In a letter to the Centers for Medicare & Medicaid Services regarding accountable care organizations (ACOs), the Medicare Payment Advisory Commission supported a two-sided financial risk model to give more ACOs more incentive to control spending.


Recently the Medicare Payment Advisory Commission (MedPAC) commented to the Centers for Medicare & Medicaid Services (CMS) that regulations governing accountable care organizations (ACOs) participating in Medicare’s Shared Savings Program should include a “two-sided” financial risk model for providers, inform Medicare beneficiaries when assigned to an ACO and possibly allow them to “opt out,” and address quality measures and spending growth targets.

MedPAC Argues Two-Sided Financial Risk Model Would Strengthen ACO Incentives

In its letter to CMS, MedPAC sets out several shortcomings of the “basic shared savings model” statute in the Patient Protection and Affordable Care Act (PPACA). The act provides that an ACO is eligible for a bonus payment if it meets quality targets and keeps spending at a target level based on the Medicare beneficiaries assigned to the ACO. This approach allocates the downside risk of relatively poor-performing ACOs to the Medicare program.

MedPAC pointed out that the “bonus-only” model has a potential flaw that needs to be addressed. Year to year, Medicare spending for a population of 5,000 (the minimum for an ACO under the Shared Savings Program) can vary randomly by a significant percentage. A 2-percent reduction in costs from one year to the next would trigger a bonus payment to an ACO, but that would result in potential waste because there is no mechanism to discriminate between random cost reductions and those actually achieved as a result of the efforts of the ACO and its affiliated providers. MedPAC suggests that setting minimum thresholds for triggering ACO bonus payments could help address this concern. However, that solution would have the unintended consequence of weakening incentives to control costs (perhaps particularly for vertically integrated ACOs). An ACO that believes a bonus payment is uncertain due to a threshold requirement is likely to be less cautious about the cost of care because, even if shared savings payments are reduced or even forfeited, at least the ACO would still receive Medicare fee-for-service (FFS) payments for those services.

Therefore MedPAC supports a two-sided model, believing that a degree of ACO exposure to financial risk in addition to the potential for financial reward is “necessary for the shared savings program to meet its potential.” Under MedPAC’s recommended approach, an ACO still would share in a portion of any savings it achieves, but also be at risk for a portion of any spending that exceeds target. In this model an ACO knows up front that its spending-reduction efforts always affect its bottom line—toward either a larger shared savings bonus payment or at least reducing any loss during a bad year. The intent is to provide the ACO with stronger incentives to control spending.

MedPAC does not advocate wholesale replacement of the basic model with the two-sided model. Rather, it opens the possibility that the two models could exist side by side, perhaps with each ACO choosing for itself which model is most appealing. MedPAC notes CMS may determine that the PPACA does not provide CMS with the regulatory authority to introduce such a two-sided risk model under the Shared Savings Program. If that is the case, MedPAC urges CMS to consider creating a program containing a two-sided risk arrangement as a demonstration to be implemented by the new Center for Medicare and Medicaid Innovation.

MedPAC Says Medicare Beneficiaries Must Know If They Are Assigned to an ACO

MedPAC warns that failing to inform beneficiaries of their assignment to an ACO “would run the risk of a repeat of the managed care ‘backlash’ experienced in the 1990s.” MedPAC advises that the backlash two decades ago “resulted from patients feeling that they were being forced into managed care by their employers and that the financial benefits were accruing to employers or health plans, not them … it behooves Medicare to pay close attention to patient notification [that the patient has been assigned to an ACO] so as not to repeat history.” (Shortly after MedPAC sent its letter, CMS officials announced they plan to provide some form of disclosure to Medicare beneficiaries assigned to ACOs. See CMS Expects to Issue ACO Rules by Mid-January 2011 for more information.)

Responding to this concern, MedPAC suggests CMS consider programs such as Medicare beneficiaries being able to participate in any shared-savings bonus, either through reduced beneficiary cost sharing or direct payments (though such may not be accomplished through rule-making). This could also increase enthusiasm for ACOs among beneficiaries. It is widely believed that in order for an ACO to function well, its assigned patients must be engaged in their own health care management, such as medication adherence and participation in wellness and other incentive programs.

Under the basic PPACA structure, as a practical matter Medicare beneficiaries assigned to an ACO could leave the ACO only if they switch to a primary care physician who is not an ACO provider. MedPAC discussed the merits and risks associated with various rules that CMS could make allowing Medicare beneficiaries to “opt out” if assigned to an ACO. While there are many possible iterations of “opt out” rules, MedPAC’s position is that Medicare should provide some opportunity to opt out, but only if, after notification of assignment to an ACO, the Medicare beneficiary feels “too uncomfortable” about being in the ACO. Another approach, which would be the least disruptive to the Medicare beneficiary, but would appear to create the risk of an administrative nightmare, would be to allow each Medicare beneficiary to “opt out” of the ACO for measurement purposes but continue to be under care from their same ACO primary care physician.

MedPAC Advocates Quality Measures

MedPAC argues the Shared Savings Program necessitates that CMS determine new quality measures applicable to ACOs in order to achieve the outcomes CMS desires. Current quality measures are tailored for the FFS payment system and may be inappropriate or inadequate for the Shared Savings Program. MedPAC suggests CMS consider ACO metrics involving the following population-based outcomes measures:

  • Emergency room use
  • Potentially preventable admission rates
  • In-hospital mortality rates, and possibly patient safety measures
  • Readmission rates

MedPAC offers an intriguing suggestion that CMS should include “patient experience” in the ACO (as measured by patients themselves) in any bundle of quality measures, which would be consistent with the stated goals of the ACO Shared Savings Program. There would also be a potential, positive side effect: Medicare beneficiaries may be more willing to stay assigned to an ACO if they know their providers’ payments are dependent on the patients’ review of the quality of care provided. However, will ACO providers then be more reluctant, for fear of generating bad reviews, to place sufficient pressure on patients to receive services within the ACO network and to use generic as opposed to more costly brand-name pharmaceuticals? Given the lack of any ACO network “lock-in,” such unintended consequences would be a concern.