Controlling oxygenation during initiation of cardiopulmonary bypass: can it improve immediate postoperative outcomes in cyanotic children undergoing cardiac surgery? A prospective randomized study.

Objective: Cardiopulmonary bypass (CPB) initiated with high oxygen levels may expose cyanotic children to reoxygenation injury. The ideal method of initiation of bypass to prevent this phenomenon still remains largely unproven. This study tested the hypothesis that controlling oxygenation during initiation of CPB improves early postoperative outcomes.

Methods: Thirty-one cyanotic children were randomized to two treatment arms of the study. In group A (intervention), CPB was initiated with fraction of inspired oxygen (Fio 2) 0.21, and after one minute of full bypass, Fio 2 was increased at increments of 0.1 per minute to reach 0.6. In group B (hyperoxemic), CPB was initiated using Fio 2 >0.6. Aortic cross clamp time (minutes), CPB time (minutes), creatine phosphokinase-MB (CPK-MB) levels (U/L), lactate levels (mmol/L), duration of ventilator support (hours), inotropic support (hours), and intensive care unit (ICU) stay (hours) as well as hospital mortality were measured.

Results: Levels of CPK-MB (group A mean = 59.6 U/L, 95% confidence interval [CI]: 45.9-73.3; group B mean = 82.6 U/L, 95% CI: 66.1-99.1, P = .016) and ventilation time (group A median = 16.5 hours; interquartile range [IQR] = 11.25-23; group B median = 27.5 hours; IQR = 17-54, P = .045) were significantly lower in the intervention group. Other parameters showed no significant differences: CPB time (group A median = 71.5 minutes, IQR = 64-100; group B median = 95.5 minutes, IQR = 58-145, P = .71), cross clamp time (group A mean = 59.2 minutes, 95% CI: 47.6-70.8; group B mean = 66.57 minutes, 95% CI: 47.6-88.5, P =.57), lactate levels (mmol/L; group A median = 1.8, IQR = 1.48-2.59; group B median = 2.1, IQR = 1.29-2.62, P = 1), inotropic support (group A median = 47.5 hours, IQR = 36-73.75; group B median = 59.5 hours, IQR = 41.75-92.5, P = .27), ICU stay (group A median = 59.5 hours, IQR = 48.25-118.5; group B median = 85 hours, IQR = 47.75-137.50, P = .21), and mortality (group A n = 2, group B n = 2).

Conclusion: A controlled oxygenation protocol was associated with significantly lower postoperative CPK-MB levels. Evaluation of other end points including ventilation times requires a study with larger sample size for validation.