Measurement of combined oximetry and cutaneous capnography during flexible bronchoscopy

P. N. Chhajed; R. Rajasekaran; B. Kaegi; T. P. Chhajed; E. Pflimlin; J. Leuppi; M. Tamm

doi:10.1183/09031936.06.00088005

Abstract

The aim of the present study was to assess the feasibility of measuring combined arterial oxygen saturation measured by pulse oximetry (S_p,O₂) and cutaneous carbon dioxide tension (P_c,CO₂) to monitor ventilation and quantify change in P_c,CO₂during bronchoscopy.

Combined S_p,O₂ and P_c,CO₂ were measured at the ear lobe in 114 patients. In four patients, the ear-clip slipped and they were excluded. In total, 11 patients had artefacts with S_p,O₂ recordings, thus, S_p,O₂ was analysed in 99 patients. Spirometry data were available in 77 patients. Multivariate analysis of covariance and logistic regression were used for statistical analyses.

Mean baseline P_c,CO₂ was 4.78±1.06 kPa (36±8 mmHg) and mean rise in the P_c,CO₂ during bronchoscopy was 1.26±0.70 kPa (9.5±5.3 mmHg), while mean P_c,CO₂ at the end of bronchoscopy was 5.85±1.19 kPa (44±9 mmHg) . Baseline P_c,CO₂ and the lowest S_p,O₂ were significantly associated with peak P_c,CO₂ and the change in P_c,CO₂ during bronchoscopy. Risk of significant hypoxaemia (S_p,O₂≤90%) was lower for a higher baseline S_p,O₂. Peak P_c,CO₂ was directly associated with significant hypoxaemia. There was no significant association in the baseline P_c,CO₂, peak P_c,CO₂, baseline S_p,O₂ or the lowest S_p,O₂ comparing patients with and without chronic obstructive pulmonary disease.

In conclusion, it is feasible to measure combined pulse oximetry and cutaneous carbon dioxide tension effectively to monitor ventilation during flexible bronchoscopy.

Flexible bronchoscopy under local anaesthesia and sedation is widely performed to diagnose and manage a variety of lung diseases. The purpose of sedation is to facilitate examination of the tracheobronchial tree, carry out the necessary diagnostic or interventional procedures, and provide patient comfort 1–4. Oximetry measures the oxygenation of blood and can detect hypoxaemia, but it cannot detect hypercarbia and hence the adequacy of ventilation 5. When bronchoscopy is performed under conscious sedation without supplemental oxygen, oxygen desaturation will occur rapidly giving a good indication of ventilatory status. It has recently been shown that pulse oximetry is a useful tool to assess ventilatory abnormalities, but only in the absence of supplemental inspired oxygen 6. However, most centres routinely offer oxygen supplementation to patients during flexible bronchoscopy 7. Due to the S-shaped curve of the oxygen dissociation curve, in such situations the alveolar carbon dioxide tension will have to increase further before significant hypoxaemia is manifested 8. The British Thoracic Society (BTS) guidelines recommend that oxygen supplementation should be used to achieve an oxygen saturation (S_p,O₂) of at least 90% to reduce the risk of significant arrhythmias during the procedure and also in the post-operative recovery period 2. Furthermore, they suggest that care should be taken to be alert to signs of respiratory failure in patients on oxygen supplementation, who may have “safe” oximetry readings, but who may be developing carbon dioxide retention 2. Therefore, a reliable substitute of carbon dioxide arterial tension (P_a,CO₂) measured noninvasively is required as it might enhance patient safety during flexible bronchoscopy.

