Published August 1, 2013
American Enterprise Institute Health Policy Outlook
The most celebrated “bend the curve” idea from the Patient Protection and Affordable Care Act (PPACA) is the Medicare accountable care organization (ACO). ACOs are intended to be integrated health plans, with hospitals, physicians, and other health services providers working together in a coordinated patient delivery system.
The Obama administration and the law’s strongest proponents repeatedly cited ACOs both during consideration of the law and in the years since its passage as evidence that Obamacare was more than just a coverage expansion measure. The public was told that ACOs, along with a few other “delivery system reforms,” would transform the way hospitals and physicians interact with patients and thus slow the rate of growth of the nation’s health spending while improving quality. In sum, ACOs were promoted as a key to a better and more affordable health system.
Yet there never was any evidence to back up these grandiose claims. Indeed, in its cost estimates for the legislation, the Congressional Budget Office (CBO) never assigned anything more than relatively trivial savings to the ACO concept. Moreover, critics noted during the law’s consideration and shortly after its enactment that structural flaws would prevent ACOs—at least as conceived in the PPACA—from ever having a meaningful impact on the delivery system or on cost growth. But that did not stop the relentless hype by the law’s advocates.
More than three years after enactment, it is clear that the ACOs of the PPACA variety will be, at best, minor blips in the nation’s $2.7 trillion health system. This is not to suggest that the real accountable care organizations—the many varied integrated care initiatives underway around the country—are without merit. In fact, these initiatives have the potential to make some real headway toward slowing cost growth and improving quality. But they have absolutely nothing to do with the PPACA, were originated by private actors, very often predate the PPACA’s passage (sometimes by decades), and would continue even if the PPACA’s ACO provisions were repealed entirely.
Before the term lost favor, these were “managed care” efforts, sometimes initiated by insurance companies, and sometimes by provider-driven health systems that have long been in the business of using better patient outcome data to continuously improve their care protocols and processes. These organizations were not invented by the PPACA, and the initiatives they are pursuing often succeed despite the provisions of the new law, not because of them.
Private-sector efforts to build high-performing health systems around the country are making real progress in some cases and should be given the room they need to succeed. But this is not the same as saying the ACO program under the PPACA is a success or that it will be successful down the road, which is very unlikely to be the case.
Fortunately, the shortcomings of the ACO model in the PPACA need not be the end of the story. It remains important to encourage the development of more integrated and cost-effective care within Medicare. And that goal can be achieved by replacing the flawed ACO model of the PPACA with a more workable approach based on incentivizing the formation of high-quality plans and on allowing the beneficiaries to share in the cost savings such plans would produce.
The PPACA’s Medicare ACOs and their Fundamental Flaws
Section 3022 of the PPACA establishes the Medicare Shared Savings Program. The section authorizes physicians and hospitals to form new integrated care networks, dubbed “accountable care organizations” (ACOs). These ACOs would be required to meet a number of requirements established by the US Department of Health and Human Services, including the capacity to oversee the cost and quality of the full spectrum of patient care services. The basic idea was to push doctors and hospitals away from the prevalent fragmented model of fee-for-service (FFS) medicine and toward a model of truly organized and well-managed care.
To incentivize the formation of these ACOs, the law established a new bonus payment system. ACOs that are able to provide care to their patients at costs below what would have otherwise occurred get to keep half of the savings. The Medicare program will make a bonus payment to these ACOs, which will then be distributed among the doctors and hospitals that helped make the cost savings possible. For the first two years of the program, these ACOs are not penalized if their costs exceed baseline projections. But in future years, they are also supposed to be accountable for cost overruns, meaning ACO doctors and hospitals could face reduced Medicare revenue if they exceed their budgets.
These organizations are essentially health maintenance organizations (HMOs), with one notable difference. Unlike HMOs, which must carry insurance licenses issued by the states, the ACOs established in the PPACA need not partner with insurers if they do not want to. Indeed, it is clear that the authors hope the ACOs set in motion by the PPACA will not be insurance companies, in large part because, in the simplistic and distorted view of some of the law’s supporters, insurers add no value to the health system and are nothing but claimspaying operations.
But the irony is that some of the models that the law’s authors point to as precursors to ACOs—such as Geisinger Health System in Pennsylvania and Intermountain Healthcare in Utah—are integrated systems of care that also have extensive experience as licensed HMOs. And the most integrated of all models—Kaiser Permanente—has been operating as a risk-bearing insurer for decades. These organizations have learned to manage the health risks of their enrolled populations through long experience. The organizations have also benefited from access to substantial capital from their insurance business to invest in the sophisticated systems necessary to establish care protocols that improve outcomes and lower costs. The lesson to be learned is that this kind of care model cannot be recreated overnight or without substantial investment.
