• Baker Institute: ACOs’ math problem

    In association with the Baker Institute’s Oct 25th conference on health care reform, I have a post up that describes a problem for ACOs that probably hasn’t occurred to you. Go read it.

    Also, videos of the proceedings of the conference are online here.


    Comments closed
    • I think it’s obvious that an individual can have only a small effect on improving outcomes. That is the reason they tasked ACOs with the work of organizing, training, and creating incentives for physicians to improve the quality of outcomes. Of course there will always be people who will attempt to “free ride” and that is a problem for the ACO to address.

      • ACO’s contain virtually the same incentives as HMO’s so I wouldn’t think that they would be good for patients or the proposed health care system. The evidence can be noted in the many suits that occurred and the studies that were written when HMO’s were believed to be the savior of the system.

        ACO’s balance hospitals, insurers, providers, physicians and government on one side with the sick patient on the other. Not a good relationship.

    • Frakt is right about the math (not surprising since that’s his business), but I’m not so sure that it follows that the math problem is incentive for clinical integration. The consensus among advisers to physicians is that specialists ought not join ACOs. Specialists with a sufficient number of patients to overcome the math problem (cardiologists, pulmonologists) run the risk of falling into the trap of being categorized as primary care (they often are for their chronically ill patients), limiting their participation to one ACO – that’s the rule. Of course, this is all academic. ACOs are unlikely to generate any significant bonuses, and the few that do won’t be distributing the “profits” to the participating physicians. Where will the “profits” go?

    • This doesn’t seem to resemble how current Medicare ACOs actually work. Most of them aren’t composed of independent physicians. It’s a partnership between a single giant physician practice organization and a hospital system. Also, the shared savings aren’t distributed at the individual physician level by the ACO. They go as a lump sum to the practice who then distributes it however it wants, if it wants to distribute it at all. The practice can fire non-performing physicians.

      With the trend of independents joining larger practices (particularly those owned by hospitals), this “math problem” may be academic.

      • “The practice can fire non-performing physicians.”

        You got the incentive and one of the potential problems, economic credentialing.

        To the ten doctors in the room. ‘You have all done a wonderful job. Dr. A you had the best record and we really like the way you work. Dr’s. B, C, D you also have done well. Dr’s E, F, G are OK but the numbers, …you know the numbers… Dr’s H, I, J. Despite the fine job you guys are doing you need improvement.

        One of the things we noted was that our overhead was too high and we think we can do better with 9 physicians than 10 so J has his two weeks. To make up for the extra work we will divide J’s future patients up equally and the income will go 1/3 to Dr. A, 1/4 to Dr’s B-C and 1/5 to Dr’s E-G. (Where does the rest go?)

        The 9 of you will meet here in 3 months when we will review job performance.

        What does anyone think happens to patient quality? Not doing necessary care frequently has no paper trail to look back upon.

    • This math assumes that the problem is individual doctor choices, not the system. I’d think plenty of improvements could be made in a better run system and doctors would probably be happier working in a system that made better sense.

      Lack of coordination causes lots of things to fall through the cracks.

    • SAO,


      The goal is to limit the range of individual physician choices, and make the behavior of physicians collectively more cost conscious. You need the group, as a whole, to go along with this, so you need a group consensus that this is a priority. Hence, you need incentives that pay off for collective behavior.

      It is true that if this were the only incentive then an individual physician could ignore the consensus guidelines and still participate in the bonus for group performance. But there is no reason for a group to put up with this. The ACO’s track the performance and compliance of each physician. Those who do not toe the line get warnings, and ultimately dismissals.

    • A saving grace may be that, as individuals, physicians may be more motivated by peer-to-peer quality comparisons than money. A more significant challenge for ACOs might be the re-setting of the financial targets every three years. In the current regulations, any progress is zeroed out and the organization receives no financial credit for improvements made in previous agreements.