Supplemental Postoperative Oxygen and Wound Infection in Morbidly Obese Patients

The investigators propose to test the hypothesis that the incidence of major complications related to infection or inadequate healing is reduced in morbidly obese patients given 80% inspired oxygen during, and for 12-18 hours after, surgery compared with patients given 80% oxygen only during surgery. The primary outcome will be a composite of major complications plausibly related to infection or healing.

Wound infection is a common and serious consequence of anesthesia and surgery. Infection is associated with major complications including intra-abdominal abscess, peritonitis without leakage, and sepsis. The average wound infection thus prolongs hospitalization by a week, doubles the need for ICU care, and doubles mortality. The first few hours following bacterial contamination constitute a decisive period during which infection is established. Techniques aimed at improving resistance to surgical wound infections are thus most effective when implemented during the decisive period - even though infections are not detected clinically until five-to-ten days after surgery. The primary defense against surgical pathogens is oxidative killing by neutrophils. This reaction critically depends on tissue oxygen tension. It is therefore unsurprising that subcutaneous tissue oxygen tension (PsqO2) is inversely correlated with the risk of surgical wound infection. As might be expected, supplemental perioperative oxygen administration increases tissue oxygen partial pressure (compared with an FIO2 of 30%) and decreases wound infection risk. While 100% oxygen is associated with atelectasis and pulmonary inflammation, we and others have shown that 80% oxygen is not. Inadequate healing is also common. Scar formation starts with transcription of pro-collagen, which undergoes post-translational hydroxylation of the abundant proline and lysine residues. This promotes the cross-linking within and between collagen strands that provides tensile strength which begins the healing process. The prolyl and lysyl hydroxylases that catalyze this reaction depend on the substrate oxygen. The Km (substrate concentration at which the velocity of the reaction is half-maximal) for O2 of prolyl hydroxylase has been variously estimated at 20, 25, and 100 mmHg. Even using the most conservative estimate, these data suggest that proline hydroxylation of collagen (e.g., scar formation) will be PO2 dependent through the range of 0 to ≈200 mmHg, with 90% of the effect occurring by 90 mmHg. In vitro and in vivo animal evidence supports this theory. Inadequate healing is associated with anastomotic leakage, wound dehiscence, and bleeding. In morbidly obese patients, the incidence of infection and major complications related to either infection or inadequate healing is 15-20% after gastric bypass based on our preliminary data and a review of records at Washington University. Roughly 30% of the American population is now obese with a body mass index (BMI) \> 30 kg/m2; about 5% of the population is morbidly obese, with a BMI exceeding 40 kg/m2. These patients are at extraordinarily high risk of wound infection, and infections in these patients are especially severe. Cardiac output, circulating blood volume, and resting oxygen consumption are all increased in the obese; however, total blood flow is sub-normal in relation to body weight. Obesity augments the size of individual fat cells without increasing blood flow. Specifically, fat tissue is hypoperfused and poorly oxygenated. Intraoperative tissue oxygenation is abnormally low in obese patients and postoperative tissue oxygenation is especially low - although it can be improved by oxygen administration. It thus seems likely that inadequate tissue oxygenation contributes to the observed extraordinary high wound infection risk in obese patients. Although supplemental oxygen during surgery and for the first two-to-six hours after surgery halves the risk of surgical wound infection, it remains unknown whether prolonged supplemental postoperative oxygen significantly augments the benefits of intraoperative oxygen. There is some evidence suggesting that prolonged postoperative oxygen administration may be helpful. For example, experimental infections in guinea pigs are reduced by administration of supplemental oxygen for 45 h. In contrast, restricting oxygen for only 45 minutes after contamination reduces tensile wound strength in rabbits no more than restricting oxygen for 5 full days. No study in humans has evaluated the effects of supplemental postoperative oxygen on surgical wound infection or associated complications. To the extent that supplemental oxygen after surgery is likely to reduce the incidence of major complications related to wound infection and healing, benefit is especially likely in morbidly obese patients because their baseline tissue oxygenation is usually sub-optimal and their risk is high. We thus propose to test the hypothesis that the incidence of major complications related to infection or inadequate healing are reduced in morbidly obese patients given 80% inspired oxygen during and for 12-18 hours after surgery compared with patients given 80% oxygen only during surgery. Patients will be enrolled at two OUTCOMES RESEARCH Group sites: the University of Louisville and the Cleveland Clinic. Patients undergoing gastric bypass will all be given 80% intraoperative oxygen and then randomly assigned to 30% or 80% oxygen until the first postoperative morning. The primary outcome will be a composite of major complications plausibly related to infection or healing. A maximum of 1,276 patients will provide a 90% power to detect a 25% reduction in the proportion of patients with at least one major complication. Supplemental oxygen is safe and inexpensive, costing only a few dollars per patient. Confirming our hypothesis would indicate that clinicians can reduce the risk of serious complications by making an essentially trivial modification in postoperative patient care.