The hospital and physician communities have reacted with a mix of wariness and fear to the ACO concept as envisioned in the PPACA. Most physician practices are far too small to have the financial capital necessary to build the sophisticated systems of managed care on their own, but they are also wary of allowing hospitals to take the lead role in ACO formation. After all, the primary cost centers of today’s overbuilt, procedure-driven system are expensive inpatient institutions.
Nonetheless, despite wariness, ACO formation has proceeded apace because of current political demands. As of early 2013, there were 106 ACOs registered with the Medicare program, supposedly serving four million Medicare beneficiaries.The hospital and physician communities have reacted with a mix of wariness and fear to the ACO concept as envisioned in the PPACA. Most physician practices are far too small to have the financial capital necessary to build the sophisticated systems of managed care on their own, but they are also wary of allowing hospitals to take the lead role in ACO formation. After all, the primary cost centers of today’s overbuilt, procedure-driven system are expensive inpatient institutions.
But it is much less clear that these ACOs are changing anything of consequence. For the most part, they are the result of larger hospital systems buying more of their affiliated physician practices and calling the resulting entity an ACO. According to a survey of ACOs that was conducted by the consulting group Leavitt Partners, more than half of ACOs are hospital-based enterprises.
This closely resembles what occurred in the late 1990s when physician-hospital organizations (PHOs) were all the craze. Then as now, hospitals bought physician practices in an effort to become more integrated, without a full understanding that genuine integration requires changing how physicians practice medicine. The result was a financial disaster, with hospitals quickly dissolving the PHOs formed only a few years earlier by selling off the physician practices they had acquired. There is no data yet available to evaluate the larger ACO experiment, but previous experience with PHOs should temper expectations.
Recently, the Obama administration tried to pump momentum into its ACO movement by touting the first-year results of the “pioneer” ACOs. These are the hand-picked group of 32 plans selected because they were, for the most, already operating like integrated systems of care long before the law passed in 2010. The administration was hoping that these organizations could pick the low-hanging fruit of costly excess in Medicare FFS and show near-term positive results for the ACO experiment, thus building momentum for the concept.
But even the pioneer ACOs have struggled to deliver on what was promised. A recent, highly preliminary report says that savings from the effort have been very modest. Of the 32 plans that started as pioneer ACOs in 2012, 9 have announced they are pulling out of the program because they fear losing money in the future, 2 increased costs for the Medicare program, and only 18 saved enough money relative to the baseline to be paid a small bonus. With several plans dropping out of the program, there is little reason to expect the “pioneers” to produce impressive results in the future.
The Medicare ACO concept, as written into the PPACA, is fundamentally flawed because the populations the ACOs are supposed to be accountable for managing most often have no idea they are even in an ACO. This was by design. The authors of the law did not want the beneficiaries to be part of the ACO enrollment equation.
So instead of explicitly asking the beneficiaries their preferences, the law has what the authors no doubt thought was a clever work-around: the Medicare beneficiaries would never actually have to agree to enroll in an ACO but would instead be “assigned” to them based on their use of physician services. Medicare administrators would comb through the massive claims database, and any beneficiary whose primary doctor joined an ACO would, de facto, also be an enrolled member of the ACO.
This “passive enrollment” system is the result of policy bias and political calculation. For the most part, the supporters of the law are opposed to using consumer choice and competition to control costs in Medicare. They did not want their new ACO model to look anything like the Medicare Advantage (MA) program, which relies on consumer choice. And so they purposefully removed beneficiaries from the shared-savings element of the program. If costs are lowered from the law’s ACOs, the only winners are the providers and the government.
The authors of the PPACA also wanted to create the illusion of allowing the beneficiaries to remain within Medicare FFS because of its popularity. The intent of the ACO program is clear enough: to herd as many beneficiaries as possible into what is essentially governmentsanctioned HMOs. But the Obama administration did not want the beneficiary community to view it this way. Thus, the ACOs would allegedly dramatically change the delivery system, but the beneficiaries would
This approach to ACO enrollment was a big mistake. It is impossible to run an effectively managed delivery system with an enrolled population that has no incentive whatsoever to stay within the system. Medicare beneficiaries assigned to ACOs are free to get care anywhere they want and will pay no more when they receive care outside of the ACOs they are enrolled in.
