medical team hear presentations from business
" data-large-file="https://i0.wp.com/keebler.health/wp-content/uploads/2025/10/iStock-525755929.jpg?fit=1024%2C683&ssl=1" class="aligncenter wp-image-1528 size-large" src="https://i0.wp.com/keebler.health/wp-content/uploads/2025/10/iStock-525755929.jpg?resize=1024%2C683&ssl=1" alt="" width="1024" height="683">
The Centers for Medicare & Medicaid Services (CMS) introduced the CMS-HCC Version 28 (v28) risk adjustment model in 2024, replacing Version 24, which had been in place for nearly a decade.
This update marks one of the most significant shifts in the history of risk adjustment. By modernizing the model to align with ICD-10 coding and refining how chronic conditions are categorized and weighted, CMS aims to improve accuracy, reduce overcoding, and ensure reimbursement better reflects patient complexity.
However, these updates have created substantial challenges for payers, providers, and risk adjustment teams who must now navigate new hierarchies, fewer qualifying codes, and altered weighting logic. Understanding what changed—and how to prepare—is critical for maintaining compliance and financial stability.
What’s New in CMS-HCC Version 28
The transition from Version 24 to Version 28 represents more than a simple update—it’s a complete redesign of how conditions map to risk. Below is a summary of the key changes and their implications.
Alignment with ICD-10
Version 28 moves away from the ICD-9 foundation that underpinned prior models and aligns fully with ICD-10. This change increases clinical precision and consistency with modern coding standards, but it also alters how conditions are grouped and scored.
For organizations accustomed to ICD-9-based mappings, this alignment introduces new documentation and coding requirements that demand system updates and training.
Reclassification and Reduction of Codes
Under v28, CMS reorganized and consolidated diagnostic categories, removing or merging approximately 2,000 ICD-10 codes. Chronic conditions such as diabetes, vascular disease, and mental health disorders have been restructured, changing how severity and comorbidities influence the risk adjustment factor (RAF).
These reclassifications mean some diagnoses that previously contributed to a patient’s RAF score under v24 no longer do under v28.
Revised Weighting Logic
The model’s weighting logic has been recalibrated to prevent double-counting of related conditions and to reflect more accurate clinical relationships.
The outcome is lower average risk scores across many patient populations—even when coding and documentation are accurate. This change particularly affects members with multiple chronic conditions that previously generated overlapping HCCs.
Updated Demographic Factors
CMS also revised demographic coefficients, including age, disability, and dual eligibility status. These updates adjust how non-clinical factors influence RAF scoring, resulting in more modest weighting for some historically higher-risk populations.
The overall goal is to make the model more representative of true clinical risk, but the operational impact is significant.
What These Changes Mean for Healthcare Organizations
CMS-HCC v28 fundamentally changes how risk is captured and scored. Organizations must adjust both their operational workflows and analytical processes to maintain accuracy and compliance.
Lower Risk Scores and Revenue Impact
The shift to ICD-10 mapping and the reduction in qualifying codes mean that many organizations will experience average RAF score decreases of 5–8%. CMS itself projected a –3.12% decline in average Medicare Advantage risk scores during the initial rollout of the v28 model in 2024, reflecting the model’s stricter weighting and code consolidation.
This can directly affect revenue for Medicare Advantage plans, ACOs, and provider groups with risk-based contracts.
Heightened Documentation Standards
Because many codes have been removed or merged, providers need to document more detailed and clinically specific information to ensure all active conditions are captured. Ambiguous or incomplete notes that previously triggered valid HCCs may no longer qualify.
Increased Dependence on Integrated Data
Fragmented systems make it difficult to identify coding gaps or monitor dropped codes under v28. Accurate performance now depends on interoperable data infrastructure and the ability to analyze both structured and unstructured clinical information in a unified environment.
How AI Helps Organizations Adapt to HCC v28
As the complexity of risk adjustment increases, artificial intelligence (AI) has become key for maintaining accuracy, efficiency, and compliance under CMS-HCC v28.
AI-driven platforms like Keebler Health are designed to manage these changes at scale—helping organizations interpret new hierarchies, find missed opportunities, and stay audit-ready.
1. Identify Code Changes Automatically
AI-based mapping tools can instantly crosswalk v24 to v28, flagging diagnoses that have been dropped, reclassified, or merged. This allows teams to forecast financial impact and prioritize retraining and process updates.
2. Detect Missed or Hidden Conditions
AI can analyze both structured data (claims, encounters) and unstructured data (clinical notes, labs, imaging reports) to identify relevant chronic conditions that still qualify under v28. This ensures comprehensive and compliant risk capture.
3. Improve Transparency and Audit Readiness
Explainable AI models, like Keebler Health, show how each condition is identified and mapped, creating a clear audit trail and building confidence in compliance processes.
4. Deliver Real-Time Provider Feedback
Integrated AI systems can provide immediate documentation guidance during patient encounters—prompting providers to include missing specificity or clarify ambiguous language before submission.
Preparing for the Future of Risk Adjustment
The adoption of HCC v28 is just the beginning. CMS has signaled its intent to continue refining risk models to better reflect clinical realities and discourage coding inflation. In fact, the agency’s 2025 Medicare Advantage Rate Announcement forecasts a 3.86% average risk score trend—a blend between older and new model methodologies—as the v28 model phases nationally.
Organizations that invest now in AI-enabled, FHIR-compliant data systems will be better positioned to adapt to future updates, minimize financial risk, and ensure sustained accuracy in coding and documentation.
Learn how Keebler Health can help you adapt to CMS-HCC v28.