General anesthesia negatively affects pulmonary function. Decreases in muscular tone and FRC lead to closure of small airways.1,2 Absorption of oxygen trapped in alveoli distal to airway closures results in the collapse of those alveoli, a condition known as absorption atelectasis.3 Oxygen is readily absorbed by pulmonary blood flow, which accelerates the onset of alveolar collapse. Oxygen is the carrier gas most frequently used for delivery of inhalant anesthetics to veterinary patients, and this presumably contributes to absorption atelectasis during anesthesia. In both anesthetized human patients and animals, the addition of nitrogen to the carrier gas mixture such that the Fio2 is 0.3 to 0.4 delays absorption atelectasis and improves alveolar aeration.4,5 Consequently, the use of nitrogen-oxygen mixtures has been recommended as a strategy to minimize atelectasis during anesthesia.
Human patients frequently develop postanesthesia hypoxemia.6 Hypoxemia during the immediate postanesthesia period can be caused by several factors including hypoventilation (from residual effects of anesthetics, postoperative administration of sedatives and analgesics, or pain), an increase in oxygen consumption secondary to shivering, or prolonged recumbency. An impaired ventilatory response may exacerbate or perpetuate hypoxemia. In human patients, atelectasis that develops during anesthesia can persist into the postanesthesia period and continue to impair gas exchange and cause hypoxemia by contributing to venous admixture.7 Therefore, a decrease in the Fio2 during anesthesia might minimize absorption atelectasis and improve oxygen exchange during the postanesthesia period.
The purpose of the study reported here was to evaluate the effects of 2 Fio2s during anesthesia on postanesthesia Pao2 and other measures of oxygen exchange in healthy dogs. We hypothesized that the addition of nitrogen to the carrier gas mixture in an amount sufficient to decrease the intra-anesthesia Fio2 from 0.9 to 0.4 would improve the Pao2 and other indices of oxygen exchange during the first hour of anesthesia recovery for healthy dogs following ovariohysterectomy.
Materials and Methods
Animals
All study procedures were reviewed and approved by the Cornell University Institutional Animal Care and Use Committee. Twenty-two client-owned, apparently healthy adult sexually intact female dogs that were admitted to the Cornell University Hospital for Animals for ovariohysterectomy by a ventral midline celiotomy were enrolled in the study after informed written consent was obtained from the owners. The dogs ranged in body weight from 15 to 35 kg and BCS from 3 to 7 on a scale8 of 1 to 9. Dogs with a history of respiratory tract or cardiac disease or aggressive behavior and those that received sedatives or analgesics prior to hospital admission were excluded from the study.
Study design
A sample size calculation was performed with a software programa prior to study initiation. Because the primary purpose of the study was to evaluate oxygenation during the immediate postanesthesia period for dogs that were maintained at an Fio2 of 0.4 while anesthetized, compared with that for dogs that were maintained at an Fio2 of 0.9 while anesthetized, the calculation was based on assumptions regarding the Pao2. We used a Pao2 value of 99.5 mm Hg for awake (unanesthetized) dogs9 and 97.3 mm Hg for dogs recovering from anesthesia.10 We wanted to be able to detect a Pao2 difference of at least 10% between the 2 treatment groups with a power of 80% and α of 0.05 on the basis of a 2-tailed test. Results indicated that 11 dogs were necessary for each treatment group.
Prior to anesthesia induction, each dog was randomly assigned to receive either oxygen (Fio2 > 0.9 [100% oxygen]; control group) or a mixture of nitrogen and oxygen (Fio2 = 0.4; 40% oxygen group) as the carrier gas for isoflurane, the inhalant anesthetic. Randomization was performed by removing labels from an opaque envelope and conducted in blocks of 4 (ie, each envelope contained 4 labels [2 labels for each treatment]).
