Official Title Optimization Strategy to Predict and Individualize Oxygen Delivery During Cardiac Surgery Under Cardiopulmonary Bypass

Cardiac surgery is a frequent procedure that can lead to serious complications, including acute kidney injury (AKI) or postoperative delirium. During the intervention, a cardiopulmonary bypass (CPB) is used to ensure tissue perfusion and oxygen delivery (DO2). The hypothesisis that an individualized strategy to optimize DO2 during CPB could decrease complications rate after cardiac surgery. Indeed, it is known that DO2 depends on CPB output, hemoglobin level and O2 tension. To this day, a DO2 threshold below 280 mL is known to be associated with postoperative AKI. From these data, a "Goal Directed Perfusion" strategy is widely used to maintain a DO2 above this threshold during CPB. But DO2 decrease in unpredictable, and the other factors influencing DO2 interindividual variability are not known. Moreover, the relation between DO2 and tissue perfusion is not well established, as DO2 requirement could differ between individuals or organs. Population approach using nonlinear mixed models is a method used to optimize drug administration with pharmacokinetics and pharmacodynamics models. Using this method, the aim of this study is to evaluate DO2 variability during CPB and develop a model to optimize GDP strategy.