4 recommendations to advance Medicare delivery system reform

Editor's note:

This analysis is part of The Leonard D. Schaeffer Initiative for Innovation in Health Policy, which is a partnership between the Center for Health Policy at Brookings and the USC Schaeffer Center for Health Policy & Economics. The Initiative aims to inform the national health care debate with rigorous, evidence-based analysis leading to practical recommendations using the collaborative strengths of USC and Brookings.

Enacted by the 114th U.S. Congress on a bipartisan basis, the Medicare Access and CHIP Reauthorization Act (MACRA) aimed at reducing the fragmentation and cost of Traditional Medicare. MACRA had three main components:

  • Ending the decades-long but unworkable Sustainable Growth Rate, which annually threatened as much as a 20+ percent cut in Medicare physician fees;
  • Encouraging increased participation by providers in alternative payment models (APMs) through incentives; and
  • Creating new rules for providers choosing to remain in fee-for-service (FFS) by combining enhanced performance measurement, rewarding high performing providers while penalizing low performers, and constraining increases in Medicare fees.

By encouraging APMs that require physicians agree to accept either bonuses or financial penalties based on their performance, MACRA envisions providers organizing themselves into efficient systems of care, where patients receive care from providers in a coordinated network accountable for quality, who also coordinate care for complex, costly patients.

However, the legislation only affected health care providers that are reimbursed based on Medicare’s Physician Fee Schedule. It did not create incentives for hospitals and other major stakeholders to participate in APMs with the physicians. To put it differently, MACRA assumes a wide range of attractive APM opportunities available to physicians, but the reality is that the opportunities are limited, especially for physicians in specialties outside of primary care.

This piece will focus on ways to improve existing APMs and create new ones that enhance opportunities for physicians and other providers to provide more efficient, higher-quality, and lower cost care. We propose:

  1. Increasing beneficiary engagement and alignment in all models;
  2. Improving design and implementation of bundled payment and their utility for systems of care;
  3. Making payments directly to organized systems of care, like accountable care organizations, so as to extend incentives for hospitals and other key providers, as well as updating the benchmarking and antitrust processes; and
  4. Supporting the development and initial funding of the costly infrastructure needed to organize systems of care among providers.

Improving Medicare Advantage (MA) is another critical piece of improving cost and quality in Medicare, and the Center for Health Policy has forthcoming pieces that will focus on ways to improve the MA program. These changes to Medicare FFS and its APM environment will better position and help to proliferate systems of care, like ACOs. Moreover, these changes aim to help systems of care accomplish what they were created to do: to increase the quality and efficiency of care, to measure performance, to control spending, to engage patients, and to promote the movement from volume to value.


The increasing need to care for beneficiaries suffering from multiple chronic illnesses and to control costs while improving quality has led to major reforms in Medicare over the past five decades. Indeed, two-thirds of today’s beneficiaries live with multiple chronic illnesses. The Patient Protection and Affordable Care Act (ACA) helped launch one of the latest reform efforts, not only creating the Medicare Shared Savings Program (MSSP, the statutory ACO program), but also providing authorities for the Center for Medicare and Medicaid Innovation to develop and test new departures from traditional FFS.

The Centers for Medicare and Medicaid Services (CMS) are experimenting with a variety of innovative models, in many cases on a large scale. MACRA built on many of these reform efforts, providing bonuses to early APM adopters and higher base fee rates in later years. In contrast to the active role private insurers play in supporting participation in their APMs, however, CMS to-date has taken a largely passive role in expanding beneficiary and provider participation in these models, maintaining voluntary participation for many and providing only limited performance feedback and resources for improvement to participants.

1. Increasing beneficiary engagement and alignment

Despite the enthusiasm for and growth of innovative models, current beneficiaries are not only unwittingly attributed to one or more providers, they have no real financial incentives to participate.

Currently, in most models, beneficiaries do not enroll into an APM, but are instead retroactively attributed to an APM based on their pattern of care consumption, leaving them—and often providers, as well—unaware of whether or not they are part of an APM. As a result, new models have largely lacked the benefit of beneficiary engagement and financial alignment.[1] In fact, the statutory provision establishing the Medicare Shared Savings ACOs expressly prohibited waiving or reducing copays or deductibles for beneficiaries. Consequently, beneficiaries are denied the opportunity to support and share in the efficiencies of the APM.

In an active model, beneficiaries would elect to participate and would receive rewards for compliance with the model. Only with beneficiaries knowingly choosing a model should they experience rewards and, potentially, penalties based on their compliance under providers they choose to affiliate with under the model. This also supports clinicians’ efforts to reinforce compliance with care recommendations and invest in new means of care delivery.

However, all of these changes to Medicare models would be for naught if supplemental Medicare insurance policies offset the model’s financial incentives. Aligning beneficiaries’ financial incentives with those of providers will improve existing reform models.