In intubated patients or patients under stable conditions without oral leakage, end-tidal carbon dioxide tension has a good correlation with P_a,CO₂ 9, 10. However, during endoscopy procedures under sedation, regular breathing is often disturbed by moving, coughing or changes between nose and mouth ventilation, causing leakage and artefacts or misinterpretation of the data acquired with end-tidal carbon dioxide measurements. These problems often restrict the use of side-stream capnography in clinical practice 10. Measurement of cutaneous carbon dioxide tension (P_c,CO₂) with a digital sensor has recently been reported 11. The P_c,CO₂values have been shown to have a good correlation with arterial blood gas measurements 12, 13. A rise in P_c,CO₂measurement has been reported during medical thoracoscopy and colonoscopies 10, 13. The present authors aimed to assess the feasibility of measuring combined pulse oximetry (S_p,O₂) and P_c,CO₂at the ear lobe to monitor ventilation and quantify the degree of change in P_c,CO₂during flexible bronchoscopy under local anaesthesia and sedation.

PATIENTS AND METHODS

Combined oximetry and P_c,CO₂ were prospectively measured in 114 patients undergoing flexible bronchoscopy in an observational fashion (monitoring system from Sentec AG, Therwil, Switzerland). The combined cutaneous digital sensor was placed on the ear lobe of all patients prior to the procedure and was removed when the patient left the bronchoscopy suite.

Flexible bronchoscopy was performed under local anaesthesia and sedation. Combined sedation was achieved with intermittent boluses of intravenous midazolam and 5 mg of hydrocodone 1. All patients received supplemental nasal oxygen at 2–4 L·min^-1. Significant hypoxaemia (S_p,O₂≤90%) during flexible bronchoscopy was successfully treated with jaw support or nasopharyngeal tube insertion 7. The duration of bronchoscopy was calculated from the administration of sedation until the flexible bronchoscope was removed from the tracheobronchial tree. Bronchoalveolar lavage (BAL) was performed in 57 patients, endobronchial biopsy in 32, bronchial washings in 19, transbronchial needle aspiration in 18, transbronchial biopsy in 15 patients and bronchial brushings in 11 patients. Spirometry data prior to bronchoscopy was available in 77 patients. Patients with chronic obstructive pulmonary disease (COPD) were defined using the American Thoracic Society/European Respiratory Society guidelines 14.

Statistical analysis

Data are presented as mean±sd. Multivariate analysis was performed using analysis of covariance to examine peak P_c,CO₂, change in P_c,CO₂ from baseline, and change in P_c,CO₂ adjusted for baseline P_c,CO₂ during bronchoscopy as dependent factors versus patient age, dose of midazolam (mg·kg^-1), baseline P_c,CO₂ and lowest S_p,O₂as independent factors. Scatterplots are provided for the association of peak P_c,CO₂ and change in P_c,CO₂ with duration of procedure. ANOVA was used to examine the association between the type of sampling procedure during bronchoscopy with the change in P_c,CO₂. Logistic regression was used to examine independent factors associated with significant hypoxaemia (S_p,O₂≤90%) and COPD. The t-test was used to compare significance between the baseline and peak P_c,CO₂ during bronchoscopy and P_c,CO₂ at lowest S_p,O₂ and the peak P_c,CO₂ during bronchoscopy.

RESULTS

The mean duration of the procedure was 19±10 min. In four patients, the ear-clip slipped off the ear lobe due to patient movement. These patients were not included in the analysis as the equilibration had to be achieved again and there was loss of data recording points. Therefore data analysis was performed in 110 patients (70 males and 40 females, mean age 61±15 yrs). In 11 patients, the S_p,O₂was not included in the analysis due to artefacts produced by coughing or a low signal whilst recording. Therefore, P_c,CO₂ measurements, for the purposes of data analysis, were available in 110 patients and both oximetry and P_c,CO₂ measurements were available in 99 patients. Therefore, for multivariate analysis only data from patients with both oximetry and P_c,CO₂ were used. Hypoxaemia during the procedure was successfully treated with jaw support or nasopharyngeal tube insertion in all patients. None of the patients developed severe respiratory depression resulting in administration of sedation reversal medication or in abandoning the procedure without completion. The indications for bronchoscopy were: 1) suspected lung malignancy, 42 (38%) patients; 2) radiological infiltrates, 31 (28%) patients; 3) haemoptysis, nine (8%) patients; 4) suspected interstitial lung disease, nine (8%) patients; 5) suspected sarcoidosis, five (5%) patients; 6) post-airway stent inspection, four (4%) patients; 7) cough, three (3%) patients; 8) post-lung surgery, two (2%) patients; and 9) evaluation of post-tracheostomy tracheal stenosis, larngeal papilloma and mucociliary dyskinesis (all n = 1).

