Urine output is closely associated with renal function and has been used as a diagnostic criterion for acute kidney injury (AKI). However, urine output during cardiopulmonary bypass (CPB) has never been identified as a predictor of postoperative AKI. Considering altered renal homeostasis during CPB, we made a comprehensible approach to CPB urine output and evaluated its predictability for AKI.
Patients undergoing cardiovascular surgery with the use of CPB, between January 2009 and December 2011, were retrospectively reviewed. AKI was defined as an increase in serum creatinine ≥0.3 mg/dL in the first postoperative 48 hours. We extrapolated a possible optimal amount of urine output from the plot of probability of AKI development according to CPB urine output. After separating patients by the predicted optimal value, we performed stepwise logistic regression analyses to find potential predictors of AKI in both subgroups.
A total of 696 patients were analyzed. The amount of CPB urine output had a biphasic association with the incidence of AKI using 4 mL/kg/h as a boundary value. In a multivariate logistic regression to find predictors for AKI in entire patients, CPB urine output did not show statistical significance. After separating patients into subgroups with CPB urine output below and over 4 mL/kg/h, it was identified as an independent predictor for AKI with the odds ratio of 0.43 (confidence interval 0.30–0.61) and 1.11 (confidence interval 1.02–1.20), respectively.
The amount of urine output during CPB with careful analysis may serve as a simple and feasible method to predict the development of AKI after cardiac surgery at an early time point.