Background: One lung ventilation (OLV) results in a ventilation-perfusion (V/Q) mismatch. Protective lung ventilation (PLV) reduces postoperative pulmonary complications following OLV. However, PLV predisposes to areas of atelectasis in the ventilated lung and worsens the V/Q mismatch. Aim of Study: To evaluate the gradient between arterial carbon dioxide tension (PaCO2) and partial pressure of end-tidal carbon dioxide gas (ETCO2) during OLV using PLV. The second objective was to see if a high gradient could be predicted based on preoperative pulmonary function tests (PFTs), American Society of Anesthesiologists Physical Status (ASA-PS) or intraoperative haemodynamic changes. Patients and Methods: The PaCO2 and ETCO2 during two lung ventilation (TLV) and OLV were noted with patient in the lateral position. The PaCO2-ETCO2 gradients during TLV and OLV were calculated. The mean values of PaCO2, ETCO2 and PaCO2-ETCO2 gradient were compared for OLV and TLV. For gradients above 8 mm Hg, PFT, ASA-PS grade and blood pressure were assessed to identify any clinical association. Results: Sixty patients were enrolled in the study. The mean values of PaCO2 were 38.17 and 44.02 mm Hg during TLV and OLV respectively. The mean values of ETCO2 were 31.31 and 34.53 mm Hg during TLV and OLV respectively. The mean PaCO2-ETCO2 gradient was 6.74 and 9.71 mm Hg during TLV and OLV respectively. These values were significantly lower during TLV than OLV. Conclusion: ETCO2 does not correspond with PaCO2 during OLV using PLV. It is not possible to predict which patients will show a higher PaCO2-ETCO2 gradient. This study could not find any clinical association between the preoperative PFT, ASA-PS grade or intraoperative haemodynamics when PaCO2-ETCO2 gradient was greater than 8 mm Hg.
Keywords: One lung ventilation, protective lung ventilation, ventilation-perfusion mismatch
|How to cite this URL:|
Pendyala SC, Date AA. Influence of protective lung ventilation on arterial-to-end tidal carbon dioxide gradient during one lung ventilation: A prospective observational study. Airway [Epub ahead of print] [cited 2023 Feb 1]. Available from: https://www.arwy.org/preprintarticle.asp?id=360437
| Introduction|| |
One lung ventilation (OLV) is routinely used to facilitate surgical access during thoracic surgery. However, OLV is known to result in ventilation-perfusion (V/Q) mismatch since the nonventilated lung continues to get perfused. The use of a higher fraction of inspired oxygen (FIO2) and higher tidal volume in an attempt to maintain adequate gas exchange can result in acute lung injury. Protective lung ventilation (PLV) strategies have been shown to reduce acute lung injury, postoperative pulmonary complications and hospital stay following OLV.
PLV strategy uses low tidal volumes to prevent over distension of alveoli and barotrauma. However, this is associated with areas of atelectasis in the ventilated lung, adding to the existing V/Q mismatch. Positive end expiratory pressure (PEEP) and recruitment manoeuvres are used to prevent atelectasis in the ventilated lung. The concept of PLV was initially introduced for the management of patients with acute respiratory distress syndrome in critical care units. Permissive hypercapnia which happens during PLV can lead to an increase in pulmonary artery pressure. In addition, carbon dioxide insufflation used during minimally invasive thoracic surgery to aid lung collapse can worsen hypercarbia. Therefore, it is important to use a reliable indicator of arterial CO2 tension during OLV.
Various studies have found that the end-tidal CO2 (ETCO2) does not correlate with arterial CO2 tension (PaCO2) during OLV.,,,, However, the difference between the ETCO2 and PaCO2 while using PLV during OLV has not yet been evaluated.
