Since the 2006 payment year, CMS has conducted Risk Adjustment Data Validation (RADV) audits on Medicare Advantage (MA) plans’ risk adjustment payments. However, CMS has been hindered by the lack of a methodology to extrapolate audit results. On February 23, 2012, CMS posted its final error calculation methodology. This methodology, if implemented, has the potential to spawn large government recoveries against plans that ripple down to plan action against downstream providers.

The good news is that CMS will start auditing only for 2011 and will not look back to the inception of the RADV program. This is interesting given that in CMS’ 2011 budget request to Congress, it asked for statutory authority to extrapolate RADV audits. The request was not implemented and CMS probably takes the position that it did not need Congress’ authority to extrapolate. CMS now cites the Improper Payments Elimination and Recovery Act (IPERA) of 2010 as its authority.

Surely plans will challenge the authority of CMS to extrapolate. CMS will most likely argue its ability to extrapolate is consistent with that of the Medicare fee-for-service program where there also is no specific statutory authority. Indeed, the new methodology provides that CMS will apply a “Fee-For-Service Adjuster” to the MA plans’ extrapolated payment error. CMS will use RADV type audits on fee-for-service claims to obtain this “adjuster” even though some MA plans have no or very limited fee-for-service enrollees. Moreover, the much larger issue from a coding perspective is that it appears CMS will apply the unwieldy CPT Code based Evaluation and Management Guidelines to MA enrollees when it would be much better for CMS to issue clearer guidance on what constitutes an appropriate visit note and follow-up for a MA enrollee. Applying these stricter standards even with an “adjuster” is clearly meant to increase RADV audit recoveries rather than improve MA enrollee care.

Another key provision in the CMS methodology is to separate the sampling into three categories: high, medium and low risk score enrollees. CMS will then select an equal number from each group to comprise the sample. Because medium score enrollees tend to be the most profitable, plans and providers will need to be mindful as to how tiered sample selection impacts error rates. There also could be room for arguments that repayments are necessary only in cases where the plan or provider profited on the enrollee’s condition insofar as the capitated amount and risk adjustment were higher than the cost of care.

It is going to a challenging time for plans and providers. Clearly concerted efforts to educate clinicians and improve documentation now will pay off in better audit results and protect MA risk adjustment payments for both plans and providers.