Notwithstanding the growth in Medicare Advantage enrollment to over a third of beneficiaries, critics still complain that the MA plans cherrypick healthy enrollees and don’t deliver a good deal for the Medicare program. But recent research, including a new study published in Health Affairs, suggests that in fact the MA program may be lowering costs for Medicare in the traditional fee-for-service arm. (HA Article) The authors used data from 2007 to 2014 to assess, at a county level, where Medicare Advantage enrollment was growing fastest, and whether more MA enrollment in an area was associated with an impact on fee-for-service Medicare spending. One theory about MA growth has been that in high-penetration areas, the plans’ tighter care management policies might spillover to a physician’s fee-for-service beneficiaries. Analyses of spending were risk adjusted and adjusted for geographic variations in input costs, like labor. The tests for spillover effects were lagged by two years from any rapid growth in MA enrollment to allow for time for spillover effects to manifest themselves. The study covered 3014 counties covering 99% of all Medicare beneficiaries.
The counties were divided into quartiles for both MA penetration in 2007 and growth in MA enrollment from 2007 to 2014. Contrary to the theory of some critics, MA penetration growth was concentrated in poorer and minority areas. The reason for this is obvious for anyone with practical MA experience: poorer areas tend to have beneficiaries in worse health, which leads to higher payments to the plans and more opportunity to save on spending by better care management, which should result in greater profits. There was a non-significant decrease in FFS Medicare spending associated with a 10% increase in MA penetration; about $33 in annual lower spending. But in the counties with the highest initial starting MA penetration, there was a significant association. In those counties a further 10% growth in MA enrollment led to $154 lower annual FFS spending. This effect was seen most clearly in counties with the highest supply of physicians, and particularly primary care physicians. This supports the spillover theory because it suggests that once a physician has a significant number of MA beneficiaries, their greater familiarity with MA care management processes impacts their care of all Medicare beneficiaries. It also suggests some competitive dynamic in areas where there are larger numbers of doctors. To the extent that it is true (and it certainly may not be true) that CMS is overpaying MA plans because of upcoding and other factors, these FFS savings more than compensate for that overpayment.