To assess for publication bias, a funnel plot was constructed by graphing the log of the OR and the standard error of the log of the OR as the x- and y-axis, respectively. As shown in Figure 6, an inverted and symmetric funnel was observed, centered around a value of 1 on the x-axis, indicating a low potential for publication bias. Heterogeneity was assessed in our meta-analysis with an I2 value of 70% (P = 0.003) (Fig. 2).
Our systematic review and meta-analysis did not find evidence that supplemental oxygen reduced surgical site infection when all published trials comparing hyperoxia with control were included; however, a post hoc analysis showed that based on the sample size available, the study was underpowered to detect the 2% (17.5%–15.5%) absolute risk reduction that we found. This finding is in contrast to prior meta-analyses performed on this topic.11,12 Al-Niaimi and Safdar11 found that high inspired perioperative oxygen therapy was associated with a reduced rate of surgical site infection (relative risk [RR] = 0.70; 95% CI: 0.52, 0.94). However, the degree of heterogeneity in this study was high (I2 = 77%), and not accounted for. One way of dealing with unexplained heterogeneity is to incorporate it into a random effects model, which, for any group of trials with heterogeneity, may potentially result in a larger CI (i.e., a more conservative estimate). When a random effects model was used, the result was no longer statistically significant (RR = 0.74; 95% CI: 0.39, 1.43).
The pooled surgical site infection rates of 15.5% and 17.5% in the hyperoxia and control groups, respectively, are nearly double the rates published by the National Healthcare Safety Network.25 Some of the discrepancy in rates may be related to the use of varying definitions for surgical site infection. For instance, the National Healthcare Safety Network excludes stitch abscesses from their definition26 whereas all of the trials included in our meta-analysis, with the exception of the trial by Meyhoff et al.,14 have a definition that is potentially inclusive of stitch abscesses. Furthermore, the mix of surgical cases varies between our meta-analysis and the National Healthcare Safety Network because our meta-analysis included more gastrointestinal procedures, which tend to be associated with higher rates of infection.
Multiple subgroup analyses were completed and the results of each of these analyses were consistent with our overall finding that hyperoxia is not associated with reduced rates of surgical site infection. The exclusion of trials that provided nitrous oxide to patients was considered because nitrous oxide has been shown to both directly suppress cell-mediated immunity and interfere with vitamin B12 and folate metabolism, leading to impaired DNA synthesis, both of which could, at least theoretically, contribute to increased rates of surgical site infection.27 We did not find that the inclusion of trials that involved nitrous oxide significantly affected the results of our meta-analysis. An evaluation of trials based on fluid management strategies was considered because both hypovolemia and overadministration of crystalloid can affect surgical site infection.28 Hypovolemia decreases blood flow to wounds, and wound hypoperfusion has been shown to aggravate infections in animals.28 Moreover, edema can prohibit oxygen delivery. We did not find evidence that the fluid management strategy significantly affected the results of our meta-analysis.
Interestingly, a subgroup analysis of trials in which patients were not given neuraxial anesthesia showed a significant benefit for hyperoxia decreasing surgical site infection. A recently published study by Chang et al.29 noted a significant correlation between the use of general (versus neuraxial) anesthesia and an increase in surgical site infection in patients having total hip or knee replacements (OR = 2.21; 95% CI: 1.25, 3.90). Although further investigation of this topic is needed (it seems that patients who received general anesthesia may have had longer procedures predisposing them to infection), Sessler30 proposes multiple mechanisms for how regional anesthesia may decrease rates of surgical site infection: neuraxial anesthesia may attenuate the inflammatory response to surgery, may improve wound perfusion via direct peripheral vasodilatation, and may provide superior analgesia thus possibly decreasing catecholamine-associated peripheral vasoconstriction. It is hypothesized that neuraxial anesthesia alone may increase tissue perfusion and tissue oxygen tension, allowing greater numbers of more effective neutrophils to migrate to the site of a surgical wound. Including such patients in a trial comparing differing levels of inspired oxygen may make it difficult to find a statistically significant benefit from high inspired oxygen therapy.
A subgroup analysis of trials in which patients underwent only colorectal surgery also showed a significant benefit of hyperoxia in decreasing surgical site infection. This result is in contrast to the findings published by Brar et al.31 Some of the discrepancy in results can be attributed to the inclusion of an additional study (Bickel et al.15) in our analysis as well as the exclusion of the trial by Myles et al.10 As discussed earlier, the study by Myles et al.10 was not designed with the primary intention of supplementing oxygen, and thus may have clouded the interpretation of any meta-analysis designed to investigate the effect of varying oxygen levels on surgical site infection. It is unclear why the colorectal patient population would stand to benefit more from hyperoxia than the general surgical population. More likely than a unique response to hyperoxia in this surgical population is the higher likelihood of finding a statistical benefit secondary to higher rates of surgical site infection in the colorectal surgical population.32 Colorectal surgery is a “clean-contaminated” procedure in which a surgical wound is exposed to a significant bacterial load. Starting with a higher baseline rate for surgical site infection would increase the power of a study to find a benefit from high inspired oxygen therapy.
