Great discussion of the JAMA Network Open manuscript demonstrating the superior performance of eCART to other early warning scores, featuring quotes from our co-founder and lead author of the study, Dana P. Edelson, MD, MS.
It's wicked hard to know how an AI model will perform in a health system. Even harder to test multiple models head-to-head. But that's just what Yale got to do in a rare benchmark analysis of 6 early warning scores, 3 of them proprietary. Epic's deterioration index was one of the worst-performing of the six, and the top-performer became the first AI tool that Yale has rolled out systemwide. Read more in STAT: