Comparison of end-tidal and transcutaneous measures of carbon dioxide during general anaesthesia in severely obese adults

J. Griffin1, B. E. Terry2, R. K. Burton1, T. L. Ray1, B. P. Keller1, A. L. Landrum1, J. O. Johnson1 and J. D. Tobias*,1

Departments of 1 Anesthesiology and 2 Surgery, The University of Missouri, Columbia, Missouri, USA

Corresponding author. E-mail: tobiasj@health.missouri.edu

Accepted for publication: March 30, 2003


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. Patients with severe obesity (body mass index (BMI) greater than 35 kg m–2) present difficulties for end-tidal carbon dioxide (FE'CO2) monitoring. Previous studies suggest that transcutaneous (TC) carbon dioxide measurements could be valuable, so we compared FE' and TC measures with PaCO2 in severely obese patients during anaesthesia.

Methods. We studied patients with severe obesity (BMI >=40 kg m–2) undergoing gastric bypass surgery. Carbon dioxide was measured with both FE' and TC devices. The difference between each measure (FE'CO2 and TC-CO2) and the PaCO2 was averaged for each patient to provide one value, and data compared with a non-paired, two-way t-test, Fisher’s exact test.

Results. We studied 30 adults (aged 18–54 yr, mean 41, SD 8.0 yr; weight: 115–267 kg, mean 162, SD 35 kg). The absolute difference between the TC-CO2 and PaCO2 was 0.2 (0.2) (mean, SD) kPa while the absolute difference between the FE'CO2 and PaCO2 was 0.7 (0.4) kPa (P<0.0001). The bias and precision were +0.1 (0.3) kPa for TC vs arterial carbon dioxide and –0.7 (0.4) kPa for FE' vs arterial carbon dioxide.

Conclusions. Transcutaneous carbon dioxide monitoring provides a better estimate of PaCO2 than FE'CO2 in patients with severe obesity.

Br J Anaesth 2003; 91: 498–501

Keywords: anaesthesia, general; arterial pressure, measurement; complications, obesity; equipment, monitors


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Alveolar ventilation is assessed by the measurement of the partial pressure of carbon dioxide in arterial blood. Although the absolute measure is intermittent arterial sampling and blood gas analysis, this is invasive and intermittent. To overcome such problems, non-invasive techniques can monitor carbon dioxide tensions. Intraoperatively, a basic monitor is end-tidal carbon dioxide (FE'CO2). However, FE'CO2 may not always correlate with the actual PaCO2.13 Patients with severe obesity (body mass index (BMI) greater than 35 kg m–2) may present difficulties in FE'CO2 measures of PaCO2 because functional residual capacity is reduced with resultant ventilation-perfusion inequalities.4

Transcutaneous (TC) measurement can also be used for the continuous non-invasive monitoring of carbon dioxide. We found TC measurements were accurate intraoperatively, in infants and children and adults.47 In the severely obese patient, excessive adipose tissue beneath the skin may possibly affect the accuracy of TC devices, and data regarding the accuracy of FE'CO2 monitoring in obese patients are limited. We compared the accuracy of FE'CO2 and TC-CO2 monitoring with PaCO2 in adult obese patients.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Approval for this study was granted by the Institutional Review Board and the Committee for the Protection of Human Subjects of the University of Missouri. We studied patients over 17 yr of age with a BMI of 40 or greater, presenting for open vertical ringed gastric bypass with roux en y (gastric bypass). Arterial cannulation was clinically necessary. Verbal consent was obtained from each patient.

The anaesthetic technique was determined by the attending anaesthesiologist. After intubation of the trachea, ventilation was controlled using a tidal volume of 800–1000 ml, an I:E ratio of 1:2, positive end-expiratory pressure of 3–5 cmH2O, FIO2 of 0.4–0.6, and the ventilatory frequency adjusted to maintain normocarbia. FE'CO2 was measured using an infrared analyser with side-stream sampling at a flow rate of 150 ml min–1 (Capnomac Ultima, Datex-Ohmeda, Madison, WI, USA). The FE'CO2 analyser was calibrated with a cylinder containing a calibration gas according to the manufacturer’s recommendations. The digital readout of the FE'CO2 is based on an algorithm that evaluates two successive waveforms and the valley between them. The FE'CO2 reported by the monitor is the maximum value from the first waveform. TC-CO2 was measured using a TCM-3 device (Radiometer, Copenhagen, Denmark). One of the authors calibrated, placed, and maintained the monitor. Before placement, the electrode was cleaned, a new membrane applied, and calibration done according to the manufacturer’s recommendations using a calibration gas supply. The working temperature of the electrode was set at 45°C and the electrode was placed on the palmar surface of the forearm. The area where the electrode was placed was swabbed with alcohol before placement to facilitate adhesion of the disk to the skin. The electrode was removed, recalibrated, and replaced in a different location on the forearm every 2 h to avoid thermal injury. Samples were taken for arterial blood gas (ABG) analysis at 37°C as clinically indicated throughout the surgical procedure. When a sample was obtained, the FE'CO2 and TC-CO2 were recorded on the data sheet. The FE'CO2 to PaCO2 and TC-CO2 to PaCO2 differences were averaged for each patient and counted as one value. This was done to avoid biasing the data by over-representation of any one patient as the number of blood gases obtained from each patient was not equal.

