Why ACOs might not make much of an impact (transcript)

This is the transcript of my recent podcast interview with Avalere Health's Erik Johnson.David E. Williams: This is David Williams, co-founder of MedPharma Partners and author of the Health Business Blog.  I’m speaking today with Erik Johnson.  He is Senior Vice President at Avalere Health.  Erik, thanks for being with me today.Erik Johnson: Thanks very much for having me.Williams:            Let’s talk about Accountable Care Organizations (ACOs).  Can you start by explaining what an ACO is.Johnson:            Well now we know what an ACO is supposed to look like given the proposal from CMS.  It’s a construct for care delivery that holds a consortium of providers across a continuum of care accountable for both the cost incurred in treating a specific population and the quality of care delivered.Williams:            Now that the proposed rules have come out, what do they cover?  What’s the scope?Johnson:            The scope is relatively ambitious.  CMS has proposed in good detail what is expected of those entities that volunteer to perform as an ACO.  It has made it clear that CMS expects to migrate risk management from the program and the trust fund itself to providers in local markets.  The rules lay out the organizational requirements and the economic savings targets that ACOs are expected to meet in order to begin to share in the savings that they generate.Williams:            The rules are hundreds of pages long with lots of detail.  Are there things that stand out as particularly notable or surprising compared with what you expected?Johnson:            Nothing exotic or esoteric compared with what we and the industry thought, but the specifics are where it gets interesting to us; specifically on the savings thresholds that ACOs are intended to attain.The idea behind the shared savings program and ACOs in general is that organizations that are being held accountable need to exceed a certain threshold of savings in order to begin to share in the first dollar. What surprised us, maybe more than anything else in these rules, is that the thresholds of savings are relatively high, particularly for the smaller ACOs that may end up forming.  While we understand actuarially why CMS did that, I do think that there is an implicit disincentive for smaller communities and community based organizations to participate in the program as a result.Williams:            In the past CMS has had demonstration projects where they’ve set out ambitious targets and the result has been that participants can’t meet them. Some companies have even gone bankrupt as a result.  Do you see parallels here?Johnson:            I don’t think so. But I think what we’re going to see is a different phenomenon, which is sort of a self-selection or self non-selection among providers.  The nearest analogy I can draw to the ACO construct is the physician group practice program that CMS has been running for the last several years.That was recently evaluated at the three-year mark. We saw that the 10 organizations participating in that met almost all of the quality reporting criteria, but only half of them met the savings criteria. And of those five that met the savings criteria, only two reaped what we would deem to be substantial savings in a way that would justify the ROI on this.I think CMS learned a lot of lessons from that, but in both the physician group practice demo as well as the prospective ACO demo, there is going to be a built-in selection bias.  Organizations that are very close to being able to do this already are the ones who are likely going to participate. Those who actually would be running a much greater risk of failure or bankruptcy or otherwise not meeting the objectives are probably not going to participate. They’re going to opt out.Williams:            I’ve heard ACOs described somewhat simplistically as being a throwback to capitation plus pay for performance quality.  Is that an accurate label to put on them?Johnson:            I don’t think it’s an accurate label, but it’s an accurate reaction to it.  Having had this conversation before with a lot of clients and health systems, the essential question is: is everything old new again in this idea of an ACO?In some respects it is a return to some of the principles of managed care, if not the actual specific design of managed care, in that it is an attempt to push risk downstream to the providers and away from the payers.  A lot of the things that managed care organizations used in the ‘80s and ‘90s would prove to be necessary in this environment too.But the environment has changed since 1985 or 1995, whichever era you want to go to. That may not be sufficient for providers to manage risk more effectively, but it’s certainly necessary.  We see the advent and adoption of health information technology, the advent and adoption of evidence based guidelines to the degree that didn’t exist in the late ‘80s and early ‘90s, and the slow migration of the payment systems to pay for performance and pay for reporting.  Those three variables are different in this era than they were 20 or 30 years ago and do suggest that health systems have found more tools to manage risk.Williams:            What would it be like for a patient who is in an ACO?Johnson:            The idea behind CMS’s thinking here is that it would be relatively transparent.  It is still by and large a fee for service environment that patients will be migrating in and they will have relatively complete freedom of choice among providers.They will have to be notified by those who are participating in the ACOs that those providers are participating in this program.  CMS will require that all of those materials that they’re given be approved by CMS.  They will have more information but under the proposed rule they will not have any restrictions on where they can get their care.Williams:            So is it more a matter of a patient being attributed to an ACO as opposed to assigned?  Is it different from someone who is in an HMO?Johnson:            Yes, very different.  Patients will be attributed to these ACOs and they will be attributed retrospectively.  