The Achilles Heel of ACOs? Shared Savings Payment Model Unlikely to Motivate Hospitals

Sometimes you read something and the full impact doesn’t hit you until hours — perhaps days — later.  As I was out mountain biking today, the importance of something I ran across yesterday suddenly hit me.

Accountable Care Organizations (ACOs) are today’s cure-du-jour for reforming the health care delivery system. Bob Berensen, MD of the Urban Institute strongly questions whether the shared savings model under current legislation provides enough economic incentive for hospitals to disrupt their existing core business of acute, inpatient care.

The dialogue took place at HSC’s 15th Annual Wall Street Comes to Washington Conference. Here’s the conversation from the transcript — I actually went back to dig this out of my trash:

Paul Ginsburg: Actually, let me just pose a devil’s advocate question. You know, I could see why ACOs might be attractive to a large —- an IPA because there’s this opportunity to reduce hospital use and be rewarded for it. But what’s the motivation for a hospital because the rewards from the ACO are likely to not measure up to the loss of volume from more efficient delivery. Is there any response to that?

Robert Berenson: I think that’s why I was going to say from a policy point of view it may be that the share — the ACA, the Affordable Care Act, actually has a Section 3022 that’s not called accountable care organizations, it’s called the shared savings model and that is the one in which there is no downside, so this some have called this the lottery. Why not participate if by luck or by creative coding you might be able to get a bonus? That is the concern from policymakers.

When I say an overreaction, I think there was a lesson from the 1900s that global capitation with hospitals as risk takers did not work out very well. Hospitals got out of it long before a lot of the physician groups. They got out of their risk taking arrangements. There are some still in Medicare Advantage arrangements where in a few cases hospitals still take risk but for the most part they don’t.

From the beneficiary side if we’re talking in the Medicare context, the lesson was that people don’t want to be locked in. They want to have their freedom of choice so now we’ve developed a model, the Shared Savings Model, that has no requirements on beneficiaries to alter their — their prerogatives are not affected at all. They have full freedom of choice and indeed in the model that’s specified in the legislation, beneficiaries don’t even know they’ve been assigned to an ACO. It’s felt that they don’t need to know because they still have freedom of choice. Some of us disagree with that judgment.

On the payment side you get an upside if you pass a threshold of savings against the target. There is no downside. At least the physician group practice sites that — there were 10 sites that were in a demonstration and a lot of those folks even though they’re — some of them are relatively positive about their experience, they say in a fee for service world if you’re maintenance fee for service you’re really not changing incentives very much and so we had a lot of difficulty engaging the physicians and other parts of our delivery system.

In direct response it’s not clear what does change here. The hospital says I get 100 percent of the revenue today for a marginal admission or I get potentially some percentage of a savings 2 years from now after my budget is analyzed. It’s the unusual hospital that’s going to say I think we’ll put some systems in to reduce that admission. It seems to me the shared savings model which doesn’t have any downside risk at all is more appropriate for a physician organization because they can try to deflect people from the hospital. I think if we want hospital organizations in this we have to have some kind of risk and some of us use the term partial capitation, which means risk with corridors to protect against really bad luck.

This is one amongst many where the devil is in the details. If we have the shared savings payment model I’m not sure it’s going to get us very far and the legislation does have the opportunity for CMS to test other payment models. The question is whether CMS will go there or whether they will deal just with the shared savings model that’s defined in the legislation.

Key Takeaways

Berensen suggests that the current ACO model is flawed. He questions whether hospitals (delivery systems) will be motivated by a shared savings model.

The economic incentives for physician groups that might lead an ACO initiative are different than for a hospital. The majority of potential savings will be from reducing hospital admissions. To a physician group, the shared savings from reducing hospital admissions is all upside.

Hospitals will view this differently. The potential savings coming from reducing hospital admissions will be seen as cannibalizing their existing core business.  Berensen questions whether hospitals will have strong enough incentives to make the ACO truly effective.

Berensen also makes a “by the way” comment about retrospective assignment to an ACO.  This is worth underscoring, and frankly I find this aspect of the ACO legislation downright squirrelly (apologies to the squirrels outside my window). We are learning that care management is a team sport among patients and care providers, and it makes no sense to assign the teams after the game is over.

Berensen ponders whether CMS should use its authority under newly enacted Center for Medicare and Medicaid Innovation (CMI)  to design ACO initiatives with much stronger incentives — e.g., capitation or partial capitation payment models where hospitals have more skin in the game and stronger economic motivation to transform their core business to one of “optimizing” admissions vs. “maximizing” admissions.  This is definitely worth pursuing, and would being to address the Achilles heel in the shared savings model.

Update October 23, 2010: Dr. Berensen expands on his views in the article “Shared Savings Programs for Accountable Care Organizations: A Bridge to Nowhere” in the American Journal of Managed Care.  Hattip to Dr. Jaan Sidorov, aka DMCB.