The use of risk prediction models to guide clinical decision making is growing due, to the incorporation of risk into clinical guidelines and the rising availability of proprietary models. In this talk, Singh will discuss how risk prediction models are evaluated, the importance of model transparency in the evaluation process, and approaches to studying their reproducibility and generalizability in real-world settings. Singh will draw on work on sepsis and acute kidney injury to highlight challenges in reproducibility and generalizability and will touch on the gradually changing regulatory landscape around the use of prediction models in clinical care.
Speaker: Karandeep Singh, MD, MMSc, Assistant Professor of Learning Health Sciences, Internal Medicine, Urology and Information
University of Michigan
To obtain Zoom link please email: epibiostat@ucsf.edu