Recorded 14 July 2022. Noa Dagan of Harvard Medical School presents "Can we improve fairness for subpopulations by utilizing medical data?" at IPAM's Graduate Summer School on Algorithmic Fairness. Abstract: Medical data can be utilized to promote proactive, predictive, and personalized care. We will discuss fairness considerations when using medical data for these purposes as part of a healthcare organization’s responsibility to continuously improve patient care. Our case study will be focused on the development and application of medical prediction models in a large Israeli healthcare organization (Clalit Health Services). We will also discuss how causality analysis can be used to increase fairness by empowering individuals to make informed decisions regarding medical interventions. We will explore the challenges of deriving causal estimates from observational data, focusing on a case study of Covid-19 vaccine effectiveness. This talk is based on several joint works with Noam Barda, Guy Rothblum, Gal Yona, Uri Shalit, Eitan Bachmat, Marc Lipsitch, Miguel Hernan, Ben Reis, and Ran Balicer among others. Learn more online at: http://www.ipam.ucla.edu/programs/summer-schools/graduate-summer-school-on-algorithmic-fairness/