CMS has told participating organizations that they will give them a best guess list of the patients that they think they may be seeing in the local marketplaces, but the actual calculations and what the ACOs will be held accountable for will all be done retrospectively, which does create a problem for the providers who participate.It is essentially telling them that they will be held accountable for the risk of a population that cannot be accurately identified at the outset of the risk period. That’s a bet that a lot of organizations may not want to make.Williams:            Is that just an interim step?  What’s behind that?  It seems like it would limit the savings. Especially if the targets are ambitious that doesn’t sound like the most ideal way to go about it.Johnson:            Yes, I would agree with that.  It isn’t the most ideal and may be the biggest hurdle to getting organizations to opt into this program.  It’s difficult to do this either prospectively or retrospectively in allocating and assigning populations as opposed to enrolling them.The idea was to steer clear of some of the less popular elements of managed care and to do this accurately.  You can certainly do retrospective assignment accurately in this environment in order to make sure that the costs that are being counted are in fact the costs being incurred.  Other than that, the policy rationale is a little thin and I’m not quite sure that retrospective assignment of populations is the best answer.Williams:            These rules are obviously coming from CMS and cover Medicare. But when people talk about ACOs they’re not necessarily limiting the discussion to Medicare.  What do you expect to occur outside of Medicare either in the private market or other publicly funded programs like Medicaid?Johnson:            In the Medicaid program we’re going to see a pediatric ACO demo announced --I believe-- this year.  That was authorized as part of the Affordable Care Act.  I think certainly states will have the ability to implement their own ACOs in their populations.  I think we will probably see some level of state-led activity in Medicaid in the near term, but I don’t think we will see it propagated widely across all 50 states any time soon.I think this is a new model.  We’re trying to do something new here.  I will give CMS all due credit with this proposal.  They did a good job of laying out a relatively elegant and theoretically consistent model for how they might like to see this happen, but we don’t know that it actually works yet.The private sector is an interesting question.  The private sector often follows Medicare in payment policy, but they usually do that only when the payment policies in Medicare are mandatory.  Since the ACO for Medicare is voluntary I’m not sure the private sector has a lot to follow.What will be interesting is that with commercial payers starting to go their own route in developing care coordination models --whether accountable care organizations or patient centered medical homes-- we are starting to see a significant uptick in activity. To the degree that those activities and local markets are robust enough I think the causality may work the other way; that providers who are used to participating in those environments with their large commercial payers may in fact be more willing to participate in the Medicare program as a result.Williams:            My mental model for an ACO is an integrated delivery network taken to the next level. Yet I have also heard that non-traditional groups are contemplating forming ACOs themselves. This could include medical societies or smaller physician organizations.  With the rules becoming clearer, is it your sense that you will see some non-traditional kinds of organizations becoming ACOs or is it more or less that you need to be an IDN in order to make it work?Johnson:            I think that’s certainly the intention of CMS, to encourage a thousand flowers to bloom in terms of the models and the type of lead actors in the ACO realm.  I think, to a certain degree, we will see that.But I’ve also noticed over the last year that providers --whether they’re integrated or not-- have migrated their attitudes. It started as ‘tell us how to be an ACO’ to ‘help us figure out whether or not we want to be an ACO’ to where we are today, which is ‘maybe we don’t want to be an ACO.  Maybe this is in fact too hard.’There is a reason that there are only a few integrated delivery networks that are successful today. It’s because it’s difficult and it’s capital intensive and it is highly dependent on local market factors.  They certainly stand the best chance of succeeding as ACOs.  I don’t think too many people disagree with that.  I don’t think they’re the only actors that can succeed in that type of environment, but they certainly start the game with a lot of advantages.Williams:            ACA is under attack both from those trying to repeal it and roll the clock back to where it was before Obama came into office, and then on the other hand you see the Ryan Plan that’s taking things in a fairly radically different direction.  Are ACOs or something like them likely to exist in any of those environments or are some more hostile or more friendly to the concept than others?Johnson:            The ACO construct is a natural evolution for the health care economy in general. The Affordable Care Act may have accelerated a bit the evolutionary path that health delivery systems were on.I don’t think that there are a lot of arguments among those knowledgeable about health care delivery that costs can be saved and outcomes can be improved if care is better coordinated across the continuum.  I think ACOs represent a very valid way to do that and I think we will see it exist in nature over the next three to ten years, regardless of whether the ACA is repealed, the Ryan Plan is enacted or any other such factors start to affect the way people think about care delivery.What is happening now with these proposed rules is the setting of an expectation that this will happen quickly. But I don’t think we are likely to see this have a dramatic impact on the health care economy in the near term simply because it takes time for evolution to play itself out.Williams:            I’ve been speaking today with Erik Johnson, Senior Vice President at Avalere Health.  We’ve been talking about Accountable Care Organizations and new proposed rules from CMS.  Erik, thank you so much.Johnson:            Thanks very much for having me.