the Medicare Payment Advisory Commission (MedPAC) described a potential approach to streamlining and harmonizing Medicare’s portfolio of alternative payment models.
In the new report, MedPAC provides specific suggestions to operationalize its June 2021 recommendation that CMS reduce the number of Medicare alternative payment models (APMs) and design models to work better together. The proposed population health-based approach would reduce the number of accountable care organization (ACO) model tracks from seven down to a smaller number of tracks that could each be geared toward provider organizations of different sizes and involve different degrees of financial risk.
MedPAC noted that the presence of multiple alternative payment models operating concurrently can create unnecessary complexity and may dilute incentives when Medicare beneficiaries are attributed to more than one model simultaneously and/or when providers participate in more than one APM at the same time.
In their June 2021 report to the Congress, MedPAC recommended that CMS reduce the number of Medicare APMs it operates and design models to work better together when combined. In particular, to reduce the complexity of CMS’s offerings, the Commission said it supports reducing the number of population-based payment model tracks available to providers.
With a smaller number of tracks, each could be geared toward provider organizations of different sizes and involve different degrees of financial risk, MedPAC says. For example, a track geared toward groups of small provider organizations (e.g., independent primary care practices) that come together to form an ACO could include the opportunity to earn modest shared savings but not hold these providers accountable for repaying any shared losses. A second track could be geared toward midsize organizations and could give them the opportunity to earn a higher percent of shared savings and be at risk for shared losses. A third track could be geared toward large provider organizations (e.g., health systems with multiple campuses) and could put them at full risk for all Part A and Part B spending generated by their attributed beneficiaries.
Alternatively, MedPAC says, a population-based payment model could have a single track, with shared savings and loss rates varying based on ACO characteristics, such as an ACO’s ability to take on financial risk. Regardless of which approach is used, MedPAC envisions allowing provider organizations of any size to move to a more advanced track involving more financial risk if they so choose.
To strengthen incentives for providers to participate in this simplified population-based payment model and to slow the growth in their spending, ACOs’ spending targets (“benchmarks”) should not be rebased every few years based on actual spending; instead, benchmarks should be updated using exogenous administrative growth factors that would be known to ACOs in advance. Moving away from “rebasing” would ensure that ACOs that succeed in lowering their spending are not penalized in subsequent years by having their benchmark “ratcheted” down based on their recent actual spending, MedPAC said.
Ideally, the commission says, a growth factor would be chosen to produce benchmarks that increase fast enough to give participating providers a reasonable chance to earn shared savings, but slow enough to give the Medicare program a high probability of realizing net savings (relative to what Medicare would have spent in the absence of this model), while avoiding significant forecasting errors.
Acknowledging that not all providers are capable of bearing financial risk under population-based payment models, MedPAC does not see a rapid transition to mandatory participation in ACOs as practical. “We do, however, encourage CMS to explore ways to strengthen incentives to participate in population-based payment models, particularly for larger provider organizations,” it wrote.
In addition to a streamlined population-based model, MedPAC also supports a national Medicare-run episode-based payment model, in which participation could be mandatory for certain providers and certain proven clinical episodes (e.g., hip and knee replacements), even if a beneficiary were concurrently attributed to an ACO. “CMS’s Innovation Center should continue testing episode-based payment for a variety of types of clinical episodes, with the goal of identifying additional types of clinical episodes that could be added to a national episode-based payment model in the future,” the report says.
To ensure the population-based payment model and the episode-based payment model envisioned here work well together, MedPAC asserts that any bonus payments resulting from reducing episode costs should be allocated in such a way that (1) episode-based providers have an incentive to furnish efficient, high-quality care; (2) providers in ACOs have an incentive to refer their attributed patients to low-cost, high-quality episode-based providers; and (3) when combined, these incentives should not be so large that they increase total Medicare spending.
MedPAC notes that these strategies would represent a shift for CMS—moving away from temporarily testing a large number of model tracks on a small scale to permanently operating a smaller number of model tracks on a large scale. The commission asserts that, designed correctly, APMs offer a promising avenue for lowering fee-for-service spending while preserving or improving care quality. The proposed changes to CMS’s APM portfolio are intended to help reach this potential by reducing the complexity and uncertainty that providers face when picking an APM, increasing provider participation in these models, and improving provider performance in these models.
