By Hamilton E. Davis
The ACO concept is in its infancy is Vermont, and to a considerable degree in the U.S. The pace is going to pick up drastically, however, as the need for cost containment is so pressing. By January of 2016 every major health care system in the country is going to be participating in the next phase of ACO movement, or looking closely at operating in a way that will cost them money if they manage badly.
Vermont will be one of the bellwethers in this campaign. Vermont’s largest ACO, OneCare, has been selected by the federal government to be one of a couple dozen ACOs that will participate in the next phase of the reform effort. The ACO phase of reform is more than simply federal, however. The Vermont Legislature assigned to the Green Mountain Care Board the responsibility for refashioning the state’s system and the use of an ACO structure was part of the state plan from the outset.
OneCare was formed in 2012 by the two tertiary hospitals that serve Vermonters.
They are the University of Vermont’s system, which dominates the western part of the state, and Dartmouth-Hitchcock Medical Center, located just over the New Hampshire border, but 40 percent of whose patients come from Vermont. Through 2014, OneCare’s membership also included the state’s 12 smaller community hospitals.
Two smaller ACOs have been operating in Vermont during the first phase of the reform effort. One is called Community Health Accountable Care (CHAC); it is made up of groups of primary care doctors that get some federal financial support. The groups are called Federally Qualified Health Centers. In 2015, CHAC accepted into membership four small hospitals from eastern Vermont—Northeastern, in St. Johnsbury; Gifford in Randolph; Springfield; and Grace Cottage in Townshend. The third ACO, called Healthfirst, is a group of 68 independent primary care physicians.
The first stage of ACO activity in Vermont is now drawing to a close. Over the last three years all three ACOs participated in some so-called “shared savings programs” with Medicare as well as Medicaid and commercial insurers. The basic idea was for doctors and hospitals to begin to get a grip on costs by assembling blocks of patients, calculating the cost of caring for them in the past, and then trying to constrain the increase in costs for those patients going forward.
If the providers kept the costs under the previous inflation trend, they saved money for Medicare and the other payers, and if they saved enough, they could keep some of that money for themselves. In the argot of the field this is called upside, or one sided, risk. It isn’t really risk: heads I win, tails you lose doesn’t meet the dictionary definition. But it was a start; the providers got a feel for the basic idea. Al Gobeille, the chair of the Green Mountain Care Board, calls this period the “training wheels” phase of the ACO movement.
The results of the effort across the three ACOs and the three payer components—Medicare, Medicaid, and Commercial—were mixed, and not particularly impressive. They saved some money for the payers, but didn’t get to keep much. The most important outcome was that most of the players in the acute segment of the state’s delivery system got their feet wet. Anyone who wants to dig into the details of the early ACO experience can do so by reading the reporting of Morgan True and Erin Mansfield on vtdigger.org.
The most important point in my view is that neither OneCare, nor CHAC, nor Healthfirst was actually operating as an ACO, if we assume that an ACO is supposed to be an integrated system. They might have been building infrastructure, and they may have been doing very good work in improving primary care by extending doctors’ ties to their local communities. But the system pretty much operated as it has in the past, and the savings racked up were more the result of luck than new lines of coordination between units. This observation is not intended as a criticism of any of the ACOs. The preliminary work they have been doing they need to do. It is just that the level of integration—the seamless management of patients from primary care up the ladder of complexity—is a shift in the medical culture of today where patients move randomly from doctor to doctor without any tight management or coordination.
The movement to date, therefore, hasn’t really moved the system closer to market discipline. You only do that by taking real risk. With real risk, you make money if you control cost and quality and you lose money if you don’t. There is still some time left for the training wheels phase. The Medicare shared-saving agreements can continue through calendar 2016, but after that the focus will be on real risk, where providers can lose money by performing poorly.
Vermont is preparing now to enter the era of real risk. The first stage. In other words, the training wheels are about to come off. And so far, it isn’t clear at all that Vermont is ready. The immediate question the system faces today is whether the two small ACOs will give up their efforts to operate ACOs on their own; and an equally important corollary—what kind of governance of a single ACO would be acceptable to them. In other words, how big a role would they have in an ACO dominated by the two big tertiary centers?
Working that issue out has been going on under the aegis of the Green Mountain Care Board for the last nine months or so. The midwife for the process has been Richard Slusky, a veteran hospital administrator, (a long-time CEO of Mt. Ascutney Hospital in Windsor) who joined the Board when it was formed in 2011. The Board has been concerned from the outset of the reform effort to keep all of the major players on board; Al Gobeille, the Board chair, likes to describe it as a “coalition of the willing.”
To that end, the Board has provided grants to all three ACOs to help them get going and to participate in the shared savings programs. The process has been grindingly slow, however; for the ACO effort is stripping off the veneer of civility that it usually obtains in these matters. One of the issues that has been working just beneath the surface is the bitterness and resentment that the “smaller” elements in the system—small hospitals and many primary care doctors—feel toward the University of Vermont Medical Center.
To get a sense for this you might read the 2013 article by Dr. Katharine Hikel of Hinesburg, a physician who writes about health care, in Vermont Woman magazine. Her basic claim is that the Burlington-based facility is too big, too expensive, too rich and too overbearing. The components of the UVM system, she wrote, “pay executive-class wages for dozens of vice presidents, revenue managers and marketing specialists whose work has little relation to patient care…A complex tertiary medical center costs more to run than a small community hospital,” she wrote. “The question is whether our tertiary-care hospitals have grown overlarge.”