The absolute difference between the FE'CO2 and the PaCO2 was compared with the absolute difference between the TC-CO2 and the PaCO2 with a non-paired, two-way t-test. A Fisher’s exact test and a two-way contingency table were used to compare the number of FE'CO2 vs TC-CO2 values whose absolute difference deviated 0.5 kPa or less and 1.0 kPa or less from the actual PaCO2 value. All data are expressed as the mean (SD) with P<0.05 considered significant.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We studied 30 adults (age from 18–54 yr, 41 (8) yr; weight from 115 to 267 kg, 161.8 (35.4) kg; BMI from 42 to 87 kg m–2, 57.3 (11.9) kg m–2). There were seven men and 23 women, and all underwent open gastric bypass. The absolute difference between the TC-CO2 and PaCO2 was 0.2 (0.2) kPa while the absolute difference between the FE'CO2 and PaCO2 was 0.7 (0.4) kPa (P<0.0001). The bias and precision were 0.1 (0.3) kPa for TC vs arterial carbon dioxide and –0.7 (0.4) kPa for FE'CO2 vs arterial carbon dioxide (Figs 1 and 2). The TC-CO2 was closer to the arterial carbon dioxide in 25 patients, the FE'CO2 was closer in four patients, and there was no difference in one patient.



View larger version (10K):
[in this window]
[in a new window]
 
Fig 1 Plot of the transcutaneous carbon dioxide minus the arterial carbon dioxide (y-axis) against the arterial carbon dioxide (x-axis). The bias and precision are labelled.

 


View larger version (10K):
[in this window]
[in a new window]
 
Fig 2 Plot of the FE'CO2 minus the arterial carbon dioxide (y-axis) against the arterial carbon dioxide (x-axis). The bias and precision are labelled.

 
Of the 30 averaged data sets, 27 of 30 (90%) TC-CO2 readings were 0.5 kPa or less from the PaCO2 while 13 of 30 (30%) FE'CO2 values were 0.5 kPa or less from the PaCO2 (P<0.00025). Thirty of thirty (100%) TC-CO2 readings were 1.0 kPa or less from the PaCO2 while 24 of 30 (80%) FE'CO2 values were 1.0 kPa or less from the PaCO2 (P=0.05).


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We found non-invasive carbon dioxide monitoring by using TC-CO2 was more accurate in patients with a BMI greater than 40 kg m–2. TC-CO2 monitoring had a lower mean absolute difference from the PaCO2 and more values within 0.5 and 1.0 kPa of the actual PaCO2. In six of the 30 patients, the average difference between the PaCO2 and the FE'CO2 was greater than 1.0 kPa. No previous studies have evaluated TC-CO2 monitoring in the severely obese. If ventilation-perfusion mismatch interferes with the gradient between the FE'CO2 and arterial CO2, TC-CO2 monitoring may be valuable.57 Data from this study further support the accuracy and clinical utility of TC monitoring during general anaesthesia in such situations.

The Bland–Altman plots show that the gradient between the FE'CO2 and the arterial CO2 was greatest with hypercarbia, and we presume that these patients had a greater degree of ventilation-perfusion inequality. We found no pattern between the alveolar-arterial oxygen gradient and the FE'CO2 to arterial carbon dioxide gradient. However, further study of the alveolar-arterial oxygen gradient and its relationship to the FE'CO2 to arterial carbon dioxide gradient may be warranted.

The success of TC-CO2 monitoring depends on monitor and patient factors. Although TC-CO2 monitoring gave a more accurate reflection of PaCO2 in most of the patients in the current study, several technical factors may affect this accuracy such as trapped air bubbles, improper placement, damaged membranes, and inappropriate calibration techniques. Patient-related factors may also affect the accuracy of monitoring, such as variations in skin thickness, the presence of oedema, tissue hypoperfusion, or the use of vasoconstricting drugs.5 810 This study shows that the TC-CO2 monitor can be used in the severely obese, who may have excessive s.c. adipose tissue. We could not measure skin thickness in our patients, but no tissue oedema was present, and no patients were given vasoactive agents or had poor skin perfusion.

To limit problems related to s.c. fat, we placed the probe on the ventral aspect of the forearm where s.c. tissue may be less than in other areas. We do not know of studies comparing the accuracy of TC-CO2 from different sites.

