The VBC Foundations series recently featured a blog unpacking how risk adjustment plays a key role in value-based care (VBC) contracts. VBC Foundations – Decoding Risk walks through how even small changes in a population’s Risk Adjustment Factor (RAF) can drive millions in losses or gains under programs like MSSP. If you have not yet read the post, I highly recommend it as it lays the foundation for understanding the math behind risk.

This blog will dive deeper into how risk adjustments are calculated. First, we must turn the page to the newest iteration of the Hierarchical Condition Categories (HCC) model, version 28. The latest update to CMS’s Risk Adjustment Model introduces some very impactful changes in its methodology and scoring. For those unfamiliar with HCC, think of this as system akin to a library that organizes patient conditions into various sections or categories, each with its own impact on healthcare planning and reimbursement. New volumes have been added, old ones retired, and some bookshelves restructured altogether. What used to count as a major work might now be a footnote and that changes everything about how care is scored and reimbursed.

Understanding HCCs

The HCC library classifies diseases and health conditions based on severity and associated treatment costs. CMS uses this system to determine Medicare reimbursements. Essentially, each condition a patient has is a book in the library, and the section it belongs to (the HCC) helps predict the cost of care, which in turn informs the reimbursement level. Each of these books has an associated score.  A patient’s total risk is essentially the total sum of the value related to their books. With the recent update, just as a library grows and evolves by adding new editions and removing outdated ones, so does the HCC model. The update expands our library significantly, introducing new volumes and editions of conditions, allowing for a more nuanced understanding of patient needs.

The update also retires certain volumes that are no longer reflective of the current healthcare landscape. With new editions added, the library's categorization becomes more precise, enabling healthcare providers to find the books that best describe their patient's health narrative, leading to a more accurate prediction of care costs.

To make this easier to grasp, we’ve built an interactive HCC Calculator that lets you see how these books add up in real time. Each diagnosis carries its own weight toward a patient’s total risk score. Add or remove conditions, and you’ll see how the overall Risk Adjustment Factor (RAF) shifts. It’s a hands-on way to see how coding accuracy directly impacts reimbursement and performance under value-based contracts.

Try it yourself! See how patient risk scores are calculated with our interactive HCC Calculator. Adjust conditions and demographics to see real-time changes in the RAF score.

What's Changed with v28?

With Version 28, CMS has reorganized the risk adjustment library to better encapsulate more recent trends in healthcare. One of the major changes involves updating the reference materials themselves. The benchmarking period used to estimate base reimbursement has changed from 2015 cost, to 2020. Since 2015 there has been a global pandemic which put significant strain on the healthcare system. The pandemic led to shifts in how people accessed healthcare services, particularly in 2020. There was a widespread reduction in non-urgent care, such as elective procedures and routine check-ups, as people delayed or avoided medical care due to fear of exposure or to reduce the burden on healthcare systems. This reduction in face-to-face encounters meant fewer diagnoses were being captured, potentially lowering patients' risk scores and impacting future payments to Medicare Advantage (MA) plans.

To account for this, CMS recognized the need to incorporate data from the pandemic years into its HCC model to reflect the actual utilization patterns and avoid inaccurate risk score calculations. It’s as if the library has pulled the dusty 2015 pricing guide off the shelf and replaced it with a 2020 edition. The new edition accounts for the pandemic’s disruption and new patterns in healthcare utilization. That updated guide recalibrates what each diagnosis costs, ensuring reimbursement reflects the world we live in now. The average cost for a patient with a risk score of 1.0 (this is the average patient cost) has increased 11.05%, up $1,035.10 from the previous version’s average. This new average is around $10k per capita.

Another key shift involves removing certain “books” from circulation. Diagnoses that were considered unreliable, poor predictors of cost, or easily gameable have been excluded or have seen decreases in reimbursement. This makes the library leaner and more accurate, focusing only on the most meaningful entries when estimating patient complexity.

The library itself has also expanded. The number of disease categories or HCCs has increased from 86 to 115, with new entries added and several categories subdivided. For instance, severe persistent asthma was not a diagnosis in the model’s previous version. It has since been added, acknowledging its distinct impact on patient care and cost. Cardiovascular diseases have been subdivided; congestive heart failure was once a single category but now is split into multiple categories to better capture the varying severity and associated cost of heart conditions.

To support these updates, the entire library has been reindexed. HCC categories have been renumbered, for example, diabetes has shifted from categories 17-19 to 36-38. Just like reassigning call numbers in a growing library, this restructuring improves navigation, reduces confusion, and prepares the system for future expansion.

With the library reorganized and more conditions now cataloged, CMS also applies a final system wide adjustment to maintain balance year-over-year, a process is called normalization. Since coding practices tend to become more thorough over time, risk scores can drift upward even if patients aren’t getting sicker. Normalization recalibrates the value of each diagnosis in the model to keep the system fair and consistent. In library terms, even if more books are being checked out now than before, CMS ensures that each one still counts for the right amount, keeping reimbursements aligned with actual risk rather than shits in documentation.

Important Implications

Our library has been reorganized and updated with new books, now what? This transition is more than just reshuffling diagnosis codes; it changes how conditions are classified, scored, and reimbursed. With the catalog expanded, sections rewritten, and outdated volumes removed, providers and provider groups need to reorient themselves to navigate this system effectively.

Some conditions that previously sat on separate shelves now share space. Under the previous version, diabetes with complications and peripheral vascular disease contributed independently to risk scores. In the current version, they are grouped under a common section which means they will not be counted twice. This restructuring is expected to reduce average risk scores by ~3.12%. Providers and provider groups accustomed to reimbursement patterns built on broad, overlapping codes may see payments decline. On the other hand, those serving patients with complex or newly recognized conditions may see increased reimbursements. The books are still there, but what counts, and how much it counts for, has changed.

As the healthcare industry adapts to CMS-HCC v28, tools that support accurate coding and clear performance tracking are essential. Azara has implemented a version of v28 in DRVS, allowing our customers to align their workflows, documentation, and analytics with the updated model. Our platform helps organizations navigate the transition confidently, monitor shifts in risk, and ensure they’re capturing the full picture of patient complexity. With Azara DRVS, you’re not just reacting to the change, you are ready for it.