Early osmotherapy exposure and subsequent incident acute kidney injury in neurocritical care patients: a MIMIC-IV retrospective cohort study.

backgroundOsmotherapy is frequently used in neurocritical care to manage cerebral edema and intracranial hypertension. However, its association with subsequent new-onset acute kidney injury (AKI), and whether early serum sodium or chloride burden materially attenuates this association, remain uncertain.

methodsWe conducted a retrospective cohort study using MIMIC-IV version 3.1. Adult first-ICU neurocritical care patients were analyzed with a 48-h landmark design. The primary risk set included patients who were alive, remained in the ICU, and were AKI-free at 48 h after ICU admission. The primary exposure was any osmotherapy, defined as mannitol or hypertonic saline administered within the first 48 ICU hours, compared with no osmotherapy. The primary outcome was incident AKI after the 48-h landmark through ICU day 7 or ICU discharge. Multivariable regression was used as the main adjusted model; propensity-score weighting, overlap weighting, doubly robust models, and modified Poisson regression were used as complementary estimators. Sodium/chloride burden analyses were conducted in the lab-complete subset.

resultsThe final 48-h landmark cohort included 2,756 adult neurocritical care patients, of whom 624 (22.6%) developed incident AKI after the landmark. Any osmotherapy was administered to 365 patients (13.2%). In the doubly robust overlap-weighted logistic model, any osmotherapy was associated with subsequent incident AKI (OR, 1.48; 95% CI 1.12-1.96; P = 0.006). The corresponding doubly robust overlap-weighted modified Poisson estimate was RR 1.30 (95% CI 1.08-1.57; P = 0.006). Sodium/chloride burden analyses included 2,455 lab-complete patients. Early sodium burden (OR, 0.99; 95% CI 0.92-1.07; P = 0.792) and chloride burden (OR, 1.01; 95% CI 0.94-1.09; P = 0.743) did not materially attenuate the observed association in exploratory adjustment models. In exploratory clinical relevance analyses, patients with subsequent AKI had higher hospital mortality and lower creatinine-based renal recovery at ICU and hospital discharge. Dose-proxy and hypertonic saline concentration-specific analyses were exploratory because dose capture was incomplete and 23.4% hypertonic saline exposure was sparse.

conclusionsAmong adult neurocritical care patients who were alive, remained in the ICU, and were AKI-free at the 48-h landmark, early osmotherapy exposure was associated with a higher risk of subsequent incident AKI. This association was not materially attenuated by early serum sodium or chloride burden in exploratory adjustment models. These findings support careful renal monitoring in patients receiving clinically indicated osmotherapy, while residual confounding by indication precludes causal interpretation.