Advertisement
Mayo Clinic Proceedings Home

Accountable Care Organization Pathways: Diverse but Ultimately Parallel

  • Len M. Nichols
    Correspondence
    Correspondence: Address to Len M. Nichols, PhD, Center for Health Policy Research and Ethics, George Mason University, 4400 University Dr, Mail stop 207, Fairfax, VA 22030
    Affiliations
    Center for Health Policy Research and Ethics, George Mason University, Fairfax, VA
    Search for articles by this author
      A prescient visitor from Mars might easily conclude that the US health reform debate is actually a diversion. Most of what he hears and reads is about whether the federal government has the authority to require all individuals to obtain health care coverage so that it can reorganize insurance markets to serve all of us and not just some of us. Now, these are indeed important issues, although they are primarily philosophical and moral, respectively, but readers of Mayo Clinic Proceedings know that the most important part of our health reform debate is over delivery system reform. Simply put, effective access for tens of millions of Americans is at risk—not to mention most other public priorities and middle-class jobs in a global economy—if we fail to improve the efficiency with which high-quality care is delivered. Excessive growth of health care costs is by far the largest contributor to our fiscal stress and stagnant middle-class standard of living. After all the Sturm und Drang, this is why the Patient Protection and Affordable Care Act (PPACA) was finally passed, although it is hardly a panacea or guarantor of desirable outcomes. Powerful interests are threatened by an emphasis on efficiency, quality, and evidence-based care, and many are finding the necessary transition—from pay-for-volume fee-for-service to any new payment system that rewards value instead—to be something between a heavy lift and a long hard slog. We are nowhere near the end of the beginning of health care system transformation in our nation.
      No single idea for transforming health care delivery in the past few years has received half as much attention as accountable care organizations (ACOs). Advocates
      • Fisher E.S.
      • McClellan M.B.
      • Bertko J.
      • et al.
      Fostering accountable care: moving forward in Medicare.
      and Administration officials
      • Berwick D.M.
      Making good on ACO's promise—the final rule for the Medicare shared savings program.
      alike repeatedly emphasized the centrality of population-based coordinated care and incentive realignment with risk bearing and sharing so that health care institutions' financial interests could be linked to the efficient delivery of high-quality care. The PPACA, reflecting its authors' strong desire to jump-start our move away from volume-based payment methods, included an unprecedented new nationwide program, as opposed to a more typical narrowly focused demonstration or pilot project, that offered new payment and quality reporting rules to any set of Medicare-participating physicians and hospitals willing to accept the brave new challenges of accountable care. Participation in how-to conferences and collaboratives spread like wildfire, for it seemed like the movement to full financial and performance risk for a defined patient population was the clear direction American health care policy was moving. But then the initial Medicare Shared Savings Program ACO proposed rule (set of regulations) was released, and enthusiasm turned to disappointment, anger, and a search for viable alternatives.
      The Center for Medicare and Medicaid Innovation (CMMI), created by the PPACA to be the catalyst for delivery system reform, as quickly as possible developed an alternative model pilot ACO project, aptly named Pioneer ACO, even as it digested the tsunami of formal comments about how to improve the Shared Savings ACO model proposed rule. The final shared savings rule was far more palatable to most observers, although only 32 and 116 organizations have signed on to become Pioneer and Medicare Shared Savings Program ACOs, respectively, at this point. While they have important differences, both ACO concepts share 3 key features for health care organizations: (1) accept clinical and financial responsibility for a defined group of Medicare beneficiaries, essentially practice population health management, (2) measure and report a wide array of quality metrics, and (3) benefit financially vs baseline if, and only if, quality and cost saving targets are met.
      The major difference in the models is the much greater degree of financial risk and reward allowed to and expected of Pioneer ACOs, for they are required to accept population-based payment risk by year 3 for certain Medicare beneficiaries and to have at least 50% of their private payer contracts also be risk based by then, in the hope of spreading the payment innovation to the health care system more generally. It is also worthy of note that the Pioneer pilot is larger than the Medicare Shared Savings Program. However, the small number of ACOs in total, after all the discussion during the legislative debate and in the year before the first proposed rule came out, is a disappointing indicator of the reality that most health care organizations are still reluctant to move from fee-for-service to outcomes-based payment without more protection and support.
      In the commentaries published in this issue of Mayo Clinic Proceedings, 6 leading health care institutions explain why they did or did not decide to participate in 1 of the 2 ACO models now operational at the Centers for Medicare and Medicaid Services (CMS). I will certainly let the authors speak for themselves, but as an economist, I should highlight some fascinating common themes.

