Secondary Logo

Journal Logo

Hypercapnia Improves Tissue Oxygenation in Morbidly Obese Surgical Patients

Hager, Helmut, MD*†; Reddy, Dayakar, MD*; Mandadi, Goutham, MD*; Pulley, Debra, MD; Eagon, J Chris, MD§; Sessler, Daniel I., MD; Kurz, Andrea, MD

doi: 10.1213/01.ane.0000229715.71464.90
General Articles: Research Report
Chinese Language Editions

Risk of wound infection is increased in morbidly obese surgical patients, in part because a major determinant of wound infection risk, tissue oxygenation, is marginal. Unlike in lean patients, supplemental inspired oxygen (Fio2) only slightly improves tissue oxygenation in obese patients. Mild hypercapnia improves tissue oxygenation in lean patients but has not been evaluated in obese patients. We thus tested the hypothesis that mild hypercapnia markedly improves tissue oxygenation in morbidly obese patients given Fio2 80% during major abdominal surgery. Thirty obese patients (body mass index 61.5 ± 17 kg/m2) scheduled for open gastric bypass were randomly assigned to normocapnia (n = 15, end-tidal Pco2 35 mm Hg) or hypercapnia (n = 15, end-tidal Pco2 50 mm Hg); Fio2 was 80%. Anesthetic management and other confounding factors were controlled. Tissue oxygen tension was measured subcutaneously at the upper arm using a polarographic probe in a silastic tonometer. Demographic characteristics, cardiovascular measurements, and Pao2 (222 ± 48 versus 230 ± 68 mm Hg in normocapnic versus hypercapnic; mean ± sd; P = 0.705) were comparable in the groups. Tissue oxygen tension, however, was greater in hypercapnic than in normocapnic patients (78 ± 31 versus 56 ± 13 mm Hg; P = 0.029). Mild hypercapnia increased tissue oxygenation by an amount believed to be clinically important and could potentially reduce the risk of surgical wound infection in morbidly obese patients.

IMPLICATIONS: Wound and tissue oxygenation are critically low in morbidly obese patients, even with supplemental oxygen administration. Mild hypercapnia combined with supplement oxygen during the perioperative period significantly increased tissue oxygenation in obese patients.

From the *Department of Anesthesiology, Washington University, St. Louis, MO; †Department of Anesthesiology and General Intensive Care, Vienna General Hospital, University of Vienna, Vienna, Austria; ‡ Department of Surgery, Washington University, St. Louis, MO; § Division of Colon-Rectal Surgery, Department of Surgery, Washington University, St. Louis, MO; ∥Department of Outcomes Research, Cleveland Clinic Foundation, Cleveland, OH; ¶Department of Anesthesiology, University of Bern, Bern, Switzerland; Outcomes Research Institute, University of Louisville, Louisville, KY.

Supported, in part, by NIH grant GM 61655 (Bethesda, MD), the Gheens Foundation (Louisville, KY), the Joseph Drown Foundation (Los Angeles, CA), and the Commonwealth of Kentucky Research Challenge Trust Fund (Louisville, KY). CRD Grant #931504, Clinical Research Division, Department of Anesthesiology, Washington University School of Medicine (St. Louis, MO). Tyco-Mallinckrodt Anesthesiology Products, Inc. (St. Louis, MO) donated the thermocouples used. None of the authors has a personal financial interest related to this research. Gilbert Haugh, MS (University of Louisville), helped with the statistical analysis and Nancy Alsip, PhD (University of Louisville), edited the manuscript.

Presented, in part, at the annual meeting of the American Society of Anesthesiologists in San Francisco, California, October 2003.

Accepted for publication May 16, 2006.

Address correspondence to Helmut Hager, MD, Department of Anesthesiology and General Intensive Care Medicine, Medical University Vienna, Austria. Address e-mail to

Obesity is a rapidly increasing health care problem worldwide. In obese surgical patients, wound infections are common and associated with considerable morbidity (1). A contributing factor is almost certainly reduced subcutaneous wound tissue oxygenation (PsqO2). The primary defense against surgical pathogens is oxidative killing by neutrophils (2); hence, infection risk is inversely related to tissue oxygen partial pressure (3–5). For instance, the incidence of infection can be increased by either perioperative cigarette smoking or hypothermia (6), both of which reduce PsqO2 (7,8).