You hear the same theme from many of the small community hospitals in the state. The idea gained some public credibility when Seven Days, the alternative newspaper in Burlington, ran a front-page story to the effect that the UVM system was swallowing up as much of the state-wide system as it could. A variation on the theme is the claim by some of the small hospitals that they are inexpensive and stand in marked contrast to what they call the over-priced UVM system.
This sort of generalized resentment is exacerbated by a genuine fear on the part of the primary care community of providers that they are in a system of strongly enhanced cost containment, and that they will get even less money than they do now. I will argue in the future that the claims that UVM, and by extension Dartmouth-Hitchcock, are too expensive are in fact not valid. The fears about primary care, however, are valid. Primary care in Vermont, and in the U.S. generally, is badly under-funded and is in a chronic crisis situation. In fact, both the Green Mountain Care Board and OneCare are committed to remedying these problems with the primary care system as soon as possible.
So, where do the Vermont ACOs go from here?
Starting with the smallest. Healthfirst, the grouping of 68 independent doctors, has been operating an ACO, in partnership with a national firm, that has been participating in the Medicare shared savings program. Amy Cooper, the director of Healthfirst, told the Legislature recently that the partnership will drop its joint effort with Medicare. The reason is that in the three-year phase of the Medicare shared savings program it was too hard to make any money. A big problem across the board is that Vermont costs are very low compared with other states so that it's very difficult to get under the trend line, and thereby get some “shared savings.”
Cooper says now that for Healthfirst the question of whether to operate an ACO in Vermont is still open. It would still be possible to participate in shared savings efforts with Vermont Medicaid or with the private insurance carriers in the state. “We just can’t tell yet what all this might look like,” she said. Cooper said whether or not Healthfirst operates an ACO the group will stay together to work for the benefit of its members.
CHAC, meanwhile, has been the center of the resistance to moving their primary care doctors into OneCare. CHAC’s central concern is to ensure that the primary care segment of the delivery system gets the resources it needs to serve its patients well, and to continue to build links into their communities that contribute fostering wellness for whole populations. CHAC’s leverage in this situation derives from the federal law that stipulates that a Vermont resident cannot be considered part of an ACO unless he or she is sent there by a primary care doctor. And CHAC has a lot of primary care doctors.
Joyce Gallimore, who manages the CHAC effort, is non-committal at this point about whether her organization will abandon its stand-alone ACO and lead its members into a single state-wide ACO.
“It’s hard to say one way or another,” she said. “Governance of the ACO is one issue. Another is the terms of the agreement with the federal government—how will the money be distributed. What will be the cost of the infrastructure and how will those costs be distributed and how will those costs be distributed to individual providers…there are so many details it’s hard to skinny down,” she concluded.
My guess is that both small ACOs will end up joining a single state-wide organization. The reality overarching the whole issue is that you can’t have a credible ACO unless you offer a full range of medical services, and all of the higher end services available in Vermont are part of OneCare. Since neither Healthfirst nor CHAC encompass a full range of services neither can take financial risk for big blocks of patients. And if you can’t do that there is no point in operating an ACO in the health care world of the future.
Beyond the simple logic, an important reason to be sanguine about eventual cooperation between the state’s providers is that Todd Moore, the president of OneCare, has been assiduous about cooperating with the smaller ACOs. He has eschewed any use of the huge muscle that OneCare has available in the state now. Primary care, for example, accounts for only three to five percent of the total health care spending in the state. And even when you subtract the spending by the four small hospitals now in CHAC, OneCare still accounts for more than 90 percent of the hospital spending for Vermonters.
Even if both small ACOs join the single state-wide unit the governance of OneCare remains problematical. Full integration of the delivery system is inherently difficult and would be even if all the players were great pals, which they are not. The culture of American medicine at least since the end of World War II has inclined strongly toward dominance by specialists over primary care physicians. Anyone who doubts that is invited to consider the money paid to each. Specialists get paid more than primary care docs: actually not more, but rather far, far more.
Health policy theorists have speculated for at least 30 years that medicine needs to reverse that pecking order and to make primary care doctors into “gate keepers,” who can manage all the care given to their patients. I think it’s a great idea, but anybody who looks inside the system should be forgiven for putting it in the “don’t hold your breath” category. If we consider only Vermont, you can safely put it in the fantasy category.
If we look out, oh, at least five years, or more likely eight to 10, market-type discipline could force an evolution along those lines. But anyone who thinks that in Vermont in the near future a five-percent tail is going to wag a 95-percent dog is dreaming.
So, the Monday morning meetings will continue, no one can say when we’ll see a resolution of the ACO governance issue. Todd Moore and his team have accepted the invitation to operate a second generation, with the caveat that they cannot begin functioning in the new wave until January of 2017.
The issue of bringing the providers now in the two small ACOs into OneCare is that OneCare will need a full year to get a good start on a truly integrated system, which could be used as the basis for risk contracts with payers. For even if Healthfirst and CHAC’s members decide to join OneCare based on acceptable governance, the units won’t be really integrated. That is, they don’t now function now in a coordinated fashion, certainly not to the degree that their efficiency would begin to resemble a seriously competent private-sector organization.
We’ve gone a long way in Vermont, farther, I would argue, than anyone else in North America, but we have only just stepped into the swamp that lies between current reality and the sustainable system that we can see on the horizon. The hardest work, in other words, has just begun.