Federal health officials are increasingly scrutinizing Medicare Advantage risk adjustment, suggesting policy changes and even clawbacks to come.
Since the inception of Medicare Advantage, from 2004 to 2013, the average risk scores for beneficiaries that are used to guide plan reimbursement have increased at a faster rate than average fee-for-service scores, a study in the Medicare & Medicaid Research Review has found.
"The increase in relative MA scores appears to largely reflect changes in diagnostic coding, not real increases in the morbidity of MA enrollees," concluded Richard Kronick, director of the Agency for Healthcare Research and Quality, and Pete Welch, HHS assistant secretary for planning and evaluation.
"Given the continuous relative increase in the average MA risk score," they wrote, "further policy changes will likely be necessary."
Between 2004 and 2013, the average Medicare Advantage risk score increased 2.2 percentage points more quickly per year than the average fee-for-service scores when using the 2004 model, Kronick and Welch found in their analysis of beneficiary data. Using the 2014 model, the rate increase was less, but still significant -- 1.6 percentage points per year.
(Source: Medicare & Medicaid Research Review)
Using the 2004 risk adjustment model, the average MA score as a ratio of the average FFS score would have increased from 90 percent in 2004 to 109 percent in 2013. Using the model that was partially adopted in 2014, the ratio would have increased from 88 percent to 102 percent over the nine-year period.
Among the conditions showing significant differences, 1.2 percent of FFS beneficiaries were coded with drug or alcohol dependence -- compared to 8.1 percent of MA beneficiaries in the 10 percent of plans with the highest level of coding intensity -- while polyneuropathy was found three times as often in the top decile of MA plans as in FFS.
On average, differences were somewhat less dramatic. In 2012, 1.7 percent of MA beneficiaries were coded with alcohol and drug dependence, compared to 1.5 percent of FFS beneficiaries; 10 percent of MA beneficiaries were coded with polyneuropathy, compared to 6.5 percent in FFS; and 7.9 percent of MA beneficiaries were coded with chest pain (angia), compared to 4.9 percent in FFS.
Still, on average, the MA risk scores are higher and increased at faster rates. Kronick and Welch have three potential explanations.
For one, "the composition of MA enrollment might have changed; for example, MA enrollees in 2013 might be older, relative to FFS, than MA enrollees in 2004, and, as a result have higher risk scores."
"Second, even if there were no change in the composition of MA enrollees, MA enrollees might have gotten sicker more quickly than FFS beneficiaries. Third, it is possible that coding intensity increased in MA."
Comprehensive risk adjustment was first implemented in Medicare Advantage in 2004, then fully phased in 2007, and it's "possible that during this phase-in period when plans were first reporting diagnosis codes from ambulatory settings, the level of coding by MA plans was not as comprehensive as FFS," Kronick and Welch wrote.
But, they continue: "If that were the case, the relatively faster growth in risk scores during this period could be attributable to plans 'catching up' to FFS, rather than coding more intensely than FFS."
While the true difference between the risks of MA and FFS beneficiaries should be the topic of future research, Kronick and Welch wrote, it does seem "that most of the reason that MA risk scores increased more quickly than FFS scores is due to increases in relative coding intensity -- measured as increases in risk scores for stayers -- with little of it accounted for by changes in enrollment mix."
The findings are fueling a bit of a firestorm in some circles -- a media outlet in Washington is suing for documents related to MA overpayments to individual insurers -- and they may spur the Centers for Medicare & Medicaid Services to pursue clawbacks and further changes.
In the proposed 2015 payment system for hospital outpatient and ambulatory surgical centers, CMS outlined a system for recovering overpayments stemming from erroneous data submitted by Medicare Advantage organizations and Part D prescription drug plan sponsors "in the limited circumstances where the plan fails to correct those data upon request," the agency wrote.