Anesth Analg 2005;101:440-443
© 2005 International Anesthesia Research Society
doi: 10.1213/01.ANE.0000156949.91614.E9
TECHNOLOGY, COMPUTING, AND SIMULATION
A Pilot Study of Continuous Transtracheal Mixed Venous Oxygen Saturation Monitoring
Wei Wei, MM,
Zhaoqiong Zhu, MB,
Lunxu Liu, MD,
Yunxia Zuo, MM, PhD,
Min Gong, PhD,
Fushan Xue, MD, and
Jin Liu, MD
Departments of Anesthesiology and Cardiothoracic Surgery, West China Hospital, Sichuan University; Department of Physics, Sichuan University, Chengdu, Sichuan, P. R. China; and the Department of Anesthesia, Plastic Surgery Hospital, Chinese Academy of Medical Sciences, Beijing, P. R. China.
Address correspondence and reprint requests to Jin Liu, MD, Department of Anesthesiology, West China Hospital, Sichuan University. No 37, Guo-xue-xiang, Chengdu, Sichuan 610041, P. R. China. Address e-mail to wuliujin{at}china.com.
 |
Abstract
|
|---|
In this study, we investigated the feasibility and the accuracy of transtracheal mixed venous oxygen saturation (Svo2) monitoring. Ten patients undergoing thoracic surgery were included in this study. A single-use pediatric pulse oximetry sensor was attached to the double-lumen tube between the tracheal and bronchial cuff. After anesthesia was induced, the double-lumen tube was inserted into the trachea and adjusted to the proper position. During surgery, the pulmonary arterial blood was sampled every 3 min for 15 min to measure the Svo2. The measurements made by the transtracheal pulmonary pulse oximeter (Sto2) were recorded at the same time that blood was sampled from the pulmonary artery for Svo2 measurements. The levels of measurement agreement between the Sto2 and the Svo2 were analyzed using the Bland and Altman method. The mean ± sd (range) oxygen saturation values during the data collecting period were 82.0% ± 4.9% (72%–91%) for the Sto2 and 82.2% ± 5.5% (71%–91%) for the Svo2, respectively. The linear correlation coefficient of the regression analysis between the Sto2 and the Svo2 was 0.934 (P < 0.05). A 95% confidence interval for absolute difference between the Sto2 and the Svo2 was 1.58%–2.09%. The mean ± 2 sd difference between the Sto2 and the Svo2 was 0.12% ± 3.97% on the Bland and Altman graph. We conclude that it is feasible to monitor the pulmonary artery oxygen saturation continuously by a transtracheal pulse oximetry technique and that it can be done so accurately.
|
Introduction
|
|---|
Continuous monitoring of mixed venous oxygen saturation (Svo2) is used in intensive care units (ICU) and operating rooms to detect potential medical problems such as respiratory failure, changes in oxygen consumption (1,2), anemia, and fluctuations of cardiac output. Svo2 is typically measured by intermittent sampling from a traditional pulmonary artery catheter (PAC) or by using a specially designed PAC fiberoptic bundle for continuous Svo2 monitoring. Placement of a PAC requires special skill, is expensive, and can result in severe complications, such as knotting of the catheter, vascular or cardiac perforation, and increased risk of infection. Thus, the clinical use of Svo2 monitoring is limited and its benefits are not fully utilized (3,4).
Oximeters have been placed in fingers, cheeks (5), nasal septum (6), tongues (7), the pharynx (8), trachea (9), and the esophagus (10,11) to monitor arterial oxygen saturation. Usually in finger oximetry, the light emitter and photodetector are aligned opposite each other, but in esophageal, pharyngeal, tracheal, and ventricular oximetry, they are arranged side by side. It is possible to obtain high quality signals using a side-by-side emitter-detector arrangement (8–11), suggesting that much of the emitted red and infrared light is reflected off adjacent tissue and onto the photodetector. Margreiter et al. (12) reported that an esophageal pulse oximetry sensor could be properly positioned under the guide of transesophageal echocardiography to monitor right ventricular oxygen saturation. Because the venous blood in the right ventricle may not be fully mixed, the pulmonary artery is usually considered as a better place to measure Svo2. The anterior wall of the carina and the right and left main bronchi are typically adjacent to the primary pulmonary artery or the left pulmonary artery. This anatomic relationship suggests that it may be possible to use a pulse oximetry sensor located in the airway to measure the oxyhemoglobin saturation of blood in the pulmonary artery. If the sensor is located in the carina facing anteriorly, the light will pass through the tracheal wall to the pulmonary artery and reflected signals will provide a noninvasive measure of Svo2. This study was designed to test the feasibility and accuracy of this technique.
