Dose-response association between intraoperative norepinephrine and postoperative acute kidney injury: a retrospective cohort study

Background Intraoperative norepinephrine is associated with postoperative acute kidney injury (AKI), but bedside-readable dose anchors remain undefined. We investigated the dose-response relationship between intraoperative norepinephrine and AKI, and its interaction with intraoperative hypotension. Methods We analysed 27 874 surgical cases from the Medical Informatics Operating Room Vitals and Events Repository. The primary exposure was intraoperative norepinephrine dose, expressed as norepinephrine base (μg kg −1 min −1 ). The primary outcome was AKI within 7 postoperative days. The maximum dose, time-weighted intensity, duration, and area under the curve (AUC) were evaluated using restricted cubic splines and tensor-product generalised additive models, with multivariable mixed-effects logistic regression with a directed acyclic graph–derived adjustment set. Results Norepinephrine was administered in 1232 patients (4.4%), and AKI occurred in 4448 (16.0%). Restricted cubic spline analyses showed significant nonlinearity for unit-rate metrics (maximum dose P =0.008; time-weighted intensity P =0.022), whereas time-integrated metrics were linear (AUC P =0.10; duration P =0.50). Two bedside-readable anchors for maximum norepinephrine dose were identified at 0.029 and 0.064 μg kg −1 min −1 . Dose-duration ( P =0.286) and dose-hypotension interactions ( P =0.298) were non-significant, supporting additive effects. A maximum norepinephrine dose of >0.029 μg kg −1 min −1 remained independently associated with AKI (odds ratio [OR] 1.76; 95% confidence interval [CI] 1.41–2.20; P <0.001); adjustment for intraoperative hypotension minimally attenuated this association (5.1%; adjusted OR 1.71; 95% CI 1.36–2.15). Conclusions Intraoperative norepinephrine exposure was independently associated with postoperative AKI, with nonlinear behaviour for unit-rate metrics and linear behaviour for time-integrated metrics, largely independent of intraoperative hypotension. The two anchors define a two-tier vigilance framework.