Statistics Seminar - 11/9/21

Abstract: Are you curious how math and statistics can improve healthcare and help individuals live the healthiest lives possible?  Come learn about the Intermountain AI and Data Science team’s strategy and also about next year’s internship opportunities with our team.  We will discuss several projects that are typical projects our interns may be asked to do:

• Implementing a C.Diff. Risk Score for Pharmacist Intervention – C.Diff. infections in hospitals are life threatening to patients and because of this treatment can also be very expensive.  To keep our patients safe and to mitigate the risk of C.Diff. infections, pharmacists can recommend changes to antibiotic or probiotic prescriptions.  These strategies are best addressed with patients at greatest risk of infection.  This portion of our talk will introduce the internally developed risk model and our strategy in getting this critical information into the hands of caregivers that can have the greatest impact on our patient’s care.
• Diversity, Equity, and Inclusion: Addressing unintended bias in AI and machine learning algorithms – There are many forms of bias that can creep into AI including: historical bias, representation bias, learning bias, deployment bias, etc. (H. Suresh and J. Guttag https://arxiv.org/pdf/1901.10002.pdf). Join us in our discussion of how our team has approached creating tools and alerts that help automatically identify potential unintended bias.