      Patients Are not Necessarily Engaged

      No one really likes the so-called patient attribution rules, mechanisms for “assigning” patients to the ACO. Shared Savings ACOs must accept retrospective attribution, which means they will not know which patients their performance is being held accountable for until after the fact. This more or less forces them to treat all patients the same, which is arguably good. However, it also risks diluting their ability to bring specific tools and process innovations to bear on patients who could benefit the most since they might be “wasted” on patients for whom it does not matter to their bottom line, and therefore there can be no savings with which to help finance the intervention that produces the savings in the first place. Basic economics has long argued that useful services will be undersupplied to the marketplace if the investor is prevented from reaping the rewards from the investment. In addition, health care institutions' ignorance about which patients will affect a practice's bottom line violates a basic precondition for markets to perform well, that both buyers and sellers have perfect knowledge about the transaction they might enter. Retrospective assignment is simply problematic on both grounds.
      Pioneer ACOs have the option of choosing prospective attribution, which means you could at least know which patients your performance is being judged on ahead of time, but the rules prohibit either the ACO or Medicare from doing anything, even providing incentives, to keep the attributed patient “in the network” of the ACO. This means that the ACOs could be penalized (or rewarded) for cost and quality performance that they literally did not contribute to.
      One must infer that the Center for Medicare and Medicaid Innovation (CMMI)/CMS/Department of Health and Human Services/Obama Administration basically made a hard judgment call that ACOs—at least in this initial phase—are more appealing to patients if the patients know they can always visit other clinicians and hospitals any time they prefer. That is surely true, but it also means that the models are far less attractive to health care organizations that are naturally fearful of accepting performance risk that they cannot control. The other serious downside of allowing absolute patient choice at all times and with no financial consequences is that it seriously missed an opportunity to test the combined effect of truly engaged patients along with new health care organization incentives. Since it is increasingly clear that for us to get health care cost growth under control, patients are going to have to assume greater responsibility for their own behavior, health, and health improvement and maintenance (through compliance as well as regular attention to diet and exercise), this seems like an omission based more on politics than health care policy. An opportunity was therefore also lost to speak clearly and honestly to the public at large about the responsibilities we must all assume to make our health care system sustainable. Given the slander about “death panels” coming from conservative opponents of reform, the White House judgment call is understandable, but a disappointing one that should be remedied in the next wave of ACO models.
      Baylor Health Care System (Dallas, Texas) had a particular problem—ultimately the main reason they did not become a Medicare Shared Savings Program ACO—with the CMMI's refusal to reconsider its insistence that patients be attributed on the basis of the tax identification number of the physician group where most care was delivered rather than on the basis of an identifier for the specific physician who provides the majority of the patient's evaluation and management services. The Baylor modification
      • Couch C.E.
      Why Baylor Health Care System would like to file for Medicare Shared Savings ACO designation but cannot.
      would be more fair in cases in which some but not all members of a given physician practice are participating ACO physicians. The CMMI method puts the ACO at risk for patients mostly seeing a physician who is manifestly not participating in the ACO activities, risks, and rewards. In trying to force ACOs to recruit more physicians to join, this rule cost the program at least one, and surely more, ACO candidates. Hopefully, different tradeoffs will be operational in the future.
      Partners HealthCare (Boston, Massachusetts) was the only organization to express anything positive about the extreme degree of patient choice in both ACO models, for they saw it as removing the stigma of primary care physician as gatekeeper.
      • Milford C.E.
      • Ferris T.G.
      A modified “golden rule” for health care organizations.
      They are also the only ACO participant to emphasize the importance of having many specialists involved in the care of patients, which is counter to the approach of limiting specialist access when possible and appropriate. There is, of course, much dispute about where to draw these lines, but Partners' support for the concept validates the other ACOs' and ACO candidates' fear of unfettered patient choice when a more limited set of health care professionals is accountable for overall quality and cost performance they cannot fully control.