For each dog, food but not water was withheld for at least 8 hours before anesthesia induction. Each dog was premedicated with dexmedetomidine (5 μg/kg, IM) and hydromorphone hydrochloride (0.1 mg/kg, IM). A catheter was aseptically placed in a cephalic vein, and the assigned treatment was administered via a face mask and anesthesia machineb for 3 minutes prior to anesthesia induction. The Fio2 was measured by means of an oxygen sensor on the anesthesia workstation (galvanic cell) located at the inspiratory limb of the circular breathing circuit. The oxygen sensor was calibrated to room air every morning prior to data collection. Anesthesia was induced with propofol (2 to 6 mg/kg, IV to effect [ie, loss of palpebral and swallowing reflexes]), and the dog was intubated with an appropriately sized cuffed endotracheal tube. The dog was then positioned in dorsal recumbency, and anesthesia was maintained with isoflurane delivered in the assigned treatment, or carrier gas, at a flow rate of 2.5 L/min by means of an adult circle rebreathing system.c Each dog was allowed to breathe spontaneously throughout the procedure. A catheter was aseptically placed in a dorsal pedal artery for collection of serial blood samples, and the patient was instrumented with a multivariable anesthetic monitoring system,d which included continuous monitoring of the ECG, pulse oximetry, capnography, respiratory rate, Vt, arterial blood pressure, and esophageal temperature. The ETISO was continuously monitored after the dogs were moved into the operating room.
A ventral midline approach was used to perform an ovariohysterectomy in all dogs. The amount of isoflurane administered to maintain an appropriate plane of anesthesia and use of other drugs to affect blood pressure or heart rate were left to the discretion of the attending anesthetist. Dogs that required positive-pressure ventilation or additional analgesics (ie, opioids or α2-adrenergic receptor agonists) during the procedure were excluded from the study.
Following completion of the surgery, each dog was positioned in left lateral recumbency, disconnected from the anesthesia machine, and transported to the anesthesia recovery room. Dogs breathed ambient air after being disconnected from the anesthesia machine, but the endotracheal tube was not removed until the swallowing reflex had returned. Immediately after the endotracheal tube was removed, each dog received hydromorphone (0.1 mg/kg, IV) and carprofen (4.4 mg/kg, SC) and then was observed but left undisturbed in a cage until fully recovered from anesthesia.
Data collection
The dose of propofol required for anesthesia induction and intubation was recorded. After tracheal intubation, the end-tidal partial pressure of carbon dioxide, ETISO, respiratory rate, and Vt were recorded every 15 minutes for the duration of the surgery. The end-tidal partial pressure of carbon dioxide and ETISO were measured by means of infrared technology within the multivariable anesthesia monitoring system, which was permanently calibrated by the manufacturer. The Vt and respiratory rate were measured by means of an ultrasonic flow sensor on the anesthesia workstation located at the expiratory limb of the circular breathing circuit. The accuracy of that sensor was verified prior to data collection with graduated gastight syringese and use of volumes ranging between 60 and 500 mL. For each measurement, minute ventilation was calculated as the Vt × respiratory rate. The MAC-hours were calculated as (mean ETISO/MAC) × the number of hours the patient received isoflurane, where MAC = 1.28%.11
An arterial blood sample was obtained during closure of the linea alba and at 5, 20, and 60 minutes after extubation. For each sample acquisition, 3 mL of blood was withdrawn from the catheter in the dorsal pedal artery and discarded, then a 1-mL sample of blood was collected into a syringe that had been flushed with heparinized saline (0.9% NaCl) solution (ie, heparinized syringe) over 30 seconds and analyzed immediately with a point-of-care device.f The accuracy of that device was confirmed with control solutions obtained from the manufacturer prior to data collection and again after data had been collected from half of the study dogs. Values for Pao2 and Paco2 were corrected to contemporary esophageal temperatures during anesthesia and to contemporary rectal temperatures after anesthesia. The Pao2:Fio2 and Pao2 – Pao2 were calculated. The Pao2 was calculated as (Fio2 × [Pb – Ph2o]) – (Paco2/0.8), where Pb is the barometric pressure measured by the point-of-care device and Ph2o is the pressure of water vapor at body temperature.
At 5, 20, 30, 45, and 60 minutes after extubation, each dog was assigned a sedation score on a scale of 0 to 3. Briefly, 0 = no sedation (dog was bright, alert, and responsive), 1 = mild sedation (dog was bright, alert, and responsive and was in a sitting position or sternal recumbency), 2 = moderate sedation (dog was in lateral recumbency but responded to verbal commands [ie, raised its head when its name was called]), and 3 = deep sedation (dog was in lateral recumbency and unresponsive to verbal commands).
Data analysis
Distribution of data for all dependent variables was evaluated with the D'Agostino-Pearson omnibus test. Results were reported as the mean (SD) for variables with parametric distributions and median (range) for variables with nonparametric distributions. Comparisons between the 2 treatment (control and 40% oxygen) groups were performed with unpaired t tests for parametric variables (weight, propofol dose, isoflurane MAC-hours, Vt, and minute ventilation) and Mann-Whitney tests for nonparametric variables (age and BCS).