|
Methods
|
|---|
With West China Hospital’s Research and Ethics Committee approval of the procedure and after written, informed patient consent, 10 patients undergoing thoracic surgery (ASA physical status I–III) were included in this study. Patients were excluded when the glottis could not be seen under the laryngoscope to prevent injury to the glottis. The edge of a single-use pediatric oximetry probe (Nellcor Puritan Bennett Inc, Pleasanton, CA) was cut to make it as small as possible and then fixed between the tracheal and bronchial cuff using a monolayer adhesive dressing (3M, Shanghai, China) in a Robertshaw size 35 double-lumen tube (DLT), in which the emitters and photodetector were facing anteriorly (Fig. 1). Anesthesia was induced with midazolam 0.15 mg/kg, fentanyl 2 µg/kg, vecuronium 0.1 mg/kg, propofol 2 mg/kg and maintained with 1
2% isoflurane and bolus injection of fentanyl and vecuronium. After anesthesia induction, the DLT with the oximetry probe attached was placed under direct laryngoscopy. Five patients were tracheally intubated with a right-sided Robertshaw DLT and the other 5 with a left-sided Robertshaw DLT. The position of the DLT was adjusted using a fiberoptic bronchoscope and auscultation. The position of the DLT was further adjusted until the best quality of pulse signal was obtained, indicating that the oximetry probe was facing toward the pulmonary artery adjacent to the airway. The depth of the inserted DLT measured from the front teeth was recorded. The patient was monitored with electrocardiogram, pulse oximetry (Spo2), and a radial arterial line was inserted for invasive monitoring of radial arterial blood pressure (ABP). After the thorax was opened, a 22-gauge catheter was directly placed into the pulmonary artery by the surgeon for pulmonary arterial pressure (PAP) monitoring and blood sampling. Oximetry probes were inserted into the trachea for transtracheal pulmonary arterial pulse oxygen saturation (Sto2) monitoring and on a finger for Spo2 monitoring. Pressure transducers for ABP monitoring from the radial artery and for PAP monitoring from the pulmonary artery were all connected to one monitor (SpaceLabs 1700; Spacelabs Medical Inc, Redmond, WA) for simultaneous recording. Five minutes later when the hemodynamic indices and Sto2 readings were stable, 0.5 mL of pulmonary arterial blood was sampled every 3 min for 15 min, and Sto2 readings were recorded at these precise time intervals. The pulmonary arterial blood was analyzed within 30 s with a calibrated blood gas machine (i-Stat; Abbott Laboratories Inc., East Windsor, NJ). The thoracic surgeries were performed using one-lung ventilation for all patients in this study. Patients were questioned about sore throats at 24 h postoperatively. Neither the recorder nor the blood analysis technician was aware of the results obtained by the other test.
View larger version (109K):
[in this window]
[in a new window]
|
Figure 1. Oximetry probe was attached anteriorly between tracheal and bronchial cuff by monolayer adhesive dressing in a right Robertshaw double-lumen tube (DLT, size 35).
|
|
Linear regression analysis was used to compare Sto2 (from transtracheal oxygen sensor) with the Svo2 (from the blood gas analysis of the pulmonary arterial samples). The levels of measurement agreement between the Sto2 and the Svo2 were analyzed with the Bland and Altman (13) method. The standard error of the mean difference (sem) was calculated by dividing the standard deviation by 
n, where "n" is the sample size. A 95% confidence interval for the absolute difference between the Sto2 and the Svo2 was calculated.
|
Results
|
|---|
The mean ± sd (range) of the patients’ age and weight were 46.4 ± 12.2 yr (30–60 yr) and 60.8 ± 8.2 kg (49–70 kg), respectively. The male/female ratio was 8:2. The depth of DLT insertion was 29.8 ± 1.2 cm (27–31 cm). The hemodynamic and respiratory variables were stable during the data collection period. The position of the transtracheal oxygen sensor was unchanged before, during, and after the data collection period in all patients. A typical view of the Sto2 wave is shown in Figure 2; the second wave was from the Sto2 oximetry.
