Follow Capnography / Bhavani Shankar Kodali     Twitter  

A Comprehensive Educational Website (established 1998)

Designed, Produced, and maintained (Edition 10, March 2019) by 

Bhavani Shankar Kodali MD

increased phase III

The slope of the phase III is increased and phase II is prolonged.  After the treatment, there is normalization of the capnogram. A similar picture results in children when a side stream analyzer with a longer response time is used.

Capnography during thoracic anesthesia

Hemodynamic effects of carbon dioxide insufflation during thoracoscopy.

Bhavani Shankar Kodali MD


As more complex thoracoscopic procedures are performed, adequate exposure becomes increasingly more important. The insufflation of CO2 has been demonstrated to aid in the compression of lung parenchyma and the effacement of subpleural lesions, and to act as a retractor when combined with changes in patient position. However, a recent study demonstrated that CO2 insufflation during thoracoscopy in the pig had adverse hemodynamic consequences. Thirty two patients undergoing thoracoscopy were studied to evaluate the effects of CO2 insufflation in the clinical setting.1 The end-tidal CO2 pressure, arterial oxygen saturation, mean arterial pressure, heart rate, and central venous pressure were monitored. Measurements were determined at baseline, at the initiation of one-lung ventilation, and at intrapleural pressures of 2 to 14 mm Hg. The authors found that the insufflation of CO2 of 2 to 14 mm Hg had no significant effect on the end-tidal CO2 pressure, arterial oxygen saturation, heart rate, or mean arterial pressure, but the central venous pressure did rise from 7.00 +/- 1.5 mm Hg to 17.30 +/- 2.53 mm Hg (p < 0.05). It is concluded from this that the insufflation of CO2 during thoracoscopy does not have adverse hemodynamic effects in the clinical setting. Therefore, a low-pressure (< 10 mm Hg) insufflation is a safe adjunct to the conduct of routine thoracoscopic surgical procedures.

Peden and Prys-Roberts2 studied 10 patients undergoing laparoscopic surgical technique for thoracic and cervical dissection of the esophagus during esophagogastrectomy. Right lung was collapsed using a double-lumen bronchial tube and carbon dioxide was insufflated into the right pleural cavity to compress the lung. Changes in hemodynamic and respiratory variables occurred. In the majority of patients airway pressure and end-tidal CO2 increased, despite alterations in ventilation. If five patients, systolic blood pressure decreased suddenly by between 15 and 35 mm Hg. In four patients, SPO2 decreased to 91% or less despite an FIO2 of 1.0. If CO2 was insufflated too fast, or the lung failed to deflate adequately, the clinical picture was that of a tension pneumothorax. One patient developed surgical emphysema and contralateral pneumothorax.

Hemodynamic instability and hypoxia can be caused by the increase in intrathoracic pressure resulting from rapid or excessive insufflation of CO2, or failure of the lung to deflate adequately in response to compression by CO2. The clinical picture which can result is that of tension pneumothorax. If this occurs the endoscope should be withdrawn immediately, the CO2 released and the patient stabilized before further cautious reinsufflation of CO2. To avoid this occurrence the CO2 should be insufflated by the surgeons at as slow rate as possible to produce the desired compression of the lung. Peden and Prys-Roberts2 suggest to ventilate lungs with 100% oxygen before deflation of the lung, and another source of oxygen is available to insufflate at 1 l.min-1 into the deflated lung, should unacceptable desaturation still occur. The lumen of the bronchial tube must be opened to air before CO2 insufflation, to allow the lung to deflate as it is compressed.

1. Wolfer RS Krasna MJ, Hasnain JU, McLaughlin JS. Hemodynamic effects of carbon dioxide insufflation during thoracoscopy. Ann Thorac Surg 1994;58(2):404-7.

2. peden CJ, Prys-Roberts C. Capnothorax: Implications for the anaesthetist. Anaesthesia 1991;48(8):664-6.



Capnography during thoracic anesthesia

Capnographic waveforms seen during thoracic anesthesia.

Bhavani Shankar Kodali MD

Abnormal CO2 waveforms that may be seen during thoracic anesthesia include:


1). Capnograms withincreased phase III slope due to a large spread V/Q ratios, as in lung disease. The initial part of the slope is represented by areas which are well ventilated with high V/Q ratios (i.e., decreased CO2 concentration), while the latter part is represented by areas which are poorly ventilated and with low V/Q ratios (i.e., increased CO2 concentration).