A change from beneficiary attribution to beneficiary enrollment in an APM is a prerequisite for other means of engagement. In contrast to the current attribution-based APM assignment, enrollment in an APM would function much like commercial health plan enrollment, where beneficiaries choose an APM provider organization based on the APM-based plan design and network. In the case of Medicare ACOs, this choice would most closely resemble enrollment into a commercial Preferred Provider Organization, where beneficiaries may pay less to see in-network providers relative to out-of-network ones but retain coverage for both in- and out-of-network providers.

Such an enrollment process in ACOs or other APMs supporting organized systems of care could be coordinated with the MA and Part D Drug Plan enrollment period. However, both initial enrollment in and any subsequent disenrollment from an APM by any unhappy beneficiaries should be easier than it is currently for either MA or Medicare Part D. For example, dis-enrolling from an ACO could require beneficiaries provide notice at any point throughout the year but at least one month in advance.

CMS should actively monitor and identify poorly performing ACOs or those that might be “gaming” the system, penalizing ACOs that have a disproportionately high disenrollment of either beneficiaries in general or of high-cost beneficiaries. Because disenrollment would be relatively quick and easy, beneficiary enrollment can also be streamlined through a provider-initiated process. One example of such a process might include CMS prospectively providing an ACO with a list of beneficiaries who have historically received a plurality of care with the ACO’s providers, and then providers obtain consent from their patients and initiate enrollment on their behalf.

With active beneficiary enrollment, the use of actuarially neutral beneficiary financial incentives becomes appropriate. Up until the point a beneficiary indicates they accept the terms of ACO participation, it hardly seems fair to impose financial penalties for seeking care from providers or in ways that otherwise do not align with the ACO. While recent APMs are experimenting with small beneficiary financial incentives,[2] they are limited solely to reductions in in-network cost sharing. This limitation is likely not only to have a lesser impact than a more comprehensive incentive structure, but it also stands to increase the actuarial cost of beneficiaries’ coverage. Without a corresponding disincentive for low-value services or out-of-ACO care to balance the reductions in in-ACO cost-sharing, Medicare may expect such in-ACO reductions to increase total costs. Such an actuarially neutral incentive structure would be new to Medicare’s APMs.

Further reforming beneficiaries’ financial incentives, CMS could allow for the creation of new Medigap plans that are specifically aligned with ACOs. Under these “Medicare Select” plans, ACO providers could agree to reduce or eliminate the deductibles and cost sharing associated for care provided by them, where the savings may be passed on by Medicare Select as lower premiums. At non-ACO facilities, however, beneficiaries enrolled in Medicare Select plans would be exposed to deductibles and other cost sharing. Additionally, Part D plans could also be aligned with the ACO, perhaps also allowing ACOs to negotiate their financial arrangements.

2. Improving design and implementation of bundles

CMS can further improve the FFS program by paying for more services based on episodes of care. Building off the example of the global surgical fee—which provides a fixed lump-sum payment for all services rendered within a set window surrounding a surgery—bundled payment involves all of the providers involved in an episode of care. In this way, bundled payments do not replace fee schedules, as they will continue to be the basis for valuing services within the episode of care, but rather they reduce the incentives to drive volume among all providers involved in the care for that episode. In fact, recent studies have found significant savings and improved, or at least maintained, quality outcome performance through the use of bundled payments for some episodes of care.

Prioritization of areas of care to bundle should consider each episode’s: variation in spending, risk-related issues, administrative complexity, and the number of differing types of providers typically involved in delivering care; provider and infrastructure readiness; alignment of clinical and financial incentives; and the ability to expand nationally. As in the case of the oncology bundle being demonstrated now, many outpatient bundle designs might be specific to certain types of episodes of care. This will increase the CMS workload, but is important to realizing the potential of bundled payment for episodes of care that do not involve inpatient stays. Comprehensive bundles give providers more flexibility for innovation on efficient service delivery.

CMS can further support APM participants by allowing bundle savings to accrue for person-level APM participants and providers. In their current form, any savings generated from a bundle by ACO participants are retained by the providers involved in the episode but is not recognized as a generated savings by the ACO. Instead, allowing a sharing of savings between the episode providers and the ACO would render bundles as an additional tool for, as opposed to a barrier to, providers engaging in broader systems of care like ACOs. In this way, providers are not pushed to avoid person-level APM participation.

3. Strengthening systems of care through direct payments and antitrust and benchmarking reforms

By switching to an active enrollment process as described in a previous section, providers are also better equipped to organize systems of care, which are multiple providers that have come together, either by common ownership or by contractual relationships, to delivery care that focuses on enrollees’ care needs over a period of time or during an episode.

With active enrollment in these models, providers in such systems can target specific beneficiaries, measure their performance against targets, and may be better positioned to allocate revenues based on personnel performance and infrastructure development strategies. Additional reforms to Medicare’s reimbursement process and to anti-kickback laws that permit the ACO to control allocation of FFS payments may further enable delivery system innovation. This would allow providers to develop systems of care and to tailor care based on their expertise and patients’ needs.