We expect that the use of PLV strategies during OLV may be associated with a greater V/Q mismatch and ETCO2 may not be an accurate indicator of PaCO2. We therefore measured the PaCO2 and ETCO2 during two lung and OLV to assess the correlation between the two. We also evaluated whether a higher gradient between PaCO2 and ETCO2 could be predicted based on preoperative pulmonary function tests (PFT), American Society of Anesthesiologists Physical Status (ASA-PS) grading or any intraoperative haemodynamic factors.
| Patients and Methods|| |
This was a prospective, single-centre, observational study. The study protocol was approved by the Institutional Scientific and Ethics Board Committee (ISEB Code C-3/18/2017). The need for patient consent was waived due to the observational nature of the study. Adult patients who were scheduled to undergo surgery lasting longer than 30 min under general anaesthesia with OLV in the lateral position were included in the study. Patients who were posted for emergency surgery, those who were to undergo OLV in the supine position and pregnant women were excluded from the study. During OLV, V/Q mismatch is found to be worse when patients are in the supine position compared to when they are in the lateral position. This is due to the effect of gravity which improves the perfusion to the dependent lung which is also the ventilated lung.
To calculate the required sample size, the difference between PaCO2 and ETCO2 was considered to be 9.7 ± 4.5 mm Hg for two lung ventilation (TLV) and 9 ± 5.2 mm Hg for OLV. Considering the expected mean difference to be 0.2 mm Hg with a standard deviation of 0.5 mm Hg at a 5% significance level and 85% power, the required sample size was calculated to be 60 patients.
All patients underwent preoperative evaluation and premedication as per the departmental policy. Demographic details of the patients were noted along with their ASA-PS grading. Blood investigations, radiological reports, echocardiography and PFT parameters (such as forced vital capacity [FVC], forced expiratory volume in the 1st second [FEV1], FEV1/FVC and diffusing capacity of the lung for carbon monoxide) were recorded. Patients were kept fasting for 6 h before surgery. In the operating room, electrocardiogram, noninvasive blood pressure and pulse oximetry monitoring were established and basal parameters were recorded. Intravenous (IV) access was secured and a balanced salt solution was started. Glycopyrrolate (0.004 mg/kg), midazolam (0.02 mg/kg) and fentanyl (2 μg/kg) were administered IV. After preoxygenation, anaesthesia was induced using IV propofol until the loss of verbal contact. Mask ventilation was confirmed and IV atracurium (0.6 mg/kg) was administered. A lung isolation device (double lumen tube or bronchial blocker) was inserted, and lung isolation was checked by the anaesthesiologist using auscultation and fibreoptic bronchoscopy or fluoroscopy. The radial artery was cannulated for invasive monitoring of blood pressure and sampling of arterial blood gases (ABG). Capnography and oxygen analyser were also used to monitor the ETCO2 and FIO2 respectively.
PLV was instituted with pressure-controlled ventilation to achieve a tidal volume of 6–7 mL/kg of actual body weight with a peak pressure not exceeding 35 cm H2O. The FIO2 was initially set to 0.5 for all patients with a PEEP of 4–5 cm H2O. After placing the patient in the lateral position, ABG analysis was done while the patient was on TLV. Lung isolation was rechecked and the time of initiation of OLV was noted. ABG analysis was done again 30 min after starting OLV. Pulse rate, blood pressure and ETCO2 at the time of ABG sampling were noted at both times. Hypoxic pulmonary vasoconstriction (HPV) is expected to set in within 30 min of initiating OLV, thereby improving perfusion to the ventilated lung and reducing the V/Q mismatch. Therefore, the ABG sample was collected 30 min after initiating OLV.
Protocols were in place to manage desaturation. In case of desaturation, the position of the lung isolation device was rechecked, blood pressure checked to ensure that it is normal, a recruitment manoeuvre was applied and if needed, TLV was resumed. PEEP to the ventilated lung was increased and continuous positive airway pressure was applied to the nonventilated lung. The values of PaCO2 and ETCO2 were noted and the difference between the two was calculated during both TLV and OLV. The mean values of the PaCO2, ETCO2 and the PaCO2–ETCO2 gradient were calculated for OLV and TLV.