The mean dose of midazolam administered was 0.06±0.03 mg·kg^-1. The mean baseline P_c,CO₂ was 4.78±1.06 kPa (36±8 mmHg) and the mean peak P_c,CO₂ was 6.11±1.19 kPa (46±9 mmHg; p<0.0001). The peak P_c,CO₂ during the procedure was recorded at a mean duration of 13±7 min. The mean rise in the P_c,CO₂ during the procedure was 1.26±0.70 kPa (9.5±5.3 mmHg). The mean P_c,CO₂ at the end of the procedure was 5.85±1.19 kPa (44±9 mmHg) and at the time of removal of the sensor was 5.85±1.19 kPa (44±9 mmHg). A rise in P_c,CO₂ was observed in all patients except one. The baseline S_p,O₂ was 97±2%. The lowest mean S_p,O₂ was 93±4%. The lowest S_p,O₂ in patients who desaturated from their baseline S_p,O₂ measurement was recorded at a mean duration of 9±6 min. The mean P_c,CO₂ measured at lowest S_p,O₂ was 5.85±1.86 kPa (44±14 mmHg) and the mean peak P_c,CO₂ recorded subsequently was 6.78±1.72 kPa (51±13 mmHg; p<0.0001).

Factors associated with peak P_c,CO₂and change in P_c,CO₂during the study

The peak P_c,CO₂ during the study was significantly associated with the baseline P_c,CO₂ (p<0.0001) and lowest S_p,O₂ (p = 0.016). A higher baseline P_c,CO₂ was associated with the peak P_c,CO₂ (β = 0.8). However, there was an inverse relationship of the lowest S_p,O₂ with the peak P_c,CO₂, thus, the manifestation of a lower S_p,O₂ was associated with the peak P_c,CO₂ (β = -0.4). Change in P_c,CO₂ (difference in baseline and peak P_c,CO₂) was also significantly associated with baseline P_c,CO₂ (p = 0.024) and lowest S_p,O₂ (p = 0.016). Both of these factors had an inverse relationship with the change in P_c,CO₂, thus, patients manifesting a lower S_p,O₂ (β = -0.4) and a lower baseline P_c,CO₂ (β = -0.2) would have a higher change in P_c,CO₂. Similarly, the change in P_c,CO₂ divided by the baseline P_c,CO₂ (adjusting the change in P_c,CO₂ for baseline) was also significantly associated with baseline P_c,CO₂ (p<0.0001) and lowest S_O₂ (p = 0.016). This analysis highlights that baseline P_c,CO₂ and the lowest S_p,O₂ are linked to the peak P_c,CO₂ as well as the rise in P_c,CO₂ during bronchoscopy. There was no significant relationship of age and midazolam dosage with the peak P_c,CO₂, change in P_c,CO₂ and the change in P_c,CO₂/baseline P_c,CO₂. There was also no association of peak P_c,CO₂ and change in P_c,CO₂ with the procedure duration (fig. 1⇓). This factor was not included in the multivariate analysis as there appear to be two groups of patients; one group with higher peak P_c,CO₂ and change in P_c,CO₂ with a short duration of procedure and the other group of patients who have a higher peak and change in P_c,CO₂ with a longer duration of procedure. To assess the influence of procedure on change in P_c,CO₂ and change in P_c,CO₂/baseline P_c,CO₂, these were categorised as follows: 1) transbronchial biopsy only and transbronchial needle aspiration only (n = 10); 2) BAL only (n = 42); 3) transbronchial biopsy, transbronchial needle aspiration, and BAL in a combination of two or more (n = 16); and 4) bronchial washings and/or endobronchial biopsy (n = 42). There was no significant association between the procedures performed and change in P_c,CO₂ (p = 0.3) and change in P_c,CO₂/baseline P_c,CO₂ (p = 0.2) during bronchoscopy.

