• Well, this is disappointing – CBO on Medicare’s Demonstration Projects

    From the CBO:

    In the past two decades, Medicare’s administrators have conducted demonstrations to test two broad approaches to enhancing the quality of health care and improving the efficiency of health care delivery in Medicare’s fee-for-service program.Disease management and care coordination demonstrations have sought to improve the quality of care of beneficiaries with chronic illnesses and those whose health care is expected to be particularly costly. Value-based payment demonstrations have given health care providers financial incentives to improve the quality and efficiency of care rather than payments based strictly on the volume and intensity of services delivered.

    In an issue brief released today, CBO reviewed the outcomes of 10 major demonstrations—6 in the first category and 4 in the second—that have been evaluated by independent researchers. CBO finds that most programs tested in those demonstrations have not reduced federal spending on Medicare.

    One way the programs were to save money was to prevent admissions to the hospital by doing a better job in the outpatient setting. Here’s how they did:

    Note: Bars with lighter shading represent programs with fewer than 400 enrollees. The estimates for those programs are less precise than the estimates for the other programs.

    On average, there was no real effect. Moreover, in pretty much every program, spending was either unchanged or increased compared to what would have occurred without them. One of the few positive notes was that programs with care managers who were assigned to have more direct interaction with physicians and patients were more likely to show spending decreases than other programs. Even then, though, those spending reductions weren’t enough to compensate for their fees.

    The conclusion:

    Demonstrations aimed at reducing spending and increasing quality of care face significant challenges in overcoming the incentives inherent in Medicare’s fee-for-service payment system, which rewards providers for delivering more care but does not pay them for coordinating with other providers, and the nation’s decentralized health care delivery system, which does not facilitate communication or coordination among providers. The results of those Medicare demonstrations suggest that substantial changes to payment and delivery systems will probably be necessary for programs involving disease management and care coordination or value-based payment to significantly reduce spending and either maintain or improve the quality of care provided to patients.

    We’ve got a lot of work to do. It would be nice to see some policymakers specifically discussing these issues, or how we might address them. Just repeating that “we will pay for quality, not quantity” isn’t enough. I’m looking at both sides here.

    • I thought the purpose of demonstration projects was to test out different ideas on how to save money. By this criteria, they found 11 programs that saved more than 6%. Other programs didn’t do as well.
      I don’t think it makes any sense to say that there was not net effect on average. Yes, that is true in the statistical sense but some of the programs did have an effect.
      While it would have been nice if all of the projects had been unqualified successes, most people would say that you should continue/expand the programs that worked and stop/redesign those that didn’t (taking lessons from those that worked). Isn’t that the purpose of “demonstration” projects?
      From the CBO report:
      The following approaches taken in various projects have been cited by observers as helpful in attaining the demonstrations’ goals:
      Gather timely data on the use of care, especially hospital admissions;
      Focus on transitions in care settings;
      Use team-based care;
      Target interventions toward high-risk enrollees; and
      Limit the costs of intervention.

      Nobody said this would be easy.

    • I would be interested in the time horizon for estimating the savings. We like to think that a Coordinated Care group possesses a magic wand that they can wave to make patients instantly more healthy. It could be that there is an up fron investment inn care that costs more money and that the real savings won’t begin accruing until several years or more in the future. How do we measure that?

    • Mark, you may wish to revisit the blog from which that graph was taken (as the graph is not labeled clearly on TIE, unfortunately). The x axis is hospital admissions, not cost. The CBO blog goes on to say that although some programs successfully reduced admissions, costs were not reduced.

      Here’s the relevant section:

      “On average, the 34 programs had little or no effect on hospital admissions. There was considerable variation in the estimated effects among programs, however (see figure below).*

      “In nearly every program, spending was either unchanged or increased relative to the spending that would have occurred in the absence of the program, when the fees paid to the participating organizations were considered.

      “Programs in which care managers had substantial direct interaction with physicians and significant in-person interaction with patients were more likely to reduce Medicare spending than other programs. But, on average, even those programs did not achieve enough savings to offset their fees.”

      *this is the figure shown on TIE.

      I am interested that they say “nearly every” program led to the same or higher cost, and hope to find more detail on this in the working papers linked at the end of the CBO blog.

      Interestingly, the CBO blog does not say that the “lessons learned” helped lead to lower costs, just were “helpful in attaining the demonstrations’ goals.” As a hospital doc myself, I can imagine how they’d improve quality, but can also see how they might cost as much as they save.

    • I thought linking to this talk might be helpful in approaching this topic… It pertains to economic incentives and how they affect performance: