Editor’s Note: This is the first in a series of Health Affairs Blog posts examining the proposed rule implementing the Medicare Shared Savings Program, issued March 31 by the Centers for Medicare and Medicaid Services. You can read subsequent posts by Douglas Hastings, Steven Lieberman, and Ron Klar.

The Centers for Medicare & Medicaid Services (CMS), the Office of the Inspector General, the Federal Trade Commission and the Antitrust Division of the Department of Justice, and the Internal Revenue Service have released proposed regulations for implementing Section 3022 of the Patient Protection and Affordable Care Act (ACA), the Medicare Shared Savings program.  The program will govern the short-term implementation of Accountable Care Organizations (ACOs) in Medicare.

The publication of these proposed regulations is an important step in enabling Medicare to help providers deliver better care for patients. The regulations represent a comprehensive and thoughtful effort to address a wide range of key issues for Medicare ACOs. Careful review and constructive analysis by the many groups who may be affected by the regulations – important work for all of us – is needed now to build on the proposed regulations.  In this commentary, we provide a brief overview of the origins and rationale for the ACO program, and some initial details on the just-released regulations. We then outline the major issues that we believe will be important to consider as a very wide range of interested groups develop responses to the proposed regulations. We expect to provide further comments on these issues as the ACO implementation process continues.

The goal of the ACO program is a simple one: to improve care for Medicare beneficiaries and thereby improve their health and lower health spending.  The approach focuses on helping physicians, hospitals, and other health care providers achieve this goal by providing more financial support when they work together to improve quality while lowering costs.

ACOs are expected to carry out a number of activities toward this end. Doing so will likely require investment in health information technology (HIT); establishing effective referral and transition procedures; and working  with nurses, pharmacists or other health professionals to prevent costly complications of chronic diseases.  These activities take time, effort, and money; the ACO regulations are intended to provide new financial support for effective investments to improve care.

As set forth in the regulations released today, the Centers for Medicare and Medicaid Services is proposing to implement two different tracks for organizations wishing to become ACOs.  The first track is a “one-sided risk model,” in which there would be sharing of savings only for the first two years and sharing of savings and losses in the third year. A second track would be for a two-sided risk model in which providers and the government alike share in savings and losses for all three years.  An ACO could opt for either model, based on its experience and capacity to implement reforms to improve care and lower costs.  Under both models, ACOs will be required to produce and then improve on 65 different quality metrics across five domains:  The patient or caregiver’s experience of care; care  coordination; patient safety; preventive health; and at-risk population health or the health of frail elderly populations. ACOs will also have to meet a number of other regulatory oversight requirements.

The Department of Justice (DOJ) and the Federal Trade Commission (FTC) also issued guidance today about allowing ACOs and other innovative health care delivery organizations to form without running afoul of antitrust laws.  For example, DOJ and FTC propose to give so-called “rule of reason” treatment to an ACO if the ACO follows similar structures and processes in the commercial market; describe a “safety zone” as well as what might be described as risky zones based on the ACO’s market share; and propose a process for potential ACOs to seek an expedited antitrust review.

Clearly the ACO program alone is unlikely to be a silver bullet to improve quality or lower costs. As we have noted previously, it is likely to work best when it is implemented along with other reforms in payment and benefits to promote better care. But depending on how the regulation and implementation process move forward, ACOs could have a substantial impact on real health care reform.


The notion of ACOs emerged in response to a growing consensus among the clinical and academic communities, including the Medicare Payment Advisory Commission and the Institute of Medicine, that shared accountability among providers could address the gaps in quality and unnecessary costs caused by fragmented and poorly coordinated care. Subsequent research on ACO-type initiatives has focused on providing financial return, or “shared savings,” to providers for a potentially wide range of effective steps for improving quality while lowering the cost of care.  As has been widely noted, under existing fee-for-service payment systems, providers who take steps to improve care quality and patient outcomes – for example, by lowering rates of hospital admission or readmissions – are often penalized financially.  The ACO shared savings program is designed to correct this problem by providing more financial support for providers who engage in efforts to improve quality, especially when it involves activities like coordinating care that are not reimbursed under fee-for-service.

For example, CMS launched the Physician Group Practice demonstration program which tested the shared savings model in a range of practice settings, including multispecialty physician groups, integrated delivery systems, and an independent practice association.  The demonstration showed that these providers were willing to participate in new payment models that required measurable performance improvement as a pre-requisite for receiving a share of savings achieved compared to other physician practices in their community.

Recognizing the potential of the shared savings model to both improve quality and reduce cost, Congress included ACOs among a range of Medicare reform proposals in the ACA.  The ACO provisions (Section 3022), will enable providers to receive additional payments if they can reduce spending growth relative to national trends, something that analysis from the Dartmouth Atlas suggests is feasible for a very wide range of practices and organizations.

As CMS and the public consider the proposed ACO regulation, it is important to recognize that more and more evidence on ACO and ACO-like payment models to support better care is emerging outside of Medicare. This is because a number of private plans and providers are implementing accountable care models ahead of regulations.

