Safety of 80% vs 30–35% fraction of inspired oxygen in patients undergoing surgery: a systematic review and meta-analysis

Katharina Mattishent, Menaka Thavarajah, Ashnish Sinha, Adam Peel, Matthias Egger, Joseph Solomkin, Stijn de Jonge, Asad Latif, Sean Berenholtz, Benedetta Allegranzi, Yoon Kong Loke

Research output: Contribution to journalArticlepeer-review

67 Citations (Scopus)
14 Downloads (Pure)


Evidence-based guidelines from the World Health Organization (WHO) have recommended a high (80%) fraction of inspired oxygen (FiO2) to reduce surgical site infection in adult surgical patients undergoing general anaesthesia with tracheal intubation. However, there is ongoing debate over the safety of high FiO2. We performed a systematic review to define the relative risk of clinically relevant adverse events (AE) associated with high FiO2.

We reviewed potentially relevant articles from the WHO review supporting the recommendation, including an updated (July 2018) search of EMBASE and PubMed for randomised and non-randomised controlled studies reporting AE in surgical patients receiving 80% FiO2 compared with 30–35% FiO2. We assessed study quality and performed meta-analyses of risk ratios (RR) comparing 80% FiO2 against 30–35% for major complications, mortality, and intensive care admission.

We included 17 moderate–good quality trials and two non-randomised studies with serious-critical risk of bias. No evidence of harm with high FiO2 was found for major AE in the meta-analysis of randomised trials: atelectasis RR 0.91 [95% confidence interval (CI) 0.59–1.42); cardiovascular events RR 0.90 (95% CI 0.32–2.54); intensive care admission RR 0.93 (95% CI 0.7–1.12); and death during the trial RR 0.49 (95% CI 0.17–1.37). One non-randomised study reported that high FiO2 was associated with major respiratory AE [RR 1.99 (95% CI 1.72–2.31)].

No definite signal of harm with 80% FiO2 in adult surgical patients undergoing general anaesthesia was demonstrated and there is little evidence on safety-related issues to discourage its use in this population.
Original languageEnglish
Pages (from-to)311-324
Number of pages14
JournalBritish Journal of Anaesthesia
Issue number3
Early online date3 Jan 2019
Publication statusPublished - Mar 2019

Cite this