RADV Audit Compliance Strategy for Medicare Advantage Plans

A CMS RADV audit verifies whether the diagnosis codes submitted by MA plans are supported by medical record documentation. These codes reflect enrollees’ health status and, together with other factors, determine the payment level from CMS to MA organizations. Since payments are higher for enrollees with diagnoses that are expected to require more intensive healthcare services, CMS is focused on ensuring the diagnoses are accurately coded.

RADV audits help CMS identify diagnosis codes not properly supported in medical records, as well as identify payment errors to MA organizations. In recent years, CMS has changed the particulars of its sampling and extrapolation audit methodology.

During a RADV audit, CMS will identify a sample of enrollees, or possibly health conditions, consistent with the methodology that CMS is using for that audit. CMS or its contracted auditors will request medical records from the selected MA organizations to support the sampled selection. The medical records will be reviewed to determine whether the diagnoses included in the audit sample are supported.

As stated in the 2026 Advanced Notice for methodology changes, CMS seeks to refine the types of diagnosis submissions that will be considered for risk score calculation. This includes the exclusion of audio-only telehealth services and unlinked chart review records (CRRs) starting in 2027.

Audio-only telehealth appointments do not meet the “face-to-face” requirement for risk adjustment and will be excluded from both model calibration and risk score calculation. CMS also proposes to exclude all diagnoses submitted via CRRs that are not linked to a previously accepted Encounter Data Record (EDR). This move addresses data integrity concerns where diagnoses were submitted without an associated record of service.

Also in the January 2026 memo, CMS indicates it will reduce the maximum number of medical records allowed per audit Hierarchical Condition Categories (HCC) from five to two. This places higher stakes on the quality of each record submitted for review during an audit.

CMS may use a sub-cohort methodology in some RADV audits as an alternative to other sampling methods. While CMS has not specified which diagnosis codes the RADV program may focus on, a review of certain published audits conducted by the U.S. Department of Health and Human Services’ Office of Inspector General (OIG) may indicate particular diagnosis groups of interest:

With spending on MA only increasing, MA plans need to be ready for what has changed and what will change in the years to come by having a sustainable RADV audit strategy to support the proposed quarterly audit schedule.

For MA plans, properly preparing for RADV audits is strategically important. MA plans should consider adopting practices to mitigate the risk of financial recoupments and maintain the integrity of their risk adjustment data submissions.

“MA plans need to be ready for an audit at any time,” says Kathryn Eshelman, M.D., MPH, Vice President of Medical Informatics at Inovalon. “If a health plan fails to substantiate the diagnoses in even a small percentage of cases, the financial repercussions can be substantial, with potential losses amounting to a significant slice of their revenue. Beyond the financial impact, there is the risk of reputational damage and the potential of reducing the quality of care for members due to inaccurate documentation of diagnoses.”

For Eshelman, MA plans need to establish a compliance program because it serves as a system of checks and balances to ensure the accuracy and integrity of coding and billing practices.

“Every health plan needs a compliance program and a dedicated coding team,” she maintains. “CMS expects plans to actively look for and fix coding errors. The goal is to prevent incorrect codes from being used from the start, which is better than trying to correct them later. This means focusing on the most relevant data and avoiding unnecessary or incorrect information that some others might include.”

According to Eshelman, MA plans should also consider directly engaging providers and deliver clear guidance on proper documentation and adherence to coding guidelines before submitting claims — all while simultaneously reviewing past claims to identify and correct any errors. Technology and analytics can aid in this process by highlighting potential issues early on and supporting ongoing compliance efforts. Collecting thorough documentation to support diagnoses can also be important since patient data is often scattered across various electronic health records.

“Organizations without advanced surveillance technology face a daunting and labor-intensive task akin to finding a needle in a haystack when preparing to respond to RADV audits,” she warns.

“However, by employing sophisticated surveillance analytics, health plans can efficiently and effectively pinpoint areas of concern within their data. For optimal audit preparedness, combining technology-driven internal audits with traditional random sampling methods helps health plans vet their coding practices.”

Eshelman suggests that internal audits should occur multiple times annually, with at least one early in the year for initial assessment. Then, perform additional audits later, particularly in the fall, to ensure any coding errors are corrected before submission to CMS.

