AIMI Research Meeting: Monitoring and Mitigating Bias in Suicide Risk Models - Duncan McElfresh, PhD

Abstract: This presentation will cover ongoing work with the VA Office of Mental Health and Suicide Prevention Program Evaluation Resource Center (PERC), toward monitoring and mitigating bias in predictive models for identifying patients at high risk for suicide.

Bio: Duncan is a Postdoctoral Fellow in Health Services Research and Development, with the VA Center for Innovation to Implementation (Ci2i) and the Stanford Department of Health Policy. His work focuses on practical approaches to algorithmic governance in healthcare, to ensure that deployed systems meet appropriate standards of quality and safety. Currently he is focusing on methods for monitoring deployed predictive models for suicide and overdose prevention. Duncan’s past projects have focused on blood donor recruitment (with Facebook), kidney exchange (with UNOS), and financial services (with FinRegLab). He holds a PhD in Applied Mathematics (AMSC) from the University of Maryland, College Park.