HCC Coding and Risk Adjustment-How AI Helps Capture More Revenue

HCC Coding and Risk Adjustment: How RapidClaims AI Helps Capture More Revenue

HCC coding plays an important role in Medicare Advantage risk adjustment, and 2026 marks a significant milestone in the transition to the updated CMS-HCC model. As CMS-HCC Model Version 28 has come into full force for the payment year 2026, the conversion of diagnosis codes into Risk Adjustment Factor (RAF) scores will no longer be done in the same old way. Proper HCC coding is now not an issue of complying with regulations, but a matter of getting paid the money the organization deserves for the patient population it serves.

This is exactly where AI-driven HCC coding software is reshaping the workflow, offering new possibilities to healthcare organizations in terms of streamlining their operations.

What Is HCC Coding, and Why Does It Matter More in 2026?

The HCC coding process involves translating the diagnoses of a patient recorded using the ICD-10-CM code to Hierarchical Condition Categories, which take into account the severity of the patient's illness and the projected cost of care. The HCC has risk scores, and the total of the risk scores of all relevant HCCs, along with the patient's demographics, gives their risk score, which is one component of the methodology CMS uses to determine Medicare Advantage payments.

Proper HCC coding is important in making sure that complicated patients with serious illnesses get adequate funding. It is easy to see why an individual with uncomplicated diabetes and another patient with diabetes and comorbid conditions like kidney disease and cardiovascular condition would have different health care needs, and HCC coding is used in capturing the differences in the system. Incomplete, outdated and poor HCC coding can lead to a situation where the health plan has an underfunded profile of patients despite the true clinical complexity of the patient.

CMS-HCC Model V28: What Changed for HCC Coding This Year

For CY 2026, the three-year phase-in of the 2024 CMS-HCC model is complete, meaning the updated model is used for 100% of applicable Medicare Advantage risk scores. The updated model was developed using ICD-10-CM diagnosis data and substantially revised the diagnosis-to-HCC mapping structure used under the earlier model.

This becomes a major change. The number of ICD-10-CM diagnosis codes mapped to payment HCCs decreased from 9,797 under V24 to 7,770 under V28, while the number of payment HCCs increased from 86 to 115. These changes can affect risk scores even when the underlying patient population has not changed.

Several structural shifts define what HCC coding teams need to understand under V28:

Greater clinical specificity is now required. Vague or unspecified diagnosis codes are far more likely to fail to map to any HCC at all under V28, which means documentation that used to generate risk adjustment value under V24 may generate nothing today.
Constrained hierarchies reduce duplicate credit. V28 introduced revised hierarchies and coefficient constraints that affect how related HCCs contribute to a beneficiary's risk score.
Condition weights have been recalibrated. Some chronic conditions, including certain diabetes complication tiers, now carry different - often lower - relative weights than they did under the previous model, which changes how individual diagnoses contribute to a patient's overall RAF score.

More importantly, this is not a one-off transition period. CMS periodically updates its risk adjustment models, mappings, and coefficients, so organizations should monitor CMS guidance and model releases for future payment years. For CY 2027, CMS has finalized continued use of the 2024 CMS-HCC model, while organizations should continue monitoring CMS guidance for future model and methodology updates.

CMS-HCC V24 vs. V28: A Side-by-Side Comparison

Feature	V24 (Previous Model)	V28 Classification / 2024 CMS-HCC Model (CY2026)
Diagnosis mapping	ICD-10 → ICD-9 crosswalk → HCC	Direct ICD-10-to-HCC mapping
Valid diagnosis codes mapped to an HCC	9,797	7,770
Number of hierarchical condition categories	~86	~115
Condition interaction	HCC hierarchies and constraints differed from the updated model	Related HCCs constrained to share a coefficient
Documentation specificity	Unspecified codes often still mapped	Certain codes no longer map to payment HCCs
Status in payment year 2026	No longer used as the sole model for CY2026	Governs 100% of MA risk scores

This shift explains why so many organizations are seeing shifts in their population's aggregate RAF scores even when the underlying patient population and clinical documentation habits haven't changed - the model itself is scoring the same clinical picture differently.

Understanding RAF Scores Under the New Model

While RAF scores continue to serve as the primary measure derived through risk adjustment coding, their underlying components have shifted dramatically in V28. A risk score is calculated using applicable HCCs, demographic factors, and other model variables. The resulting score is incorporated into the Medicare Advantage risk adjustment payment methodology. A higher risk score generally indicates higher expected healthcare costs relative to the applicable model population and is incorporated into the Medicare Advantage payment methodology.

The RAF scores under V28 become much more dependent upon the specificity of the documentation as compared to the V24 system. An issue which would have been considered in general terms may fail to contribute towards the RAF score of a particular patient regardless of the clinical condition itself being precisely the type of complexity that the risk adjustment models are meant to address.

This serves to underscore the importance of proper documentation and accurate coding for a particular plan's bottom line. Risk adjustment generally relies on diagnoses supported by current-year documentation, making annual review and appropriate recapture workflows important.

Why Manual HCC Coding Struggles to Keep Up

Even prior to the advent of V28, HCC coding was an elaborate and meticulous process. Manual chart review, the most commonly used method, becomes increasingly inefficient in the face of such complexities. Coders reviewing patients' records manually have fewer and fewer options available – in terms of proper codes, updated condition weights, and required specificity. In addition, they have to work with the same volume of patients and do all of that under the same deadlines. As a result, many of them miss HCCs that could have been detected – or use coding strategies that were popular during V24 and do not provide any value at all for the purposes of risk adjustment anymore.

The increased regulatory scrutiny also creates a new challenge for coders.