For the data showing normal distribution, the statistical significance of the results was calculated using paired Student's t-test. For data that were not normally distributed, the Wilcoxon matched-pairs test was used for the comparison of the data. Differences were considered statistically significant when P < 0.05. In patients with PaCO2-ETCO2 gradient >8 mm Hg, the ASA grade, PFT and blood pressure at the time of sample collection were assessed to identify any clinical association.
| Results|| |
A total of 60 patients were enrolled in the study (44 male and 16 female). The mean age of the patients was 54.57 ± 13.01 years. Of these, nine patients belonged to ASA-PS 1, 40 to ASA-PS 2 and 11 to ASA-PS 3. The surgeries performed are summarised in [Table 1]. The results of PFTs as a fraction of predicted values are given in [Table 2].
The mean ETCO2 during OLV and TLV are shown in [Table 3]. ETCO2 was found to be significantly lower during TLV as compared to OLV (P < 0.001, Wilcoxon matched pairs test). The mean PaCO2 was also lower during TLV as compared to OLV (P < 0.001, Wilcoxon matched pairs test). The mean PaCO2-ETCO2 gradient during TLV was found to be significantly lower than during OLV (P < 0.001, Wilcoxon matched pairs test). The PaCO2-ETCO2 gradient noted on OLV and TLV was statistically evaluated using the Pearson's correlation coefficient [Figure 1]. Pearson's correlation coefficient (r) was noted to be 0.21 (P = 0.11) which was not statistically significant.
|Figure 1: Correlation between PaCO2-ETCO2 gradient during OLV and TLV. OLV: One lung ventilation, TLV: Two lung ventilation|
Click here to view
|Table 3: Comparison of mean arterial carbon dioxide tension, end tidal carbon dioxide, arterial carbon dioxide tension-end tidal carbon dioxide gradient during one lung ventilation and two lung ventilation|
Click here to view
The PaCO2-ETCO2 gradient under anaesthesia is expected to be up to 5 mm Hg. In this study, we found that the PaCO2-ETCO2 gradient was >5 mm Hg in 38 patients during TLV and 49 patients during OLV. Patients with a PaCO2-ETCO2 gradient >8 mm Hg were individually studied to evaluate if they had poor preoperative PFT, higher ASA grade or low intraoperative blood pressure. This study could not identify any of these clinical factors to be associated with a greater PaCO2–ETCO2 gradient.
| Discussion|| |
This study aimed to evaluate the gradient between PaCO2 and ETCO2 during OLV using PLV. It can be seen from this study that ETCO2 does not truly reflect PaCO2 during OLV using PLV. The mean PaCO2-ETCO2 gradient during TLV and OLV was 6.74 ± 3.59 mm Hg and 9.71 ± 5.75 mm Hg, respectively. This gradient was significantly greater (P < 0.05) during OLV than during TLV. Various other groups have also made similar observations as summarised in [Table 4].,,, However, these studies were not conducted using PLV strategies. Ip Yam et al. used a high tidal volume of 10 mL/kg and observed no difference in the PaCO2-ETCO2 gradient during TLV and OLV. We expected that the use of low tidal volume during PLV would worsen the V/Q mismatch created by OLV and therefore increase the PaCO2-ETCO2 gradient.
|Table 4: Comparison of mean arterial-to-end tidal carbon dioxide gradient during one lung ventilation and two lung ventilation with values reported in previous studies|
Click here to view
The PaCO2-ETCO2 gradient is <5 mm Hg in healthy adults undergoing general anaesthesia. Razi et al. have found a good correlation between ETCO2 and PaCO2 during different modes of ventilation. This gradient has been shown to increase when there is an increase in dead space ventilation or intrapulmonary shunting. OLV results in a right-to-left intrapulmonary shunt as the lung that is not ventilated continues to get perfused. It is estimated that approximately 80% of pulmonary blood flow gets diverted to the ventilated lung due to HPV. After induction of anaesthesia and positioning, areas of atelectasis develop in the ventilated lung due to the compression of the dependent lung by the mediastinum, the abdominal contents and surgical manoeuvres. The use of low tidal volumes aggravates atelectasis in the ventilated lung. This may cause an increase in the intrapulmonary shunting of blood, resulting in a greater PaCO2-ETCO2 gradient. The mean PaCO2 and ETCO2 were high during OLV as compared to TLV. Despite the fact that HPV is expected to set in after 30 min of initiating OLV, carbon dioxide remained higher during OLV than during TLV.