There are several limitations to our work. All meta-analyses have well-recognized limitations in part because of an inability to control for differences among studies. The results of this meta-analysis cannot be extrapolated to conclude that all levels of high inspired perioperative oxygen are unlikely to be beneficial. The choice of 80% and 30% oxygen concentrations in the studies used in this meta-analysis are somewhat arbitrary, and the use of other concentrations of oxygen may produce different results. Variables that differed among studies in this meta-analysis include perioperative antibiotic use, surgical operations performed, the absence or presence of neuraxial anesthesia, the use of nitrous oxide, the definition of surgical site infection, fluid management strategy, patient populations studied, and the duration of perioperative oxygen supplementation. A change in any of these variables might theoretically influence the overall outcome. Lastly, the Surgical Care Improvement Project was implemented in 2006; thus, Surgical Care Improvement Project measures (appropriate antibiotic selection and timing, appropriate hair removal, etc.) exerted a variable influence on the trials included in this meta-analysis, which may have influenced our results.
In summary, a meta-analysis of all trials that met our inclusion criteria did not show that high inspired perioperative oxygen therapy is beneficial for preventing surgical site infections. This is in contrast to 2 previously published meta-analyses11,12 that suggest an association between a decrease in surgical site infection and the use of high inspired oxygen therapy. Interestingly, 2 subgroup analyses (trials that excluded neuraxial anesthesia and trials that involved colorectal surgery only) showed a benefit for high inspired oxygen therapy decreasing surgical site infection. The positive results of the above-mentioned subgroup analyses point to a need for additional studies to further investigate this intervention.
1. Hall MJ, DeFrances CJ, Williams SN, Golosinskiy A, Schwartzman A. National Hospital Discharge Survey: 2007 summary. Natl Health Stat Report 2010;29:1–20, 24
2. Klevens RM, Edwards JR, Richards CL Jr, Horan TC, Gaynes RP, Pollock DA, Cardo DM. Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep 2007;122:160–6
3. Anderson DJ, Kaye KS, Chen LF, Schmader KE, Choi Y, Sloane R, Sexton DJ. Clinical and financial outcomes due to methicillin resistant Staphylococcus aureus surgical site infection: a multi-center matched outcomes study. PLoS One 2009;4:e8305
4. Dellinger EP. Increasing inspired oxygen to decrease surgical site infection. JAMA 2005;294:2091–2
5. Roos D, van Bruggen R, Meischl C. Oxidative killing of microbes by neutrophils. Microbes Infect 2003;5:1307–15
6. Babior BM. Oxygen-dependent microbial killing by phagocytes (first of two parts). N Engl J Med 1978;298:659–68
7. Allen DB, Maguire JJ, Mahdavian M, Wicke C, Marcocci L, Scheuenstuhl H, Chang M, Le AX, Hopf HW, Hunt TK. Wound hypoxia and acidosis limit neutrophil bacterial killing mechanisms. Arch Surg 1997;132:991–6
8. Belda FJ, Aguilera L, García de la Asunción J, Alberti J, Vicente R, Ferrándiz L, Rodríguez R, Company R, Sessler DI, Aguilar G, Botello SG, Ortí R. Spanish Reduccion de la Tasa de Infeccionq Uirurgica Group. Supplemental perioperative oxygen and the risk of surgical wound infection: a randomized controlled trial. JAMA 2005;294:2035–42
9. Greif R, Akça O, Horn EP, Kurz A, Sessler DI; Outcomes Research Group. Supplemental perioperative oxygen to reduce the incidence of surgical-wound infection. N Engl J Med 2000;342:161–7
10. Myles PS, Leslie K, Chan MT, Forbes A, Paech MJ, Peyton P, Silbert BS, Pascoe E; ENIGMA Trial Group. Avoidance of nitrous oxide for patients undergoing major surgery: a randomized controlled trial. Anesthesiology 2007;107:221–31
11. Al-Niaimi A, Safdar N. Supplemental perioperative oxygen for reducing surgical site infection: a meta-analysis. J Eval Clin Pract 2009;15:360–5
12. Qadan M, Akça O, Mahid SS, Hornung CA, Polk HC Jr. Perioperative supplemental oxygen therapy and surgical site infection: a meta-analysis of randomized controlled trials. Arch Surg 2009;144:359–66
13. Gardella C, Goltra LB, Laschansky E, Drolette L, Magaret A, Chadwick HS, Eschenbach D. High-concentration supplemental perioperative oxygen to reduce the incidence of postcesarean surgical site infection: a randomized controlled trial. Obstet Gynecol 2008;112:545–52
