Overcoming AI Implementation Challenges at the University of Michigan Health System

Jesse chats with Dr. Andrew Wong, Research Fellow at the National Clinician Scholars Program Institute at the University of Michigan Institute for Healthcare Policy and Innovation and Clinical Instructor in the Department of Internal Medicine at the University of Michigan, and Dr. Mike Burns, Associate Professor of Anesthesiology and Associate Chief Medical Information Officer for AI at the University of Michigan. Together, they discuss what led Michigan to report on the underperformance of Epic's sepsis prediction model, considerations for health systems as they look to set up a structure for AI governance, market diversity and competition in today's clinical AI sector, the University of Michigan Health System's AI workflow from model output to clinician decision support to measurable patient benefit, how model implementation can influence workflow, thresholds for healthcare AI models and acceptable trade-offs for clinicians, guardrails to prevent harm and what to monitor after go-live, and much more. Jesse also reflects on this key headline: OpenAI publishes wish list for healthcare AI. In our Resource Link segment, we list one valuable resource you might want to check out. To view this link, subscribe, or find out more information about our podcast, visit www.HealthcareAIPioneers.com. Want to be a sponsor, marketing partner, or guest, or provide feedback on the podcast? Email us at [email protected].  Music by Turning Pages from Pixabay.