The Medicare Shared Savings Program is intended to lower costs and improve quality. If program participants succeed in meeting these goals, they are able to share in the savings — calculated by comparison to expected Medicare FFS expenditures for the same population if folks were not enrolled in an ACO.

Many program participants have complained that the criteria for being able to share in the savings were too tough. and the latest version of the MSSP (ACO) regulations adding some flexibility in that department were finalized earlier this year. (See my post on the proposed ACO regulations; you may read the presser that links to the final version, too, as well as earlier posts on the ACO phenomenon.)

This week, CMS released data on ACO performance in 2014 (so the effects of the latest regs are obviously not reflected in this data). While the federales try to put a good face on it, the fact of the matter is that only about 1/4 of ACOs are in shared savings territory. The total savings (Pioneer ACOs and other MSSPs) only came to about $1B. One billion dollars sounds like a lot, but since the total Medicare spend is about $600B and about 15% of Medicare beneficiaries are enrolled in the MSSP program, that’s not a very impressive savings figure (it’s on the order of 1%, and does not demonstrate a sea change of the sort that we need).

The savings figures are inching up, and MSSPs (and, in particular, Pioneer ACOs) that have been at it longer seem to do better. However, a number of the Pioneer ACOs dropped out of the program, so even the more “advanced” participants are not uniformly delivering the best possible results.

There are certainly a number of alternative approaches to payment reform that have been floated, and CMMI continues to crank out new ideas, but there has been significant emphasis placed on the MSSP program as embodying CMS’s approach to value-based payment in the context of the broad effort to shift away from fee for service medicine.

One of the benefits of central planning on the scale of the Medicare program ought to be the ability to learn from successes, cull failures, and engage in an ongoing process of improvement. It seems that CMS is content to engage in watchful waiting at this point. (In theory, I believe that physician-led ACOs ought to be able to deliver better cost and quality improvements, but I have not seen data broken down by type of ACO in a way that would allow for confirmation of that hypothesis.)

It remains to be seen whether the revised regulations will allow MSSPs to deliver better results on the cost and quality fronts. Tune in again next year!

David Harlow
The Harlow Group LLC
Health Care Law and Consulting

Image credit: Colin Dunn via Flickr CC

11 replies on “Accountable Care Organizations: The Show So Far”

  1. Nice post and analysis. The thing I can’t figure out with ACOs is if they’re really changing the cost of healthcare or if they’re mostly a game of good accounting and reporting. Basically, does the measures they’re require really cause organizations to change how they care for patients or does it just change how organizations document and report what they’re doing? I’d love to hear your thoughts.

  2. Thanks, John. I think that the ones that are succeeding in improving quality and reducing costs are really doing it. I think the problem is that these successes not necessarily reproducible. Not everyone understands why what works in one place won’t work somewhere else. The demo that inspired the ACO/MSSP rule had 10 sites over 5 years and if you look at the financial results you would think that the demo was a failure overall. See: link to . I don’t think anyone can explain why the Marshfield Clinic succeeded and the others not so much, or not at all. Another problem is the tying-one-hand-behind-the-back problem of refusing to limit patient freedom of choice (in selecting providers). Same issue faced by managed care. Without narrow networks this whole concept doesn’t really work.

    1. Hello David,
      Actually I believe the Payor involved on the commerical side of an organization contributes greatly to how well an ACO performs on the Medicare side. Using your example, Marshfield clinic has a stable payor in Security Health who for years has looked at their population like you would in population health management. There is VERY little churn in rural WI so Marshfield and Security have partnered long before it was in vogue. (The same could be said of closed model organizations like Kaiser). This payor relationship yields a culture of mutual benefit versus one of zero sum gain. The opposite can be said of commercial payors who “say” they want lower cost and higher quality but when you look at their business models, lower cost is a direct conflict to their success. Even when given clear quality data that shows higher quality in a particular provider network they will not steer patients aggressively to it.
      Look at the culture of the primary commercial payor in a given marketplace and the churn rate they have historically planned for coupled with the culture and partnership of the major providers in that marketplace and I believe you will find a direct correlation to ACO success.

      1. Matt, I thought it was something in the milk ….

        I like your point, but you could make a similar argument about Geisinger, though, and they did really poorly by comparison.

        1. It would be interesting to see Geisinger’s churn rate and their actual commitment to the population health/capitation mentality and when it seriously began to influence provider behavior. I have personal knowledge that as far back as 2002 Security Health and Marshfield behaved differently from every other client I had in the Midwest in terms of how they managed chronic disease (Diabetes). Then again, maybe it’s a one rat study:).

          1. Thanks for the insight, Matt. Clearly we need to be doing something different in the population health management arena in order to make serious improvements on both the quality and cost sides…. At the very least, we could use some more rats ….

  3. And the initial premise of your article which is to see if the DNA of seemingly look alike rats is different and hence we can find something that’s reproducible. I always enjoy your work.

    1. So, from the trenches of an actual medical practice, I report the following:

      1. Nobody pays for population health, but we all pay for population illness/disease.
      2. Population health benefits (financially) whom, exactly? Payers, right? So, do they pay providers more, or decrease premiums? Nope. Maybe there is a decrease in ER visits – I wouldn’t bet on that.
      3. I maintain this is a excellent arena for the game of Limbo – how low can you go, also known as diminishing returns. As a set of providers in an ACO gets more efficient and cuts costs, they get paid a proportion of what they save, then the bar goes down to the new efficiency level, and they have to get more efficient. What’s the endgame?

  4. And look at this:
    Quality Improvement: ‘Become Good At Cheating And You Never Need To Become Good At Anything Else
    Gary Claxton and Lerry Levitt in The Kasier Family Foundation
    “The good news about hospital readmissions, much like the home run totals of sluggers from the steroid era, might have to come with an asterisk. A Health Affairs blog post points out that the much-touted drop in readmissions among Medicare beneficiaries has coincided with a rapid rise in “observing” these patients. Observation counts as an outpatient service an not an inpatient admission, so in addition to not counting against the readmission penalty it leaves patients responsible for the bill. This, Health Affairs said, essentially eradicates whatever quality gain hospitals claimed by reducing readmissions.” – Brian Eastwood

Comments are closed.