      Risk-Reward-Investment Tradeoffs Are Hard for CMMI to Calibrate

      The high infrastructure cost of being able to report the requisite quality metrics and to institute the information system plus clinical and financial management tools to improve performance over time were some of the main reasons many organizations decided not to become ACOs. Reducing the number of quality metrics required from 65 to 33 surely helped, as did permitting Shared Savings ACOs to avoid downside risk altogether (if they want to accept lower reward upsides). In addition, permitting the boldest Pioneer ACOs to gain if they reduce cost growth below the target rate and if their level of cost is lower than the fixed population-based payment also encouraged those who believe they can manage risk to enter and adopt the more aggressive risk-reward options. However, the CMMI cannot be a force for quality improvement without demanding far more quality measurement than is the norm today, and Medicare has to keep some of the savings to minimize taxpayer pain in the long run, so it is hard to see how much more it can relax these parameters in not-yet participating institutions' favor.

      Long-run Visions of Participants and Nonparticipants Are Remarkably Similar

      No author and organization whose views are reflected in the commentaries published in this issue of Mayo Clinic Proceedings thinks that the goals of the PPACA or the ACO experiments are misguided. Those that chose not to participate are proceeding to develop ACO-like capacities, Baylor Health Care System quite rapidly because of its more sophisticated information technology infrastructure and long-time quality improvement experience; SSM Health Care St Louis
      • Porter R.G.
      • Tosto A.P.
      The SSM Health Care approach to achieving “true north”: improving health care quality while reducing costs.
      and Health Choice
      • Mayzell G.
      A physician hospital organization's approach to clinical integration and accountable care.
      more slowly but about as fast as they think they can, which is faster than many organizations are moving today. The question for all will be, can they find payers willing to move toward value- and performance-based payment at their preferred pace, for without payment reform and aligned incentives among the full range of physicians and hospitals, real-care transformation cannot occur and the goals of more efficient high-quality care and better health will not be realized. Health Choice clearly prefers Medicare Advantage plans to Medicare FFS (fee-for-service) plans as a payer, since those plans can provide actionable data much faster than the CMS. This is clearly a problem that should be remedied as soon as possible.
      Interestingly, while all participants hope for short-term success, they also recognize that there are long-term gains to participating regardless of specific captured rewards. Atlantic Health System (Morristown, New Jersey) is already happy to report that many more clinicians than before are energized to work on quality improvement daily.
      • Shulkin D.J.
      Building an accountable care organization for all the wrong reasons.
      Ascension Health Partners will use its exemplary Pioneer ACOs—Seton Health Alliance ACO (Austin, Texas) and Genesys Physician Hospital Organization (Flint, Michigan)—to teach the rest of its large system about both medical and financial management as well as strategies for engendering physician engagement around values and shared business goals.
      • Anderson R.D.
      • Aderholdt E.
      • Chenven N.
      • et al.
      Ascension Health partners with Centers for Medicare and Medicaid Services' to provide patient-centered care through the Pioneer Accountable Care Organization model.
      Partners HealthCare is clearly desirous of encouraging specialists to share in the gains from reform and thereby avoid the zero-sum nature of delivery reforms in the 1990s.
      • Milford C.E.
      • Ferris T.G.
      A modified “golden rule” for health care organizations.
      This common vision is an important takeaway, for we really do face an urgent task of reining in cost growth as humanely as possible. We must and surely will find a way to put our fiscal house in order. The question is, will the successful way entail large benefit and price cuts and much human suffering or incentive realignments and quality improvements as envisioned in the PPACA and ACO models? Given the direction of Medicare's movement, we are starting to see a groundswell of parallel incentive realignments in the private sector as well—medical homes, bundled payment arrangements, and even full-scale ACOs.
      • Ignani K.
      Health plan leadership in delivery system reform (slide presentation) America’s Health Insurance Plans Summit on Shared Accountability.
      • Higgins A.
      • Stewart K.
      • Dawson K.
      • Bocchino C.
      Early lessons from accountable care models in the private sector: partnerships between health plans and providers.
      Transforming care delivery.
      These new incentive alignments, like the CMMI ACOs, are just as or more likely to be led by independent physician groups as by integrated health care systems. So the good news is, the practice vision of incentive realignment is starting to spread across the entire and diverse health care sector, not just in pockets of institutions that happen to have highly integrated systems.
      However, if the law is repealed next year despite the favorable ruling by the Supreme Court of the United States, many of the tools for incentive realignment, but more importantly the momentum in this triple-aim direction, will have been lost. The Atlantic group
      • Shulkin D.J.
      Building an accountable care organization for all the wrong reasons.
      explicitly wrote that market forces in New Jersey are not strong enough to produce these incentive realignment tools in the absence of government ACOs. It is hard to see how the movement can continue to overcome the myriad barriers and spread without continued pressure from and the specific example of the largest payer in the country, the Medicare program. While a bare-bones premium support model like US House Budget Chairman Paul Ryan (R-WI) has proposed would exert even more pressure on public-sector health care spending by getting Medicare out of the fee-for-service business, his plan would undermine the potential transformative power Medicare has had as a buyer before (diagnosis-related groups, resource-based relative value scale) and could have again (ACOs, bundled payments, patient-centered medical homes). Post-2012, a bipartisan consensus around ACO-like payment reform could arise, but it's hard to be optimistic about bipartisan anything these days. Therefore, working to retain the useful tools that we have to tame the cost growth beast seems like a prudent aim, even as we also work to learn from, modify, and improve those tools so more clinicians and patients may ultimately benefit from them.

