KFF's Business of Health with Chip Kahn

Caroline Pearson, executive director of the Peterson Health Technology Institute (PHTI) and the Peterson Center on Healthcare, joins Chip to discuss who really benefits as AI moves into health care's back office. A quiet arms race is underway — providers deploying AI to code, bill, and capture revenue and insurers deploying it right back to review, deny, and hold the line. The open question is whether any of this lowers what the country actually spends or just speeds up the fight over the dollar. Pearson brings a rare vantage: PHTI runs rigorous, independent evaluations of health technology — from a now-famous assessment that found digital diabetes tools didn't lower the total cost of care, to its own examination of the administrative AI arms race in prior authorization and medical billing. She and Chip dig into whether AI can deliver better care and real savings, or simply more activity — and how incentives and policy decide which way it breaks.

In a conversation focused on the technology underlying AI in health care, Chip is joined by Seema Verma, former administrator of the Centers for Medicare & Medicaid Services (CMS), and now Executive Vice President and General Manager at Oracle Health and Life Sciences, the second largest electronic health records (EHR) platform in the U.S. Seema shares her insights on the evolution of EHRs and why it’s necessary to redesign these systems to better integrate AI capabilities and improve the quality of care.

Chip talks with Dr. Toyin Ajayi, co-founder and CEO of Cityblock Health, which delivers value-based care to more than 100,000 Medicaid and dual-eligible members across ten states, many of them people of color managing chronic conditions. Ajayi makes a pointed case: Roughly 60 percent of health care AI investment goes to billing, coding, and risk adjustment — making sure someone gets paid — while only a fraction goes to delivering care. If we continue to concentrate AI there, she warns, it will drive up cost without improving outcomes. She says there is a better way — AI built for care can lower costs by improving care for those hardest to reach. She and Chip discuss what that looks like and how Cityblock is using AI now to improve care and the patient experience for its members.

In this episode, Dr. Ziad Obermeyer joins Chip to talk about AI bias in patient management, including how far the health care industry has come since his groundbreaking research that revealed alarming biases in a widely used algorithm that underestimated the health needs of Black patients. The discussion emphasizes that while AI can analyze data with remarkable precision, it is the context surrounding that data that ultimately determines its efficacy, usefulness, and fairness. Dr. Obermeyer is an emergency department physician, researcher, co-founder of Nightingale Open Science and Dandelion, and associate professor at the UC Berkeley School of Public Health and the College of Computing, Data Science, and Society.

How can AI determine who gets matched to new therapies, who is identified for clinical trials, and how patient tracking is scaled across large populations? Chip is joined by Dr. A.J. Blood, a practicing cardiologist at Brigham and Women's Hospital and the co-founder and Chief Executive Officer of AIwithCare, a startup company that delivers AI-enabled solutions for research, clinical operations, and patient care. They discuss the role of AI in identifying patients for clinical trials and new therapies—which is typically a critical bottleneck in drug development—as well as how to ensure clinical trials are representative. Also, Dr. Blood shares insights from his extensive research background and the tool, RECTIFIER (RAG-Enabled Clinical Trial Infrastructure for Inclusion Exclusion Review), designed to enhance patient recruitment for clinical trials by efficiently sifting through complex medical data.

How do we know if AI use in health care actually makes patients better? Chip talks with Dr. David Bates — a veteran physician leader at Mass General Brigham and the Brigham and Women's Hospital and Co-director of the Center for Artificial Intelligence and Bioinformatics Learning Systems — about measuring patient outcomes reliably and in real time to create a strong foundation for everything else in health care administration — clinical deployment, payment reform, consumer transparency, and accountability. David has spent his career at the center of this challenge and his research helped inform To Err Is Human: Building a Safer Health System, the landmark 1999 report by the U.S. Institute of Medicine that revealed the high frequency of preventable medical errors in health care.