Oxygenation indicators were stratified into intra-anesthesia and postanesthesia values for statistical analyses. Unpaired t tests were used to compare the mean intra-anesthesia Pao2, Pao2:Fio2, Pao2 – Pao2, and Paco2 between the 2 treatment groups. The effects of treatment, sample acquisition time after extubation (time), and interaction between treatment and time on postanesthesia Pao2, Pao2:Fio2, Pao2 – Pao2, Paco2, and sedation scores were evaluated with mixed-effect models and Tukey post hoc tests. A random effect for dog was included in each model to account for repeated measures within the study subjects. Instances of hypoxemia (Pao2 < 80 mm Hg) were identified. All analyses were performed with statistical software,g,h and values of P ≤ 0.05 were considered significant.
Results
Age, weight, BCS, propofol dose required for anesthesia induction, isoflurane MAC-hours, Vt, minute ventilation, duration of anesthesia, and intra-anesthesia Pao2:Fio2 did not differ significantly between the 40% oxygen and control groups (Table 1). The mean intra-anesthesia Pao2 (P < 0.001), Pao2 – Pao2 (P < 0.001), and Paco2 (P = 0.05) for the 40% oxygen group were significantly lower than the corresponding mean values for the control group.
Summary demographic and intra-anesthesia data for 22 healthy adult female dogs that received either oxygen (Fio2 > 0.9; control group; n = 11) or a mixture of nitrogen and oxygen (Fio2 = 0.4; 40% oxygen group; 11) as the carrier gas for isoflurane while anesthetized for ovariohysterectomy.
Variable | Control group | 40% oxygen group | P value* |
---|---|---|---|
Age (y) | 1 (0.8–4) | 2 (1–5) | 0.15 |
Weight (kg) | 20.4 (4) | 21.7 (4) | 0.5 |
BCS† | 4 (4–7) | 5 (3–7) | 0.9 |
Propofol dose (mg/kg) | 2.4 (0.7) | 2.4 (0.6) | 0.9 |
Isoflurane MAC-hours (h) | 1.8 (0.3) | 1.7 (0.3) | 0.17 |
Vt (mL/kg) | 11 (2.8) | 10 (2.6) | 0.4 |
Minute ventilation (mL/kg/min) | 134 (41) | 117 (28) | 0.2 |
Duration of anesthesia (min) | 151 (28) | 136 (35) | 0.2 |
Pao2 (mm Hg) | 410 (62) | 166 (12) | < 0.001 |
Pao2:Fio2 | 427 (71) | 417 (31) | 0.6 |
Pao2 – Pao2 (mm Hg) | 181 (77) | 46 (13) | < 0.001 |
Paco2 (mm Hg) | 57 (7) | 50 (8) | 0.05 |
Values represent the mean (SD) or median (range). Arterial blood gas variables were measured during closure of the linea alba. The respective values for VT and minute ventilation represent the means for the duration of anesthesia.
Comparisons between the 2 treatment groups were performed with unpaired t tests for parametric variables (weight, propofol dose, isoflurane MAC-hours, Vt, and minute ventilation) and Mann-Whitney tests for nonparametric variables (age and BCS); values of P ≤ 0.05 were considered significant.
Body condition score was assigned on a scale of 1 to 9 as described.8
The mean postanesthesia Pao2, Pao2:Fio2, Pao2 – Pao2, and Paco2 did not differ significantly between the 40% oxygen and control groups at any time after extubation (Figure 1). For both groups, the mean Pao2 – Pao2 at 60 minutes after extubation was significantly (P = 0.04) greater than that at 5 and 20 minutes after extubation, and the mean Paco2 at 20 and 60 minutes after extubation was significantly (P < 0.001) lower than that at 5 minutes after extubation. Hypoxemia (Pao2 < 80 mm Hg) was diagnosed in 1 dog in the control group (Pao2 = 74 mm Hg) and 1 dog in the 40% oxygen group (Pao2 = 76 mm Hg) at 5 minutes after extubation, and in another dog in the control group (Pao2 = 79 mm Hg) at 20 minutes after extubation. The mean sedation score did not differ significantly (P = 0.70) between the 2 groups at any time during the postanesthesia period, although it did decrease significantly over time for both groups (Figure 2).
Discussion
Results of the present study indicated that indices of arterial oxygenation did not differ significantly between healthy dogs in which the Fio2 was maintained at > 0.9 (control group) and those in which the Fio2 was maintained at 0.4 (40% oxygen group) while anesthetized for ovariohysterectomy. Thus, the addition of nitrogen to the carrier gas for the inhalant anesthetic conferred neither an advantage nor disadvantage in regard to oxygenation during the first hour of anesthesia recovery.
Absorption atelectasis has been well described for over 50 years.3,12 Absorption of gasses trapped within alveoli contributes to the development of atelectasis, and because the rate of absorption for oxygen is greater than that for nitrogen, high concentrations of oxygen in alveolar gas accelerate alveolar collapse. Blood that perfuses collapsed alveoli is unable to exchange carbon dioxide for oxygen, which can lead to pulmonary venous admixture and systemic hypoxemia. In human patients, nitrogen is commonly used as a carrier gas to decrease Fio2 during inhalation anesthesia and minimize the adverse effects associated with absorption atelectasis, and blood gas exchange (ie, exchange of carbon dioxide for oxygen) is better for patients in whom the Fio2 is maintained at 0.3 to 0.4, compared with patients in whom the Fio2 is maintained at 1.0.13,14 Likewise, maintenance of an Fio2 of 0.4 during anesthesia helps preserve alveolar aeration and minimize atelectasis in both dogs and cats, compared with maintenance of an Fio2 of 1.0 (administration of 100% oxygen).5,15 We cannot comment on any differences in alveolar aeration during anesthesia for the dogs of the present study because thoracic imaging was not performed.
The inevitable consequence of the inspiration of nitrogen is a decrease in Fio2, which is not without risk. The FRC represents a reservoir of oxygen that is available to a patient if the supply of oxygen to the lungs is temporarily interrupted (eg, apnea or anesthesia equipment malfunction). Although the addition of nitrogen to the carrier gas to decrease the Fio2 during anesthesia may limit atelectasis, a high Fio2 (> 0.9) might enable dogs to retain a greater proportion of oxygen in their lungs for gas exchange despite the decrease in FRC (eg, atelectasis). Moreover, dogs with a high Fio2 will have greater arterial oxygen concentrations than dogs with a decreased Fio2 (0.4), as evidenced by the fact that the mean intra-anesthesia Pao2 for the control group of the present study was 2.5 times that for the 40% oxygen group. Thus, although a high Fio2 may accelerate the onset of absorption atelectasis, it may also protect the patient from hypoxemia during anesthesia.16 Addition of nitrogen (medical air) to the inspired gas can result in the inadvertent administration of a gas mixture with an oxygen concentration lower than that intended if the Fio2 is not monitored. Ideally, an anesthesia machine should be equipped to measure the Fio2 and alert anesthetists if it falls below a predetermined value. Unfortunately, not all anesthesia machines have that capability. Similarly, not all gas analyzers measure oxygen; many monitors are equipped to measure only carbon dioxide or anesthetic agents. Therefore, the ability to safely deliver oxygen-nitrogen mixtures should be critically assessed before this technique is implemented.
In human patients, atelectasis that develops during anesthesia persists into the postanesthesia period and negatively affects the Pao2 and Pao2 - Pao27; however, the association between intra-anesthesia Fio2 and postanesthesia pulmonary function remains unclear. Multiple studies17–20 have failed to detect differences in postanesthesia pulmonary function between patients with a high Fio2 (≥ 0.8) and those with a low Fio2 (0.3) during anesthesia. Additionally, in a large multicenter study21 involving 1,400 patients, the incidence of postanesthesia atelectasis, pneumonia, or respiratory failure did not differ between those that received 80% and 30% oxygen during anesthesia. Results of the present study indicated that decreasing the Fio2 did not cause any appreciable clinical benefit for dogs. In fact, if absorption atelectasis developed more frequently or to a greater extent in the control group, compared with the 40% oxygen group, that magnitude was not sufficient to affect the intra-anesthesia Pao2:Fio2 or any of the postanesthesia variables related to arterial oxygenation during the first hour of anesthesia recovery. Thus, these results raised questions regarding the risk-benefit relationship of decreasing the Fio2 during anesthesia, at least for this population of dogs.
