Medicare HMOs, now with less cream skimming

We analyzed utilization patterns of enrollees in Medicare Advantage health maintenance organization (HMO) plans compared to matched samples of people in traditional Medicare during 2003–09, to ascertain whether the HMO enrollees demonstrated different levels of use of services, which can be a hallmark of more integrated care. We found that utilization rates in some major categories, including emergency departments and ambulatory surgery or procedures, generally were 20–30 percent lower in Medicare Advantage HMOs in all years. Medicare Advantage HMO enrollees initially had lower rates of ambulatory visits and hospitalizations, although these rates converged by 2008; they also received about 10 percent fewer hip or knee replacements. In contrast, HMO enrollees underwent more coronary bypass surgery than patients in traditional Medicare. These findings suggest that overall, Medicare Advantage HMO enrollees might use fewer services and be experiencing more appropriate use of services than enrollees in traditional Medicare.

That’s from Bruce Landon and colleagues in the latest issue of Health Affairs. Were that the only article you read in that issue, you might be tempted to dismiss these findings as evidence of favorable selection (cream skimming) by plans. However, another paper in the issue, by Joe Newhouse et al., documents improvements in risk adjustment, suggesting plans are not as advantaged by selection of favorable risks as they used to be.

In sum, favorable selection in the Medicare Advantage program in the 1990s meant that Medicare spent more per beneficiary who enrolled in Medicare Advantage than if the enrollee had remained in traditional Medicare. In the mid-2000s Medicare took a number of actions to mitigate selection, including introducing diagnosis-based risk adjustment, a lock-in period, and an expanded array of plan types. These actions were associated with reduced favorable selection.

Unsurprisingly, Michael McWilliams, John Hsu, and Joe Newhouse report similar findings using other methods in yet another paper in the same issue.

My comments:

  1. I recognize I am not providing many details. I recommend you read the papers if you can get your hands on them. You can at least click through to the abstracts.
  2. As the authors point out, similar risk-adjustment mechanisms will be applied to accountable care organizations and health insurance exchanges. Thus, if they are improving, that bodes well for these other insurance- and insurance-like arrangements.
  3. Improving risk adjustment doesn’t mean selection is no longer an issue. It means only that it is less of an issue than it otherwise might have been.
  4. All in all, this body of work takes some of the bite out of standard critiques of premium support: private, managed care plans seem to reduce utilization; they seem to cream skim less than thought. This by no means implies that premium support is devoid of any other trade-offs.
  5. I wonder what all this means for the use of Medicare Advantage bids as estimates for the cost of care in a premium support regime. If sub-FFS bids don’t reflect selection as much as thought, then they’re a more accurate reflection of true costs. On the other hand, if plans are able to manage utilization to as great an extent as suggested above, bids at the 95% level of FFS Medicare (PDF) seem high. (There are other considerations and distortions. I’m just flagging this issue.)
  6. Of course there are limitations to the work above, as is true of all work. The authors acknowledge them.


Hidden information below


Email Address*