Health Affairs

JANUARY 31, 2012

Next Steps for ACOs

Will this new approach to health care delivery live up to the dual promises of reducing costs and improving quality of care?

What's the issue?

Accountable care organizations (ACOs) are networks of physicians and other providers that are held accountable for the cost and quality of the full continuum of care delivered to a group of patients.

Although the ACO model is being adopted in the private sector, industry observers are keeping a close eye on how it is being implemented within the Medicare program. Under contracts to the Centers for Medicare and Medicaid Services (CMS) authorized by the Affordable Care Act, which will go into effect in April 2012, ACOs will work to improve Medicare enrollees' health while simultaneously constraining costs and will earn annual bonus payments if they succeed.

This new approach is already affecting how other health plans pay providers and resulting in a number of ACO contracts between providers and private health plans. According to the American Medical Group Association, more than 100 of its member medical groups are well positioned to become ACOs under Medicare's Shared Savings Program, and many other providers are likely to be interested in exploring the ACO concept.

This Health Policy Brief provides an overview of ACOs, their origins, and the current status of adoption by Medicare and private health insurance plans.

What's the background?

Most insurance programs, including Medicare and private plans, pay for health care on a fee-for-service basis. This means that individual doctors, hospitals, and other providers are paid for each service they furnish to a patient. Critics of this system have complained that it rewards providers financially for delivering as many services as possible while driving up costs for patients and payers.

EARLY EFFORTS AT COORDINATION: Over the years, there have been many attempts to encourage primary care physicians, specialists, and hospitals to coordinate among themselves to manage the overall care of their patients. Commonly cited prototypes include Kaiser Permanente, Mayo Clinic, and Cleveland Clinic. Their centralized organizations have allowed providers within the system to work together to improve quality and efficiency--for example, by developing and adhering to practice guidelines. Frequently, the providers in these systems are paid on a salaried basis, removing incentives to drive up volume that are inherent under a fee-for-service system.

In the 1970s, some physician groups and joint ventures between physicians and hospitals tried to operate as health insurers on their own, or contracted with insurers, to provide total care to an enrolled population. Consumers, however, came to resist these network arrangements because they restricted their choice of providers. During the 1980s and '90s, health maintenance organizations (HMOs) gained prominence as another approach to managing care and controlling costs. But patients, often encouraged by their physicians, objected to the perceived intrusion of HMOs into their health care decisions, and HMOs have become less popular.

Elliott Fisher, a physician and researcher at the Dartmouth Institute for Health Policy and Clinical Practice, is credited with developing the concept and conducting the research that led him and Glenn Hackbarth, chair of the Medicare Payment Advisory Commission, to jointly coin the term "accountable care organizations" at a commission meeting, leading to its first published use in a 2007 Health Affairs article.

Rather than recreating HMOs, Fisher envisioned that freestanding hospitals and physicians could be organized into virtual organizations that could be held accountable for the cost and quality of the full continuum of care delivered to their patients. The ACO model has since evolved into a number of different forms, but overall, the term has come to mean an organization, virtual or real, that agrees to take on the responsibility for providing care for a particular population while achieving specified quality objectives and constraining costs.

ENCOURAGING QUALITY PERFORMANCE: When private health insurers enter into ACO-type agreements with providers, the providers are held accountable for providing high-quality care to their usual patient population while reducing the unnecessary use of resources. Organizations that meet agreed-upon performance levels on a range of specific quality measures are rewarded financially. The idea is to encourage further steps to improve care management, leading to a steady evolution toward fully coordinated care systems.

At least eight private health insurance plans have entered into ACO agreements with providers using a "shared risk" payment model. These arrangements make providers eligible for bonuses if they keep costs below a certain threshold but assess financial penalties against them if they exceed spending targets. Examples include Blue Cross Blue Shield plans in Illinois, Massachusetts, New Jersey, and North Carolina as well as Aetna and Anthem/WellPoint.

In addition, as many as 27 private health plans have entered into "shared savings" contracts, which make providers eligible for bonuses but do not put them at financial risk. Some other private ACO arrangements plan to employ "partial capitation"--a mechanism in which providers receive preset payments per patient in return for providing whatever services are needed, combined with payments based on actual services performed.

COLLABORATIONS: Health care organizations are only beginning to enter into these types of agreements with private insurers. Still, dozens of major health systems and provider groups have joined learning collaboratives to explore what it might take to become an ACO such as those convened by the Premier healthcare alliance, or by the Brookings Institution and Dartmouth Medical School. In addition, in 2011 the National Committee for Quality Assurance issued a list of proposed ACO capabilities, which form the basis of a voluntary ACO accreditation program.