The result is that the beneficiaries by and large have no idea they are in an ACO, and even if they do know, they have no incentive to behave any differently than they did before the ACO was initiated. And yet the PPACA makes the ACOs accountable for all expenses incurred by enrolled beneficiaries, including expenses provided outside the ACO network.
The ACOs are thus in an impossible position. The ACOs are further hamstrung by the continued reliance on the FFS payment systems. The hope is that the ACOs will begin to treat the revenue associated with caring for Medicare patients like a fixed budget that must be managed to maximize patient health outcomes. But that would be a far easier culture shift to encourage if the providers participating in the ACO program were forced to get their reimbursements for services from the central ACO structure instead of the Medicare program. But this is not how the law works. Instead, providers in an ACO continue to get paid just as they did before the PPACA: directly from the Medicare program, whenever they render services to Medicare patients.
The only difference under the PPACA is that each year, there is supposed to be a comparison drawn between the total spent for care provided to beneficiaries “enrolled” in an ACO and what would have been spent on them in a non-ACO context, with bonuses and penalties assigned accordingly. The delayed and tenuous connection between ACO operations and what ACO providers get paid is much less effective than the tightly written contacts between HMOs and their provider networks, and therefore also much less effective in influencing provider practice patterns.
A Better Model: Make Plans Accountable to the Beneficiaries
It is ironic that the PPACA’s authors chose to call what was enacted in the law “accountable care” because the ACOs it set in motion are in no way accountable to the beneficiaries that should be rendering judgment on them. If anything, the ACOs are accountable only to the federal government, with its lengthy rules and cumbersome quality reporting requirements. The beneficiaries are never given a real, transparent enrollment choice to join an ACO or alternative coverage options such as MA plans.
Some of the early proponents of the ACO concept as enacted in the PPACA are belatedly regretting the law’s flawed design and are now admitting that the “passive enrollment” model will not work to achieve the goals of reduced cost growth. The public was emphatically told that the PPACA had the provisions necessary to bring in a new era of higher quality and lower cost care, largely because of ACOs. It is certainly convenient to admit now that the concept will not work as advertised.
The solution, however, is not more federal micromanagement of the ACO program, as some are suggesting. The solution is a model that incorporates provider-driven managed care into a larger system of genuine consumer choice and marketplace competition.
The starting point for reform should be the realization that Medicare already had an ACO program before enactment of the PPACA. The MA program gives beneficiaries the option to get their Medicare coverage from private insurance plans, many of which are genuine HMOs with varying levels of coordination and integration.
Low-cost, high-value, provider-driven organizations that also happen to partner with insurance plans or operate as licensed insurers themselves can already compete for Medicare business in the MA program. And many do so today, including Geisinger and other models that the Obama administration touts as exemplars. Recent studies confirm that, on an apples-to-apples basis, the best-performing MA HMOs can offer the Medicare benefit package at premiums that are below the cost of FFS in most parts of the country.
The fastest, surest way to reward innovative ACO-like organizations would be to make it even easier for low-cost, high-value delivery systems to build market share by competing on a level playing field with Medicare’s dominant FFS insurance system. This is the premise of the “premium support” model of Medicare reform. Under premium support, private insurance options—the MA plans of today, plus whatever new models may emerge—would compete directly with the government-administered FFS option on a regional basis, and the beneficiaries would select their coverage from the competing options. The government’s contribution to coverage would be set based on the submitted premium
bids of the competing plans.
For instance, some versions of premium support would set the government’s contribution based on the weighted average of bids from the plans, including FFS Medicare. Importantly, the government’s contribution would not increase if a beneficiary chose a relatively expensive coverage option. The additional premium for such plans would be paid entirely by the Medicare enrollee, not the government. This would provide a powerful incentive for beneficiaries to seek out high-value, low-cost plans.
A recent series of AEI reports made the case that a premium support approach to reform within Medicare is necessary both because it is an essential step toward improving the productivity of the broader health delivery system and because it would substantially reduce federal Medicare spending in both the short and long run. And, of course, premium support ensures genuine accountability: beneficiaries can disenroll from plans that provide poor quality care or charge too much.
For the moment, the premium support model is unlikely to be adopted because the Obama administration prefers a more regulatory approach with Medicare. But that need not stop all progress. Short of full premium support, it is still possible to move toward a system in which the beneficiaries are directly engaged in all of their coverage options. Doing so would provide a strong boost to provider-driven integrated care organizations that ACO supporters say they want to encourage in Medicare.