The first means of supporting system development could be through making payments directly to an APM entity, such as an ACO, on behalf of all participating and affiliated providers, incorporating hospitals and other stakeholders into the financial incentives of the APM. Under this method of payment, the APM entity, in turn, would allocate its revenues to personnel, ACO-affiliated providers, and infrastructure based on its strategic priorities. This would enable the ACO to better align provider compensation with its internal quality improvement initiatives, incentivize hospitals and other health care stakeholders, contract with other provider groups, and invest in health technology and facilities based on strategic care delivery coordination and outreach plans.

A second, complementary means of system strengthening is through reforms to anti-kickback legislation that would allow providers to steer patients to preferred external health care entities based upon their professional judgment and collaboration. If the goal is to improve coordination among providers that are at financial risk, it would follow that, at least in some cases, it would be desirable to have providers steer patients to a post-acute care provider with a track record of good patient outcomes and collaboration with the patient’s primary care provider, for example. In the case of lab services and medical devices, providers may be better able to negotiate for lower-than-Medicare-fee-rates in exchange for channeling volume, which would further strengthen providers’ ability to develop strong systems of care. This could be accomplished by allowing Medicare pay the ACO directly, which would, in turn, allocate payments to coordinating entities based on CMS-approved contractual agreements. The anti-kickback policies were conceived for traditional FFS delivery; they are less needed where incentives are broader than FFS and delivery is more organized, such as in MA and in ACOs.

Changes to the benchmarking process for APMs to establish earlier transitions from historical-based benchmarks to regional benchmarks may also strengthen systems of care. One example of this need for updated benchmarking is in Medicare’s Shared Savings Program (MSSP). There, MSSP ACOs’ cost performances had been measured against historical utilization patterns for the defined population. As many others have written, this historical benchmarking process places providers that had low per capita spending at a disadvantage. The recently announced exit of one of Medicare’s earliest participants (an institution with which creators of the accountable care concept are affiliated) was another exit by potentially driven by this problem. In response, CMS issued a final rule in 2016 that incorporated regional spending into benchmarking calculations for an ACO’s subsequent contract. Given the voluntary nature of the program and the modest penetration to date, this may be as far as CMS could go under current law.

4. Supporting the development of infrastructure for organized systems of care

Finally, CMS can support dissemination of and participation in its APMs through increasing investment in infrastructure-building mechanisms for providers of all sizes. Unlike private insurers—which have taken an active role in disseminating new payment and delivery reforms among in-network providers and providing data on performance—CMS has taken a passive role in rolling out its APMs. Owing partly to retrospective attribution, retroactive quality reporting of some data, and quarterly or annual reporting periods, the performance feedback data provided by CMS are often not timely or actionable. Moreover, CMS invests modestly in technical assistance and training efforts that would otherwise enable small provider groups to interpret the data and develop responsive delivery strategies.

CMS should organize a source of “working capital” that provides funds, management expertise, and technical expertise to providers, better supporting the development of systems of care like ACOs where they do not currently exist. In fact, MACRA recognized this need both in its explicit inclusion of $500 million in funding for technical assistance for small and rural providers, as well as in its creation of a Physician-Focused Payment Model Technical Advisory Committee. Using the allotted MACRA funds, CMS could solicit bids and establish a panel of eligible technical assistance and management suppliers. ACOs and other system-level ACO participants could then use a CMS-provided voucher to “purchase” services from this panel of suppliers, mitigating issues where small ACOs are highly disadvantaged. With this centralized resource of technical assistance and management expertise in place, CMS can better support participation in its APMs among providers in small practice and low-resource settings.


Improving the traditional Medicare program will take ongoing efforts to create comprehensive systems of care. Improving bundles is also a valuable step. By consolidating “individual units” of goods and allowing bundle savings to accrue for provider systems of care, this redesigned reimbursement can support accountable provider reform efforts rather than stand in their way.

Progress will require increased beneficiary engagement in APMs, flexibility in service delivery for APM participants, and support for APM infrastructure development particularly in low-resource settings. Additionally, the performance measurement and benchmarking processes used to evaluate accountability and performance will require further development.

With these reforms to APMs, CMS can improve existing APMs, as well as their viability moving forward. Most importantly, these reforms improve APM participants’ ability to achieve their original goal of improving care quality and efficiency.

[1] Medicare’s Next Generation Accountable Care Organization prospectively assigns beneficiaries and provides the option for beneficiaries to voluntarily enroll; for those enrolled, the ACO may reduce in-ACO beneficiary cost-sharing for some services.

[2] The financial incentives for beneficiaries in Medicare’s Next Generation ACO include reduced in-ACO cost sharing; yet, beneficiary cost-sharing for out-of-ACO services is held at normal rates.