The underlying mechanism driving ACOs is the use of financial bonuses that groups can receive if they meet quality and cost benchmarks. This in turn gives providers an incentive to coordinate their patients' care to reduce duplication of services, invest in infrastructure such as health information technology, redesign care processes, and practice with greater adherence to clinical evidence of what treatments work best.

FEDERAL ACO EFFORTS: From 2005 through 2010, CMS carried out the Medicare Physician Group Practice demonstration, involving nine multispecialty group practices and one physician-hospital organization. These 10 organizations were eligible to retain a portion of the savings they generated for Medicare, relative to a projected spending target, and they could increase their share of savings depending on how well they performed on a set of 32 quality measures. Although the results were mixed (see below), the Medicare Physician Group Practice demonstration is widely considered to be an immediate predecessor to ACOs.

To encourage the transition from demonstration to implementation, the Affordable Care Act authorized Medicare to enter into contracts with ACOs in what is called the Medicare Shared Savings Program. The health care law left it to CMS to decide on the rules and standards for this program, and the agency issued final regulations in October 2011.

CMS's Center for Medicare and Medicaid Innovation, which was also created by the health care reform law, is testing alternative ACO models in addition to the Medicare Shared Savings Program. In May 2011 the Innovation Center announced that it would test a new Pioneer ACO model, targeted to organizations that already have a track record of managing financial risk and developing systems for being accountable for quality-related performance.

In December 2011, 32 health care organizations were selected to participate in the Pioneer program. Just as in the Medicare Shared Savings Program, participating providers will be eligible to receive bonuses based on cost savings they generate for Medicare, compared to spending targets. Pioneer ACOs will be eligible for larger bonuses if they slow the growth in their health care expenditures, but they will be at risk to pay Medicare substantial financial penalties if they end up accelerating the growth in their spending. CMS hopes the savings in the Pioneer ACO initiative will reach $1.1 billion over five years.

The ACO approach may appeal to many more health plans than the prior attempts described earlier because it provides an intermediate form of delivery. Providers in ACOs aren't paid entirely through volume-increasing fee-for-service payments, nor do they operate within tightly managed, prospectively defined, capitated budgets that place providers at full financial risk for all spending for their enrolled populations. There is great hope that the balance struck between these two payment mechanisms will lead to the best care at an affordable cost.

How will Medicare ACOs work?

In typical managed care systems, an often-distant insurance company oversees diagnosis and treatment decisions. Providers that belong to Medicare ACOs, however, will be free to make these decisions but encouraged to exercise self-control through payment incentives that encourage prudence in service delivery. As noted, Medicare ACOs will also be monitored for quality outcomes to ensure that they do not skimp on needed care. And as with most fee-for-service systems of care, patients in ACOs will be free to seek additional services from other clinicians or facilities.

Medicare ACOs differ from existing health plans and provider arrangements in three major ways:

  • Shared savings. Both the Medicare Shared Savings Program and private ACO contracts typically pay providers on a fee-for-service basis and then add bonus payments if the providers have met certain objectives. The key objective is for overall spending on patients in the ACO to be less than a projected amount based on the providers' own historic spending, regardless of whether that spending was high or low. The size of the bonuses depends, in part, on how much savings the ACO produces. As a result, the bonuses represent "shared savings" that are split between the providers and the payer, whether the payer is CMS or a private health plan. Under the Medicare Shared Savings Program, ACOs can receive 50-60 percent of the savings they generate in the form of bonuses, and bonuses are capped at 10-15 percent of their spending target. For Medicare ACOs that decide to accept financial risk--that is, they agree to pay back amounts exceeding their spending projections--their shared losses are capped at 10 percent of their spending target and phased in over three years.

  • Accountability for quality. An ACO's performance on numerous quality metrics is central to determining whether it qualifies for shared savings bonuses and the amount of savings it will receive. In the Medicare Shared Savings Program, CMS will monitor 33 quality measures across four domains: patient/caregiver experience, care coordination/patient safety, preventive health, and at-risk populations (Exhibit 1). ACOs that do not perform at a specified level of quality on a certain number of measures will not be eligible to share in any savings and will have one year to improve performance or be terminated from the program. ACOs that perform better on these measures are eligible for higher shares of savings. There are no efficiency or resource-use measures, presumably because the payment model itself provides incentives for ACOs to be cost conscious.

    Exhibit 1
    Download Powerpoint Slide






  • Choice of providers. Patients assigned to a Medicare ACO are free to seek care from any health care provider of their choosing. There is no enrollment and patients are not forced to see particular providers within a designated provider network.
What are the issues?

The results from the five-year Medicare Physician Group Practice demonstration suggest that ACOs will be able to improve the quality of care they deliver (at least as measured by process-oriented clinical quality measures) but may have a harder time generating savings. These findings are the best systematic evidence about how the ACO model works, and they point to ways the model might be modified for greater impact.

SMALL SAVINGS: Of the 10 large medical groups participating in the demo, three received no financial bonus at all. Of those that did earn a bonus, the average annual amount was $5.4 million and ranged from a few hundred thousand dollars to about $16 million. Only two participants reduced health spending enough to receive bonuses in all five years.

To put these numbers in context, if new ACOs participating in the Medicare Shared Savings Program were to make the same initial investment that the demonstration participants did ($1.7 million in their first year on average), they will need to turn a 20 percent profit to break even over their first three-year ACO contract with Medicare--a highly unrealistic outcome. Medicare also accrued relatively modest savings from the demo. On net, the demonstration, which covered 220,000 Medicare beneficiaries in a select group of large group practices, saved the Medicare program only $26.6 million or approximately $121 per beneficiary over five years.

The bottom line is that the Physician Group Practice demonstration did not meaningfully reduce spending growth. However, that's not surprising because the existing fee-for-service system penalizes providers for doing what was asked in this demo: namely, to reduce the volume of services they deliver through better care coordination and to pay greater attention to evidence of effectiveness. Moreover, it may have been unreasonable to expect significant changes in behavior in such a limited timeframe. In any case, all 10 participants in the demonstration are continuing to test the model in an extension of the program.

QUALITY IMPROVEMENT: The 10 physician organizations in the Medicare Physician Group Practice demonstration were able to meet performance benchmarks for the vast majority of the quality measures they were held to, which grew from 10 diabetes measures in the first year to 32 measures covering diabetes, coronary artery disease, congestive heart failure, hypertension, and cancer screening by the fifth year.

It is unclear whether using these measures alone will ensure that cost savings are not achieved at the expense of quality. The quality measures selected by CMS for the Medicare Shared Savings Program address some areas that until now have not been given sufficient attention, such as care coordination and care of at-risk populations. On the other hand, they do not cover appropriate referral to specialized expertise and care not available within the ACO.

Other features of the Medicare Shared Savings Program are aimed at improving quality. For example, ACOs must submit written plans to CMS explaining how they will promote beneficiary engagement, coordinate care, promote evidence-based medicine, and measure quality. ACOs must also include a Medicare beneficiary on their governing board or provide an alternative means of assuring meaningful opportunity for beneficiaries to participate in ACO governance.

IMPACT ON HEALTH DISPARITIES: Although ACOs will focus on improving adherence to evidence-based medicine to improve provider performance, it is unclear whether this focus will improve or worsen health disparities among racial and socioeconomic subgroups. On the one hand, minorities may benefit from ACOs' increased attention to keeping patients in good health, which could "raise all boats" and thus shrink the current disparities in care delivery. On the other hand, ACOs may end up primarily forming in geographic areas where a higher proportion of the population has private insurance, and providers are therefore reimbursed more generously. This could inadvertently worsen health disparities if racial subgroups are left behind as other populations are targeted by ACOs.

Perhaps because of these issues, the final CMS regulations go further than their proposed rule initially did to encourage ACOs to form in areas likely to have fewer resources and lower-income patients. For example, federally qualified health centers (usually located in underserved areas), rural health centers, and certain critical-access hospitals (located in remote areas) are allowed to form ACOs. The ultimate impact of these provisions on health disparities remains to be seen.

IMPACT ON PRIVATE INSURERS: ACOs are intended to reduce fragmentation and improve coordination between different providers, which could lead to lower health care use. But they could also produce higher prices as hospitals and physicians consolidate and become more powerful negotiators. This could worsen existing problems: Studies exploring why US health care spending far exceeds that of other countries have already found that substantially higher prices are more important in explaining higher spending than overuse of common services, such as doctor visits and hospitalizations.

ANTITRUST ISSUES: Because of the concern that newly formed ACOs could use their newfound market power to demand and receive higher payment rates from private insurers, the Department of Justice and the Federal Trade Commission had originally proposed a mandatory antitrust review for ACOs that met certain thresholds for provider concentration, for instance, if two or more providers participating in an ACO offer a common service to patients from the same Primary Service Area and have a combined market share of 50 percent or more. The objective was to allow ACOs that are large enough to become accountable for quality and cost, but not so large that they could demand high prices from private health plans because of their market dominance.

Many would-be ACOs opposed this mandatory review for various reasons, including that it would be bad public policy to change the nature of antitrust enforcement from law enforcement to administration of a regulatory regime. In the final rule, CMS no longer requires a letter from a reviewing antitrust agency, but CMS still recommends that prospective ACOs seek a voluntary review by such an agency. As suggested in comments submitted to CMS, a different antitrust enforcement approach would focus on an ACO's actions, not its size and configuration--such as by monitoring per capita costs Medicare ACOs charge for non-Medicare patients.