In this study, we found that 11 patients (18%) had a PaCO2-ETCO2 gradient <5 mm Hg during OLV, while 22 (36%) had a PaCO2-ETCO2 gradient <5 mm Hg during TLV. In the study conducted by Tobias, 6 out of 15 patients (40%) were found to have a PaCO2-ETCO2 <5 mm Hg during OLV, while this number was 13 (86.6%) during TLV. Our study had a greater number of patients with a PaCO2-ETCO2 gradient >5 mm Hg that may be attributed to the use of PLV during both TLV and OLV.
It is evident from the present study and those of others that ETCO2 cannot be used to titrate ventilatory settings during OLV in most patients. However, it would be helpful to identify a parameter that could be used to predict which patients are likely to show a higher PaCO2-ETCO2 gradient during OLV.
Ali et al. found that the ETCO2 is not reflective of the changes in the PaCO2 during controlled hypotensive anaesthesia. As the pulmonary blood flow decreases due to the fall in the cardiac output, there is a V/Q mismatch which increases the amount of shunted blood. We observed the effect of changes in blood pressure on the PaCO2-ETCO2 gradient. Although we hypothesised that low blood pressure may be associated with a higher PaCO2-ETCO2 gradient, such a clinical association was not seen during TLV or OLV.
Wahba and Tessler published a review of the correlation between ETCO2 and PaCO2 during anaesthesia. They reported that patient factors such as age, smoking, ASA physical status, OLV and change in the cardiac output are associated with a high PaCO2-ETCO2 gradient. They also highlighted that the gradient determined initially will not be constant throughout the surgery. Therefore, ETCO2 does not accurately reflect the trend in PaCO2. Fletcher and Jonson reported that increasing the tidal volume without changing the frequency of ventilation can reduce the PaCO2-ETCO2 gradient from 4.5 to 2.5 mm Hg, implying that lower tidal volumes result in a higher PaCO2-ETCO2 gradient. Cox and Tobias found that patients with underlying lung disease had a significantly greater PaCO2-ETCO2 gradient as compared to those with normal pulmonary function.
Among the patients who had a PaCO2-ETCO2 gradient >8 mm Hg during either OLV or TLV, no clinical association was observed with a higher ASA-PS grade, poor PFT or low intraoperative blood pressure measurements. Therefore, we could not identify any factors to help us predict which patients are likely to have a greater PaCO2-ETCO2 gradient during OLV.
Yamauchi et al. have shown that the PaCO2-ETCO2 gradient is affected by FIO2 in anaesthetised and ventilated patients. They proposed that this may be attributed to the increase in alveolar dead space, caused by the high FIO2 which results in a reduction in vascular resistance. This causes the redistribution of blood from the areas of low perfusion to the well perfused areas. Other mechanisms postulated are absorption atelectasis and a fall in cardiac output caused by high FIO2. All these factors increase intrapulmonary shunt, thus increasing the PaCO2-ETCO2 gradient. A limitation of our study was that the FIO2 level was not evaluated while assessing the PaCO2-ETCO2 gradient. Although the initial FIO2 was 0.5, it was increased during the surgery as per the individual requirement. Another limitation of the study was that only one ABG reading was taken during OLV. It may have been possible to observe the trends in the gradient if multiple samples were analysed throughout the course of OLV.
| Conclusion|| |
To conclude, our study found that ETCO2 does not correlate with PaCO2 during OLV using protective lung ventilatory strategies. We also found that it was not possible to predict preoperatively which patients are likely to show a higher PaCO2-ETCO2 gradient during OLV. Therefore, repeated ABG analysis or transcutaneous CO2 monitoring may still be considered more reliable indicators for adjusting ventilatory settings during OLV. Transcutaneous CO2 monitoring cannot be used as an alternative to ETCO2 since capnography is also used to identify other problems such as tube misplacement, circuit disconnection and pulmonary embolism. It may, however, be used as an additional monitor.
