TY - JOUR T1 - Designing COVID-19 mortality predictions to advance clinical outcomes: Evidence from the Department of Veterans Affairs JF - BMJ Health & Care Informatics JO - BMJ Health Care Inform DO - 10.1136/bmjhci-2020-100312 VL - 28 IS - 1 SP - e100312 AU - Christos A Makridis AU - Tim Strebel AU - Vincent Marconi AU - Gil Alterovitz Y1 - 2021/06/01 UR - http://informatics.bmj.com/content/28/1/e100312.abstract N2 - Using administrative data on all Veterans who enter Department of Veterans Affairs (VA) medical centres throughout the USA, this paper uses artificial intelligence (AI) to predict mortality rates for patients with COVID-19 between March and August 2020. First, using comprehensive data on over 10 000 Veterans’ medical history, demographics and lab results, we estimate five AI models. Our XGBoost model performs the best, producing an area under the receive operator characteristics curve (AUROC) and area under the precision-recall curve of 0.87 and 0.41, respectively. We show how focusing on the performance of the AUROC alone can lead to unreliable models. Second, through a unique collaboration with the Washington D.C. VA medical centre, we develop a dashboard that incorporates these risk factors and the contributing sources of risk, which we deploy across local VA medical centres throughout the country. Our results provide a concrete example of how AI recommendations can be made explainable and practical for clinicians and their interactions with patients.Data comes almost exclusively from the internal VA Corporate Data Warehouse (CDW) database and none of the information leaves the VA premises. This project has been completed in partnership between the National Artificial Intelligence Institute at the Department of Veterans Affairs and the local Washington DC VA medical center. ER -