PAYMENT OPTIONS: CMS is implementing several payment approaches, including offering either shared savings (bonus only) or shared risk (bonus and penalty) in the Medicare Shared Savings Program and shared risk with a transition to partial capitation in the Pioneer ACO demonstration. These should eventually permit an assessment of which payment models are best able to achieve the desired reorientation of clinical practice to improve value for patients and taxpayers.

BENEFICIARY ENROLLMENT: As noted, in the Medicare Shared Savings Program, beneficiaries are free to seek care from any health care provider they choose. But if a beneficiary obtains most of his or her primary care from a provider who belongs to an ACO, that beneficiary's total health care spending, along with quality metrics, will be used to calculate that provider's bonus payments. ACOs will be sent lists of beneficiaries for whom they are likely to be held accountable under CMS's assignment algorithm on a quarterly basis. Then, at the end of the year, CMS will calculate ACOs' shared savings bonus payments based on a reassessment of where those beneficiaries actually ended up receiving a plurality of their primary care services.

CMS calls this approach preliminary prospective assignment. Beneficiaries cannot opt out of having their data used to measure the performance of their provider's ACO, but they will have the opportunity to decline to allow their clinical information to be shared with the ACO to which they are likely to be assigned, for privacy reasons. For their part, private ACO contracts suggest that there is not yet a consensus on the best way to attribute patients to an ACO.

What's next?

As mentioned, the CMS Innovation Center has announced 32 health care organizations that will participate in its Pioneer ACO program, a three-year pilot under which the groups are eligible to earn higher shared savings bonus payments than under the Medicare Shared Savings Program, but the groups will be at risk of paying back larger amounts if they increase spending above projections. In the third and final year of the Pioneer ACO experiment, groups that meet a specified level of savings will be eligible to move a substantial portion of their payments to a population-based model in which they could receive a dollar amount per beneficiary per month--true capitation--instead of continuing to layer ACO bonus payments on top of traditional fee-for-service reimbursement.

The Innovation Center will also allow some ACOs participating in the Medicare Shared Savings Program, including small physician practices and rural community hospitals, to take out loans from CMS to pay for infrastructure investments, such as purchasing electronic health records and hiring nurse care managers. These loans would be deducted from any future shared savings payments the ACO might qualify for from CMS.

CMS estimates that the Medicare Shared Savings Program will generate up to $940 million in net federal savings in its first four years, assuming 270 ACOs sign up to participate. So far, the reception to CMS's final regulations has been positive. By the end of 2012, it should be clear how successful the program has been in attracting provider interest in the ACO model and how extensively the private sector plans to experiment with this payment model. Within a few years after that, there should be a much stronger evidence base about how to improve quality and reduce costs using ACO-style payment arrangements.

Resources

Berenson, Robert A. and Rachel A. Burton, "Accountable Care Organizations in Medicare and the Private Sector: A Status Update," Urban Institute, November 2011.

Fisher, Elliott S., Douglas O. Staiger, Julie P.W. Bynum, and Daniel J. Gottlieb, "Creating Accountable Care Organizations: The Extended Hospital Medical Staff," Health Affairs 26, no. 1 (2007): w44-57 (published online December 5, 2006).

Meyer, Harris, "Accountable Care Organization Prototypes: Winners and Losers?" Health Affairs 30, no. 7 (2011): 1227-31.

Sebelius, Kathleen, "Report to Congress: Physician Group Practice Demonstration Evaluation Report," September 2009.

Shortell, Stephen M., Lawrence P. Casalino, and Elliott S. Fisher, "How the Center for Medicare and Medicaid Innovation Should Test Accountable Care Organizations," Health Affairs 29, no. 7 (2010): 1293-98.

About Health Policy Briefs

Written by
Robert A. Berenson
Institute Fellow
Urban Institute

Rachel A. Burton
Research Associate
Urban Institute

Editorial review by
Ted Agres
Senior Editor
Health Affairs

Anne Schwartz
Deputy Editor
Health Affairs

Susan Dentzer
Editor-in-Chief
Health Affairs

This brief is based on a longer analysis, "Accountable Care Organizations in Medicare and the Private Sector: A Status Update," by Robert A. Berenson and Rachel A. Burton of the Urban Institute, November 2011. The research was funded by the Robert Wood Johnson Foundation.

Health Policy Briefs are produced under a partnership of Health Affairs and the Robert Wood Johnson Foundation.

Cite as:
"Health Policy Brief: Next Steps for ACOs," Health Affairs, January 31, 2012.