The accountable care organization movement took a hit when eight of the 32 pioneer ACOs turned back, applying instead as regular shared savings model ACOs or, in one case, dropping out of the program. The health care blogosphere (which I believe now accounts for 17 percent of the gross national product all by itself) lit up with spinning and spitting on the program from ACO boosters and critics alike. Why did some pioneers (arguably the smartest, best-qualified managed care players in the country) turn back from a program that they themselves helped to shape? Well, as they say on Facebook, when it comes to relationship status: It's complicated.
I talked to a number of the CEOs and industry notables involved, and share my synthesis of their off-the-record and candid insights.
ACO Pioneers: What We Know
A good starting point in understanding the ACO pioneer story is to read my friend and fellow futurist Jeff Goldsmith's excellent piece (with attached references) that he posted on Aug. 15. Jeff describes the similar flaws (some say fatal) of ACOs and the physician group practice demonstrations of 2005 through 2010. As one veteran health plan CEO told me: "It's tough to be responsible for the care of a defined population when you don't know who they are until after you deliver the care, they don't know and don't care they're in it, and you let them go anywhere they want." These were violations of three of Morrison's 10 Laws of Accountable Care that I shared with readers when ACOs were launched.
The other key insight from Jeff's piece is the surprising resilience of Medicare Advantage. As I also will argue, the key takeaway from all this pioneer experience should be that ACOs are not an end state but, I hope, a means to an end. At their best, they are training wheels for capitation. The end game should be clinically integrated organizations at risk for the care of a defined population, with member incentives to stay in network and provider incentives to coordinate care, based on true and fair risk-adjusted capitation. The result is Medicare Advantage or some modernized derivative for all Medicare beneficiaries (and, actually, while you are at it, maybe all Americans; but that is a ways off).
Why Become Pioneers Anyway?
When the ACO program first was launched, there was significant blowback from advanced managed care organizations that really would have much preferred full, risk-adjusted capitation for a new cohort of Medicare recipients. That was not being offered; indeed, the perceived "subsidies" to Medicare Advantage would be trimmed under the Accountable Care Act. Instead, ACO shared savings models were offered as a new alternative.
Against this backdrop, the pioneer program was born as a place for the ninja warriors of coordinated care to do their thing. As one industry veteran told me: "The pioneer program was born out of patriotism and a little bit of greed." In other words, the pioneers were trying to be supportive of the ACA and the ACO movement.
Even though the specifics of the program weren't what the pioneers wanted, they thought they could make it work because they were confident in their ability to manage care. They thought they could make it work financially in the short run; in the long run, they were confident they could convert happy Medicare beneficiaries from fee for service to Medicare Advantage. And let's be clear: If many of those pioneer organizations were given full, risk-adjusted capitation for the newly attributed, they probably would have done just that.
The other key point that the Centers for Medicare & Medicaid Services and the Obama administration wanted to make clear was that beneficiaries always had a choice to go wherever they wanted. ("Member choice and flexibility was always the mantra from CMS," one CEO told me.) Many of us thought being able to skate away to the Mayo Clinic and having the ACO financially responsible for that was a goofy idea from the beginning; but it was and is centrally important to the politics, optics and positioning of the ACO program.
Why Some Pioneers Struggled
All ACO pioneers reportedly improved quality slightly and, overall, the ACO pioneers saved a little bit of money (about $80 million in the first year). This is consistent with the original ACO program estimate of approximately $800 million over 10 years. (Just to put that into perspective, the 2 percent sequestration cuts on hospitals took $43 billion out of hospitals over a decade. So if you were trying to control Medicare costs, which idea works better? Hum, let's see … .)
There are a number of reasons ACO pioneers struggled on the financials.
It's tough to get better when you are already really good. Several of the ACO pioneers that dropped out or dropped back from pioneer status had among the lowest Medicare costs in the country, with low bed days per thousand, high levels of existing care coordination and sophisticated care management systems. For them, making financial improvement was a tougher bar. "There were no low hanging fruit left," one CEO told me.
Advanced managed care markets likely will have older attributed patients. It is no accident, in my view, that many of the pioneers who turned back were in areas with an extremely high penetration of Medicare Advantage plans: Southern California, New Mexico, Minnesota and Massachusetts in particular have Medicare Advantage penetration rates around 40 percent. In those states it was much more likely that the patients who were attributed to the ACO were older and sicker and more likely to be dual eligibles (a population that has a 30 to 35 percent higher utilization than traditional beneficiaries). Several CEOs confirmed that the proportion of frailer, older patients was significantly higher than the benchmarks CMS had estimated.