View larger version (88K):
[in this window]
[in a new window]
|
Figure 2. A typical Sto2 signal. From top to bottom were the signals of Spo2, Sto2, radial arterial blood pressure (ABP), and pulmonary arterial pressure (PAP). In terms of time phases, the relationship of the waves between Spo2 and ABP for each cardiac contract was the same as between Sto2 and PAP; the relationship of the waves between Spo2 and Sto2 was the same as between RAP and PAP. PAP wave and Sto2 wave appeared earlier than ABP wave and Spo2 wave in each circle on the screen, respectively.
|
|
The mean ± sd (range, sem) of oxygen saturation during the data collection period were 82.0% ± 4.9% (72%–91%, 0.7) for the Sto2 and 82.2% ± 5.5% (71%–91%, 0.78) for the Svo2, respectively. The linear correlation coefficient of the regression analysis between the Sto2 and the Svo2 was 0.934 (P < 0.05). The Bland and Altman graph for Sto2 versus Svo2 in 10 patients is presented in Figure 3. The mean difference ± 2 sd between the Sto2 and the Svo2 was 0.12% ± 3.79%. A 95% confidence interval for absolute difference was 1.58%–2.09%. None of the patients had sore throats 24 h postoperatively.
View larger version (19K):
[in this window]
[in a new window]
|
Figure 3. Bland and Altman graph comparing the difference between Sto2 and Svo2 versus the mean oxygen saturation by the two methods.
|
|
|
Discussion
|
|---|
This study demonstrated that transtracheal pulse oximetry is a simple method for monitoring the Svo2 noninvasively in tracheally intubated patients. The Sto2 wave detected by transtracheal oximetry is similar to the digital pulse oximetry as a single wave (Fig. 2). The measurement of the Sto2 using transtracheal oximetry differed from Svo2 measured directly by 1.58%–2.09%. This difference indicates that the Svo2 could be monitored continuously and accurately by transtracheal pulse oximetry using a standard monitor. This technique is less expensive and carries less risk when compared with placing a PAC.
It might be argued whether the Sto2 signals observed in this study were indeed obtained from pulmonary arterial blood. There are several reasons why this is likely the case. First, the probes were directed at the pulmonary artery and a red light on the pulmonary artery emanating from the oximetry probe was seen by both surgeons and anesthesiologists during surgery. Second, in a study of transtracheal oximetry, Brimacombe et al. (9) demonstrated that the tracheal oximetry readings are not derived from the tracheal mucosa but are primarily derived from deeper anatomic structures such as the aorta or left common carotid artery. In our study, the pulse oximetry probes were located in the carina or the right (or left) main bronchus where the pulmonary artery is adjoining the wall of the trachea and main bronchus. According to Brimacombe et al., the pulse signals we monitored should be from deeper structures, most likely from the adjacent pulmonary artery. Third, in terms of time phases, the relationship of the waves between the Spo2 and the ABP for each cardiac contract was almost identical to that between the Sto2 and the PAP; the relationship of the waves between the Spo2 and the Sto2 was almost identical to that between the ABP and the PAP. Like the relationships between the PAP wave and the ABP wave, the Sto2 wave appeared earlier than the Spo2 wave in each cycle on the screen, which may indicate a shorter distance from tricuspid to pulmonary artery than from aortic valve to radial artery. These findings suggest that the Sto2 signals and the PAP waves should be from the same source, most likely the pulmonary artery. Finally, and most importantly, the Sto2 readings matched closely with the Svo2 measured from the blood gas analysis of the pulmonary artery blood samples. It is unlikely that the Sto2 signals were severely contaminated by any other tissue or blood. This study shows that the waves and corresponding readings of the Sto2 are from the pulmonary artery and that a noninvasive continuous Sto2 monitoring is feasible, reliable, and accurate for Svo2 monitoring.