2.)Biphasic capnogram: ‘Phase III’ of the capnogram represents mixed alveolar gas at the CO2 sampling site. Therefore, if the lungs have distinctly different V/Q ratios and exhalation time constants, then a ‘Biphasic’ waveform can be seen, as in lateral decubitus position.1 The upper lung (non-dependent) has a low airway resistance, high V/Q ratio (secondary to gravity dependent blood flow) and a low CO2 concentration compared with the lower, dependent lung. The earlier part of the biphasic CO2 waveform is due to the expired gases from the upper lung containing lower PCO2 and the later part of the biphasic waveform is predominantly due to the expired gases containing high PCO2 from the lower lung.

A similar capnogram can occur following a single lung transplant Some patients with COPD may also display a slight biphasic expiratory plateau if they have, throughout both lungs, two distinct populations of alveoli with very different time constants. In this situation, rapidly exchanging alveolar spaces are overinflated during inspiration (their compliance is high) so that their CO2 concentration is low, whereas slower exchanging alveoli empty only during the later part of exhalation, releasing a higher CO2 content.1,2


3.) Reverse phase III capnogram: Occasionally seen in patients with emphysema. The slope of phase III can be reversed in patients with emphysema where there is marked destruction of alveolar capillary membranes and reduced gas exchange.


1. Benumof JL. Anesthesia For Thoracic Surgery. 2nd edition. W B Saunders company, 1995;245-50.

2. Carlon GC, Ray c, Miodownik S, Kopec I, Groeger JS. Capnography in mechanically ventilated patients. Crit Care Med 1988;67:579-81.

previous next




Capnography during thoracic anesthesia


 Evaluation of ventilation/perfusion status of each lung

(Dual lung capnography)

Bhavani Shankar Kodali MD

Conventional capnography samples mixed expired gases, ie., gases from both the lungs. When there is a pathophysiologic defect in one lung, this approach would not be useful. Dual capnography overcomes this limitation by sampling (measuring) gases from each lumen of double-lumen tube (DLT).1,2

Conventional method of ETCO2 monitoring

Dual lung capnography



Dual capnography can be achieved by using two end-tidal CO2 monitors to analyze CO2 wave forms from each lumen of a DLT.1 Since using two monitors to measure ET-CO2 in the operating room may be inconvenient and not practical, a device as suggested by Bhavani-Shankar et al,2 can be used as shown in the figure. It has a three way stopcock with connections to the two CO2 sampling adapters that are interposed between each lumen and its corresponding limb on the Y-double lumen tube adapter. This allows gas sampling from either each lung or two lungs simultaneously. The standard method of PETCO2 sampling via DLT (capnograms of individual lung obtained via interruption of ventilation to contralateral lung) can also reveal the ventilation/perfusion status of lung. However, dual capnography allows one to assess the ventilation/perfusion status of each lung in individuals who cannot tolerate one-lung ventilation long enough to reach a new steady state.


In the absence of a DLT, a FOB (fiber optic bronchoscope) can be used to obtain PETCO2 samples from each lung.3 This technique allows one to assess the ventilation/perfusion status of each lung in patients with or without an endotracheal tube. Although there is no published comparison of this approach and standard dual capnographic techniques, this method, at least theoretically, may be useful to determine the PETCO2 of expired gases from each lung. It is possible that bronchial sampling via FOB from one lung could be contaminated with expired gases from the contralateral lung, particularly when expired flow rate decreases to less than 150 ml/min (sampling flow rate of capnograph). Such contamination, however, could be detected by the shape of capnograms. Contamination is an unlikely factor in the presence of a normal shape of CO2 waveforms from each lung.4 Therefore under these circumstances, a difference between the heights of the alveolar plateau of CO2 waveforms from each lung suggests a difference between the V/Q status lungs.

Uses of Dual capnography:

1.Bronchospy is probably the best method to verify correct placement of double-lumen tubes. However, capnography can also be used to confirm correct DLT placement, detect malposition, unilateral obstruction or accidental displacement of a DLT.1
  Normal Misplaced Normal
Bronchial-lumen capnogram /Thoracic/normaldlt.gif /Thoracic/dltnormal.gif /Thoracic/normaldlt.gif
Tracheal-lumen capnogram /Thoracic/normaldlt.gif

One of the advantages that is inherent to capnography is that it is a non invasive continuous monitor. Correct placement of double-lumen tubes can be checked by analyzing the CO2 wave form from each lung and also during clamping and unclamping procedures of each lung. Further, the CO2 wave forms can be examined periodically from each lung. A change in end-tidal concentration or CO2 wave form could give early warning of a misplaced double-lumen tube or inadequate ventilation and CO2 elimination from the lungs.

