[A meta-analysis of perioperative hyperoxia for the surgical site infections in patients with general surgery].

Objective To systemically analyze the effect of perioperative hyperoxia for the surgical site infections in patients with general surgery. Methods Electronic databases consisting of Pubmed, Embase, Cochrane Library, Wanfang Database, CNKI and VIP were systemically searched from established time to November 18, 2015. The randomized controlled trials about perioperative high and low-concentration oxygen for the surgical site infections in patients with general surgery were screened strictly and analyzed by the software of Revman 5.3. The included trials were stratified according to the colorectal or non-colorectal surgery, the duration of oxygen inhalation and the kinds of mixed gas to perform subgroup analyses. Sensitivity analysis was conducted by removing the low-quality trials, etc. The outcome was the surgical site infections. Results There were 989 relevant articles were searched out. At last, 9 randomized controlled trials consisting of 3 281 patients were included. The 80% oxygen group and 30% oxygen group consists of 1 644 and 1 637 patients, respectively. The rates of surgical site infection were 15.1% (248/1 644) and 17.5% (286/1 637) in the two group. Heterogeneity existed between the included trials and random-effect model was used, the RR=0.80, 95%CI: 0.60-1.08, P=0.15. Therefore, statistically significant difference was not found for the surgical site infections in the general surgery between the perioperative high and low-concentration oxygen. However, the results of subgroup analyses showed that the perioperative hyperoxia decreaced the surgical site infections significant in the subgroups of colorectal surgery and intraoperative plus postoperative 6 h oxygen inhalation. Conclusions Perioperative hyperoxia does not reduce surgical site infections in patients with general surgery. However, the results of two subgroup analyses (colorectal surgery and intraoperative plus postoperative 6 h oxygen inhalation trials) show a significantly benefit for perioperative hyperoxia in decreasing surgical site infections. Key words: Hyperoxia; Meta-analysis; General surgery; Surgical site infection