The Centers for Medicare & Medicaid Services, or CMS, is the federal agency that runs Medicare and works with states to manage Medicaid. A key part of their job is to make sure payments to health plans are fair and accurate. To do this, they use a system called a risk adjustment model.

Think of a risk adjustment model as a way to predict a person's future healthcare costs. It looks at their age, gender, and health conditions to calculate a "risk score." A higher score means the person is expected to need more medical care. Health plans that cover sicker, higher-risk patients receive larger payments to help cover those costs. This prevents plans from only trying to enroll the healthiest people.

CMS V28 is the latest version of this model, specifically for Medicare Advantage plans. It's an update designed to make those predictions more accurate based on current medical knowledge and data.

CMS regularly updates its risk models to keep pace with changes in healthcare. The 2026 release of V28 introduces some significant changes aimed at improving how patient health is measured.

One of the biggest changes is a full transition to the ICD-10 coding system. For years, the models used a mix of older and newer diagnostic codes. V28 will rely exclusively on the more detailed ICD-10 system. This means the model can understand health conditions with greater specificity.

Additionally, the update recalibrates how different health conditions are weighted. Certain conditions may have a greater or lesser impact on a patient's risk score than before, reflecting new data on treatment costs. The model also introduces new Hierarchical Condition Categories (HCCs), which are used to group related diagnoses.

The architecture of the CMS V28 model can be understood as a clear, multi-step process. It's designed to take raw patient information and transform it into a single, meaningful risk score.

First, the system takes in basic patient data. This includes demographic information like age and sex, as well as all of the patient's diagnosed medical conditions, which are submitted as ICD-10 codes. Next, a clinical classification software maps these codes to the appropriate HCCs.

Once the patient's conditions are categorized, the model applies specific weights, or values, to each HCC and demographic factor. The sum of these values produces the final risk score, which is the system's primary output.

This overview provides a foundation for what CMS V28 is and what major changes it brings. The shift to a more modern, data-driven model aims to create a more equitable and accurate payment system for Medicare Advantage.