Health Affairs March 13, 2025
As Medicare payment models evolve, reconciling differences in coding intensity between Medicare Advantage (MA) and traditional Medicare (TM) is a persistent challenge. High coding intensity in MA and undercoding in TM—often due to resource constraints and limited incentives—creates a risk gap that distorts payments and resource allocation and affects patient care.
This article contends that the ethical and transparent deployment of artificial intelligence (AI) can help close this risk gap by improving coding completeness and accuracy across MA and TM. When properly integrated, AI can improve coding integrity and accuracy by capturing patient health status and streamlining administrative tasks. This process can lead to more equitable reimbursement for MA and TM, ensuring better access to high-quality care.
The Risk Gap...







