Time–dose response of oxygen delivery during cardiopulmonary bypass predicts acute kidney injury

Abstract Objective Previous studies have reported that nadir oxygen delivery during cardiopulmonary bypass is associated with the occurrence of postoperative acute kidney injury (AKI). However, these measurements only considered the bottom point of the oxygen delivery (DO 2 ) but did not consider the duration of DO 2 . We aimed to examine whether the time–dose response of DO 2 during cardiopulmonary bypass can be used to estimate the risk for postoperative AKI. Methods We evaluated 112 patients who underwent cardiac surgeries with cardiopulmonary bypass. We analyzed the perfusion parameters recorded every 20 seconds. To obtain time–dose response of DO 2 index (DO 2 i), the area under the curve was calculated as below the 300 mL/min/m 2 threshold, which accounts for depth and duration of cumulative oxygen debt. In addition, the cumulative time below DO 2 i 300 mL/min/m 2 was also calculated. Receiver operating characteristic analysis, univariate regression analysis, and multivariate regression analysis were used to evaluate associations between perioperative variables and postoperative AKI. Results Patients who developed AKI had larger area under the curve below the 300 mL/min/m 2 threshold (1581 vs 632; P 2 i 300 mL/min/m 2 (34.7 vs 15.3 minutes; P 2 i was not significantly different between the non-AKI and AKI groups (263.4 vs 247.0 mL/min/m 2 ; P  = .291). Conclusions The time–dose response of DO 2 i during cardiopulmonary bypass is a better indicator than nadir DO 2 i in evaluating AKI risk. Maintaining DO 2 i levels >300 mL/min/m 2 may result in decreased risk for postoperative AKI.