Impact of Intraoperative Hyperoxia Versus Normoxia on Mortality and Organ Injury in Cardiac Surgery: A Systematic Review and Meta-analysis of Randomized Clinical Trials

Article summarized by Gabija Valauskait?, MD

DOI: https://doi.org/10.1053/j.jvca.2025.08.053

Published: Journal of Cardiothoracic and Vascular Anesthesia, February 2026

Authors: Iago T.C. Grillo, Evelyn S.P. de Santana, Felipe S. Passos, Larissa E. Tanimoto, Jeane C. de Melo, Amanda M. Kondo, Ricardo E. Treml, Tulio Caldonazo

Article description

Summary:
  • This systematic review and meta-analysis aimed to compare mortality and organ injury associated with intraoperative hyperoxia versus normoxia in patients undergoing cardiac surgery with cardiopulmonary bypass or off-pump procedures.
  • Systematic review and meta-analysis of randomized controlled trials.
  • The study included 19 randomized controlled trials comprising 2,001 patients, of whom 982 (49.1%) received hyperoxia.
  • Hyperoxia was generally defined as PaO? 180–550 mmHg or FiO? 0.6–1.0, whereas normoxia corresponded to PaO? 50–150 mmHg or FiO? 0.21–0.5, according to the definitions used in the included trials.
  • Primary outcomes showed no significant difference in mortality between hyperoxia and normoxia during cardiac surgery, with comparable in-hospital mortality rates (1.1% vs 1.3%; RR 0.84, 95% CI 0.36–1.94) and 30-day mortality rates (0.4% vs 1.4%; RR 0.41, 95% CI 0.08–2.28).
  • Secondary outcomes showed that normoxia was associated with significantly better postoperative oxygenation, reflected by a higher PaO?/FiO? ratio (MD 31.49; 95% CI 15.85–47.13; p < 0.01), whereas hyperoxia was linked to increased myocardial and oxidative injury markers, with higher CK-MB at the end of surgery (SMD 1.65; 95% CI 0.12–3.18; p = 0.03; not significant on POD1) and elevated MDA levels (SMD 3.77; 95% CI 2.99–4.55; p < 0.01).
  • No significant differences were found in major postoperative clinical outcomes, including ICU or hospital length of stay, time to extubation, arrhythmia, MI, stroke, AKI, renal replacement therapy, infection, or reoperation.
Conclusions:
  • In cardiac surgery, intraoperative hyperoxia, compared with normoxia, does not lower in-hospital or 30-day mortality, but is associated with reduced postoperative PaO?/FiO? ratios and increased markers of oxidative stress and myocardial injury (MDA and CK-MB).
  • Current evidence is insufficient to mandate practice change, and further trials are needed to determine whether biomarker-guided oxygen titration can reduce organ injury and improve outcomes.