Perioperative oxygen therapy in patients undergoing surgical procedures: an overview of systematic reviews and meta-analyses.

Background: Perioperative oxygen administration has been proposed as a strategy to reduce postoperative complications. However, uncertainty exists as to which strategies are the most clinically effective. Objectives: To provide an overview on the effectiveness of perioperative oxygen therapy and formulate recommendations to inform clinical decision-making and research. Methods: We followed the Preferred Reporting Items for Overviews of Reviews guidelines. We searched key databases for systematic reviews (from inception to September 2021) and randomised controlled trials (from April 2018 to March 2022) comparing perioperative oxygen strategies. Reviews with the most comprehensive coverage of literature were chosen as anchoring reviews. We assessed risk of bias for each anchoring review using the Risk of Bias in Systematic Reviews tool. We updated meta-analyses from anchoring reviews with data from recent randomised controlled trials and conducted subgroup analyses and meta-regression. We assessed the certainty of evidence using grading of recommendations assessment, development and evaluation framework and conducted trial sequential analysis. We used grading of recommendations assessment, development and evaluation informative statements to communicate our findings. Our advisory panel reviewed mapping of studies and interpretation of evidence. Results: We identified 59 systematic reviews and selected 5 anchoring reviews. A high fraction of inspired oxygen may result in a slight reduction in surgical site infection compared with a low fraction of inspired oxygen (risk ratio 0.91, 95% confidence interval 0.78 to 1.05; risk difference 1.2% lower, 2.9% lower to 0.7% higher, low-certainty evidence). This effect may be modified by type of surgery, oxygen delivery method or study quality. The evidence suggests that a high fraction of inspired oxygen results in a large increase in the incidence of atelectasis (risk ratio 1.47, 95% confidence interval 1.20 to 1.79; risk difference 6.5% higher, 2.8% higher to 10.9% higher, low-certainty evidence) and may increase postoperative pulmonary complications slightly (risk ratio 1.06, 0.77 to 1.46; risk difference 1.1% higher, 4.1% lower to 8.2% higher) but the evidence is very uncertain. A high fraction of inspired oxygen may result in little to no difference in mortality, nausea and vomiting, and length of hospital stay. Postoperative high-flow nasal oxygen may reduce the need to escalate respiratory support compared with conventional oxygen therapy (risk ratio 0.61, 0.41 to 0.91; risk difference 7.8% lower, 11.7% lower to 1.8% lower) but the evidence is very uncertain. High-flow nasal oxygen may result in little to no difference in mortality and reintubation rate. Compared with conventional oxygen therapy, postoperative non-invasive ventilation may decrease postoperative pulmonary complications (risk ratio 0.62, 0.44 to 0.87; risk difference 12.2% lower, 18% lower to 4.2% lower) and probably results in a slight reduction in the incidence of acute respiratory distress syndrome (risk ratio 0.70, 0.53 to 0.93; risk difference 1.2% lower, 1.9% lower to 0.3% lower). Non-invasive ventilation results in little to no difference in mortality, pneumonia or reintubation rate. Grading of recommendations assessment, development and evaluation certainty in evidence was low for most outcomes. Trial sequential analysis revealed further studies are required to provide conclusive evidence on the effectiveness of perioperative oxygen therapy. Conclusions: There is no clear evidence that either a high or a low fraction of inspired oxygen improves outcomes in surgical patients. Existing evidence is insufficient for recommending routine use of non-invasive ventilation or high-flow nasal oxygen. Future work: Future randomised controlled trials should stratify participants by type of surgery, anaesthesia technique and documented risk factors for postoperative complications, such as body mass index. Study registration: This study is registered as PROSPERO CRD42021272361. Funding: This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR132987) and is published in full in