Risk Adjustment Coding Errors That Cost Plans Millions

Risk Adjustment Coding Errors That Cost Health Plans

The health plans and physician groups that participate in the Medicare Advantage program, the ACA marketplace plan, and value-based Medicaid programs depend on one metric for survival – risk score. This metric is calculated via risk adjustment coding, and even minor and consistent mistakes with risk adjustment coding can cost a health plan millions of dollars per year – whether because of being underpaid and losing money it should be paid or being overpaid, which results in CMS audits and other penalties.

This article intends to provide answers to the following questions: What is risk adjustment coding? How does the HCC risk adjustment coding work? What mistakes do health plans make while doing it? And how AI-powered platforms help health plans and physician organizations avoid such mistakes before they become a big issue in 2026.

What Is Risk Adjustment Coding?

Now that we have that out of the way, let's move on to the most common risk adjustment coding errors. First off, what is risk adjustment coding? Risk adjustment coding is the process of analyzing clinical documentation and assigning proper ICD-10-CM diagnosis codes in order to accurately represent a complete list of chronic disease conditions for a patient, so that health plans are paid according to the real cost of taking care of that particular population.

Risk adjustment coding is not the same thing as fee-for-service coding, because the latter applies to a single encounter, while risk adjustment coding involves a whole population over time. Every diagnosed condition (or lack of diagnoses) in a calendar year will impact an individual member's RAF. The CMS, followed by some commercial and Medicaid payers, will take all of those codes and calculate a RAF score, which in turn will be used to determine the payment for that particular member.

In other words, risk adjustment coding ensures that health plans providing coverage for sicker members are fairly compensated compared to those health plans that have healthy individuals under their coverage.

What Is HCC Risk Adjustment Coding?

What most people are asking when they inquire about risk adjustment coding in the Medicare Advantage world refers primarily to HCC risk adjustment coding. HCC stands for Hierarchical Condition Categories, and it is a classification process developed by CMS that combines several thousand ICD-10-CM diagnosis codes into around 115 condition categories (using CMS-HCC Version 28 under the full Phase In approach for Payment Year 2026).

The process of HCC risk adjustment coding is accomplished through the identification of qualifying diagnoses and assigning them to specific HCCs that are weighted for risk. If a member has diabetes with chronic kidney disease and congestive heart failure, for example, he or she will be assigned to several HCCs, each contributing to the total RAF score.

This is because HCCs use a hierarchy system in that only the highest severity of the condition in the related disease category can be considered (in most hierarchies); HCC risk adjustment coding needs to be accurate rather than voluminous. Assigning less serious versions of the conditions when more serious cases exist or failing to assign any at all leads to an understatement of the risk score.

Why Risk Adjustment Coding Errors Are So Costly in 2026

Financial risks associated with risk adjustment coding have been heightened even further going forward in 2026 due to several overlapping factors:

The CMS-HCC model V28 changed weights and eliminated or reclassified many diagnosis codes from the older V24 model. This makes it inevitable for health plans using the old risk adjustment coding or using old crosswalks to mis-code their members.
Broadened RADV audits. The rules were finally passed by CMS permitting extrapolation of the RADV audit results beyond the sampled number of records, but throughout the entire Medicare Advantage membership of the plan in question. Any pattern of error in the coding process discovered through an audit will result in a financial penalty through a much larger membership than the sample.
Increased scrutiny on undocumented diagnosis codes. OIG and CMS have been continuously warning of diagnosis codes which may appear on the claims data, but are not supported by any medical evidence in the medical record. That is the key mistake of manual risk adjustment coding.
Star Ratings and quality payments are increasingly overlapping with risk adjustment.

The Most Common Risk Adjustment Coding Errors

The following are the mistakes that repeatedly occur in health plan audits and internal coding reviews, causing significant financial risks.

1. Unsupported diagnoses. The diagnosis code is used for risk adjustment coding purposes, but there is no indication in the medical record that the disease was examined, monitored, assessed, or treated (MEAT/TAMPER criterion) during that encounter.

2. Missed chronic diseases. Providers mention the existence of a condition in free text or problem lists but fail to turn it into a codable diagnosis for risk adjustment coding review. As a result, some of the true risk is not captured, and the health plan gets an inadequate payment for its services.

3. Wrong HCC categorization. There is a correct diagnosis in terms of ICD-10-CM codes, but it is mapped incorrectly to the HCC due to outdated crosswalk logic. It is becoming a problem in the transition between V24 and V28 models for many health plans.

4. Mistakes of copy-forward. The diagnosis is copied from one year to another, and it is not reassessed, making the coding invalid according to CMS rules, which require face-to-face risk adjustment coding in the current year.

5. Lack of hierarchy interaction. In the HCC risk adjustment code, missing out on the hierarchy interaction between the conditions may lead to over-counting or under-counting of the most severe condition.

Common Risk Adjustment Coding Errors and Their Financial Impact

Error Type	Root Cause	Typical Financial Consequence
Unsupported diagnosis	Missing MEAT/TAMPER documentation	RADV repayment, extrapolated clawback
Missed chronic condition	Condition in notes but never coded	Underpayment, inaccurate RAF score
Incorrect HCC mapping	Outdated crosswalk logic (V24 vs. V28)	Misaligned reimbursement, audit flags
Copy-forward diagnosis	Prior-year code reused without re-evaluation	Compliance risk, potential overpayment
Specificity gap	Unspecified code used instead of specific one	Understated risk score, lost revenue
Hierarchy miscoding	Misapplied HCC hierarchy rules	Overcounted or undercounted RAF score

How These Errors Add Up to Millions

On its own, one risk adjustment coding mistake could equate to hundreds or thousands of dollars being incorrectly reimbursed. Yet, health plans conduct their risk adjustment coding on the total population of members they cover, which could easily number in the hundreds of thousands of patients, with each inaccurate RAF score carrying over year after year because risk adjustment coding is reassessed on an annual basis by calendar year.

