Three Audits. Three Alarm Bells.
In March 2026, the OIG published findings from three Medicare Advantage audits simultaneously. The combined results: 857 sampled enrollee-years, error rates between 81% and 91%, and over $15 million in estimated overpayments across just two payment years per organization. The findings weren’t anomalies. They were consistent patterns of documentation failure across three different organizations of different sizes in different regions.
The consistency is what matters most. If error rates varied widely, the failures could be attributed to organization-specific problems. Instead, all three fell in the 81% to 91% range, suggesting a systemic industry problem. The same documentation failures appeared in all three audits: history-of conditions coded as active, acute diagnoses carried forward without current management evidence, and chronic conditions documented without adequate MEAT support.
What the Error Patterns Reveal About RADV Exposure
The OIG audits specifically targeted high-risk diagnosis categories. These are the same categories RADV audits prioritize for sampling because they represent the largest potential overpayments. Acute stroke and myocardial infarction hit 100% error rates in multiple audits. Breast cancer categories also reached 100% failure. These aren’t marginal failures. They’re total documentation collapse in the categories that carry the highest regulatory stakes.
The failure mode is consistent: diagnoses that entered the system during an acute event and were carried forward in subsequent years without re-validation against current clinical documentation. The original coding was likely defensible. The carried-forward coding is what fails, because nobody checked whether the condition still has active management documentation.
For RADV preparation, this means plans should priority-audit their submitted codes in the same high-risk categories OIG targeted. If your internal review of acute stroke, MI, and cancer HCCs finds similar documentation gaps, those gaps will produce similar results when CMS examines the same categories in RADV.
The Pre-Audit Self-Assessment
Plans can run their own version of the OIG’s methodology. Select a random sample of 100 to 200 enrollee-years from your submitted data. Over-sample high-risk categories (acute conditions, single-occurrence diagnoses, morbid obesity). Evaluate each sampled HCC against MEAT criteria using the same standard CMS auditors apply. Calculate your internal error rate.
If your internal error rate approaches 80%, your RADV results will be similarly unfavorable. If it’s below 20%, your coding and documentation practices are strong. The number itself tells you where you stand and how urgently you need to remediate.
This self-assessment should run annually, timed to precede anticipated RADV cycles. It’s the most direct way to predict your own audit outcomes and the most efficient way to identify which diagnosis categories need enhanced validation before CMS examines them.
Using OIG Findings as RADV Preparation
The three March 2026 OIG audits are public. The error patterns are documented. The high-risk categories are identified. Plans that use this information to run internal assessments and fix documentation gaps before their own radv audit in risk adjustment arrives are converting public enforcement data into proactive compliance action. Plans that wait for their own audit results to identify the same problems will discover them under deadline pressure with recoupment on the line.
