A new rule proposed by the Centers for Medicare and Medicaid Services (CMS) on October 26, 2018, would revise the way the agency validates the risk adjustment data and collects repayments from Medicare Advantage (MA) organizations. With the new methodology, CMS is expecting to return $4.5 billion in savings to the Medicare Trust Fund over 10 years, according to an October 26 CMS news release.
Additionally, CMS announced that it completed a review of records supporting fee-for-service (FFS) Medicare Part A and Part B claims and found an absence of systematic bias in payments to, and recoveries from, MA organizations. As a result, CMS decided against including a FFS Adjuster in the methodology for determining risk adjustment data audit recoveries.
Background
CMS’ payments to MA plans are adjusted based on the health status and demographic characteristics of the enrolled population—the sicker the enrollees, the higher health care costs are expected to be, resulting in larger payments. CMS uses ICD-10-CM diagnosis coding to prospectively determine health risk scores for enrollees and adjust payments accordingly.
The accuracy of the risk-adjustment methodology used in the MA program depends on accurate coding, which the ICD-10-CM guidelines say ‘“cannot be achieved” without “consistent, complete documentation in the medical record,”’ according to the CMS proposal. In other words, the extent of documentation in the medical record will determine the “truthfulness” in any MA organization’s risk adjustment data submission.
The risk adjustment model used to determine MA plan payments categorizes ICD-10-CM codes into disease groups called Hierarchical Condition Categories (HCCs), which include clinically related diagnosis codes. MA organizations submit data used by CMS to assign the HCCs, which contribute to enrollee risk scores. To check the accuracy of MA plans’ self-reported data, CMS annually selects MA plans for risk adjustment validation (RADV) audits. CMS looks through hospital and provider medical records for diagnoses related to the HCC profile. When an enrollee’s HCCs used in payment don’t match the information found in the audit—a risk adjustment discrepancy—that is considered a payment error.
CMS has been inconsistent with its payment validation audits; since 1999, audits were conducted sporadically. Any resulting payment adjustments were limited to payments associated with a sample of enrollees pulled for the audits. CMS notes that the agency could use the payment error rate found in the sample and extrapolate it to calculate a contract-level payment error rate. The methodology for extrapolation was published in 2012. CMS used the methodology to conduct audits in payment years 2011, 2012 and 2013 but never finalized the audits—and never used the extrapolation process.
Proposed Action
In the proposed rule, CMS is again announcing its intent to apply extrapolation for 2011, 2012 and 2013. CMS is also proposing to use a sub-cohort methodology for sampling and extrapolation to calculate improper payments to MA plans. A sub-cohort could comprise enrollees for whom a particular HCC was reported, such as those that appear to raise programmatic concerns, according to CMS. The agency believes the smaller, sub-cohort sample size would enable CMS to spread audit resources across a wider range of MA contracts and still generate statistically significant recoveries.
CMS seeks comment on both the previously proposed contract-level audit methodology and the extrapolated audit methodology based on sub-cohorts of enrollees. CMS also seeks comment on whether the audits already conducted but not finalized for payment years 2011, 2012 and 2013 should be revised.
For payment year 2011 and subsequent years, CMS plans to finalize and retroactively apply one or more of the RADV payment error methodologies under consideration, except that the sub-cohort-based methodology would not be used before 2014. Retroactive application of the finalized methodology may require exercising statutory authority, which could be possible if CMS were to show that failure to do so would be contrary to the public interest. The potential numbers are huge. CMS estimates that $14.35 billion was improperly paid to MA plans in 2017 and CMS projects $650 million will be recovered from MA organizations for 2011, 2012 and 2013. CMS is seeking comment on this consideration too.
If the new, finalized contract-level audit methodology differs from the one published in 2012, CMS will consider whether to apply the new methodology to 2011 through 2013. Alternatively, CMS would begin applying the new methodology to payment year 2014 and, for the earlier years, apply the methodology published in 2012. CMS said, regardless, “we believe it is vitally important for the health of the MA program to have extrapolated recoveries available for future audit years.”
In 2012, CMS stated its intent to address concerns about possible erroneous diagnosis codes in the FFS Medicare Part A and Part B data used to calculate payments to MA plans. At the time, CMS said it would apply an FFS Adjuster to account for error in FFS Medicare data and limit the RADV audit recoveries accordingly. The amount of the adjuster would be determined through a review of records submitted to support FFS claims data.
On October 26, 2018, CMS released its report on the FFS data review and concluded “errors in FFS claims data do not have any systematic effect on the risk scores calculated by the CMS-HCC risk adjustment model, and therefore do not have any systematic effect on the payments made to MA organizations.” Further, CMS believes it would be inequitable to correct any systematic errors in the payments made to audited MA plans and not to apply it to unaudited MA plans. As a result, CMS dropped its plan to include a FFS Adjuster in the methodology used for determining risk adjustment data audit recoveries.
Regarding the potential inequity between MA payments and FFS Medicare payments, CMS said in a footnote that the agency is aware of the September 7, 2018, decision in United Healthcare Insurance Co. v. Azar and is considering its response (see previous Healthcare Law Insights posting at https://www.healthcarelawinsights.com/2018/10/court-decision-on-overpayment-rule-leaves-uncertain-future-for-medicare-payment-methodology-and-pending-justice-department-lawsuits/). In the footnote, CMS also said the ruling “was made on the basis of the administrative record before the court, which did not include the results of our study.”