Fig. 1—

Scatter plot of duration of procedure in a) peak cutaneous carbon dioxide and b) change in cutaneous carbon dioxide.

Factors associated with significant hypoxaemia and COPD

In total, 15 patients had significant hypoxaemia during bronchoscopy. A lower baseline S_p,O₂ (p = 0.003; odds ratio (OR) 0.6) and the peak P_c,CO₂ (p = 0.041; OR 1.1) were significantly associated with significant hypoxaemia. Therefore, with a higher baseline S_p,O₂ the risk of significant hypoxaemia is lower and the peak P_c,CO₂ is directly associated with significant hypoxaemia. Patient age, duration of procedure, midazolam dose and baseline P_c,CO₂ were not associated with significant hypoxaemia. In total there were 25 patients with COPD. The mean forced expiratory volume in one second in patients with mild, moderate, severe and very severe COPD was 90±10% (n = 4), 64±8% (n = 13), 39±5% (n = 6) and 25±3% (n = 2), respectively. There was no significant association in the baseline P_c,CO₂, peak P_c,CO₂, baseline S_p,O₂, or the lowest S_p,O₂ comparing patients with and without COPD.

DISCUSSION

Recently, P_c,CO₂ measurements obtained using the device used in this study have been shown to have a good correlation with arterial carbon dioxide values when >100 samples were compared (R = 0.95) 13. The findings of the present study show that ventilation can be effectively monitored during flexible bronchoscopy using combined oximetry and P_c,CO₂measurement. In a study of 22 patients, an increase in P_c,CO₂ has been shown to occur during flexible bronchoscopy 5. However, Evans et al. 5 concluded that the technology they used for P_c,CO₂ monitoring was complex. Measurement of P_c,CO₂ in the current study was simple and did not lead to a delay in the start of the procedure or any complications related to the equipment. The potential complication when using this device is burning of the skin as the sensor is heated to a temperature of 42°C. The recommended maximum duration for which a sensor can be placed cutaneously at 42°C is 8 h 15. None of the patients in the present study had an apparent burn on the skin. The BTS guidelines caution physicians to monitor signs of respiratory failure in patients undergoing flexible bronchoscopy on oxygen supplementation that may have safe oximetry readings, but may be developing carbon dioxide retention. Patients with a higher baseline P_c,CO₂ are at a greater risk of developing a higher peak P_c,CO₂. Furthermore, the peak P_c,CO₂ was significantly associated with the lowest S_p,O₂ during bronchoscopy. The results of the current study show that it is feasible to measure continuous P_c,CO₂during bronchoscopy, which has been shown to be a good surrogate of P_a,CO₂ 2, 12, 13. Based on the findings of the present study, no specific trend was observed in the association between peak P_c,CO₂ and change in P_c,CO₂ with procedure duration, both a short and a longer procedure duration were associated with the peak P_c,CO₂ and change in P_c,CO₂. The ability to effectively measure P_c,CO₂ is an important step in patient monitoring as diagnostic bronchoscopies are performed using routine sedation and also drugs, such as propofol, which are considered as general anaesthetics are being increasingly used also in the absence of a trained anaesthetist 16–19.

