Renal regional oxygenation during pediatric cardiac surgery predicts acute kidney injury: a prospective cohort study with model comparisons
摘要
Acute kidney injury (AKI) is a common complication after pediatric cardiac surgery using cardiopulmonary bypass (CPB). This study evaluated renal regional oxygen saturation (R-rSO₂), measured via near-infrared spectroscopy (NIRS), as an AKI predictor.
Methods120 pediatric patients undergoing CPB-assisted congenital heart surgery were prospectively enrolled. Continuous intraoperative R-rSO₂ monitoring was performed. Four multivariable logistic regression models (adjusted for age, pre-SCr, pre-hemoglobin, RACHS-2, cyanosis, and CPB time) assessed distinct R-rSO₂ metrics during CPB for predicting AKI (pRIFLE criteria).
ResultsAKI incidence was 35.8% (n = 43). The model using the area under the curve (AUC) for R-rSO₂ decrease ≥5% during CPB showed superior predictive performance (C-index = 0.854) and fit. Within this model, a greater AUC for R-rSO₂ decrease ≥5% during CPB was independently associated with increased AKI risk (OR 1.02, 95% CI 1.00–1.03, P = 0.014). Prolonged CPB duration (OR 1.02, 95% CI 1.00–1.04, P = 0.028) and lower preoperative serum creatinine (OR 0.87, 95% CI 0.76–0.99, P = 0.031) were also significant predictors. AKI correlated with prolonged ventilation (P < 0.001) and higher costs (P < 0.001).
ConclusionRenal tissue desaturation during CPB, quantified as the AUC for R-rSO₂ decrease ≥5%, is significantly associated with postoperative AKI in children. Higher preoperative creatinine (mature function) was protective, while longer CPB time increased risk.
ImpactThe cumulative burden of renal desaturation specifically during CPB is the outstanding intraoperative predictor of postoperative AKI in children undergoing congenital heart surgery. This study defines a quantitative metric—the cumulative area under the curve for renal regional oxygen saturation decrease ≥5% during CPB—which demonstrates high specificity and provides a potential, actionable intraoperative monitoring threshold for predicting AKI risk.