The team employed machine learning to develop two computerized models that can “fine tune” the screening of individual pediatric patients who using other screening methods had already been classified as being at low risk for adverse effects from anesthesia by another scoring system.