The dogs of the present study underwent ovariohysterectomy at a veterinary teaching hospital, which can be a fairly lengthy abdominal surgery because it is typically performed by veterinary students with limited experience. The dogs were positioned in dorsal recumbency for the duration of the surgery, and positive airway pressure maneuvers designed to inflate the lungs were not performed. We thought that those conditions might promote airway closure and absorption atelectasis and thereby emphasize the disadvantages associated with a high Fio2. However, none of the oxygenation indices evaluated during the postanesthesia period differed significantly between the control and 40% oxygen groups. Three dogs developed hypoxemia (Pao2 < 80 mm Hg) during the postanesthesia period at 5 (1 dog in the control group and 1 dog in the 40% oxygen group) and 20 (1 dog in the control group) minutes after extubation; however, the condition was mild and transient and resolved without intervention for all 3 dogs. Nevertheless, the fact that hypoxemia developed in any of the study dogs underscored the importance of monitoring the Pao2 during the postanesthesia period. We have not evaluated the effects of different Fio2s during anesthesia for dogs with pulmonary disease or obese dogs; therefore, we cannot comment on those subpopulations. Oxygenation is impaired in obese dogs following sedation, likely owing to compression of the lungs and small airways by excessive body fat.22 It is unclear whether the addition of nitrogen to inspired gases during anesthesia would have a beneficial effect for obese dogs, but it almost certainly would increase the risk for hypoxemia. Until such data are available, results of the present study should be not extrapolated to other subpopulations of dogs.
In the present study, the mean Pao2 and Pao2:Fio2 did not differ between the control and 40% oxygen groups at any time during the postanesthesia period. The mean Paco2 decreased over time for both groups, which coincided with a decrease in mean sedation score. It is likely that the depressant effects of the anesthetic and analgesic agents administered to the dogs decreased over time, which in turn improved ventilation and decreased Paco2. Thus, it is likely that the decrease in Paco2, rather than a decrease in Pao2, was responsible for the increase in Pao2 – Pao2 that was observed over time during the postanesthesia period. The Pao2 – Pao2 was frequently negative for the dogs of the present study, which might have been the result of inaccuracies in the Pao2 calculation. That calculation involved use of a respiratory quotient of 0.8, which was assumed but not confirmed, and the Paco2 rather than alveolar partial pressure of carbon dioxide. Hence, calculations can have errors introduced by values that are assumed rather than measured. For that reason, and because the mean Pao2 did not differ significantly between the control and 40% oxygen groups during the postanesthesia period, we chose not to evaluate other indicators of pulmonary function, such as the shunt fraction, which involve assumptions including that for the Pao2.23
The dogs of the present study were allowed to breathe spontaneously while anesthetized, and the mean minute ventilation did not differ between the 2 treatment groups. The use of positive pressure ventilation could alter pulmonary function, and the findings of the present study should not be extrapolated to dogs that are mechanically ventilated while anesthetized. In human patients who are mechanically ventilated while in supine position (ie, dorsal recumbency), lack of spontaneous diaphragm activity appears to contribute to atelectasis formation in the dependent areas of the lungs.24 Therefore, it is possible that the severity of atelectasis in anesthetized dogs may vary if positive pressure ventilation is used.
In the present study, indices of intra-anesthesia and postanesthesia arterial oxygenation did not differ significantly between healthy dogs in which the Fio2 was maintained at > 0.9 (control group) and those in which the Fio2 was maintained at 0.4 (40% oxygen group) while anesthetized for ovariohysterectomy. Therefore, the addition of nitrogen to inspired gasses during anesthesia did not appear to have any clinical benefit for healthy dogs in regard to oxygenation during the first hour of anesthesia recovery.
Acknowledgments
The authors thank Ms. Olivia L. Tyrrell for assistance with subject recruitment.
ABBREVIATIONS
BCS | Body condition score |
ETISO | End-tidal isoflurane concentration |
Fio2 | Fraction of inspired oxygen |
FRC | Functional residual capacity |
MAC | Minimum alveolar concentration |
Pao2 | Alveolar partial pressure of oxygen |
Pao2:Fio2 | Pao2-to-Fio2 ratio |
Pao2 – Pao2 | Alveolar-arterial gradient in partial pressure of oxygen |
Vt | Tidal volume |
Footnotes
G*Power, version 3.1.7, Heinrich Heine University Düsseldorf, Düsseldorf, Germany.
Narkomed GS anesthesia machine, North American Dräger, Telford, Pa.
Corr-A-Tube reusable corrugated tubing, Teleflex Medical, Morrisville, NC.
Cardell Touch Veterinary Monitor, Midmark Corp, Versailles, Ohio.
Model S-500, Hamilton Robotics, Reno, Nev.
i-STAT 1 Analyzer, Abbott Point of Care Inc, Princeton, NJ.
Prism, version 6.05, GraphPad Software Inc, La Jolla, Calif.
