More Rigorous Audits Coming to Medicare Advantage


Sarah Snider, Andrew Lay, Sam Wu, and Eric Williams

8 min read

Insurers offering Medicare Advantage plans need to ready themselves for more intense reviews by federal auditors. The Centers for Medicare and Medicaid Services in late January finalized a controversial rule laying out a new approach for how it will recoup improper risk adjustment payments made to MA plans. CMS’ 2018 proposed rule on the audit program garnered significant criticism from Medicare Advantage Organizations and the final rule is no different.

Under the new methodology, CMS expects to collect $4.7 billion in improper payments over the next 10 years, a significant escalation from previously projected recoveries. The first payments will be due in 2025.

Below we spotlight three areas that will be significantly impacted under the final rule and what it all means for Medicare Advantage Organizations.

Error extrapolation retroactive to 2018

For the past decade, CMS audited MA contracts by selecting a sample of beneficiaries, reviewing diagnosis codes submitted by each MAO against beneficiaries’ medical records and estimating the total risk adjustment payment error for the entire contract. It did so by extrapolating the rate of unsupported diagnoses from the sample size to the full contract population.

Under the 2023 final rule, CMS is moving to contract-level payment recoveries based on error rate extrapolation and applying this methodology retroactively, going back to 2018. In the proposed rule, CMS suggested extending retroactive audits to 2011. While the final rule doesn’t extend contract-level extrapolation that far back, plans will be responsible for overpayments identified in samples from 2011-2017. CMS said it will notify affected plans on a rolling basis of audit findings dating to 2011, providing them with an understanding of their overpayment risk going forward.

The shift to retroactive contract-level extrapolation poses significant challenges for MAOs, including:

Uncertain statistical methodology: CMS is not publishing its audit methodology, stating it will “rely on any statistically valid method for sampling and extrapolation that is determined to be well-suited to a particular audit.” While saying it is not required to do so, CMS committed to providing “information sufficient to understand” how improper payments were calculated. The rule also gives CMS freedom to adopt more aggressive methodology, even differing from year to year. This will make it difficult for MAOs to predict potential overpayment recoveries since they won’t have transparency into CMS’ methodology.

Impact of sampling error: In the final rule, CMS argued that a small sample size — as few as 201 MA beneficiaries per contract based on the previous guidance from 2012— is sufficient to extrapolate the contract-level payment error rate. This small sample size could lead to major problems when applied to a full contract. Moreover, CMS alluded to using a sub-cohort of beneficiaries for whom particular diagnoses were reported as a part of its methodology. This idea was raised in the 2018 proposal but not codified in the final rule, other than CMS saying it was a way to provide the industry with transparency on its approach. In the proposed rule CMS also said it would “focus on cohorts of enrollees that appear to raise programmatic concerns.” Targeting high-risk, high-intensity diagnoses could result in larger, nonrepresentative overpayment rates being applied to full MA contracts.

Removal of the Fee-for-Service Adjuster

Implemented when RADV started in 2012, the FFS Adjuster accounts for the fact that CMS’ model predicts costs based on fee-for-service Medicare claims data, which have an inherent level of payment errors. Since these payments are not subject to RADV oversight, MAOs have only been accountable for errors above the level seen in fee-for-service Medicare.

CMS in the final rule determined that the errors “did not have any systematic effect on the risk scores … and, therefore, did not have any systematic effect on the payments made to MAOs.” Even if a systematic effect existed, CMS said that an adjustment factor would not be appropriately applied in the RADV context.

Eliminating the FFS Adjuster creates several concerns for MAOs, including:

Higher overpayment liabilities: The new RADV methodology could increase recoveries sought by CMS since overpayment amounts starting with plan year 2018 will be recouped without applying the offsetting FFS Adjuster. This is compounded by the small RADV sample size issue mentioned above.

Compliance costs: Without the FFS Adjuster, MAOs are essentially held to a 0% error rate as opposed to bearing responsibility only for error above fee-for-service levels. A zero-error standard creates a significant burden on MAOs and requires increased compliance oversight and outlays on such operations as coding integrity and chart retrieval to limit financial risk. This is particularly burdensome for new market entrants and smaller regional plans that typically operate on thin margins.

Strategic impact of 2023 RADV final rule

Beyond the specific impacts laid out above, the additional uncertainty and financial liability driven by the new RADV methodology creates broader strategic challenges for MAOs, including:

Redefining risk adjustment operations: MAOs have typically relied on retrospective reviews conducted in the year following each plan year — 2022 for 2021 date of service claims — to ensure all applicable diagnoses are captured for each member. Under the new methodology, MAOs will be incentivized to shift their focus and increase prospective and concurrent review activities to improve documentation and better engage with providers closer to the date of service. MAOs may also increase the proportion of hierarchical condition categories codes they delete due to lack of documentation given the new zero-error standard, creating further financial headwinds in bid mechanics.

Adjusting benefit offerings: More importantly, an increase in RADV audit recoveries and an associated increase in risk adjustment conservatism presents significant financial anxieties that will quickly play a role in MA bid mechanics. MAOs may need to shift the financial risk of future audit results to enrollees in the form of higher premiums. Or MAOs may look at reducing supplemental benefits, diminishing Part B premium give backs, and instituting higher enrollee cost sharing. CMS will likely target MAOs with historically high error rates, so these firms need to consider restructuring their risk adjustment practices. These firms may need to adjust benefit packages to offset higher go-forward audit recoveries, while MAOs with lower historical error rates will be less affected by the new methodology.

Don’t wait, prepare now

Although the ink is barely dry on the RADV final rule, it has already become a lightning rod for the MAO community. Most major insurer advocacy groups have questioned the legality of the rule, particularly whether it falls under the narrow circumstances wherein retroactive rulemaking is permitted. Critics contend that the rule is not only unlawful, but may lead to higher costs, reduced benefits, and fewer plan options. As of this writing, some MAOs and their trade associations were weighing legal action.

While this debate unfolds, MAOs must prepare now to compete and succeed under the new RADV methodology. An enhanced focus on compliance, robust risk adjustment operations and clinical integration, and smart bid economics must be prioritized to win in the MA marketplace of tomorrow.

To learn more contact Matthew Weinstock, Senior Editor, Health and Life Sciences.

  • Sarah Snider,
  • Andrew Lay,
  • Sam Wu, and
  • Eric Williams