One approach to jumpstarting reform would be to combine changes in today’s supplemental insurance system with a better integrated care option. Unmanaged FFS insurance does not work well if the beneficiaries are not part of the cost equation. With no check on use of services coming from the providers in a FFS environment, only beneficiary cost sharing can slow down overuse of services.
Medicare law requires large cost sharing on the part of the beneficiaries, but it is completely ineffective because the vast majority of FFS enrollees—90 percent, according to data from the Medicare Payment Advisory Commission—have secondary insurance that pays for the all of the costs that Medicare does not cover. This secondary insurance comes in the form of Medigap plans, employer-sponsored wraparound coverage, or Medicaid. Beneficiaries enrolled in these secondary insurance plans pay almost nothing at the point of service, and thus there is very little reason for them not to agree to the extra procedures and tests that FFS medicine encourages.
The predictable result is a steady and unrelenting rise in the volume and intensity of services provided to Medicare FFS enrollees. CBO has estimated that between 1997 and 2005, the real price per capita paid for physician services declined by about 5 percent, but overall spending per FFS enrollee rose nearly 40 percent because of an explosion in the use of services.
A first-step Medicare reform could combine reform of the supplemental insurance system with a better integrated care option than the current ACOs offer. This approach would involve three steps.
First, the ACO concept should be replaced with a new integrated care option in Medicare FFS. Beneficiaries would not be passively enrolled in these plans as they are in ACOs but would instead be given the option to enroll in these plans at the same time that MA plans are open for enrollment.
Second, these integrated plans would compete with each other and with MA plans on price and quality. They would be required to provide to Medicare’s administrators estimated cost-reduction amounts that would be shared with the beneficiaries in the form of reduced Medicare premiums. The reduced premiums paid by these beneficiaries when they enrolled in an integrated plan would not result in a financial loss for the government. Instead, the government would reduce the amounts paid for services rendered by the integrated plan’s physicians, hospitals, and other service providers to cover, at a minimum, the total amount of reduced premiums paid by the plan’s enrollees (this policy could be calibrated to provide savings to the government, too).
Thus, the government would be assured of spending no more than it would without the new integrated care option, and the beneficiaries would have strong financial incentives to enroll in more efficient care arrangements. Under this approach, the integrated plans would be paid using the same fee-for-service payment schedules in use today, but the plans would have the option to direct the payments to a centralized entity, not to the individual providers. Plans electing this option would be free to design their own provider compensation programs without meddling from the Medicare program.
Third, supplemental insurance plans offered by employers and insurers would be prohibited from entirely filling in Medicare’s cost sharing, except for beneficiaries enrolled in integrated care options. This rule would apply to the health benefit programs for retired federal employees and for retired members of the military, both of which serve as secondary insurance to Medicare for millions of former government workers and military personnel. Beneficiaries who wish to remain in unmanaged fee-for-service could do so, but they would be required to pay some of the cost at the point of service, up to a reasonable out-ofpocket limit.
These reforms would ensure that low-cost, high-value, and provider-driven managed care options would be given ample room to compete and thrive in Medicare. The beneficiaries would have strong incentives to enroll in them to reduce their out-of-pocket costs, and the integrated plans would have the control they need over their enrolled populations to make real and lasting changes in how health care is delivered.
The authors of the PPACA overpromised on a number of fronts, including ACOs. They sought a reform that would involve no political controversy but that would somehow magically herd millions of seniors into tightly controlled managed care arrangements. That is never going to happen.
A better approach is to treat seniors as the responsible citizens they are. They should be given real choices and options, and the opportunity to share in the savings from a more productive delivery system. A program built on accountability to patients and consumers will have far more success than one focused on pleasing the government.
James C. Capretta is a senior fellow at the Ethics and Public Policy Center and a visiting scholar at AEI.
1. The CBO’s cost estimate for the final legislation projected cost savings from the Medicare ACO concept at $4.9 billion over 10 years. See Congressional Budget Office, “Letter to the Honorable Nancy Pelosi,” March 20, 2010, www.cbo.gov/sites /default/files/cbofiles/ftpdocs/113xx/doc11379/amendreconprop.pdf, table 5.