PsqO2 in healthy lean and obese awake volunteers is approximately 60 mm Hg (9). PsqO2 remains near that concentration during anesthesia and surgery in lean patients, but it is dramatically decreased—in some patients by almost half—in obese patients during anesthesia and surgery (10). Furthermore, in obese patients providing 80% inspired oxygen during surgery only slightly increases perioperative PsqO2 (10), whereas it doubles PsqO2 in lean patients (11).

Two factors contribute to diminished PsqO2 in obese patients. One is that arterial oxygen (Pao2) is reduced in obese patients at any given inspired oxygen concentration (10). The second factor is that obesity increases the size of individual fat cells without changing blood flow (12). Although cardiac output, the amount of circulating blood, and resting oxygen consumption are all elevated in obese patients (13), tissue perfusion remains abnormally low in relation to body weight (14,15) and, therefore, poorly oxygenated. Thus, even at a given Pao2, PsqO2 is significantly less in obese patients than in lean patients (10). All the above factors result in PsqO2 that is critically low in the obese individual undergoing anesthesia and surgery, i.e., approximately 40 mm Hg (10), leading to an increased risk of wound infections (4).

Mild hypercapnia (i.e., an end-tidal Pco2 of 50 mm Hg) increases cardiac index (16) and may induce peripheral vasodilation (17). Hypercapnia thus significantly increases PsqO2 in lean volunteers (16) and surgical patients (18) and does so at no cost and with minimal side effects. Because supplemental oxygen is relatively ineffective in obese patients (10), hypercapnia remains an attractive option for improving PsqO2 in this high-risk population. It remains unknown, though, whether hypercapnia can overcome the relative hypoperfusion of the adipose tissue. We, therefore, tested the hypothesis that hypercapnia (end-tidal Pco2 of 50 mm Hg versus 35 mm Hg) increases PsqO2 in obese patients given 80% inspired oxygen.

Back to Top | Article Outline


The study was performed at Washington University, St. Louis, MO. With approval of the local human studies committee and written informed patient consent, we enrolled 30 patients undergoing open gastric bypass surgery. A patient had to have a body mass index (BMI) of more than 50 kg/m2 to qualify for the study. A single surgeon performed all operations. Exclusion criteria were documented coronary or peripheral artery disease, insulin-dependent diabetes mellitus, history of recent cigarette smoking, and any symptoms of infection or sepsis.

Back to Top | Article Outline


General anesthesia was induced with sodium thiopental (3–5 mg/kg) and fentanyl (1–3 μg/kg). Vecuronium (0.1 mg/kg) or succinylcholine (0.8–1 mg/kg) was given to facilitate endotracheal intubation. After induction of anesthesia, patients were randomly assigned to one of two groups: normocapnia (n = 15) or hypercapnia (n = 15). The patients randomized to normocapnia were maintained at an end-tidal carbon dioxide partial pressure (ETco2) of 35 mm Hg, which we considered routine care; those randomized to hypercapnia were maintained at an ETco2 of 50 mm Hg.

The lungs of the patients in both groups were ventilated with a tidal volume of 6–8 mL/kg with a respiratory rate of approximately 10 breaths per minute. Positive end-expiratory pressure was maintained at 5 mm Hg, and peak ventilatory pressure was kept <45 mm Hg. We induced hypercapnia by removing the soda lime from the rebreathing system and increasing or decreasing the fresh gas flow as necessary to maintain the ETco2 partial pressure near 50 mm Hg. ETco2 partial pressure was continuously monitored using infrared side stream capnography.

Anesthesia was maintained with sevoflurane (1%–2%) in 80% oxygen, balance nitrogen. Sevoflurane administration was adjusted to maintain mean arterial blood pressure within 20% of the preinduction value. A supplemental bolus dose of fentanyl (100 μg) was given when heart rate or arterial blood pressure exceeded 120% of their baseline values. No vasoactive drugs were given.

In all patients, normal body weight in kg was estimated according to an average normal BMI of 22.5 kg/m2. A fluid bolus of 10 mL/kg normal body weight was given before induction of anesthesia. Subsequently, patients were given a maintenance dose of at least 10–12 mL·kg normal body weight−1·h−1. Further boluses of colloids were administered as necessary to maintain mean arterial blood pressure within 20% of baseline and to maintain urine output more than 1 mL·kg normal body weight−1·h−1. Blood loss was replaced with crystalloid at a 4:1 ratio or colloid at a 2:1 ratio. Allogenic blood was administered as necessary to maintain hematocrit more than 26%. A forced-air warmer was placed over the lower body of each patient to keep core temperature near 36°C.

