Written by: Dr. Auren Weinberg, Chief Medical Officer, Veradigm and Aaron Winkel, Director, Veradigm Risk Analytics
In March 2023, the Centers for Medicare & Medicaid Services (CMS) issued the “Final Notice,” in which they finalized the 2024 risk adjustment Hierarchical Condition Category (HCC) model known as V28. This model is expected to significantly change Medicare Advantage (MA) capitation rates.
Currently, Medicare determines payment amounts for MA plans using risk scores calculated based on patients’ disease burdens together with demographics and geographic location data. Risk scores enable CMS to pay MA Organizations higher rates for patients diagnosed with multiple conditions or more severe conditions, as these patients have a higher anticipated cost of care.
The V28 model will significantly change existing methodologies for calculating risk adjustment scores, including changes to HCC codes, disease mappings, and disease coefficient values. These changes will impact risk adjustment factor (RAF) scores for MA Organizations, which will, in turn, change the amounts CMS pays per beneficiary.
The new HCC model V28 will be phased in over 3 years:
In this article, we consider what these changes mean for medical practices—and how, as a busy practice, you can best prepare.
One reason behind the recent changes was the discovery that some MA plans have engaged in upcoding—changing a patient’s designated HCC code to increase CMS’s calculated payment amount for that patient—resulting in overpayments. In the past few years, multiple Office of the Inspector General (OIG) audits of MA plans have revealed cases where the diagnoses were not supported by the clinical record.
Another reason for the change: The current risk adjustment model was based on the use of ICD-9-CM coded data—but the industry transitioned to using ICD-10-CM codes in 2015. The V28 model is designed to incorporate that change, reflecting the ICD-10-CM’s specific, detailed diagnoses.
CMS calibrated the V28 model using years of updated fee-for-service data (2018 diagnoses and 2019 expenditures versus the current model, which is based on 2014 diagnoses and 2015 expenditures to more accurately reflect the costs and coding patterns seen in current fee-for-service Medicare.
Ultimately, the goal is for the updated HCCs to capture more complete and accurate data about the health statuses of patients with chronic conditions so that healthcare practices will better understand patients’ health needs and be better able to help their patients effectively manage their conditions.
The V28 model renumbers and changes numerous HCC categories, including eliminating HCCs in the V28 model for conditions that failed to accurately predict costs, conditions with small coefficients, uncommon conditions, and conditions without clear diagnostic coding criteria.
Some of the significant diagnoses removed from V28 include:
CMS also mapped 268 new ICD-10 diagnosis codes to HCCs. Significant added diagnoses include:
Overall, the number of HCCs increases from 86 in V24 to 115 in V28; but V28 removes 2,294 ICD-10-CM diagnosis codes that previously mapped to a payment HCC in V24, so they no longer map to a payment HCC in V28.
CMS predicts that the proposed changes will increase risk-adjustment payments to MA plans by an average of 3.56% in 2024, or an estimated $11.4 billion. However, others predict the V28 changes are most likely to decrease overall risk scores.
For instance, diabetes is a commonly reported condition among MA beneficiaries, and its risk adjustment values are decreased in V28.
Diabetes is a revealing example of a strategy called “constraining”, which CMS uses to determine HCC risk adjustment coefficients in V28. Constraining means that related HCCs are given the same coefficients. For diabetes, this means that uncomplicated diabetes, diabetes with acute complication, and diabetes with chronic complications all have the same coefficient, which is higher than the current coefficient for uncomplicated diabetes but lower than those for diabetes with complications. Overall, this results in a decreased RAF for patients with acute or chronic diabetes complications—which will have a significant financial impact on payments for these patients.
V28 is also expected to have a significant financial impact because it requires medical practices to update their coding practices and risk adjustment models. Making these changes will require a major investment of time and resources from practices. Although the 3-year transition period to full use of V28 is intended to make the change proceed more smoothly, providers will likely find the transition challenging because it will require practices to manage both the V24 and V28 systems simultaneously. Conditions that map to an HCC category in one version may not have an associated HCC in the other, and even for diagnoses that map to HCCs in both versions, the specific HCC and the associated RAF may differ.
The critical point for providers to remember as they transition to use of the V28 model is that the basics have not changed: They need to continue to follow ICD-10-CM conventions and guidelines, taking care to assign codes with as much accuracy and specificity as possible, and ensuring diagnoses are supported with sufficient documentation. Accuracy and specificity in diagnostic coding and documentation have always been the basis for accurate risk adjustment. V28 will require even greater specificity to capture the full severity of each MA patient’s condition.
It’s important to note that ICD-10 guidelines require providers to code not only for the condition behind a patient’s visit but for all other documented conditions that coexist at the time of the visit and affect the patient’s care or treatment in any way. At the same time, though, providers must also take care not to code for conditions that no longer exist.
A recent risk adjustment case study published by the American Academy of Professional Coders (AAPC) reveals that providers and coders are not well educated in ICD-10-CM coding guidelines in the areas of combination coding or selecting unspecified versus higher-specified codes.
In this study, AAPC Services performed risk adjustment audits to review providers’ and coders’ use of ICD-10-CM diagnosis codes, checking for accuracy and supporting documentation. They found that 14% of HCC codes, or approximately 211 of 1500 individual HCC conditions, were assigned without adequate supporting documentation, so the diagnosis code was deleted. Considering that every HCC code has an estimated value of $2,500, these 211 deleted HCC codes reflect a loss of about $527,500. This type of error also poses the risk of over-reporting patient diagnoses, potentially making organizations targets for government and payer audits.
As the CMS-HCC risk adjustment model grows and changes, it will be critical to stay current on the latest requirements. Consider having a provider in-service training on HCC coding and the use of codes with changed or deleted risk scores. By proactively adopting these changes, your practice will be positioned to receive the financial support needed to continue providing quality patient care. In light of the increased legal action from CMS, and promises of more aggressive audits of MA plans in the future, it’s become essential for provider organizations to ensure that their ICD-10-CM coding and documentation will stand up to review.
Investing in technologies that enable accurate and efficient coding for large volumes of clinical documents will be critical for MA Organizations and other stakeholders seeking to effectively manage risk adjustment in the changing environment. Is your organization prepared for the changes? Contact us today to talk about ways to improve the accuracy and efficiency of your risk adjustment coding.