Comments on the 2026 Medicare Part C and D Advance Notice

Matthew Fiedler commented on the proposed 2026 Part C and D Advance Notice published by the Centers for Medicare and Medicaid Services (CMS). In the Advance Notice, CMS states that it may begin calibrating its risk adjustment models using Medicare Advantage (MA) encounter data rather than Traditional Medicare (TM) claims data. Fiedler’s letter argues that this change would not achieve what CMS seems to hope it would—and could have the opposite of the desired effect. He makes four main points: 

• Contrary to what CMS suggests in the Advance Notice, a coding pattern adjustment would remain necessary if the risk adjustment model were calibrated using MA encounter data
• Calibrating the model using MA encounter data could either increase or decrease the appropriate coding pattern adjustment, depending on how the spending of “upcoded” beneficiaries compares to their peers
• CMS would bear a heavier evidentiary burden if it wished to apply a coding pattern adjustment after calibrating the model using MA encounter data, and it could face challenges to its authority to apply any such adjustment
• Calibrating the model using MA encounter data would not necessarily produce better measures of relative risk

Fiedler notes these points do not necessarily imply that CMS should avoid using the encounter data for calibration. However, he argues that they do imply that CMS should proceed cautiously, at least if its goals are to mitigate the payment consequences of plans’ coding efforts or improve its measures of relative risk. Namely, CMS should carefully assess whether calibrating the model using encounter data would actually advance these goals, not simply assume it would do so, and it should also develop a viable plan for applying a coding pattern adjustment in the changed environment.

Read the full comment letter here