News & Commentary

The MA Cup Runneth Over: Study Shows Spillover Effect on FFS Medicare

This was not supposed to happen. The Congressional Budget Office in 2010 projected that enrollment in Medicare Advantage (MA) would decline from 10.9 million (or 24% of total Medicare enrollment) to 8.2 million (14% of total) by 2015.

But a recent study published in Health Affairs noted that, instead, MA enrollment grew to 17.8 million in 2015, about a third of all Medicare enrollees.

At the same time, spending on traditional fee-for-service (FFS) Medicare slowed, also confounding expectations. Per capita Medicare spending in 2014 was estimated to be about $11,170, which is $1,200 less than what the CBO projected in its 2010 projection.

The spillover effect of MA on FFS expenditures has been studied and talked about for quite some time. MA plans use cost-control tools such as prior authorization, narrow physician networks, and utilization management. Of course, physicians and hospitals that take care of MA enrollees also treat Medicare FFS beneficiaries, so some of the practice patterns for MA enrollees spread to their care of FFS beneficiaries, reducing expenditures.

But Garret Johnson, the lead author of this study, and his colleagues at the Harvard T. H. Chan School of Public Health have added some important nuances to the understanding of the spillover effect. For example, they found that the spillover effect was most apparent in counties with highest baseline MA penetration back in 2007, which was defined as 17.2% to 65% of Medicare beneficiaries belonging to MA plans. In those counties, a 10 percentage-point increase in MA penetration from 2007 to 2014 was associated with an annual decrease in per capita FFS expenditures of $154. When all counties are included, the effect is a rather modest $33 decrease in per capita FFS expenditures (although a decrease of any kind is notable).

The explanation offered by Johnson and company is that perhaps providers start to change the way they practice for all their patients only after a sizable percentage are in MA plans.

MA enrollment has grown so much that high penetration is becoming the norm. According to these researchers, almost two thirds of the counties in the country now have MA penetration above 17.2%, the cutoff for the highest quartile in 2007.

Another interesting finding in this study was that the evidence for a spillover effect on FFS expenditures was strongest in counties with greatest per capita supply of primary care physicians. An abundance of primary care providers may amplify the influence of MA plans.

The Harvard researchers used data from the Geographic Variation Public Use File, which contains information on the MA penetration for each U.S. county, as well as the average spending for FFS Medicare from 2007 through 2014. The researchers said theirs is only the second study based on national data from the years after the ACA passed.

One misgiving about MA plans has been a concern that they are skimming the actuarial cream off the top by disproportionately enrolling Medicare beneficiaries in affluent areas with a healthier—and whiter—mix of patients. But to their surprise, the Harvard researchers found that the counties that saw the largest MA growth were poorer and had larger percentages of black residents than counties that experienced the lowest growth.

MA plans have also been controversial because of calculations by MedPAC and others that they are overpaid relative to what the federal government would have paid if the beneficiaries had stayed in traditional FFS Medicare. But Johnson and his colleagues found the spillover effect may more than offset the overpayment and, so, may have a good overall effect on Medicare spending.