• Medicare Advantage’s favorable selection

    A new paper in Health Services Research by Michael Morrisey and colleagues examines favorable selection experienced by Medicare Advantage (MA) plans. Though there is much more in the paper, their Figure 1 provides a summary of some of it.

    The figure tells us that in every year for the 1999-2008 decade, Medicare beneficiaries switching out of MA to traditional (aka, fee for service or FFS) Medicare were much costlier relative to beneficiaries that remained in FFS. Also, relative to those in FFS Medicare, beneficiaries switching into MA were less costly.

    By necessity, cost is measured as FFS costs during the six months after (before) switching out of (into) MA. Comparable claims and payment data are not available for MA, preventing researchers from studying utilization and spending for the program directly. However, this is not as large a limitation as it may seem. If we could analyze experience of MA directly, it would be confounded by care management by plans. That is, beneficiaries’ costs while enrolled in MA may be lower both due to favorable selection and due to care management. By comparing FFS experience for switchers and non-switchers, the effect of MA care management is removed.

    The conclusions are clear:

    [B]y the end of our analysis period in 2008, the 6 months prior claims experience of new enrollees remained only 91 percent of that of traditional Medicare enrollees in the same county over the same time period. We also found that the average claims experience of those disenrolling relative to those always in traditional Medicare was substantially higher in the last years of our study period. […]

    [Our study] indicates that the MA population continues to have claims experience that is favorable relative to those in traditional Medicare.

    The clear implication is that studies that take MA bids as reflective of plan costs for an average beneficiary may be overstating the potential savings of premium support.* A substantial portion, if not all, of the degree to which MA bids are below FFS costs may be due to favorable selection.

    * UPDATE: The sense in which savings may be overstated is that if, under premium support, private plans do not experience selection as favorable as MA does, then they may need to charge more than their current MA bids suggest. This would happen if a substantially greater proportion of beneficiaries enrolled in private plans, a not unlikely outcome of premium support.


    • What a mess.

      Insurance companies have two key functions: 1) Identifying the risk level of patients.
      2) Defining the benefits these patients should receive.

      The two dirty truths are: Some element of the health system is going to have to engage in #2 or else we will continue to pay for grossly ineffective treatments.

      The other dirty truth is that “risk identification” will only be a means of improving health care when insurance companies must guarantee coverage. Without that element, all the resources put into “risk identification” are simply used to off-load problems onto others. When insurance companies compete on choosing the best (i.e., cheapest/healthiest) patients, everyone loses.

      • Whoops, I lied. Not everyone loses. Some insurance companies win and walk away with tons of money, without having to care for anyone sick.

    • Two thoughts/questions. 1) If I am reading this correctly, this is being driven by the beneficiaries, or can I make that conclusion? Was there any pressure being exerted by MA plans to get people to leave?

      2) Risk adjustment seems to work fairly well in some European countries. What do they do differently?


      • 1) My guess is that the pressure to leave is by the nature of the plan design. Of course beneficiaries exit voluntarily.

        2) I don’t know. Do you have a good source that demonstrates that RA works better in European countries? Whenever I’ve looked into it I get frustrated by an apples/oranges problem: different metrics applied across countries and systems and/or RA being used for different functions.

      • MA insuree costs per month are low. Their ongoing beneficiary utilization costs are high (e.g. co pay per visit etc.) If the insured is well they stick to the MA plan. Once sick, costs start adding up and they switch at year end to the FFs model.

        MA insurance company benefits, customer benefits. FFS insurer and Medicare loses out on a healthier base to spread out expenses.