Under extrapolation of RADV audit technique by CMS, one could see that any coding error rate revealed in just a few hundred records audited can be extrapolated onto the total Medicare Advantage membership of the health plan, creating an eight-figure or even nine-figure repayment liability. From the other side, consistently under-reporting actual chronic conditions due to inadequate risk adjustment coding implies that health plans are never getting paid what they deserve for the populations they actually care for.

How AI Is Reducing Risk Adjustment Coding Errors in 2026

Health plans and the provider groups and RCM vendors working for them are increasingly embracing AI-enabled solutions to deal with the identified error patterns at scale. The risk adjustment coding technology of today, including the one offered by RapidClaims, emphasizes certain core functionalities:

Natural language processing (NLP) chart review. Using AI, the solution analyzes entire medical records rather than problem lists, finding those chronic diseases that are described but never converted to a coded form.

MEAT/TAMPER validation. AI solutions learn how to find those diagnoses that lack the necessary supporting documentation to be submitted in the first place, lowering the number of unsupported diagnoses reported in the future.

Automated HCC crosswalk updates. Since CMS-HCC model updates, such as the one from V24 to V28, occur regularly and with a certain schedule, an AI-enabled solution will be able to update the corresponding mapping logic automatically.

Suspecting and gap closure. With the help of AI-based suspecting models, the tool identifies undiagnosed or undercoded chronic conditions using claims, labs, and pharmacy data, leading to chart review or provider outreach.

Audit readiness scores. The confidence score and documentation for each code recommended by the AI can be established, ensuring that the compliance team has an audit-ready history, which is important considering the increased extrapolation by CMS RADV.

Best Practices for Health Plans in 2026

Reconcile the risk adjustment coding methodology with the latest version of the CMS-HCC model at least once a year and following any update to the CMS model.
Give priority to prospective chart review and supporting documentation rather than retrospective "code chasing".
Utilize AI-powered risk adjustment coding together with certified human coder supervision, especially when it comes to high-risk and/or high-RAF impact charts.
Run internal MEAT/TAMPER audits continuously and not just ahead of any upcoming CMS audit cycle.
Monitor HCC-level accuracy statistics and not just RAF score trends.

Final Thoughts

Ultimately, what is risk adjustment coding? Risk adjustment coding is what makes sure health plans are being reimbursed fairly for their populations. For Medicare Advantage plans, in particular, HCC risk adjustment coding is the main driver of financial risk and financial opportunities. The risks associated with unsupported diagnoses, missing conditions, mapping issues, copying of diagnoses forward, lack of specificity, and problems with hierarchy are not the rare exceptions that may happen once a year. They occur regularly during health plan audits and become more expensive than ever before under new RADV extrapolation guidance issued by CMS.

For health plans, combining strong documentation standards and the latest technology like RapidClaims platform for risk adjustment coding can be the way to mitigate these risks and protect from audits in 2026.

Whether you have an existing and mature HCC risk adjustment coding program or just start asking yourself about risk adjustment coding, the basics for 2026 remain the same: proper documentation, updated crosswalk logic, and AI-based chart review.

FAQs

What is risk adjustment coding, in simple terms?

What is risk adjustment coding, answered plainly: it's the process of documenting and coding every chronic condition a patient has so a health plan's payment reflects the real cost of caring for that population, rather than a fee tied to a single visit.

How is HCC risk adjustment coding different from standard diagnosis coding?

Standard diagnosis coding supports a single claim. HCC risk adjustment coding is population-level and cumulative — every qualifying diagnosis captured across the year feeds into a member's RAF score, and CMS's hierarchy rules mean only the most severe condition in a category typically counts.

Why do risk adjustment coding errors matter more in 2026 than in past years?

The combination of the CMS-HCC V28 model rollout and expanded RADV audit extrapolation means a documentation gap that once affected a handful of charts can now be projected across an entire health plan's Medicare Advantage membership, dramatically raising the financial stakes of routine risk adjustment coding mistakes.

Can AI fully automate risk adjustment coding?

Not entirely, and it shouldn't. AI is highly effective at surfacing missed conditions, flagging unsupported diagnoses, and keeping HCC crosswalks current, but CMS still expects human coder and compliance oversight on final code submission, particularly for high-RAF-impact charts.

What's the first step for a health plan trying to reduce risk adjustment coding errors?

Start with an internal audit measuring MEAT/TAMPER compliance and HCC mapping accuracy against the current CMS-HCC model version — this typically surfaces the same error patterns CMS auditors look for, before they show up in a formal RADV review.

Get the next analysis in your inbox

The Revenue Cycle Briefing — operator-grade research, every other Wednesday. 12,400+ subscribers.

About the author

Muyied Ulla Baig

Medical Coder · RapidClaims

Muyied Ulla Baig is a dedicated medical coder with 1 year of experience in E/M Outpatient, HCC, and Dental coding, supporting accurate risk adjustment and claims integrity through detailed and compliant coding processes at RapidClaims.