Brookings and Dartmouth have provided assistance to some of these payer-provider collaborations, which reflect a wide range of provider organizational structures, including two integrated delivery systems, a physician-owned hospital, and two physician practices. Although  their particular ACO contracts are adapted to local markets and organizational considerations, each shares the set of core ACO principles.

Other payer-provider collaborations include the Alternative Quality Contract, established as a partnership between Blue Cross Blue Shield of Massachusetts and a growing number of provider organizations, and pilots implemented by other insurers like Cigna, Aetna and Blue Cross Blue Shield of Illinois. Still others, like THINC, Taconic Health Information Network and Community, are bringing together payers and providers to align payment incentives in regional collaboratives to support better care coordination at the regional level, through HIT infrastructure such as HIEs and regional quality improvement initiatives. Medicaid programs in New Jersey, Colorado, and other states are also moving forward with ACOs for their beneficiaries.

The implementation of ACOs under Medicare offers an opportunity to both learn from and build upon these private and Medicaid models. Indeed, as the Centers for Medicare and Medicaid Innovation has already shown in its patient-centered medical home pilots, multi-payer collaboration can provide much greater support for health care reform than actions by individual payers alone, even Medicare.

It is also important to keep in mind that ACOs are not managed-care insurers. Although this is not the first time that payment reforms have sought to change the incentives facing health care providers, these recent ACOs do not turn provider groups into insurers, or simply change financial incentives and hope for the best. Previous efforts not only shifted risk imprecisely – often including insurance risk, not just accountability for feasible steps to lower costs – but also without a clear plan for how care could be improved and meaningful clinical support to help make it happen.

Indeed, many such initiatives included little or no accountability for measurable quality improvement.   Beneficiaries often found themselves locked into specific primary care physicians who were set up as gatekeepers, rather than partners with their other providers, creating conflicts among providers and fears among patients that they are not receiving needed care.

Recent ACO implementations outside of Medicare seek to address all of these concerns and provide useful models for how Medicare can proceed.  They emphasize shared risk, and the use of sound actuarial principles to make sure that the provider groups bear risk for things that they can control and that the level of risk is reasonable.   Quality measurement with increasingly robust measures is a critical component.  Financial incentives are structured so as to support care changes that lower costs by improving quality, with some models making any share of savings fully contingent upon demonstrable improvements in quality.

The approach means that ACOs typically involve implementing clinical improvement plans in conjunction with a range of changes in financial support going beyond shared savings alone, to help address the “chicken and egg” problem of needing payments that do a better job than fee-for-service of supporting the hard work of real reforms in care. And beneficiaries retain freedom of choice of providers.

Many concerns have been raised about ACOs’ potential weaknesses for Medicare beneficiaries in particular, and about unintended consequences such as increased provider market power that may emerge in Medicare applications.  In Exhibit 1, we summarize some of the major issues that have surfaced, and how program design options could affect these potential problems.  We recognize this list is not exhaustive.  We hope this exhibit provides the preliminary guideposts to those who are setting out to read through and assess all of the details of regulations that were recently released.

Successfully addressing these issues could entail some tradeoffs. In general, more stringent and specific requirements – especially in areas that conflict with or add to existing ACO models’ design, or go beyond their core goal – will increase costs and lead to lower participation and a more restricted range of reforms, potentially reducing the impact of ACOs on quality and costs. Conversely, fewer structural requirements could create additional payouts by Medicare in the absence of real care improvements. This underscores why thoughtful comments on ways to address these concerns will be critical if the ACO program is to be implemented successfully. For example, thoughtful analysis and collaboration during the comment period could identify ways to assure meaningful accountability (e.g., through strong performance measures focused on meaningful outcomes, and better data flows), more effective financial support (e.g., promoting coordination with private and Medicaid ACOs, and other Medicare payment reforms), and a well-developed framework that supports ongoing evaluation and timely learning.

Success will also depend on meaningful clinical leadership. The experience of prior Medicare demonstration projects and the growing number of ACO payment reforms around the country have shown that meaningful engagement on the part of physicians and other health professionals is essential for real success in health care reform. In theory, ACOs are intended to accommodate a wide variety of physician practice settings, and the proposed regulations seek to do so as well.

Making sure that this wide range of effective ACOs actually develops in practice – a key objective of the proposed regulations – will depend upon whether physicians and other providers are full participants in the design and development of delivery reforms that can really improve health and save money – and what is needed in the ACO program to help achieve these aims.

Although clinical leadership will be essential, successful ACO implementation is really all about the bottom line of improving care and avoiding unnecessary costs for Medicare beneficiaries. We will frame our own analysis of and comments on the regulations, in collaboration with a broad range of stakeholders, by focusing on the simple but challenging question: How best can we refine the proposed regulations so that physicians, hospitals, nursing homes, payers and consumers can get much better support than they receive today when they work together and take steps to improve care and slow spending growth?

Those who care deeply about health care reform all have a common interest in the success of ACOs as a way of avoiding more classic fee-for-service payment cuts to providers; access problems for patients; and the other adverse consequences of rising health care costs and inefficient care. Medicare’s ACO initiative, if implemented successfully, can help make a difference. The success of this program is largely up to how we all respond not only to these regulations, but to the opportunities created by ACOs and many other public and private initiatives to advance health care reform.