The mean duration to the manifestation of lowest S_p,O₂ was earlier than the mean duration to the peak P_c,CO₂. Therefore, it seems impossible to predict the occurrence of lowest S_p,O₂ based on rising P_c,CO₂ values. The lowest S_p,O₂ manifested was significantly associated with peak P_c,CO₂. The hypothesis for this observation is that hypoventilation during flexible bronchoscopy performed under sedation with oxygen supplementation is multifactorial, and not only due to central respiratory depression, in which case a rise in P_c,CO₂ would precede significant hypoxaemia due to the nature of the oxygen dissociation curve. Hypoventilation during flexible bronchoscopy is therefore due to a combination of upper airway obstruction, alveolar hypoventilation caused by sedative medication and the procedure itself 7. The effective treatment of hypoxaemia with jaw support or nasopharyngeal tube insertion supports the role of upper airway obstruction leading to acute hypoxaemia during flexible bronchoscopy 7. The rise in P_c,CO₂ manifested thereafter reflects alveolar hypoventilation, which is likely to be due to the combined effect of sedatives (central respiratory depression), as well as the procedure itself potentially leading to ventilation perfusion mismatch. The findings of the current study show that after effective treatment of upper airway obstruction, there continues to be a rise in P_c,CO₂ reflecting hypoventilation. Therefore, S_p,O₂ alone is not sufficient to monitor the complete ventilation status of the patient during bronchoscopy, thus highlighting the value of combined oximetry and cutaneous capnography during bronchoscopy.

The permissible level of P_c,CO₂ rise during bronchoscopy is not known. The BTS guidelines state that the use of sedation in patients with severe COPD has increased risks relating to potential carbon dioxide retention 2. The findings of the present study show that patients with COPD who are not in hypercapnic respiratory failure are at an equal risk of hypercapnia as well as hypoxaemia compared with patients without COPD. Based on the findings of the current study (mean rise in P_c,CO₂ of 1.33±0.66 kPa), the present authors speculate that a rise of 2.0 kPa in P_c,CO₂ from baseline may be an indication to limit further administration of sedatives and needs to be confirmed in future studies.

In conclusion, the current study demonstrates the feasibility of measuring combined oximetry and cutaneous capnography effectively to monitor ventilation during flexible bronchoscopy. Patients with COPD who are not in hypercapnic respiratory failure have a similar risk of hypoxaemia and hypercapnia compared with those without COPD during bronchoscopy. The peak P_c,CO₂ is higher in patients with a higher baseline P_c,CO₂ and is associated with lowest S_p,O₂ during bronchoscopy. Significant hypoxaemia is associated with the peak P_c,CO₂ and baseline S_p,O₂. With the current study results, it is not possible to predict and accurately select a group of patients who should undergo capnography during bronchoscopy. Patients with elevated baseline P_a,CO₂ or a low baseline S_p,O₂ will have the benefit of better monitoring of ventilation using such a device. Specifically, patients with a longer duration of procedure did not show a different rise in P_c,CO₂ compared with the others. Furthermore, the current authors expected an association of the sedative dose with rise in cutaneous carbon dioxide tension, but that was also not the case. Therefore, future studies need to analyse the different patient groups with regards to the baseline cutaneous carbon dioxide tension to potentially limit the application of cutaneous capnography to only a specific group of patients.

Acknowledgments

The authors would like to thank F. Baty and A. Schötzau for their assistance with performing the statistical analysis. The authors would also like to thank Sentec AG (Therwil, Switzerland) for providing the combined cutaneous carbon dioxide and oximetry monitoring system.

Received July 28, 2005.
Accepted April 10, 2006.

