Artificial intelligence may be able to use EHR data to predict the prevalence of postoperative complications for patients to the same degree as physicians, according to a May 16 study published in JAMA Network Open.
The research team from the University of Florida Gainesville created a platform that used data from EHRs and fed it into the artificial intelligence platform as well as taking interactions with medical professionals into account. The total sample consisted of almost 60,000 patients who were admitted to the University of Florida Health Gainesville for 74,417 operations. The algorithm would display its predictions on a mobile app especially built for physicians.
The results showed that the artificial intelligence system could accurately predict postoperative complications using automated data collection from EHRs. The system had an accuracy rate that matched that of physicians. The researchers believe that the AI model created is the first to accurately predict postoperative complications using automated data acquisition as well as putting the results onto a mobile app that could easily be used in hospital settings.