RADV Audit Extrapolation: How It Changes the Game for MA Plans

What Is RADV Extrapolation

RADV audit exposure is increasing as CMS expands enforcement and extrapolation methodologies. Risk Adjustment Data Validation (RADV) is the process CMS uses to verify that diagnosis codes submitted by Medicare Advantage plans are supported by clinical documentation in the medical record. When an HCC lacks supporting documentation, CMS determines that the associated payment was improper and seeks recovery.

Historically, RADV recoveries were limited to the specific members whose charts were audited — typically a sample of approximately 200 enrollees per contract. If CMS found that 20 of those 200 members had unsupported HCCs, the plan repaid only the overpayment attributable to those 20 individuals. The remaining membership was unaffected regardless of whether similar documentation gaps existed across the broader population.

Extrapolation changes this fundamentally. Under extrapolated RADV, the error rate discovered in the audited sample is applied statistically to the plan's entire enrolled population. A 10 percent error rate in a 200-member sample is no longer a $200,000 problem — it becomes a projection applied to 100,000 or more members, potentially yielding recoveries in the tens or hundreds of millions of dollars.

The Regulatory Timeline

The path to RADV extrapolation spans more than a decade of regulatory development, legal challenges, and industry lobbying.

• 2012 Proposed Rule: CMS first proposed using extrapolation in RADV audits, drawing immediate industry opposition. Plans argued that extrapolation without a Fee-for-Service adjuster would overstate MA error rates because Traditional Medicare was not subject to equivalent documentation scrutiny.
• 2014-2022 Legal and Comment Period: Multiple rounds of comments, industry lawsuits, and Congressional inquiries delayed finalization. CMS conducted pilot RADV audits for payment years 2011-2013 using non-extrapolated methodology during this period.
• January 2023 Final Rule: CMS published the final rule confirming extrapolation methodology for payment year 2018 forward. The rule eliminated the FFS adjuster that the industry had sought, meaning MA error rates would not be reduced by an estimated Traditional Medicare baseline error rate.
• 2024-2025 Implementation: CMS began conducting RADV audits under the new methodology for payment years 2018 and 2019. Initial audit notifications were sent to selected contracts with results expected to be finalized by late 2026.
• 2026 and Beyond: Extrapolated RADV becomes a standard, ongoing audit mechanism. CMS has indicated it will expand the number of contracts audited annually and may conduct audits for multiple payment years simultaneously, compounding financial exposure for plans with persistent documentation issues.

How Extrapolation Works

Understanding the statistical mechanics of extrapolation is essential for quantifying your plan's actual financial exposure. The process follows a defined sequence.

• Sample Selection: CMS draws a statistically valid random sample of enrollees from the audited contract — typically 200 members. The sample is designed to be representative of the contract's overall population.
• Medical Record Review: For each sampled member, CMS requests medical records supporting every HCC that contributed to the member's RAF score. Certified coders review each record to determine whether the documentation supports the submitted diagnosis.
• Error Rate Calculation: CMS calculates the per-member overpayment for each sampled enrollee where unsupported HCCs are found. The average overpayment per sampled member becomes the basis for extrapolation.
• Extrapolation to Population: The sample error rate is applied to the full contract enrollment using standard statistical methodology. CMS calculates a confidence interval and applies the lower bound as the recovery amount — a concession to statistical uncertainty that still produces massive recovery figures.
• Financial Example: Consider a plan with 80,000 enrolled members. CMS audits 200 and finds an average overpayment of $1,200 per audited member. Without extrapolation, the recovery is 200 × $1,200 = $240,000. With extrapolation at the lower confidence bound, the recovery could be 80,000 × $900 (conservative estimate) = $72 million — a 300-fold increase from the same audit findings.

Plans should reference the RADV audit checklist for detailed preparation steps aligned with CMS methodology.

Impact on Medicare Advantage Plans

The financial and operational impact of extrapolated RADV extends far beyond the recovery payment itself. Plans must recalibrate their entire approach to risk adjustment compliance.

• Financial Reserves: Plans must now reserve for potential extrapolated RADV recoveries in their actuarial calculations. Auditors and rating agencies are requiring MA organizations to disclose RADV exposure in financial statements, affecting creditworthiness and parent company valuations.
• Contract-Level Risk: Extrapolation is applied at the contract level, not the plan level. A single high-error contract within a multi-contract organization can generate disproportionate recovery exposure even if other contracts perform well.
• Provider Network Impact: Plans that delegate risk adjustment coding to provider organizations face amplified exposure. If a delegated entity's documentation practices produce high error rates, the plan — not the provider — bears the extrapolated financial consequence.
• Historical Exposure: Because the rule applies to payment year 2018 forward, plans face retroactive exposure for years of historical submissions. Documentation practices that were acceptable under non-extrapolated RADV may prove financially devastating under the new methodology.
• Competitive Dynamics: Plans with strong documentation and coding compliance will face lower RADV exposure relative to competitors. This creates a potential competitive advantage for organizations that have invested in pre-submission audit programs and rigorous documentation standards.

Why This Changes Everything

Before extrapolation, the economics of RADV compliance were manageable. A plan could accept a modest error rate knowing that the financial consequence was limited to a small sample. The cost of achieving near-zero error rates often exceeded the potential recovery from a non-extrapolated audit. That calculus has inverted completely.

• Error Tolerance Is Gone: Under non-extrapolated RADV, a 5 percent HCC error rate on 200 sampled members meant recovering overpayments for 10 individuals — perhaps $50,000 to $100,000. Under extrapolation, that same 5 percent rate applied to a 100,000-member contract could produce $30 million to $75 million in recoveries. The acceptable error threshold has effectively dropped to zero.
• Investment Justification Shifts: Compliance programs that previously appeared expensive relative to RADV risk now look like insurance. A $2 million annual investment in pre-submission chart review that reduces your HCC error rate from 8 percent to 2 percent could prevent $40 million or more in extrapolated recoveries.
• Documentation Culture Must Change: Risk adjustment can no longer be treated as a back-office revenue optimization function. It must be embedded in clinical operations, with providers understanding that every documented diagnosis must be supportable by the medical record because any chart could end up in a RADV sample.
• Vendor Accountability Increases: Plans using third-party vendors for chart reviews, health risk assessments, or retrospective coding must hold those vendors to higher standards. Vendor-introduced errors are indistinguishable from plan-generated errors in a RADV audit.
• Board-Level Visibility: RADV extrapolation has elevated risk adjustment compliance from a departmental concern to a board-level financial risk. Plans should ensure their governing bodies understand the magnitude of extrapolated exposure and the adequacy of their mitigation programs.

Preparation Strategies

Proactive preparation is the only effective response to extrapolated RADV. Plans that wait for an audit notification to begin remediation have already lost the most valuable preparation window.

• Internal Pre-RADV Audits: Conduct annual internal audits using the same statistical sampling methodology CMS employs. Audit at least 200 members per contract with independent medical record review to identify your true error rate before CMS does. Use findings to reduce RADV exposure systematically.
• Prospective Chart Review: Review medical records before encounter data is submitted to CMS rather than after. Prospective review catches unsupported diagnoses at the point of submission, preventing them from entering the RADV-auditable dataset entirely.
• Provider Education Programs: Train network providers on documentation standards that meet RADV scrutiny. Focus on the specific elements CMS coders look for: condition specificity, clinical indicators, treatment plans, and linkage between documented conditions and ICD-10 codes.
• HCC-Level Risk Scoring: Use analytics to identify which specific HCCs across your population carry the highest documentation risk. Concentrate pre-submission review on high-value, high-risk HCC categories where unsupported diagnoses are most likely and most financially impactful.
• Vendor Audit Programs: If you use third-party vendors for chart reviews or health risk assessments, audit their work product independently. Verify that vendor-submitted diagnoses meet the same documentation standards you apply to provider-generated encounters.
• Technology Investment: Deploy a risk adjustment analytics platform that continuously monitors documentation quality across your entire population rather than relying on periodic sample-based reviews. Continuous monitoring catches emerging documentation issues before they compound into systemic error patterns. Building a comprehensive risk adjustment analytics program is the foundation for sustained RADV readiness.