Accountable care organizations (ACOs) are widely regarded as part of the solution to a fragmented health care system — one plagued by duplicative services, avoidable errors, and other impediments to efficiency and quality. But 20 years of reform efforts have led to a wave of provider consolidation that has made little headway in efficiently coordinating care. Providers continue to follow a strategy that has shown minimal evidence of success.

We should admit that we don’t know what works and, instead, test a variety of potential solutions that could address fragmentation. Before I explore the concrete steps we can take to encourage that kind of innovation, let me provide some important historical context.

Payment Reform’s First Life

Early efforts to promote coordinated care emphasized payment reform. Toward that end, managed-care and health maintenance organizations used payment schedules and gatekeeper physicians to create provider networks. In addition, the Clinton administration introduced proposals to implement “pay for performance” and dedicated quality-improvement initiatives, suggesting that financial pressures might force the coordination and rationalization of care. But Congress rejected payment-focused reform, and market preferences eliminated managed-care pressures.

Commentators then suggested that payment reform could happen only in conjunction with provider-based reforms, and the Institute of Medicine later issued a series of reports calling for pairing payment solutions with structural reform. Then, when the Affordable Care Act instituted Medicare’s Shared Savings Program in 2010, it invited providers to create ACOs and to accept changes in reimbursement that allowed them recoup part of any savings they generated. Eventually, however, prospective ACOs were given the option ofcontinuing under Medicare’s traditional fee-for-service payments. In other words, providers were encouraged to pursue structural reform while being permitted to avoid any constraint from payment reform.

INSIGHT CENTER

The Disappointments of Provider Reform

The continued failure of payment-driven reform has sadly given provider-based reform a blank check. The U.S. health sector has been in a merger-and-acquisition frenzy for nearly 20 years, and much of the integration has been justified as an effort to construct ACOs. Buzz phrases such as “clinical integration” and “eliminating fragmentation” are routinely paraded before regulators who scrutinize proposed mergers.

The problem, of course, is that after waves of acquisitions, most hospital markets are now highly concentrated and lack meaningful competition. And, consistent with basic economic theory, hospital systems that acquired dominant market shares dramatically increased prices for health care services. Perhaps even worse is that these large entities have shown little capacity for achieving the efficiencies they promised through coordinated care. Newly integrated delivery systems retain their inefficiencies and bring higher prices without any evident reduction in costs or errors.

We don’t know exactly why efforts at integration have not yielded efficiencies, and it seems we simply didn’t think very hard about it. The health reform debate focused primarily on a handful of success stories we all can repeat in our sleep: Kaiser, Geisinger, Intermountain. The plan was to have other hospital systems mimic them. That is like instructing all high-tech companies to mimic Apple, as if what makes Apple successful is an easy-to-follow cookbook for large-scale structural change.

It is a curiosity about the U.S. health system that producers with better outcomes and lower costs than their competitors cannot dominate the market. Kaiser, for example, has tried but failed to enter more local markets. But it is foolhardy to think that the systems that have not achieved Kaiser’s success can replicate it simply with the help of government regulators. This duplication strategy at best seems mindless, and at worst smacks of a Khrushchev-era economic policy.

The truth is, despite a glut of business press and how-to manuals, we still understand very little about why certain organizations succeed and others do not. With all the complexities of delivering medical care, we should expect even more variation among health care providers than among manufacturers. We likewise should be very hesitant to claim we understand what works and prescribe nationwide structural reforms.

Concrete Steps for the Future of ACOs

Precisely because we don’t know what works at this juncture, we cannot continue encouraging the formation of vast integrated systems that are difficult to disentangle. Until we have more evidence that integration yields efficiencies, regulators should continue to halt mergers that harm competition.

But scrutinizing mergers will only prevent further damage. We also must improve our delivery system, and we cannot give up on the ACO as a potential source of innovative configurations. Specifically, we should:

1. Redefine and broaden our concept of an ACO. Too much ACO formation has emphasized linking hospitals with other providers. Instead of this top-down approach, we should work from the bottom up by linking providers withconsumers and payors, so that the focus is on serving patients’ needs and managing budgets.
2. Encourage nontraditional parties — such as social workers, professionals who help people navigate the health care system (often called “navigators”), and IT companies — to lead efforts at ACO formation. These parties would be well equipped to construct networks that provide accountability, given their expertise in connecting consumers to complex organizations and advocating on behalf of those consumers.
3. Use contract-based and virtual provider collaborations instead of relying on mergers. Joining providers under common ownership might not be necessary. Electronic health records (EHRs) and other information technologies have the potential to create platforms that enable coordination without incurring the high costs of integration. EHR tools can also allow patients to control their own information and tailor collaborations to individual patients’ needs.
4. Entertain disruptively innovative reconstructions of the health care delivery system — ones that make use of mobile health, medical tourism, and informatics. Many technology companies that traditionally have not participated in the health sector are now offering improvements to our delivery system. Because business scholarship tells us that outsiders frequently introduce the most valuable innovations to a market, we should ensure that regulatory barriers do not preclude participation from unconventional participants.
5. Perhaps most important, we cannot pursue structural reform withoutpayment reform. We will distinguish valuable provider reforms from ineffective ones only if sustained revenue pressures force ACOs to be truly “accountable” to consumer demands and other economic realities.

Not every solution we try will work, but we’re likely to have more success letting providers figure out what works than telling them how to do it.