Back to Top | Article Outline


Demographic data, ASA physical status, preoperative laboratory values, and type and duration of surgery were recorded. All routine anesthetic, respiratory, and hemodynamic variables during the surgery were also recorded. Detailed records of fluid management, including urine output, were kept. Inspired oxygen, end-tidal sevoflurane, and ETco2 concentrations were measured continuously during anesthesia. Oxygen saturation was measured with a pulse oximeter throughout anesthesia. Intraoperative core temperature was measured in the distal esophagus (Mon-a-therm; Tyco-Mallinckrodt Anesthesiology Products, St. Louis, MO).

After induction of anesthesia, a 20-g cannula was inserted into a radial artery. Pressure from this catheter was monitored continuously during surgery and in the postanesthesia care unit. Cardiac output was measured noninvasively using the partial rebreathing Fick method (NICO, Novametrix Medical System Inc, Wallingford, CT). Arterial blood for gas analysis was obtained intraoperatively in each patient at least once during the study during stabilized conditions approximately 1 h after the beginning of surgery.

After induction of anesthesia, a silastic tonometer was inserted into the lateral left upper arm for measurement of subcutaneous PsqO2 and temperature. The silastic tonometer consisted of 15 cm of tubing filled with hypoxic saline; 10 cm of the tubing was tunneled subcutaneously. A polarographic oxygen sensor and thermistor (Licox, Integra Life Science, CA) were inserted into the subcutaneous portion of the tonometer as previously described (19). PsqO2 was recorded at 10-min intervals during the intraoperative period, and the probe was removed before discontinuation of anesthesia.

In vitro accuracy of the oxygen sensors is ±3 mm Hg for the range from 0 to 100 mm Hg and ±5% for 100 to 360 mm Hg in a water bath at 37°C. Temperature sensitivity is 0.25%/°C. However, thermistors were incorporated into the sensors, and temperature compensation was included in the PsqO2 calculations. Oxygen sensor calibration remains stable (within 8% of baseline value for room air) in vivo for at least 8 h. The electrodes were individually factory-calibrated, but calibration was confirmed by exposing the electrode to room air (ambient Po2 of 154 mm Hg); in all cases, measurements in air were within 10% of 154 mm Hg. To exclude a significant drift of the oxygen sensor, probes were again exposed to room air after each surgery; none differed by more than 10% from its baseline value.

Forearm-minus-fingertip temperature gradient was measured as an indicator of peripheral arteriovenous shunt vasoconstriction. As in previous studies (20), we defined a gradient more than zero an indication of significant vasoconstriction.

The major purpose of our study was to test the hypothesis that mild hypercapnia increases intraoperative subcutaneous oxygen partial pressure in morbidly obese patients given 80% inspired oxygen. Based on the results of a previous study in lean volunteers (16), we estimated that hypercapnia would increase subcutaneous oxygen partial pressure by 15 mm Hg, with a standard deviation of 15 mm Hg. A sample-size estimate based on these data suggested that 13 patients per group would provide 90% power to detect a significant difference in PsqO2 at an α level of 0.05. Therefore, we studied 30 morbidly obese patients (15 per group).

Demographic and morphometric factors and major and minor outcomes in the two groups were compared with unpaired, two-tailed Student’s t-test or Fisher exact test. Intraoperative values were averaged over the entire operation for each patient, and then averaged among patients in each group. Nonparametrically distributed data were analyzed by Wilcoxon Rank Sum tests. Results are presented as means ± sd or medians (25th percentile, 75th percentile); P < 0.05 was considered statistically significant.

Back to Top | Article Outline


All 30 enrolled patients completed the study according to protocol. Demographic and morphometric factors were similar in the two groups (Table 1). Intraoperative anesthetic management was similar in each group (Table 2), with one exception: only 2 of 15 hypercapnic patients required colloids to maintain hemodynamic stability whereas 9 of 15 normocapnic patients were given colloids (P = 0.008). No blood transfusions were needed to keep the hematocrit above 30%.

Rebreathing CO2 induced mild hypercapnia without causing any adverse effects. All hemodynamic variables were comparable in the two groups (Table 3). ETco2 partial pressures were 35 ± 2 mm Hg in the normocapnic group and 52 ± 6 mm Hg in the hypercapnic group (P < 0.001). Paco2 also was significantly less in the normocapnic patients (42 ± 8 mm Hg) than in the hypercapnic patients (58 ± 7 mm Hg; P < 0.001); pH was significantly less in the hypercapnic patients (Table 3).