JMP Pro, version 12.0.1, SAS Institute Inc, Cary, NC.
References
1. Wahba RW. Perioperative functional residual capacity. Can J Anaesth 1991;38:384–400.
2. Hedenstierna G, McCarthy G, Bergström M. Airway closure during mechanical ventilation. Anesthesiology 1976;44:114–123.
3. Burger EJ Jr, Macklem P. Airway closure: demonstration by breathing 100 percent O2 at low lung volumes and by N2 washout. J Appl Physiol 1968;25:139–148.
4. Joyce CJ, Baker AB, Kennedy RR. Gas uptake from an unventilated area of lung: computer model of absorption atelectasis. J Appl Physiol 1993;74:1107–1116.
5. Staffieri F, Franchini D, Carella GL, et al. Computed tomographic analysis of the effects of two inspired oxygen concentrations on pulmonary aeration in anesthetized and mechanically ventilated dogs. Am J Vet Res 2007;68:925–931.
6. Moller JT, Johannessen NW, Berg H, et al. Hypoxaemia during anaesthesia—an observer study. Br J Anaesth 1991;66:437–444.
7. Lindberg P, Gunnarsson L, Tokics L, et al. Atelectasis and lung function in the postoperative period. Acta Anaesthesiol Scand 1992;36:546–553.
8. Laflamme D. Development and validation of a body condition score system for dogs. Canine Pract 1997;22(4):10–15.
9. Haskins S, Pascoe PJ, Ilkiw JE, et al. Reference cardiopulmonary values in normal dogs. Comp Med 2005;55:156–161.
10. Campbell VL, Drobatz KJ, Perkowski SZ. Postoperative hypoxemia and hypercarbia in healthy dogs undergoing routine ovariohysterectomy or castration and receiving butorphanol or hydromorphone for analgesia. J Am Vet Med Assoc 2003;222:330–336.
11. Steffey EP & Howland D Jr. Isoflurane potency in the dog and cat. Am J Vet Res 1977;38:1833–1836.
12. Nunn JF, Payne JP. Hypoxaemia after general anaesthesia. Lancet 1962;2:631–632.
13. Agarwal A, Singh PK, Dhiraj S, et al. Oxygen in air (FIO2 0.4) improves gas exchange in young healthy patients during general anesthesia. Can J Anaesth 2002;49:1040–1043.
14. Rothen HU, Sporre B, Engberg G, et al. Prevention of atelectasis during general anaesthesia. Lancet 1995;345:1387–1391.
15. Staffieri F, De Monte V, De Marzo C, et al. Effects of two fractions of inspired oxygen on lung aeration and gas exchange in cats under inhalant anaesthesia. Vet Anaesth Analg 2010;37:483–490.
16. Gleed RD, Dobson A. Improvement in arterial oxygen tension with change in posture in anaesthetised horses. Res Vet Sci 1988;44:255–259.
17. Staehr AK, Meyhoff CS, Henneberg SW, et al. Influence of perioperative oxygen fraction on pulmonary function after abdominal surgery: a randomized controlled trial. BMC Res Notes 2012;5:383.
18. Joyce CJ, Baker AB. Effects of inspired gas composition during anaesthesia for abdominal hysterectomy on postoperative lung volumes. Br J Anaesth 1995;75:417–421.
19. Joyce CJ, Baker AB, Chartres S. Influence of inspired nitrogen concentration during anaesthesia for coronary artery bypass grafting on postoperative atelectasis. Br J Anaesth 1995;75:422–427.
20. Akça O, Podolsky A, Eisenhuber E, et al. Comparable postoperative pulmonary atelectasis in patients given 30% or 80% oxygen during and 2 hours after colon resection. Anesthesiology 1999;91:991–998.
21. Meyhoff CS, Wetterslev J, Jorgensen LN, et al. Effect of high perioperative oxygen fraction on surgical site infection and pulmonary complications after abdominal surgery: the PROXI randomized clinical trial. JAMA 2009;302:1543–1550.
22. Mosing M, German AJ, Holden SL, et al. Oxygenation and ventilation characteristics in obese sedated dogs before and after weight loss: a clinical trial. Vet J 2013;198:367–371.
23. Cruz JC, Metting PJ. Understanding the meaning of the shunt fraction calculation. J Clin Monit 1987;3:124–134.
24. Bynum LJ, Wilson JE III, Pierce AK. Comparison of spontaneous and positive-pressure breathing in supine normal subjects. J Appl Physiol 1976;41:341–347.