      References

        • Fisher E.S.
        • McClellan M.B.
        • Bertko J.
        • et al.
        Fostering accountable care: moving forward in Medicare.
        Health Aff (Millwood). 2009; 28: w219-w231
        • Berwick D.M.
        Making good on ACO's promise—the final rule for the Medicare shared savings program.
        N Engl J Med. 2011; 365: 1753-1756
        • Couch C.E.
        Why Baylor Health Care System would like to file for Medicare Shared Savings ACO designation but cannot.
        Mayo Clin Proc. 2012; 87: 723-726
        • Milford C.E.
        • Ferris T.G.
        A modified “golden rule” for health care organizations.
        Mayo Clin Proc. 2012; 87: 717-720
        • Porter R.G.
        • Tosto A.P.
        The SSM Health Care approach to achieving “true north”: improving health care quality while reducing costs.
        Mayo Clin Proc. 2012; 87: 729-731
        • Mayzell G.
        A physician hospital organization's approach to clinical integration and accountable care.
        Mayo Clin Proc. 2012; 87: 727-728
        • Shulkin D.J.
        Building an accountable care organization for all the wrong reasons.
        Mayo Clin Proc. 2012; 87: 721-722
        • Anderson R.D.
        • Aderholdt E.
        • Chenven N.
        • et al.
        Ascension Health partners with Centers for Medicare and Medicaid Services' to provide patient-centered care through the Pioneer Accountable Care Organization model.
        Mayo Clin Proc. 2012; 87: 714-716
        • Ignani K.
        Health plan leadership in delivery system reform (slide presentation).
        America’s Health Insurance Plans Coverage Web site, Washington, DCOctober 18, 2011 (Accessed April 24, 2012)
        • Higgins A.
        • Stewart K.
        • Dawson K.
        • Bocchino C.
        Early lessons from accountable care models in the private sector: partnerships between health plans and providers.
        Health Aff (Millwood). 2011; 30: 1718-1727
      1. Transforming care delivery.
        AHIP Issue Brief. 2012;