Predictive Models, Bias, Health Equity, and COVID-19

Michael Pencina, PhD, Vice Dean for Data Science and IT & Professor of Biostatistics and Bioinformatics, Duke University School of Medicine. Modern health systems stand to derive substantial benefit from employing algorithm-based clinical decision supposed (ABCDS) tools. However, adoption has been slow and recent high profile failures have raised concerns about bias and inequity, lack of transparency and trust, as well as poor evaluation standards. Dr. Pencina discusses an objective framework for evaluation of ABCDS tools aimed at overcoming these barriers.

AI | predictive models | health equity | clinical decision support