There were two limitations with the oximetry sensor used in this study. First, the sensor used was adapted to measure a finger Spo2 with a design of light emitters and photodetectors placed opposite to each other. Our device was self-made, with the light emitter and photodetector aligned side-by-side after it was attached to the DLT. Because of the difficulty of the technique, we could not shorten the distance between the light emitters and photodetector to adapt Sto2 monitoring. Second, the normal range of Svo2 was approximately 75%, but the oximetry probe used in this study was designed to monitor arterial Spo2 (normally >90%). Tachibana et al. (14) reported that the pulse oximetry at low levels of saturation (Sao2 less than 80%) was not as accurate as at a higher saturation level (Sao2 greater than 80%). But from the Bland and Altman graph comparing the difference between the Sto2 and the Svo2 in this study (Fig. 3), it appears that the agreement of the two variables is almost the same between the range of 70%–80% and the range of 80%–90%. The quality of the Svo2 monitoring might be improved by designing an oximetry sensor made specifically for transtracheal Svo2 monitoring. Although we did not observe any unexpected adverse events or effects in our study, some precautionary aspects, such as the safety of the tracheal mucosa and the heart when electrocautery or defibrillation is used, are still unknown. Therefore, this method for transtracheal Svo2 monitoring is not recommended until more data and new apparatus are available.
In summary, it is feasible and accurate to measure the Svo2 continuously by transtracheal pulse oximetry when the oxygen saturation sensor is positioned properly. This technique merits further evaluation and development for intubated critically ill or anesthetized patients in critical care units, ICUs, and operating rooms.
|
Footnotes
|
|---|
Accepted for publication January 5, 2005.
|
References
|
|---|
- Vaughn S, Puri VK. Cardiac output changes and continuous mixed venous oxygen saturation measurement in the critically ill. Crit Care Med 1988;16:495–8.[ISI][Medline]
- Engler EL, Holm K. Perspectives on the interpretation of continuous mixed venous oxygen saturation. Heart Lung 1990;19:578–80.[Medline]
- Norfleet EA, Watson CB. Continuous mixed venous oxygen saturation measurement: a significant advance in hemodynamic monitoring? Clin Monit 1985;1:245–58.
- Ahrens T. Continuous mixed venous (Svo2) monitoring: too expensive or indispensable? Crit Care Nurs Clin North Am 1999;11:33–48.[Medline]
- O’Leary RJ Jr, Landon M, Benumof JL. Buccal pulse oximeter is more accurate than finger pulse oximeter in measuring oxygen saturation. Anesth Analg 1992;75:495–8.[Abstract/Free Full Text]
- Ezri T, Lurie S, Konichezky S, et al. Pulse oximeter from the nasal septum. J Clin Anesth 1991;3:447–50.[Medline]
- Jobes DR, Nicolson SC. Monitoring of arterial hemoglobin oxygen saturation using a tongue sensor. Anesth Analg 1988;67:186–8.[Free Full Text]
- Agro F, Carassiti M, Cataldo R, et al. Pulse oximetry with the laryngeal mask airway. Resuscitation 1999;43:65–7.[Medline]
- Brimacombe J, Keller C, Margreiter J. A pilot study of left tracheal pulse oximetry. Anesth Analg 2000;91:1003–6.[Abstract/Free Full Text]
- Vicenzi MN, Gombotz H, Krenn H, et al. Transesophageal versus surface pulse oximetry in intensive care unit patients. Crit Care Med 2000;28:2268–70.[ISI][Medline]
- Kyriacou PA, Powell SL, Jones DP, et al. Evaluation of esophageal pulse oximetry in patients undergoing cardiothoracic surgery. Anaesthesia 2003;58:422–7.[Medline]
- Margreiter J, Keller C, Brimacombe J. The feasibility of trans-esophageal echocardiograph-guided right and left ventricular oximetry in hemodynamically stable patients undergoing coronary artery bypass grafting. Anesth Analg 2002;94:794–8.[Abstract/Free Full Text]
- Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.[ISI][Medline]
- Tachibana C, Fukada T, Hasegawa R, et al. Accuracy of a pulse oximeter during hypoxia [in Japanese]. Masui 1996;45:479–82.[Medline]
Top
Abstract
Introduction
