Hong JC, Eclov NCW, Dalal NH, et al. - Given the frequent necessity for acute care (emergency department evaluation or hospitalization) among patients undergoing outpatient radiotherapy (RT) or chemoradiation (CRT), researchers here investigated if machine learning (ML) can identify high-risk patients and direct mandatory twice-weekly clinical evaluation to decrease acute care visits during treatment. They conducted a single-institution randomized quality improvement study evaluating 963 outpatient adult courses of RT and CRT initiated between January 7 and June 30, 2019, by an ML algorithm. Among these, ML identified 311 courses as high risk (> 10% risk of acute care during treatment); randomization of these was done to standard once-weekly clinical evaluation (n = 157) or mandatory twice-weekly evaluation (n = 154). Per outcomes, patients undergoing RT and CRT were accurately triaged using ML, hence, directing clinical management with decreased acute care rates vs standard of care. This prospective study thereby establishes the potential advantage of ML in healthcare and provides opportunities to improve care quality and decrease healthcare costs.