Table 1

Table 1

Table 2

Table 2

Table 3

Table 3

Pao2 was virtually identical in the two treatment groups. However, PsqO2 was significantly greater in patients assigned to mild hypercapnia (78 ± 31 mm Hg) than in those assigned to normocapnia (56 ± 13 mm Hg, P < 0.029, Table 3 and Fig. 1).

Figure 1.

Figure 1.

Patients in both groups were slightly vasodilated, but forearm-minus-fingertip temperature gradients were similar in the hypercapnic patients (−1.0°C ± 2.6°C) and the normocapnic patients (−1.0°C ± 2.8°C, P = 0.99).

Back to Top | Article Outline


Wound infection risk is inversely related to PsqO2 (4). Supplemental (80%) oxygen, which doubles PsqO2 in lean patients, halves the risk of surgical wound infection in this patient population (11). However, supplemental oxygen only slightly increases PsqO2 in morbidly obese patients (10). It is thus apparent that providing supplemental oxygen alone will only partially address the overwhelming infection risk that occurs in obese patients.

Our results suggest an additional strategy: we found that in morbidly obese patients having major abdominal surgery, mild intraoperative hypercapnia significantly increased PsqO2 by 22 mm Hg from 56 ± 13 to 78 ± 31 mm Hg (∼40%; P = 0.03). This is a clinically important increase, as PsqO2 near 80 mm Hg almost eliminates the risk of wound infection (4). Our results are notable in that our patients were all given 80% oxygen; our data thus indicate that mild hypercapnia provides a substantial incremental improvement in PsqO2 beyond that provided by supplemental oxygen alone.

Nearly all mechanistic, animal, and clinical evidence suggests that PsqO2 is an important determinant of infection risk. The exception is a study by Pryor et al. (21); however, this study was under-powered and had methodological problems. Thus, the results are difficult to interpret (11,22,23) and should not be considered proof that PsqO2 is unimportant.

Mild hypercapnia increases cardiac index in lean patients (16) and may induce peripheral vasodilation (17). Which factor contributes most to improved PsqO2 remains unknown. In contrast, in our study, hypercapnic and normocapnic obese patients had similar cardiac output during the study period. A possible, but unlikely, explanation is that the NICO partial rebreathing Fick system, which has not been validated in morbidly obese patients, may simply have missed a true increase. Another possibility is that cardiac output does not increase in the morbidly obese during hypercapnia, possibly because cardiac reserve is limited in these patients or they are normally hyperdynamic (24). It is also possible that cardiac output in the normocapnic patients was relatively high as a result of supplemental colloid administration. However, in the hypercapnic patients, hypercapnia itself probably increased cardiac output and, consequently, they did not need additional fluids to maintain arterial blood pressure. Thus, arterial blood pressure, heart rate, and cardiac index were comparable in each treatment group.

As expected, arterial pH was slightly less in the hypercapnic patients, but the reduction was of questionable clinical importance. There was thus little or no evidence of hypercapnia-induced toxicity in our patients.

We provided mild intraoperative hypercarbia by removing the soda lime CO2 absorbent from the circle system of the anesthesia machine. We used this approach so that minute ventilation and respiratory rate would be comparable in each treatment group. However, it is obviously easier to simply reduce the respiratory rate until the desired end-tidal Pco2 accumulates. This is presumably the approach that would be used clinically. Hypercapnia by hypoventilation also increases cardiac output by increasing venous return as a result of the decreased mean intrathoracic pressure. Because no drugs or special equipment is required, there is hardly any cost associated with inducing mild hypercapnia with either approach.

There are many ways to improve PsqO2 during surgery, including maintenance of normothermia, adequate pain treatment, regional anesthesia, and adequate intravascular volume and fluid substitution. Fluid administration in morbidly obese patients might be difficult to evaluate. However, in critical patient populations prone to wound infections, such as morbidly obese patients, a combination of all the above treatments might be necessary to adequately increase PsqO2 and, thus, decrease wound infection rate.

We studied a population of morbidly obese patients who had a BMI near 63 kg/m2. Such patients have a frequent incidence of sleep apnea and, thus, presumably normally maintain abnormally high arterial Paco2 concentrations. Therefore, an ETco2 of 50 mm Hg may have been normal for our study participants while 35 mm Hg was too low. However, these characteristics are typical of the morbidly obese population and do not reduce the validity of our results or the conclusions we draw from them.

The BMI in our hypercapnic patients was 67 kg/m2 whereas the BMI in the normocapnic patients was 60 kg/m2. This difference was not significant and is not likely to be clinically important. Furthermore, because a greater BMI reduces PsqO2, a totally equal BMI in the hypercapnic and normocapnic group would have increased the difference in PsqO2 and further strengthened our results.

