Effects of oxygen on post-surgical infections during an individualised perioperative open-lung ventilatory strategy: a randomised controlled trial

Abstract

Background: We aimed to examine whether using a high fraction of inspired oxygen (FIO 2 ) in the context of an individualised intra-and postoperative open-lung ventilation approach could decrease surgical site infection (SSI) in patients scheduled for abdominal surgery. Methods: We performed a multicentre, randomised controlled clinical trial in a network of 21 university hospitals from June 6, 2017 to July 19, 2018. Patients undergoing abdominal surgery were randomly assigned to receive a high (0.80) or conventional (0.3) FIO 2 during the intraoperative period and during the first 3 postoperative hours. All patients were mechanically ventilated with an open-lung strategy, which included recruitment manoeuvres and individualised positive end-expiratory pressure for the best respiratory-system compliance, and individualised continuous postoperative airway pressure for adequate peripheral oxyhaemoglobin saturation. The primary outcome was the prevalence of SSI within the first 7 postoperative days. The secondary outcomes were composites of systemic complications, length of intensive care and hospital stay, and 6-month mortality. Results: We enrolled 740 subjects: 371 in the high FIO 2 group and 369 in the low FIO 2 group. Data from 717 subjects were available for final analysis. The rate of SSI during the first postoperative week did not differ between high (8.9%) and low (9.4%) FIO 2 groups (relative risk [RR]: 0.94; 95% confidence interval [CI]: 0.59e1.50; P¼0.90]). Secondary outcomes, such as atelectasis (7.7% vs 9.8%; RR: 0.77; 95% CI: 0.48e1.25; P¼0.38) and myocardial ischaemia (0.6% [n¼2] vs 0% [n¼0]; P¼0.47) did not differ between groups. Conclusions: An oxygenation strategy using high FIO 2 compared with conventional FIO 2 did not reduce postoperative SSIs in abdominal surgery. No differences in secondary outcomes or adverse events were found. Clinical trial registration: NCT02776046.

CON: Routine hyperoxygenation in adult surgical patients whose tracheas are intubated

“…In the most recent meta‐analysis, peri‐operative hyperoxygenation did not significantly reduce the incidence of surgical site infections when compared with normoxia, OR (95%CI) 0.89 (0.73–1.08), p = 0.23 . These findings were corroborated in another methodologically robust trial that was published recently and showed no significant difference in the incidence of surgical site infections between an inspired oxygen fraction of 0.3 and 0.8 . So far, over 13,000 patients have been randomly allocated to receive hyper‐ or normoxia and still no clear benefit from hyperoxygenation has been shown.…”

Perioperative Hyperoxyphobia: Justified or Not? Benefits and Harms of Hyperoxia during Surgery

“…We will first examine mortality, as the ultimate clinical endpoint of any adverse effect. Interestingly, the point estimates of all but two [14,15] of the clinical trials on perioperative high versus normal FiO 2 indicated lower, rather than higher, short-term mortality in patients given 0.80 FiO 2 [16][17][18][19][20]. However, each of these trials had low mortality rates with wide confidence intervals, including unity.…”

“…The most likely clinical consequence of pulmonary oxygen toxicity due to ROS, as well as resorption atelectasis, would be an increased incidence of postoperative pulmonary complications. Multiple studies report on adverse effects such as pneumonia or composite measures of respiratory complications [14,15,18,26]. A meta-analysis of all but the most recent of these studies demonstrated no increase in respiratory AE after 30 min to 4 h perioperative administration of 0.80 versus 0.30-0.35 FiO 2 , nor was there a difference in the incidence of postoperative atelectasis [12].…”

“…This trial was well designed, with double-blind outcome assessment limited to culture proven infections, and assessment of oxygen delivery by measurement of PaO 2 and PtO 2 . Subsequent trials did not meet this standard and could not consistently reproduce these results [14][15][16]20]. The issue was further complicated when concerns on the validity of some of the trials arose [7][8][9][10].…”

“…Many of the explanations for the observed difference in efficacy of hyperoxia on SSI rely on the idea that the application of high FiO 2 alone does not guarantee adequate tissue oxygenation. Adequate gas exchange, as well as sufficient macro-and microcirculatory flow, are needed to actually deliver oxygen to the tissues; this is also suggested by the results of the most recent trial on perioperative hyperoxia, published after the aforementioned systematic review, that showed no beneficial effect of 0.80 versus 0.30 FiO 2 in abdominal surgery [15]. In this study, both groups were ventilated using lung-protective settings, including modulation of end-expiratory pressure to optimise pulmonary compliance.…”