As Vice President for Evidence Generation and Translation, Dr. Margo Edmunds leads AcademyHealth's portfolios ... Read Bio
This introductory seminar provided an overview of a variety of predictive analytics methods for public health and health services and policy researchers to use in their own work.
Developed with graduate students and junior investigators in mind, the seminar explored a pragmatic approach to designing predictive models with a focus on decisions regarding predictor inclusion and modeling strategy, and meaningful measures to identify biases for health services research.
Dr. Ernest Moy of the Department of Veterans Affairs (VA) provided context for the webinar by introducing and describing the value of real-world predictive analytics. Dr. Jodie Trafton of the VA provided an overview of the development and deployment of the VA’s predictive model to improve opioid safety and prevent overdose and suicide, which featured working closely with clinicians who were familiar with treatment models. Dr. Suzanne Tamang of the VA explained a variety of analytic strategies to test these models for different biases, such as inherent assumptions reflecting racial and ethnic bias. Dr. Alyce Adams from Kaiser Permanente served as a discussant for the webinar, highlighting the value of collaborations between research and clinical teams to obtain the best results for patients.
This webinar is brought to you by AcademyHealth through the HSRProj program. HSRProj is a joint effort of AcademyHealth and the Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, funded by the National Library of Medicine, National Institutes of Health, U.S. Department of Health & Human Services.
- Define predictive modeling
- Identify practical considerations in the implementation of predictive modeling
- Describe ways big data can be applied to enhance public health and health services research
Watch Recorded Webinar
To access the webinar’s slide deck, please download the presentation »