2. Detection of perfusion defect-secondary to thromboembolism, tumor, hemothorax:- Bhavani-Shankar et al3 have reported a case where a patient scheduled for right sided pneumonectomy was unable to tolerate OLV due to hypoxemia and arterial blood gas analysis showed a large PaCO2-PETCO2 gradient. A dual capnography revealed the PaCO2-PETCO2 gradient more than two fold greater in the left than in the right lung. Surgery was restricted to pleurectomy and a further work-up later via echocardiography and pulmonary angiogram confirmed the presence of a large thrombus in the left pulmonary artery.

In another report,5 information obtained via dual capnography allowed the anesthesia team to render an informed opinion to the surgical team regarding the status of the flow in a major pulmonary artery, after the surgeons expressed concern about
accidentally stapling the vessel during a technically complex thoracic
operation. Despite the use of one lung ventilation (right
lung), the left pulmonary artery proved difficult to visualize, and the
possibility arose that this vessel had been accidentally occluded (stapled)
during the wedge resection. The anesthesiologists hypothesized that if such occlusion had indeed occurred, it would produce a major unilateral perfusion deficit, characterized by a marked decrease in CO2 delivery to the left lung, and a correspondingly low end tidal CO2 from the left lung that could be detected using dual capnography. Double lung ventilation was commenced using the set up described above to measure and record CO2 waveforms from each lung during
two lung ventilation.3 The end-tidal PCO2 values of the individual lungs
were found to be similar (36 mmHg on left, vs 38 mmHg on right), and the
waveforms from the two lumens of DLT were similar. Therefore it was interpreted
from these findings that there was no major blood flow limitation in the left main pulmonary artery. Confirming this hypothesis was a subsequent intraoperative Doppler study which showed that the left pulmonary artery was indeed patent with normal flow velocity.
 Conventional method of ETCO2 monitoring Dual lung capnography
   Note decreased PETCO2 and increased PaCO2-PETCO2 from right lung

3.Other uses includes- To explain abnormal wave forms on capnograms. Example: Biphasic capnogram wave forms are seen in patients with single lung transplant.6 Dual capnography would have confirmed the etiology of biphasic capnogram as it could have isolated the capnograms of each lung thereby showing the individual characteristics of respective lungs.


1. Shafieha M.J, Sit.J, Kartha.R et al- End Tidal CO2 analyzers in proper positioning of the double lumen tubes. Anesthesiology 1986; 64: 844.

2. Shankar K.B, Moseley H.S.L, Kumar A.Y.- Dual end tidal CO2 monitoring and double lumen tubes. Can. J. Anaesth 1992; 39: 100-1.

3. Shankar K.B, Russell Roger, Aklog Lishon, Mushlin Phillip- Dual capnography facilitates detection of a critical perfusion defect in an individual lung. Anesthesiology, Vol 90 (1), Jan 1999, 302-304.

4. Brampton WJ, Watson RJ. Arterial to end-tidal carbon dioxide tension difference during laparoscopy. Magnitude and effect of anaesthetic technique. Anaesthesia. 1990 Mar;45(3):210-4.

5. Russell R, Bhavani Shankar K, Mushlin PS. Another application of dual capnography. Anesthesiology 2000; 92:1:288-9.
6. Lynne Williams, W.Scott Jellish, Paul A. Modica et al – Capnography in a patient after a single lung transplant. Anesthesiology Vol 74, No.3, Mar1991, 621-622.