Choosing the right technology partner is another factor in ensuring accurate data collection and analysis as well, especially given the evolving nature of RADV audits to date and into the future.

“The ideal technology partner is committed to continuously refining its solutions, utilizing a dedicated clinical team and coding team to ensure all systems and data are accurate,” says Eshelman. “An innovation partner should leverage artificial intelligence (AI) effectively, benefiting from extensive data sets to refine their models while recognizing that AI requires appropriate clinical and coding oversight and cannot yet operate independently. In an evolving healthcare landscape, it’s critical that their systems remain effective, accurate, and trustworthy for health plans preparing for RADV audits.” Rigorous RADV audit preparation ensures compliance and strategic success and requires robust verification, continuous training, and technology integration to safeguard against financial and reputational risks.

To prepare for RADV audits, MA plans can help their provider partners be successful in completing the documentation required to substantiate the diagnosis codes submitted for risk adjustment.

Potential strategies include:

Standardizing data collection. Start by developing and implementing standardized documentation practices across all care settings — for example, requiring providers to submit what service was provided, the date, the diagnosis, procedures conducted, the cost, and other key pieces of data. By determining which data entries need to be captured, MA plans can increase consistency and completeness as they collect data.

Standardizing your data will also help streamline the ingestion and integration process as well, as standardized data will need less time to clean and normalize, and will be ready faster for analysis.

Conducting regular education sessions. MA plans should consider conducting regular education sessions for healthcare providers on the importance of data accuracy. During education, highlight the necessity for the data accuracy that’s required for risk adjustment and RADV audits.

Conducting regular education sessions can also help build relationships between plans and providers, and open further communication around the value of good data practices. Ensuring that everyone is “on the same page” will increase buy-in to data accuracy and decrease error risk as well.

Providing feedback to providers from internal audits. If an MA plan identifies coding irregularities or errors during an internal audit, the MA plan could develop feedback to providers and/or incorporate such feedback into education sessions.

Leveraging digital tools for data accuracy. Electronic health records (EHR) contain essential data that MA plans can use to round out their data collection and compliance. However, ingesting and integrating EHR data into risk adjustment reporting has traditionally been time consuming, manual, and error prone. With the help of AI, plans can transform data extraction and data analysis to better capture, ingest, and apply EHR data in risk adjustment reporting.

Collecting, cleaning, managing, and maintaining data is thankfully no longer a manual process, and emerging technologies can greatly assist MA plans in the efficiency and accuracy of their RADV audit preparation. The following technologies and advances can help implement increased data management and governance systems:

Artificial intelligence, machine learning (ML), and natural language processing (NLP). MA plans looking to increase their data management and compliance can use AI or ML algorithms to identify unusual patterns or anomalies in their data submissions, potentially flagging inaccuracies or areas requiring further investigation. AI can also automate and speed up various steps of the audit process, from initial data analysis to identifying potential risk areas, reducing the manual effort required for reviewers. Additionally, MA plans can use NLP to simplify the review of collected provider documentation, improving the identification of necessary information for validating diagnosis codes. Such use of AI, ML, and NLP should be reviewed by individuals with expertise in risk adjustment and documentation requirements.

Advanced data analytics. Plans can incorporate analysis tools into the audit approach as well, which could include integrated analytics or implementing healthcare-focused analysis software to support accurate risk adjustment and reduce the risks associated with audits. Advanced data analytics and predictive models can forecast future audit risks, enabling a more proactive compliance strategy and earlier recognition of corrective measures.

With CMS seeking to recover billions in overpayments, MA plans must prioritize RADV audit preparation to avoid financial losses and reputational damage, and ensure their members receive appropriate care.

A compliance program with a dedicated coding team is essential to maintaining the accuracy and integrity of data submissions. Taking proactive measures through provider engagement and advanced surveillance analytics can streamline the audit preparation process. Conducting internal audits throughout the year can further strengthen a health plan’s compliance posture.

Partnering with a technology provider committed to ongoing innovation and improvement, particularly one that integrates AI with clinical oversight and coding expertise, will help MA plans keep pace with regulatory changes and safeguard against the risks associated with RADV audits.

For more strategies to prepare for RADV audits, read our blog on identifying potentially over-coded conditions early on.