2. For a description of various integrated care systems that long predate the PPACA, see Anthony Shih et al., Organizing the U.S. Health Care Delivery System for High Performance (The Commonwealth Fund, Commission on a High Performance Health System, August 2008), www.commonwealthfund.org/~/media/Files/Publications/Fund%20Report/2008/Aug/Organizing%20the%20U%20S%20%20Health%20Care
%20Delivery%20System%20for%20High%20Performance/Shih_organizingushltcaredeliverysys_1155%20pdf.pdf, appendix A2.
3. Several prominent integrated care plans participate in Medicare’s private health plan option, Medicare Advantage. The PPACA reduces payments to Medicare Advantage plans by $156 billion over the period 2013–22. Those cuts will lower enrollment in Medicare Advantage plans, and thus also induce a migration out of integrated plans and back to Medicare fee-for-service. See Congressional Budget Office, “Letter to the Honorable John Boehner,” July 24, 2012, www.cbo.gov/sites /default/files/cbofiles/attachments/43471-hr6079.pdf.
4. Office of the Legislative Counsel for the Use of the US House of Representatives, “Compilation of the Patient Protection and Affordable Care Act,” May 2010, http://housedocs.house.gov/energycommerce/ppacacon.pdf.
5. Centers for Medicare & Medicaid Services, “More Doctors, Hospitals Partner to Coordinate Care for People with Medicare: Providers Form 106 Accountable Care Organizations,” press release, January 10, 2013, www.cms.gov/apps/media/press/release.asp?Counter=4501.
6. David Muhlestein et al., Growth and Dispersion of Accountable Care Organizations: June 2012 Update (Salt Lake City, UT: Leavitt Partners, June 2012), http://leavittpartners.com/wp-content/uploads/2013/03/Growth-and-Dispersion-of-ACOs-June-2012-Update-Download.pdf.
7. Melinda Beck, “Mixed Results in Health Pilot Plan,” Wall Street Journal, July 16, 2013.
8. The secretary of the US Department of Health and Human Services was given the authority in the statute—in the new section 1899(c) of the Medicare law created by the shared savings program—to establish a method of assigning Medicare beneficiaries to ACOs based on their use of primary care physician services.
9. See Steven M. Lieberman, “Reforming Medicare Through ‘Version 2.0’ of Accountable Care,”Health Affairs 32, no. 7 (2013): 1258–64.
10. Gretchen Jacobson et al., Transforming Medicare into a Premium Support System: Implications for Beneficiary Premiums (Kaiser Family Foundation, October 2012), http://kaiserfamilyfoundation.files.wordpress.com/2013/01/8373.pdf.
11. See James C. Capretta, The Role of Medicare in Inefficient Health Care Delivery (Washington, DC: AEI, April 2013), www.aei.org/files/2013/04/15/-the-role-of-medicare-feeforservice-in-inefficient-health-care-delivery_141413376272.pdf; Joseph Antos, Plan Competition and Consumer Choice in Medicare: The Case for Premium Support (Washington, DC: AEI, April 2013), www.aei.org/files/2013/04/15/-plan-competition-and-consumer-choice-in-medicare-the-case-for-premium-support_141511443922.pdf; and Roger Feldman, Bryan Dowd, and Roger Coulam, A Competitive Bidding Approach to Medicare Reform (Washington, DC: AEI, April 2013), www.aei.org/files/2013/04/15/-a-competitive-bidding- approach-to-medicare-reform_141610273790.pdf.
12. Medicare Payment Advisory Commission, Report to Congress: Medicare and the Health Care Delivery System (Washington, DC, June 2012), 6, www.medpac.gov/chapters/Jun12_ch01.pdf.
13. Congressional Budget Office, Factors Underlying the Growth in Medicare’s Spending for Physician Services (Washington, DC, June 2007), www.cbo.gov/sites/default/files/cbofiles/ftpdocs/81xx/doc8193/06-06-medicarespending.pdf.
14. The Bipartisan Policy Center (BPC) recently released a plan with similar features to this proposal, titled “Medicare Networks.” The BPC plan relies more on differential provider reimbursement rates to incentivize the formation of the Medicare networks and less on direct and transparent premium competition among the networks to drive enrollment. The plan also differs in how beneficiaries would share in the savings, including the provision of a $60 per beneficiary rebate in the first three years regardless of the performance of the program. Finally, the plan does not include supplemental insurance reform. For a full description of the proposal, see Bipartisan Policy Center, A Bipartisan Rx for Patient-Centered Care and System-Wide Cost Containment (Washington, DC, April 2013), http://bipartisan policy.org/sites/default/files/BPC%20Cost%20Containment %20Report.pdf.