Abstract Body (Do not enter title and authors here): Background: Neonates undergoing cardiac surgery with cardiopulmonary bypass (CPB) may be exposed to supraphysiologic oxygen (O₂) concentrations, termed hyperoxia, which has been associated with worse outcomes in single center studies. We aimed to describe variation in oxygen exposure during CPB within and across centers and determine if hyperoxia was associated with worse outcomes in a multicenter cohort of neonates undergoing cardiac surgery.
Methods: We performed a retrospective study across 29 centers enrolled in CoRe-PCICS: Collaborative Research from the Pediatric Cardiac Intensive Care Society. Neonates (< 30d) who underwent Society of Thoracic Surgery STAT category 2–5 surgery with CPB between 01/2021 and 12/2022 were included. Clinical outcomes of interest were operative mortality and the composite outcome of major adverse cardiovascular events (MACE), including: cardiac arrest, extracorporeal support, stroke, and operative mortality. Logistic regression was performed to determine association between median PaO₂ during CPB and outcomes. Post-hoc subset analyses were performed on neonates with single ventricle (SV) anatomy.
Results: We reviewed 1,175 neonates (median 48 per center), including 357 with SV anatomy. Variation in O₂ exposure during CPB is shown in Figure 1. There were 54 mortalities (5%) (Figure 2) and 203 MACE (17%) in total and 35 mortalities (10%) and 111 MACE (31%) in patients with SV anatomy. We observed no mortalities in neonates with median PaO₂ < 200mmHg (n=161). Analysis identified a median PaO₂ cut point of 221 mmHg as having modest predictive value for mortality (96% sensitivity, 19% specificity, Youden’s index score 0.16). In multivariate regression, hyperoxia (median PaO₂ > 221mmHg) was associated with mortality and MACE, controlling for birth weight, ventricular anatomy, non-cardiac anomalies, and CPB duration, however, not associated when also controlling for center. In a similar model among neonates with SV anatomy, hyperoxia was associated with mortality (p=0.022) and MACE (p=0.008) (Table 1).
Conclusions: In a multicenter study of neonates who underwent cardiac surgery, variation of O₂ exposure during CPB was apparent within and across centers and supraphysiologic PaO₂ values were common. Hyperoxia was associated with worse morbidity and mortality, particularly in neonates with SV anatomy. These findings underscore the need for a trial to determine if conservative oxygen titration during CPB leads to improved outcomes.