UCSF professor: AI can make value-based payment feasible but payment design matters
Dr. Robert Walker told PTAC that generative AI can read charts, synthesize literature, and support longitudinal measurement — potentially making better risk adjustment and outcomes measurement possible — but warned AI could raise costs under fee-for-service and urged pilots and safety safeguards.
The full story
Dr. Robert Walker (University of California, San Francisco) told PTAC that recent advances in generative AI create an opportunity to move long-delayed changes in payment and measurement. Demonstrating an AI clinical reasoning example, Walker argued that AI can quickly summarize long medical records and improve real-time decision making and population management. "AI now can do this review 600page chart in a few seconds," he said, noting that LLMs can also improve risk adjustment and quality measurement by reading clinical notes rather than relying solely on billing data.
Sign up to keep reading.
Your free newsletter brings you what your city council, county, and school board discussed and decided.