References

↵
Stolz D, Chhajed PN, Leuppi JD, Brutsche M, Pflimlin E, Tamm M. Cough suppression during flexible bronchoscopy using combined sedation with midazolam and hydrocodone: a randomised, double blind, placebo controlled trial. Thorax 2004;59:773–776.
OpenUrl Abstract/FREE Full Text
↵
British Thoracic Society Bronchoscopy Guidelines Committee. British Thoracic Society guidelines on diagnostic flexible bronchoscopy. Thorax 2001;56: Suppl. 1 i1–i21.
OpenUrl Abstract/FREE Full Text
Ereth MH, Stubbs SE, Lennon RL. Bronchoscopic pharmacology and anesthesia. In: Prakash UBS, ed. Bronchoscopy. New York, Raven Press, 1994; p. 91
↵
Chhajed PN, Aboyoun C, Malouf MA, et al. Management of acute hypoxemia during flexible bronchoscopy with insertion of a nasopharyngeal tube in lung transplant recipients. Chest 2002;121:1350–1354.
OpenUrl CrossRef PubMed Web of Science
↵
Evans EN, Ganeshalingam K, Ebden P. Changes in oxygen saturation and transcutaneous carbon dioxide and oxygen levels in patients undergoing fibreoptic bronchoscopy. Respir Med 1998;92:739–742.
OpenUrl CrossRef PubMed Web of Science
↵
Fu ES, Downs JB, Schweiger JW, Miguel RV, Smith RA. Supplemental oxygen impairs detection of hypoventilation by pulse oximetry. Chest 2004;126:1552–1558.
OpenUrl CrossRef PubMed Web of Science
↵
Chhajed PN, Glanville AR. Management of hypoxemia during flexible bronchoscopy. Clin Chest Med 2003;24:511–516.
OpenUrl CrossRef PubMed Web of Science
↵
Chhajed PN, Heuss LT, Tamm M. Cutaneous carbon dioxide tension monitoring in adults. Curr Opin Anaesthesiol 2004;17:521–525.
↵
Miner JR, Heegaard W, Plummer D. End-tidal carbon dioxide monitoring during procedural sedation. Acad Emerg Med 2002;9:275–280.
OpenUrl CrossRef PubMed Web of Science
↵
Heuss LT, Chhajed PN, Schnieper P, Hirt T, Beglinger C. Combined pulse oximetry/cutaneous carbon dioxide tension monitoring during colonoscopies: pilot study with a smart ear clip. Digestion 2004;70:152–158.
OpenUrl PubMed Web of Science
↵
Kocher S, Rohling R, Tschupp A. Performance of a digital PCO2/SPO2 ear sensor. J Clin Monit 2004;18:75–79.
OpenUrl PubMed
↵
Chhajed PN, Langewitz W, Tamm M. (Un)explained hyperventilation. Respiration 2005: Epub ahead of print
↵
Chhajed PN, Kaegi B, Rajasekaran R, Tamm M. Detection of hypoventilation during thoracoscopy: combined cutaneous carbon dioxide tension and oximetry monitoring with a new digital sensor. Chest 2005;127:585–588.
OpenUrl CrossRef PubMed Web of Science
↵
Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004;23:932–946.
OpenUrl FREE Full Text
↵
Skin temperature at the pulse oximeter probe. In: Medical and Electrical Equipment-Particular Requirements for Basic Safety and Essential Performance of Pulse Oximeter Equipment for Medical Use. Geneva, International Organisation for Standardisation, 2003–2004
↵
Chhajed PN, Wallner J, Stolz D, et al. Sedative drug requirement during flexible bronchoscopy. Respiration 2005;72:617–621.
OpenUrl PubMed Web of Science
Chhajed PN, Aboyoun C, Chhajed TP, et al. Sedative drug requirements during bronchoscopy are higher in cystic fibrosis after lung transplantation. Transplantation 2005;80:1081–1085.
OpenUrl CrossRef PubMed Web of Science
Heuss LT, Schnieper P, Drewe J, Pflimlin E, Beglinger C. Risk stratification and safe administration of propofol by registered nurses supervised by the gastroenterologist: a prospective observational study of more than 2000 cases. Gastrointest Endosc 2003;57:664–671.
OpenUrl CrossRef PubMed Web of Science
↵
Heuss LT, Froehlich F, Beglinger C. Changing patterns of sedation and monitoring practice during endoscopy: results of a nationwide survey in Switzerland. Endoscopy 2005;37:161–166.
OpenUrl CrossRef PubMed Web of Science