Hyperoxia During Neonatal Cardiopulmonary Bypass is Associated with Worse Clinical Outcomes: A Multi‐Institutional Study

Background Exposure to supraphysiologic oxygen concentrations, or hyperoxia, during neonatal cardiopulmonary bypass (CPB) has been associated with worse outcomes in single‐center studies. We aimed to describe variation in oxygen exposure during CPB and determine if hyperoxia is associated with worse outcomes in a multicenter cohort of neonates undergoing cardiac surgery. Methods We conducted a retrospective study of neonates who underwent surgery with CPB between January 2021 and December 2022 at 29 centers. Primary outcomes were operative mortality and major adverse cardiovascular events (MACE), which included CPR, extracorporeal support, stroke, and mortality. Logistic regression assessed the associations between median PaO 2 during CPB and outcomes in the entire cohort and subset of patients with single ventricle (SV) anatomy. Results We analyzed 1,175 neonates, including 357 with SV anatomy. Median PaO 2 during CPB was 296 mmHg (range: 54 – 800), with significant variation across centers (p<0.001). There were 54 deaths (5%) and 203 MACE (17%). No patients with PaO 2 < 200mmHg (n=161) died, and PaO 2 > 221 demonstrated modest mortality prediction (96% sensitivity, 19% specificity). In multivariable analysis, before adjustment for center, PaO 2 > 221 mmHg was significantly associated with mortality and MACE in all patients and patients with SV anatomy. After adjustment for center, associations between hyperoxia and mortality (p=0.053) or MACE (p=0.08) were not significant in the entire cohort but remained significant in patients with SV anatomy. Conclusions We observed significant variation in oxygen exposure during neonatal CPB across centers and significant associations between hyperoxia and worse outcomes, particularly in neonates with SV anatomy.