14. Meyhoff CS, Wetterslev J, Jorgensen LN, Henneberg SW, Høgdall C, Lundvall L, Svendsen PE, Mollerup H, Lunn TH, Simonsen I, Martinsen KR, Pulawska T, Bundgaard L, Bugge L, Hansen EG, Riber C, Gocht-Jensen P, Walker LR, Bendtsen A, Johansson G, Skovgaard N, Heltø K, Poukinski A, Korshin A, Walli A, Bulut M, Carlsson PS, Rodt SA, Lundbech LB, Rask H, Buch N, Perdawid SK, Reza J, Jensen KV, Carlsen CG, Jensen FS, Rasmussen LS; PROXI Trial Group. Effect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial. JAMA 2009;302:1543–50
15. Bickel A, Gurevits M, Vamos R, Ivry S, Eitan A. Perioperative hyperoxygenation and wound site infection following surgery for acute appendicitis. Arch Surg 2011;146:464–70
16. Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, Tugwell P, Klassen TP. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998;352:609–13
17. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, McQuay HJ. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1–12
18. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 1999;354:1896–900
19. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58
20. Bakri MH, Nagem H, Sessler DI, Mahboobi R, Dalton J, Akça O, Roselli EE, Insler SR. Transdermal oxygen does not improve sternal wound oxygenation in patients recovering from cardiac surgery. Anesth Analg 2008;106:1619–26
21. Whitney JD, Heiner S, Mygrant BI, Wood C. Tissue and wound healing effects of short duration postoperative oxygen therapy. Biol Res Nurs 2001;2:206–15
22. Wilson AP, Treasure T, Sturridge MF, Grüneberg RN. A scoring method (ASEPSIS) for postoperative wound infections for use in clinical trials of antibiotic prophylaxis. Lancet 1986;1:311–3
23. Mayzler O, Weksler N, Domchik S, Klein M, Mizrahi S, Gurman GM. Does supplemental perioperative oxygen administration reduce the incidence of wound infection in elective colorectal surgery? Minerva Anestesiol 2005;71:21–5
24. Pryor KO, Fahey TJ III, Lien CA, Goldstein PA. Surgical site infection and the routine use of perioperative hyperoxia in a general surgical population: a randomized controlled trial. JAMA 2004;291:79–87
25. Edwards JR, Peterson KD, Mu Y, Banerjee S, Allen-Bridson K, Morrell G, Dudeck MA, Pollock DA, Horan TC. National Healthcare Safety Network (NHSN) report: data summary for 2006 through 2008, issued December 2009. Am J Infect Control 2009;37:783–805
26. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309–32
27. Myles PS, Leslie K, Silbert B, Paech MJ, Peyton P. A review of the risks and benefits of nitrous oxide in current anaesthetic practice. Anaesth Intensive Care 2004;32:165–72
28. Kabon B, Akça O, Taguchi A, Nagele A, Jebadurai R, Arkilic CF, Sharma N, Ahluwalia A, Galandiuk S, Fleshman J, Sessler DI, Kurz A. Supplemental intravenous crystalloid administration does not reduce the risk of surgical wound infection. Anesth Analg 2005;101:1546–53
29. Chang CC, Lin HC, Lin HW, Lin HC. Anesthetic management and surgical site infections in total hip or knee replacement: a population-based study. Anesthesiology 2010;113:279–84
30. Sessler DI. Neuraxial anesthesia and surgical site infection. Anesthesiology 2010;113:265–7
31. Brar MS, Brar SS, Dixon E. Perioperative supplemental oxygen in colorectal patients: a meta-analysis. J Surg Res 2011;166: 227–35
32. Gaynes RP, Culver DH, Horan TC, Edwards JR, Richards C, Tolson JS. Surgical site infection (SSI) rates in the United States, 1992–1998: the National Nosocomial Infections Surveillance System basic SSI risk index. Clin Infect Dis 2001;33:S69–77
33. Anaya DA, Dellinger EP. The obese surgical patient: a susceptible host for infection. Surg Infect (Larchmt) 2006;7:473–80
34. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for Prevention of Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control 1999;27:97–132
APPENDIX 1: SEARCH STRATEGY USED FOR LITERATURE SEARCH
- MEDLINE search: oxygen[Title/Abstract] AND infection [Title/Abstract] Limits: Randomized Controlled Trial, Case Reports
- Cochrane Collaboration's CENTRAL database search: oxygen in Keywords and infection in Keywords in Cochrane Database of Systematic Reviews
- EMBASE search: ‘oxygen'/exp AND ‘infection'/exp AND ([Cochrane review]/lim OR [controlled clinical trial]/lim OR [randomized controlled trial]/lim) AND [humans]/lim AND [1-1-1945]/sd NOT [12-7-2010]/sd
APPENDIX 2: FURTHER DISCUSSION OF FIGURE 4
The P value calculated in Figure 4 without correction for multiple comparisons and large heterogeneity. The P value was calculated using the Cochrane Collaboration's RevMan version 5.0.25 using the procedure described by DerSimonian and Laird for fitting the random effects model for meta-analysis. DerSimonian and Laird obtain an estimate for τ2 using the method of moments. Alternative methods may be preferable for calculating the treatment effect for Figure 4 on the smaller sample size included in this subgroup analysis. On the basis of the stated statistical method uncertainty, the results depicted may be unreliable.