We thank Ms Amrita Date for her help in preparation of the manuscript.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Şentürk M, Slinger P, Cohen E. Intraoperative mechanical ventilation strategies for one-lung ventilation. Best Pract Res Clin Anaesthesiol 2015;29:357-69.
Marret E, Cinotti R, Berard L, Piriou V, Jobard J, Barrucand B, et al.
Protective ventilation during anaesthesia reduces major postoperative complications after lung cancer surgery: A double-blind randomised controlled trial. Eur J Anaesthesiol 2018;35:727-35.
Grande B, Ganter MT. What is the best strategy for one-lung ventilation during thoracic surgery? J Thorac Dis 2018;10:6404-6.
Lee MS, Lee YH, Kim HK. The relationship between end tidal and arterial carbon dioxide tension during endobronchial anesthesia. Korean J Anesthesiol 1993;26:478-82.
Tobias JD. Noninvasive carbon dioxide monitoring during one-lung ventilation: End-tidal versus transcutaneous techniques. J Cardiothorac Vasc Anesth 2003;17:306-8.
Cox P, Tobias JD. Noninvasive monitoring of PaCO (2) during one-lung ventilation and minimal access surgery in adults: End-tidal versus transcutaneous techniques. J Minim Access Surg 2007;3:8-13.
Ip Yam PC, Innes PA, Jackson M, Snowdon SL, Russell GN. Variation in the arterial to end-tidal PCO2 difference during one-lung thoracic anaesthesia. Br J Anaesth 1994;72:21-4.
Jin HC, Seo JW, Kim SH, Chae WS, Lee JS, Kim YI. Is end-tidal carbon dioxide tension useful to predict arterial carbon dioxide tension during one lung ventilation? − A comparison with during two lung ventilation. Korean J Anesthesiol 2008;54:609-13.
Slinger P, Campos J. In: Gropper M, editor. Miller's Anesthesia. 9th
ed. Philadelphia: Elsevier; 2020. p. 1648-716.
Razi E, Moosavi GA, Omidi K, Khakpour Saebi A, Razi A. Correlation of end-tidal carbon dioxide with arterial carbon dioxide in mechanically ventilated patients. Arch Trauma Res 2012;1:58-62.
Satoh K, Ohashi A, Kumagai M, Sato M, Kuji A, Joh S. Evaluation of differences between PaCO2 and ETCO2 by age as measured during general anesthesia with patients in a supine position. J Anesthesiol 2015;2015:710537.
Wahba RW, Tessler MJ, Béïque F, Kleiman SJ. Changes in PCO2 with acute changes in cardiac index. Can J Anaesth 1996;43:243-5.
Ali SS, Dubikaitis A, al Qattan AR. The relationship between end tidal carbon dioxide and arterial carbon dioxide during controlled hypotensive anaesthesia. Med Princ Pract 2002;11:35-7.
Wahba RW, Tessler MJ. Misleading end-tidal CO2 tensions. Can J Anaesth 1996;43:862-6.
Fletcher R, Jonson B. Deadspace and the single breath test for carbon dioxide during anaesthesia and artificial ventilation. Effects of tidal volume and frequency of respiration. Br J Anaesth 1984;56:109-19.
Yamauchi H, Ito S, Sobue K. Dependence of the gradient between arterial and end-tidal Pco2 on the fraction of inspired oxygen. Masui 2012;61 Suppl: S159-64.
Aparna Avinash Date,
Department of Anesthesiology, Kokilaben Dhirubhai Ambani Hospital and Medical Research Institute, Four Bunglows, Andheri (West), Mumbai - 400 053, Maharashtra
Source of Support: None, Conflict of Interest: None
[Table 1], [Table 2], [Table 3], [Table 4]