Ugwuowo U, Yamamoto Y, Arora T, et al. - Given that algorithms that predict high risk of acute kidney injury (AKI) are of great interest, but no investigations have included such an algorithm into the electronic health record (EHR) to help with clinical care, so researchers performed this prospective observational cohort analysis to report the experience of implementing such an algorithm. This study included 2,856 hospitalized adults with an algorithm-predicted risk of AKI in the next 24 hours exceeding 15%. AKI within 24 hours of pre-AKI alert (AKI₂₄) was the outcome. The development of AKI ₂₄ was reported in 18.9% of patients. This population had generally poor outcomes. In those who developed AKI ₂₄ and in those who did not, the inpatient mortality was identified to be 29% and 14%, respectively. Experts noted that those who encountered AKI ₂₄ more often had systolic BP < 100 mmHg, heart rate > 100 bpm and oxygen saturation <92%. In those who did vs did not develop AKI ₂₄, a difference was evident in only hyaline casts on urine microscopy as well as fractional excretion of urea nitrogen, of all biomarkers measured. Overall, successful integration of a real-time AKI risk model into the EHR was done.