Scientific Reports (Feb 2023)

Updated meta-analysis on intraoperative inspired fraction of oxygen and the risk of surgical site infection in adults undergoing general and regional anesthesia

  • Yoann El Maleh,
  • Charlotte Fasquel,
  • Christophe Quesnel,
  • Marc Garnier

DOI
https://doi.org/10.1038/s41598-023-27588-2
Journal volume & issue
Vol. 13, no. 1
pp. 1 – 13

Abstract

Read online

Abstract This updated meta-analysis aims at exploring whether the use of systematic high vs low intraoperative oxygen fraction (FiO2) may decrease the incidence of postoperative surgical site infection during general (GA) or regional anesthesia (RA). PubMed, Cochrane CENTRAL, ClinicalTrials.gov databases were searched from January 1st, 1999 and July, 1st 2022, for randomized and quasi-randomized controlled trials that included patients in a high and low FiO2 groups and reported the incidence of SSI. The meta-analysis was conducted with a DerSimonian and Laird random-effects model. Thirty studies (24 for GA and 6 for RA) totaling 18,055 patients (15,871 for GA and 2184 for RA) were included. We have low-to-moderate-quality evidence that high FiO2 (mainly 80%) was not associated with a reduction of SSI incidence compared to low FiO2 (mainly 30%) in all patients (RR 0.90, 95%CI 0.79–1.03). Moderate inconsistency existed between studies (I2 = 38%). Subgroup analyses showed a moderate protective effect in patients undergoing GA (RR 0.86, 95%CI 0.75–0.99) (low level of evidence), while high FiO2 was not associated with a reduction of SSI in patients undergoing RA (RR 1.17, 95%CI 0.90–1.52) (moderate level of evidence). Sensitivity analyses restricted to patients ventilated without nitrous oxide (n = 20 studies), to patients operated from abdominal surgeries (n = 21 studies), and to patients suffering from deep SSI (n = 13 studies), all showed the absence of any significant effect of high FiO2. As a conclusion there is no compelling evidence that high FiO2 can improve postoperative patient’s outcome on its own when good SSI prevention practices are properly applied. Recent well-designed and adequately powered randomized controlled trials add further weight to these results.