Using Artificial Intelligence to Identify COVID-19 Patients at High Risk of Clinical Deterioration

Peter Winkelstein, MD, MS, MBA, FAAP, Professor of Clinical Pediatrics, VP and CMIO, Kaleida Health, Executive Director, University at Buffalo Institute for Healthcare Informatics, and CMIO, UBMD; and Randall Wald, PhD, Senior Data Scientist, Cerner. The COVID-19 pandemic has challenged hospitals to make careful use of resources, especially for patient subpopulations which are clinically complex.

One such subpopulation is admitted COVID-19 patients who are more likely to deteriorate (either moving onto ventilation or dying) within the near future. Learn how Dr. Winkelstein utilized collecting patient deterioration data and built models from admitted COVID-19 patients to identify risk of ventilation and mortality, and how that information can be incorporated into care team workflows.