previous next


Capnography during thoracic anesthesia

Capnography as a non invasive-monitor of PaCO2 during thoracic anesthesia

Bhavani Shankar Kodali MD

The arterial to end-tidal PCO2 difference is dependent on the underlying pulmonary disease. The most common reason for thoracotomy in adults is surgical treatment of lung cancer. Patients with lung cancer are almost exclusively smokers, and therefore have airways disease, and hence the PaCO2-PETCO2 gradient during anesthesia will be increased even in the supine position.1 In the 17 patients studied by Werner et al,2 mean PaCO2-PETCO2 from each lung was approximately 5 mm Hg in the supine position at a PaCO2 of 28 mm Hg and ventilatory frequency of 10/min. The lungs receive an approximately equal share of ventilation in the supine position, whereas, the ventilation of the upper lung is increased in the lateral position.2 However, the blood flow to the lungs is gravity dependent resulting in a decreased blood flow to the upper lung. This results in a decrease in the PETCO2 from the upper lung in the lateral position. Therefore, the PaCO2-PETCO2 as measured by Werner et al2 was zero for the lower lung and 11 mm Hg for the upper lung. Since the ventilation of the upper lung increases in the lateral position, it may surmised that if PETCO2 measured in the combined expirate of both lungs, the PaCO2-combined PETCO2 difference would be large.1,2

An incision of the chest wall produces an increase in mean pulmonary artery pressure2 which produces an increase in the CO2 elimination of the upper lung and a decrease in physiological dead space. The upper lung PaCO2-PETCO2 gradient is reduced when the mean pulmonary artery pressure increases as a result of surgical stimulation.

Furthermore, opening of the pleura increases CO2 elimination in the upper lung and thereby decreases PaCO2-PETCO2. Retraction of the lung produces exactly opposite effects. The lower lung gradient is not greatly effected with these maneuvers; it remains small. If combined end-tidal CO2 monitoring is performed, the PaCO2-PETCO2 would decrease when the pleura is opened, and increase again during retraction.

One lung ventilation:1,2

When the upper lung ventilation is discontinued to facilitate surgery, perfusion to the upper lung does not cease completely unless the pulmonary artery is clamped. In the absence of such clamping, there is a right to left shunt to the upper lung, the effect of which is to increase PaCO2, thus PaCO2-PETCO2 increases. However, this PaCO2-PETCO2 may not be any greater than the original combined two lung PaCO2-PETCO2.

Affect of prolonged expiratory maneuvers on PaCO2-PETCO2 during thoracotomy:3

In 16 patients undergoing thoracoabdominal esophagectomy, the affect of two prolonged expiration maneuvers to improve prediction of PaCO2 from PETCO2 were studied. PCO2 at the end of a simple prolonged expiration (PE1CO2), and PCO2 at the end of a prolonged expiration preceded by sustained hyperinflation of the lungs (PE2CO2), were measured during laparotomy, in the lateral thoracotomy position during two-lung ventilation, and after transition to one lung ventilation. PaCO2-PETCO2 was 9.75 (SD 3) mm Hg during laparotomy and this remained stable throughout the study. Both maneuvers decreased the mean arterial to peak expired PCO2 difference particularly during one lung ventilation. These results are in agreement with the results obtained by Bhavani-Shankar et al, where squeeze PETCO2 decreased PaCO2-PETCO2 in pregnant patients undergoing laparoscopic surgery.

Table from reference 3

Measurement mean (SD) mm Hg PaCO2-PETCO2 PaCO2-PE1CO2 PaCO2-PE2CO2
End of abdominal procedure 9.75 (3) 6 (3.7) 4.5 (3.7)
TLV for 20 min 10.5 (3.7) 7.5(4.5) 3 (4.5)
OLV for 20 min 9.75 (3) 3 (3.7) - 0.75 (3.7)
OLV for 50 min 10.5 (3.7) 2.2 (4.5) -1.5 (3.7)
TLV after skin closure 9 (3.7) 3.7 (5.2) 0.75 (4.5)

However, the end-expiratory PCO2 obtained with each maneuver during laparotomy and thoracotomy agreed poorly with PaCO2. The authors3 suggest that these maneuvers should no longer be recommended to improve estimation of PaCO2 from PETCO2 during anesthesia.

Arterial to end-tidal CO2 difference after bilateral lung transplantation.

