An AI risk score for failing kidneys in hospital changed nothing
A randomized clinical trial tested a machine learning risk score meant to trigger early nephrology consults for hospitalized patients at high risk of acute kidney injury. The result was clear: no meaningful improvement in clinical outcomes. A useful reminder that spotting risk earlier does not help if the action that follows fails to change the patient's actual course of care.
practitioners › Randomized trial published July 10, 2026 in JAMA Network Open, on an AI decision support score for acute kidney injury (AKI).
JAMA Network Open →