How AI-Powered HCC Coding Software Helps Close the Gap

These are the conditions under which HCC coding software becomes not merely convenient but crucial. AI-powered HCC coding software like RapiClaims, can use natural language processing and machine learning to analyze clinical documentation, identify potentially relevant diagnoses, and help map supported diagnoses to applicable HCCs under the current model.

Below is a list of features RapidClaims’ HCC coding software delivers in 2026:

Automated chart review at scale. AI models can scan clinical notes across an entire patient population far faster than manual review, surfacing likely HCC-relevant diagnoses that might otherwise be missed in unstructured documentation.
V28-aware mapping logic. Rather than relying on outdated crosswalks, modern risk adjustment coding tools are built around the current model's direct ICD-10-to-HCC structure, helping coding teams apply current diagnosis-to-HCC mappings and applicable model logic.
Specificity prompts for clinicians. Under V28, certain diagnosis codes that previously contributed to risk adjustment under earlier models may no longer map to payment HCCs, making accurate and sufficiently specific documentation important.
Recapture tracking. AI-powered systems can flag chronic conditions that were coded in a prior year but haven't yet been re-documented in the current year, helping close the recapture gap before year-end.
Supporting traceability and documentation review. Given the current RADV and OIG scrutiny on risk adjustment coding, AI platforms that log the clinical rationale behind each suggested HCC give coding teams a defensible audit trail rather than a black-box output.

Crucially, this technology isn't meant to replace certified coders - it's meant to give them a faster, more accurate starting point. Complex or ambiguous charts still require human clinical judgment, but AI-assisted HCC coding software removes the volume burden of routine chart review so coding staff can focus their expertise where it matters most.

The Direct Revenue Impact of Getting HCC Coding Right

The relationship between HCC accuracy and revenue generation is immediate under the risk-adjustment payment structure. When a supported diagnosis is not appropriately captured, the resulting risk score may not fully reflect the beneficiary's documented health status, which can affect risk-adjusted payment. Add that up over the whole patient base, which can be in the tens of thousands, and even slight discrepancies in the RAF scores add up to real money lost on the plan year.

AI-based HCC coding software solves this problem through two mechanisms at once. Firstly, it helps with more complete coding, identifying legitimate HCCs that were missed during the manual process, especially in patient populations with a large amount of documentation or cases with multiple HCCs. Secondly, it mitigates the risk of compliance problems by ensuring that coding adheres to the proper V28 logic and hierarchy rules, rather than the old-fashioned V24 coding methods that no longer hold any merit.

For organizations still calibrating their coding workflows to V28, this combination - completeness plus compliance - is what separates AI-assisted risk adjustment coding from simple automation of the old process. Automating an outdated workflow can reproduce outdated coding assumptions at scale. The value of modern HCC coding software lies in building V28's rules into the foundation of the workflow itself.

What to Look for in HCC Coding Software in 2026

For organizations evaluating a risk adjustment coding partner or platform this year, a few questions matter more than general AI marketing claims:

Is the mapping logic built natively around V28, not adapted from a V24 foundation? Given how structurally different the two models are, this distinction significantly affects accuracy.
Does it support both prospective and retrospective coding workflows? Capturing HCCs at the point of care and correcting gaps through retrospective chart review both can play a role in a comprehensive risk adjustment workflow.
Does it flag documentation specificity issues, not just code suggestions? Given V28's stricter mapping rules, prompting clinicians toward more specific documentation is often more valuable than the code suggestion itself.
Does it produce an audit-ready rationale for each HCC? In an environment with active RADV audits and high OIG error findings, defensible documentation trails are not optional.
Does it keep coders in the loop for complex cases? The goal is augmenting coder judgment on ambiguous charts, not replacing it entirely.

The Bottom Line

HCC coding in 2026 operates under a fundamentally different model than it did just two years ago. With V28 now fully governing Medicare Advantage risk adjustment, the organizations that succeed will be the ones that treat accurate, specific, and compliant HCC coding as a continuous operational discipline rather than a once-a-year documentation push. AI-powered HCC coding software, applied correctly, gives coding teams the scale and precision to keep up with V28's tighter mapping rules,accurately represent the supported health conditions of a patient population within the applicable risk adjustment model. For plans and provider groups navigating this shift, that combination of speed, accuracy, and defensibility is exactly what turns risk adjustment coding from a compliance obligation into a real driver of appropriately captured revenue.

FAQs

1. What is risk adjustment coding in healthcare?

Risk adjustment coding is the process of accurately documenting and coding a patient’s relevant health conditions so they can be appropriately represented within an applicable risk adjustment model. ICD-10-CM diagnosis codes may map to Hierarchical Condition Categories (HCCs), which are then used by the model to calculate risk scores.

2. How does HCC coding software support risk adjustment?

HCC coding software can help identify potentially relevant diagnoses, map ICD-10-CM codes to applicable HCC categories, and support consistent review of clinical documentation. Automated tools can also help coding teams work with current model mappings and rules as CMS updates its risk adjustment models and software.

3. What are RAF scores, and why are they important?

RAF scores are calculated using applicable risk adjustment model variables, including HCCs and demographic factors, and are used within the Medicare Advantage risk adjustment payment methodology.

4. How does accurate risk adjustment coding affect RAF scores?

Accurate, supported diagnosis coding helps ensure that relevant conditions are appropriately represented within the applicable risk adjustment model. Because mappings and coefficients vary by model, incomplete or inaccurate coding can affect the resulting risk score.

5. What should organizations look for in HCC coding software?

A strong HCC coding software solution should support current CMS model mappings, identify relevant diagnosis-to-HCC relationships, streamline chart review, and help coding teams maintain documentation and coding accuracy. It should also accommodate model updates, since CMS periodically releases updated risk adjustment software and ICD-10 mappings.