Variable Time after lung transplantation
Mean (SD) 10 min 1 hr 3 hr 12 hr 24 hr
PaCO2-PETCO2 mm Hg 16 (5) 14 (5) 9 (4) 6 (3) 5 (3)
(PaCO2-PETCO2)/PaCO2 0.36 (0.13) 0.33 (0.13) 0.21 (0.12) 0.15 (0.07) 0.11 (0.06)

The time course of the arterial to end-tidal PCO2 difference suggests a rapid improvement of this post-transplant ventilation/perfusion mismatch. After 24 hrs, the values were close to the physiological range, which is supposed to be 4-5 mm Hg. A possible explanation of these findings could be ischemia-reperfusion injury that affects microcirculation. An impaired distribution of pulmonary blood flow with unperfused alveoli would clinically appear as alveolar dead space as is seen immediately following lung transplantation. The ventilation/perfusion mismatch normalized in about 24 hrs indicating a redistribution of pulmonary blood flow and recovery of microcirculation.

Arterial to end-tidal carbon dioxide difference during anesthesia for thoracoscopy:

Presently, thoracoscopy is the procedure of choice in most patients requiring thoracic surgery, especially those patients with severe underlying lung disease.6,7 Srinivasa et al7 studied the difference between PaCO2 and PETCO2 during thoracoscopic surgery in ten patients scheduled for elective thoracoscopic procedures. All patients had general anesthesia induced by IV propofol, fentanyl and vecuronium. A double lumen endo-tracheal tube was positioned with the aid of a fiber optic bronchoscope. Anesthesia was maintained with 100% oxygen, desflurane, fentanyl and vecuronium. Ventilation was kept at, tidal volume (TV) 10 ml/kg, respiratory rate (RR) of 10 bpm and an I:E ratio of 1:2 while on two lung ventilation. TV was 7 ml/kg, RR of 10 bpm and an I:E ratio of 1:2 while on one lung ventilation (OLV). The FiO2 during all measurements was kept at 1.0. Arterial blood gas was sampled 10 min after the patient was in lateral decubitus position while on two-lung ventilation (T1). The second sample (T2) was10 min after the introduction of trocars (OLV). The last sample (T3) was taken 10 min after restarting two-lung ventilation.
Results: The mean FVC was 2.6 ± 0.8 L (79 ± 19% predicted) and the FEV1 was 1.9 ± 0.7 L (76 ± 26% predicted). Table1 shows demographic data of the patients. The mean PaCO2 to PETCO2 difference at times T1, T2 and T3 were 5.5 ± 4, 7.4 ± 5, 6.8 ± 4 mm Hg respectively. The lowest value noted for PETCO2 was 27 mm Hg and the highest value for PaCO2 was 52 mm Hg during the study (Fig.2). The difference between the PETCO2 and PaCO2 during the various time intervals was not statistically significant.

Time PETCO2 PaCO2 Difference
T1 32 ± 3 38 ± 5 5.5 ± 4
T2 35 ± 4 42 ± 5 7.4 ± 5
T2 32 ± 3 39 ± 6 6.8 ±



(1). Fletcher R: The Arterio-End-Tidal CO2 difference during cardiothoracic surgery. J. cardiothorac. Anesthesiol. 4:105-117, 1990.

(2). Werner O. Malmkvist G, Beckman A, et al: CO2 elimination from each lung during endobronchial anaesthesia. Br. J. Anaesth. 56: 995-1001, 1984.

(3). Tavernier B, Rey D, Thevenin J, Triboulet P, Scherpereel P. Can prolonged expiration manoeuvres improve the prediction of arterial PCO2 from end-tidal PCO2? Brit J Anaesth 1997;78:536-40.

(4). Bhavani Shankar K, Steinbrook R, Mushlin PS, Freiberger D. Transcutaneous carbon dioxide monitoring during laparoscopic surgery in pregnancy. Canadian J Anaesth 1998;45:164-9.

(5) Jellinek H, Hiesmayr M, Simon P, Klepetko W, Haider W. Arterial to end-tidal CO2CO2 tension difference after bilateral lung transplantation. Crit Care Med 1993;21:1035-40.

(6) Horvath KA: Thoracoscopic transmyocardial laser revascularization. Ann.Thorac.Surg. 1998; 65: 1439-41.

7) Kotloff RM, Tino G, Bavaria JE, Palevsky HI, Hansen-Flaschen J, Wahl PM, Kaiser LR: Bilateral lung volume reduction surgery for advanced emphysema. A comparison of median sternotomy and thoracoscopic approaches. Chest 1996; 110: 1399-406.

(8) Srinivasa V, Kodali BS, Bean T, Hartigan PM. Arterial to end-tidal carbon dioxide difference during thoracoscopic surgery. Anesthesiology ASA abstracts 2004;A1556.