Abstract Background The impact of hyperoxia exposure on postoperative respiratory complications remains unclear, and cyanotic and non-cyanotic neonates may respond differently to hyperoxia. Therefore, we investigated the relationship between hyperoxia during cardiopulmonary bypass (CPB) and postoperative respiratory complications, with a stratified analysis based on cyanotic status. Methods This retrospective study included 340 neonates who underwent open-heart surgery at our hospital. Univariable and multivariable analyses were used to assess the relationship between hyperoxia exposure (PaO₂ 300 mmHg) during CPB and postoperative respiratory complications. The multivariable model was adjusted for confounding factors, including body surface area, anemia, cyanosis, use of lower body circulatory arrest, and CPB duration. The maximum PaO₂ was evaluated as a continuous variable and in quartiles, and hyperoxia exposure was categorized as no hyperoxia, hyperoxia once, twice, or ≥3 times. Stratified analysis was performed based on cyanosis status. Additionally, preoperative and postoperative levels of 12 inflammatory cytokines were analyzed in 108 patients, and differences in cytokine levels were compared among patients with varying frequencies of hyperoxia exposure. Results Multivariate analysis revealed that for every 100 mmHg increase in PaO₂ Max, the risk of postoperative respiratory complications significantly increased (OR=1.561, 95% CI: 1.206–2.055, p=0.001). Compared with neonates without hyperoxia exposure, those with hyperoxia occurring ≥2 times (OR=2.145, 95% CI: 1.126–4.175, p=0.022) or ≥3 times (OR=3.528, 95% CI: 1.660–7.696, p=0.001) had a markedly higher risk of postoperative respiratory complications. Quartile analysis of PaO₂ Max showed that neonates in the highest quartile (Q4) had a significantly higher incidence of respiratory complications than those in the lowest quartile (Q1) (OR=2.211, 95% CI: 1.085–4.110, p=0.028), with a significant trend across quartiles (p for trend = 0.040). Stratified analysis demonstrated that maximum PaO₂ significantly increased the risk of postoperative respiratory complications in cyanotic patients, but no such association was observed in non-cyanotic patients. Cytokine analysis demonstrated a significant postoperative increase in IL-6, IL-8, and IL-10 levels (p0.001). Further investigation into hyperoxia exposure frequency showed that higher hyperoxia frequency was associated with increased postoperative IL-4 (p=0.015), IL-5 (p=0.037), and IFN-γ (p=0.006) levels. Conclusion Our findings indicate that hyperoxia during neonatal CPB is strongly associated with postoperative respiratory complications, with a more pronounced effect in cyanotic ineonates. Additionally, hyperoxia may contribute to adverse outcomes by promoting inflammatory responses. These results underscore the need for individualized oxygen management strategies during CPB to mitigate postoperative risks.
Liu et al. (2025) studied this question.