Medicare Plans Face Massive Overpayment Recoveries Under New CMS Methodology

On February 24, the Centers for Medicare and Medicaid Services issued a Notice of Final Payment Error Calculation Methodology for Part C Medicare Advantage Risk Adjustment Data Validation Contract-Level Audits ("Notice"). Beginning with payment year 2011, CMS will calculate payment recovery at the Medicare Advantage plan level based on extrapolated estimates. If a plan is selected for an RADV audit under the new methodology, it will be critical for the plan to respond quickly and thoroughly in order to minimize an extrapolated overpayment recovery.

Previously, CMS had limited payment recovery only to the sample size of a RADV audit—a tiny fraction of a plan's total enrollees, resulting in quite modest overpayment recoveries by CMS. Under the new extrapolation model, CMS will seek hugely increased recoveries from plans for alleged overpayments. The Notice also details the sampling method CMS will use for RADV audits beginning in payment year 2011. For a Medicare Advantage plan with as few as 15,000 enrollees, the overpayment recoveries would increase by nearly 7,500%. As a result, Medicare Advantage plans selected for a RADV audit should engage counsel so they are primed to challenge CMS's overpayment determinations.

The risk adjustment payment model is designed so that payments to a Medicare Advantage plan are tied to the health status of its enrollees, and a plan receives a larger payment for a sicker enrollee than a healthier one. Under the Medicare Advantage payment model, CMS's payments are based specifically on the diagnoses the plan submits for each enrollee. CMS groups certain diagnosis codes into separate disease categories known as Hierarchical Condition Categories ("HCCs") and calculates a risk score for each enrollee based on his or her HCCs. The risk score in turn determines the risk adjustment payment to be made to a plan on behalf of the enrollee. In a RADV audit, each Medicare Advantage plan selected for the audit is required to submit medical record documentation for each HCC on which payment is based. CMS contractors then review these medical records to determine if they believe the records support the diagnoses submitted by the plans. Those diagnoses that the CMS contractor concludes are not validated by the medical records during the RADV audit are then used to calculate a payment error rate that will now be extrapolated by CMS over the entire plan population to recover claimed overpayments.

The risk adjustment system contains a significant amount of built-in error. CMS currently estimates the national payment error rate at 11 percent for fiscal year 2011 (in prior years, CMS has estimated an HCC error rate of over 30 percent). CMS expects that RADV audits will both recover overpayments made to Medicare Advantage plans and positively impact the quality and accuracy of the risk adjustment data submitted by Medicare Advantage plans. CMS estimates that the first audit applying the new methodology will recover approximately $370 million in overpayments.

The Methodology Introduces New Sampling Methods

For each RADV audit, CMS selects a number of Medicare Advantage plans to be audited and a sample of beneficiaries from each audited plan's RADV-eligible enrollees. Under the new methodology, to select those beneficiaries, CMS will first rank a plan's beneficiaries from lowest to highest based on their community risk score, and then divide the enrollees into three equal groups of the highest, middle, and lowest risk scores.

Next, CMS will select up to 201 enrollees for the RADV from each plan being audited – 67 enrollees from each of the three designated groups. (CMS will select a proportionately smaller sample for plans with fewer than 1,000 RADV-eligible beneficiaries.) CMS will then calculate an enrollee weight for each of the three groups by dividing the total number of RADV-eligible enrollees in each group by the number of enrollees sampled from that group. CMS will use these weights to extrapolate the sample payment error determination to the entire plan population.

The Medicare Advantage plans selected for the RADV audit will then submit medical records to support each HCC. For the first time with the calendar year 2011 RADV audit, CMS will allow Medicare Advantage plans to submit multiple medical records for each HCC being validated. This change will almost certainly increase the likelihood of CMS validating a particular HCC during the RADV audit, resulting in a lower payment error calculation for the audited plans. CMS has noted, however, that the "one best medical record" policy will still apply during the RADV dispute and appeal process to which audited plans are entitled under 42 C.F. R. § 422.311. CMS has indicated that it will provide more detail regarding the RADV audit procedures to the Medicare Advantage plans selected for each RADV.

New Payment Error Calculation Methodology

CMS will calculate each Medicare Advantage plan's payment error based on the results of the medical record validation. For each enrollee included in the RADV sample, CMS will calculate a corrected payment based on the HCCs validated by the RADV audit and then a payment error for each enrollee based on the difference between the original payment and the RADV-corrected payment. The enrollee-level payment error will then be extrapolated to determine the payment error for the entire plan by multiplying the enrollee's sampling weight by the annual payment error for each sampled enrollee. CMS will sum the weighted enrollee payment error across all enrollees in the sample to determine an estimated payment error for each audited Medicare Advantage plan. Finally, CMS will calculate a 99 percent confidence interval for the estimated payment error for each audited plan, which along with a Fee-for-Service adjuster amount as an offset, will determine the ultimate payment recovery amount. The Fee-for-Service adjuster takes into account that the documentation standard for RADV audits is different from the documentation standard used to develop the Medicare Part C risk adjustment model.

Implications of New Methodology

CMS's first effort at a risk adjustment extrapolated plan-level payment adjustment was through a False Claims Act litigation in which CMS effectively short-circuited the RADV audit process and proceeded directly to litigation based on the RADV audit results. That case, U.S. v. Janke, Case No. 09-14044-CIV-MOORE/LYNCH (S.D. Fla.), for which the Pillsbury attorneys authoring this Client Alert were lead counsel, is the only known risk adjustment litigation. There, the United States relied on the RADV audit results to try to establish False Claims Act liability, and also relied for the first time on an extrapolated result in its effort to seek dramatically increased damages for alleged overpayments. (The case ultimately settled before trial.)

With CMS's universal implementation of the extrapolation methodology to the RADV payment error process, administrative proceedings and related litigation will very likely increase as CMS's new extrapolation calculations subject Medicare Advantage plans to payment recoveries many multiples greater than in the past. The consequences and risks to a plan that is chosen for a RADV audit are significantly increased. By way of example, if a Medicare Advantage plan has 15,000 RADV-eligible enrollees, the sample size will consist of three groups of 67, for a total of 201 enrollees, and each enrollee's sampling weight would be 74.627 (5,000 enrollees per group divided by 67). Assuming an average overpayment of just $500 for each sampled enrollee, the new extrapolated plan-level overpayment amount would total over $7,500,000 ($500 average overpayment x 74.627 sampling weight x 201 enrollees), as compared to the result under the old methodology of an overpayment of just $100,500 based solely on the sample size.

With this new methodology, it is even more important for Medicare Advantage plans to attempt to protect themselves from large payment error recoveries by submitting medical records to support each HCC during the RADV audit, and by taking advantage of CMS's new policy that more than one medical record may be submitted for each diagnosis. Furthermore, because each alleged discrepant diagnosis takes on such significance under the new extrapolation model, once the RADV results are announced, it is critical that affected Medicare Advantage plans be positioned to appeal and aggressively challenge these determinations. The extrapolation model could also lead to repercussions in the mergers and acquisitions context, where acquirers and lenders need to account in their diligence for the potential significant exposure an acquired plan may face as a result of a large, extrapolated overpayment recovery.

Download Medicare Plans Face Massive Overpayment Recoveries Under New CMS Methodology