In summary, intraoperative mild hypercapnia (ETco2 = 52 mm Hg) significantly increased PsqO2 by 22 mm Hg (∼40%) in morbidly obese patients having major abdominal surgery. Hypercapnia is easy to provide, involves no additional costs, and causes minimal side effects. We thus conclude that intraoperative mild hypercapnia could be a useful addition to supplemental oxygen in morbidly obese patients at risk for surgical wound infection. However, further studies are needed to evaluate the impact of our results on patient outcome.

Back to Top | Article Outline


1. Dindo D, Muller MK, Weber M, Clavien PA. Obesity in general elective surgery. Lancet 2003;361:2032–5.
2. Babior BM. Oxygen-dependent microbial killing by phagocytes (first of two parts). N Engl J Med 1978;298:659–68.
3. Gottrup F. Physiology and measurement of tissue perfusion. Ann Chir Gynaecol 1994;83:183–9.
4. Hopf HW, Hunt TK, West JM, et al. Wound tissue oxygen tension predicts the risk of wound infection in surgical patients. Arch Surg 1997;132:997–1004; discussion 5.
5. Van Esbroeck G, Gys T, Hubens A. Evaluation of tissue oximetry in perioperative monitoring of colorectal surgery. Br J Surg 1992;79:584–7.
6. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group. N Engl J Med 1996;334:1209–15.
7. Jensen JA, Goodson WH, Hopf HW, Hunt TK. Cigarette smoking decreases tissue oxygen. Arch Surg 1991;126:1131–4.
8. Sheffield CW, Sessler DI, Hopf HW, et al. Centrally and locally mediated thermoregulatory responses alter subcutaneous oxygen tension. Wound Rep Reg 1997;4:339–45.
9. Hiltebrand LB, Kaiser HA, Niedhart DJ, Kurz A. Obesity does not decrease subcutaneous tissue oxygenation [abstract]. Presented at the Annual Meeting of the American Society of Anesthesiologists. New Orleans, LA, 2005: A365.
10. Kabon B, Nagele A, Reddy D, et al. Obesity decreases perioperative tissue oxygenation. Anesthesiology 2004;100:274–80.
11. Greif R, Akca O, Horn EP, et al. Supplemental perioperative oxygen to reduce the incidence of surgical-wound infection. Outcomes Research Group. N Engl J Med 2000;342:161–7.
12. Jansson PA, Larsson A, Smith U, Lonnroth P. Glycerol production in subcutaneous adipose tissue in lean and obese humans. J Clin Invest 1992;89:1610–7.
13. de Divitiis O, Fazio S, Petitto M, et al. Obesity and cardiac function. Circulation 1981;64:477–82.
14. Cheymol G. Drug pharmacokinetics in the obese. Fundam Clin Pharmacol 1988;2:239–56.
15. Di Girolamo M, Skinner NS Jr, Hanley HG, Sachs RG. Relationship of adipose tissue blood flow to fat cell size and number. Am J Physiol 1971;220:932–7.
16. Akca O, Doufas AG, Morioka N, et al. Hypercapnia improves tissue oxygenation. Anesthesiology 2002; 97:801–6.
17. Mas A, Saura P, Joseph D, et al. Effect of acute moderate changes in PaCO2 on global hemodynamics and gastric perfusion. Crit Care Med 2000; 28:360–5.
18. Akca O, Liem E, Suleman MI, et al. Effect of intra-operative end-tidal carbon dioxide partial pressure on tissue oxygenation. Anaesthesia 2003;58:536–42.
19. Hopf HW, Viele M, Watson JJ, et al. Subcutaneous perfusion and oxygen during acute severe isovolemic hemodilution in healthy volunteers. Arch Surg 2000;135:1443–9.
20. Kurz A, Xiong J, Sessler DI, et al. Desflurane reduces the gain of thermoregulatory arterio-venous shunt vasoconstriction in humans. Anesthesiology 1995;83:1212–9.
21. Pryor KO, Fahey TJ 3rd, 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.
22. Greif R, Sessler DI. Supplemental oxygen and risk of surgical site infection (letter). JAMA 2004;291:1957.
23. Akca O, Sessler DI. Supplemental oxygen and risk of surgical site infection. JAMA 2004;291:1956–7.
24. Kanoupakis E, Michaloudis D, Fraidakis O, et al. Left ventricular function and cardiopulmonary performance following surgical treatment of morbid obesity. Obes Surg 2001;11:552–8.
© 2006 International Anesthesia Research Society