If you didn't tighten down the provider network, you might be in trouble. Some successful pioneers were able to make money even in advanced managed care markets by tightening down the provider network. In other words, by limiting the number of primary care physicians in the ACO to a restricted number of high performers, the risk of lower performance was minimized. Some pioneer ACOs that had bigger networks of physicians struggled more.
DRG+SNF=BAD. So you are a pioneer ninja: You have fewer bed days per thousand than almost anyone in the nation and you have incredibly low lengths of stay. You achieve this by managing discharges and readmissions and emphasizing high utilization of skilled nursing facilities and alternate-site resources. These are all good things for patients and taxpayers. If you were capitated for these attributed patients on a fair risk-adjusted basis, you would be making money, receiving gold stars and getting national awards. But if you are a pioneer receiving a diagnosis-related group payment for the inpatient care of attributed patients, and you have all the skilled nursing facilities and alternate-site costs on top, you don't look so good to CMS. This is crazy. It is a classic case of no good deed goes unpunished.
Risk adjustors are key, and they are still inadequate. Risk adjustment is improving, but it still does not adequately level the playing field to encourage true and effective care coordination. It is still way easier in America to dodge risk than it is to manage care (as we are finding out in ACO pilots, and as we will find out in the exchanges). This problem will not go away until we have fair and valid risk-adjustment methodologies. Why don't we conscript all those Wall Street quants who brought us credit default swaps to work on this?
The data infrastructure is inadequate, too. Everyone I talked with said something like: "The data sucked; it was late, bad, dirty, not actionable; and we couldn't integrate it with our clinical systems." To be honest, most CEOs didn't spend a lot of time talking about the data problems. It's like the Scots and bad weather: You just take it for granted. I always told my kids not to expect any sunshine when visiting Scotland. The best you can hope for is "periods of brightness." There are precious few periods of brightness in the ACO data infrastructure.
CMS personnel try hard, but they need more capacity. I heard incredibly flattering statements about the hard-working people at CMS and their eagerness to help. But I also heard from many sources that, while they are nice people who are working hard, they are a little overwhelmed and out of their depth. In fairness to CMS, a polarized Congress refuses to provide adequate resources to implement Obamacare generally; but when it comes to ACOs, it is not just a resource problem. As one CEO put it: "CMS is just not culturally or managerially equipped to be an effective health plan partner in the ACO business. At least in Medicare Advantage you have sophisticated national health plan intermediaries."
The end game for true pioneers is Medicare Advantage, and ACOs look like a distraction. Some CEOs pointed to the core attraction of Medicare Advantage (even with the cuts in rates embedded in ACA) because of the alignment of incentives. While they participated in the pioneer program for the good of the managed Medicare movement, the pioneer program may be a distraction from the central work of growing Medicare Advantage enrollment.
Look, there may be some real positives in this pioneer story. First, some pioneers in some markets made it work, and those that follow them (the settlers) will start to bring coordinated care to many unmanaged Medicare markets. National health plans in the ACO business will be key because they have the data to identify where to play and with whom in each of the low-hanging fruit markets. They will be able to harvest the excess utilization and waste in uncoordinated markets and may succeed in further bending the overall curve of health care, nationally.
Second, the power of ACO experimentation is that different organizations went after this in different ways. Some focused on nursing homes, some focused on care transitions, some focused on restricted networks of primary care physicians. We need the policy and academic communities to carefully evaluate all this and draw some conclusions about how to create true accountable care organizations. The information should be helpful.
Third, while a dysfunctional political system makes it impossible to change the ACA statute at least until after the 2014 mid-term elections, there are promising policymaking signals about the future of Medicare. In particular, coalitions of organizations, including the California Association of Physician Groups, are working on proposals for Medicare that recommend a third way to Medicare that moves toward a world of fairer, risk-adjusted payment models that would engage beneficiaries and provide a transitioning model from fee-for-service Medicare. More to come.
Fourth, the message to the hospital and health systems field: Risk is not dead. Long live risk. Large health systems taking risk for the health and health care of a defined and engaged population is the future of Medicare. This may take policy and payment change, this may involve working with new partners in new ways, and it may look a lot like Medicare Advantage.
Harvard polls show that while the current elderly (those older than 65) prefer fee for service over a managed care plan by a two-to-one margin (57 percent to 29 percent), the reverse is true for those 50 to 64 years old. Four million baby boomers who grew up on managed care are about to enter the program each year for the next decade, and they are going to pick managed Medicare. Let's give them what they want, but do it right.
We'll thank the pioneers, even the ones that turned back, for their sacrifice and energy, and we're hoping we'll discover new ways to deliver high-quality accountable care as a result of their journey.