The type and manufacturer of the TC-CO2 monitor and the intrinsic ‘calibration factor’ used by the monitor can affect the accuracy of the technique. The TC-CO2 monitor of Radiometer, used in the current study, corrects for the alterations in carbon dioxide related to heating the skin to 45°C. With heating of the skin, the uncorrected TC-CO2 could be significantly greater than the PaCO2, which is measured at 37°C, because of the increased carbon dioxide production from the elevated local tissue metabolism and the altered solubility of carbon dioxide at higher temperatures.11 12 More information concerning these issues is available in the Radiometer Handbook (Radiometer, Copenhagen, Denmark).

Previous reports have described TC-CO2 monitoring in adults in the following circumstances: (i) intraoperatively during neurosurgical procedures and laparoscopy; (ii) mechanical ventilation in the ICU; (iii) non-invasive mechanical ventilation; (iv) sleep studies; (v) fibre-optic bronchoscopy; and (vi) in the immediate postoperative period.7 11 1316 Our results support previous studies using intraoperative TC-CO2 monitoring. Reid and colleagues17 compared FE'CO2 and TC-CO2 during general anaesthesia in 22 adults. The minute ventilation was adjusted at three different levels. Sixty-six data sets with the PaCO2 ranging from 28 to 62 mm Hg showed an FE' to arterial difference of 7.0 (3.1) mm Hg, and a TC to arterial difference of 2.3 (2.4) mm Hg (P<0.05). The differences were greatest at the higher PaCO2 values. Phan and colleagues18 also compared FE' and TC-CO2 during general anaesthesia in 24 adults. The bias and precision for the FE' to arterial comparison were –7.8 (6.1) and –1.6 (4.3) mm Hg, respectively, for the TC to arterial comparison.

The carbon dioxide monitors compared in the current study are not mutually exclusive. Although TC-CO2 is more accurate than FE'CO2, it needs more time to set up and has a longer response time (20–30 s) compared with the FE'CO2 device. Gas monitoring is widely available and provides additional useful information aside from the carbon dioxide value such as correct position of the tracheal tube, a ventilator disconnection, and analysis of the waveform.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Bhavani-Shankar K, Moseley H, Kumar AY, et al. Capnometry and anaesthesia. Can J Anaesth 1992; 39: 617–32[Abstract]

2 Pansard JL, Cholley B, Devilliers C, et al. Variation in arterial to end-tidal CO2 tension differences during anesthesia in the ‘kidney rest’ lateral decubitus position. Anesth Analg 1992; 75: 506–10[Abstract]

3 Grenier B, Verchere E, Meslie A, et al. Capnography monitoring during neurosurgery: reliability in relation to various intraoperative positions. Anesth Analg 1999; 88: 43–8[Abstract/Free Full Text]

4 Oberg B, Poulsen TD. Obesity: an anaesthetic challenge. Acta Anaesthesiol Scand 1996; 40: 191–200[ISI][Medline]

5 Tobias JD, Meyer DJ. Noninvasive monitoring of carbon dioxide during respiratory failure in toddlers and infants: end-tidal versus transcutaneous carbon dioxide. Anesth Analg 1997; 85: 55–8[Abstract]

6 Nosovitch MA, Johnson JO, Tobias JD. Noninvasive intraoperative monitoring of carbon dioxide in children: endtidal versus transcutaneous techniques. Paediatr Anaesth 2002; 12: 48–52[CrossRef][ISI][Medline]

7 McBride DS, Johnson JO, Tobias JD. Noninvasive carbon dioxide monitoring during neurosurgical procedures in adults: end-tidal versus transcutaneous techniques. South Med J 2002; 95: 870–4[ISI][Medline]

8 Rithalia SVS, Ng YN, Tinker J. Measurement of transcutaneous PCO2 in critically ill patients. Resuscitation 1982; 10: 13–18[ISI][Medline]

9 Rithalia SVS, Clutton-Brock TH, Tinker J. Characteristics of transcutaenous carbon dioxide tension monitors in normal adults and critically ill patients. Intensive Care Med 1984; 10: 149–53[ISI][Medline]

10 Shoemaker WC. Physiologic and clinical significance of PTC-O2 and PTC-CO2. Crit Care Med 1981; 9: 689–90[ISI]

11 Hasibeder W, Haisjackl M, Sparr H, et al. Factors influencing transcutaneous oxygen and carbon dioxide measurements in adult intensive care patients. Intensive Care Med 1991; 17: 272–5[ISI][Medline]

12 Drummond KJ, Fearnside MR, Chee A. Transcutaneous carbon dioxide measurement after craniotomy in spontaneously breathing patients. Neurosurgery 1997; 41: 361–7[ISI][Medline]

13 Bhavani-Shankar K, Steinbrook RA, Mushlin PS, et al. Transcutaneous PCO2 monitoring during laparoscopic cholecystectomy in pregnancy. Can J Anaesth 1998; 45: 164–9[Abstract]

14 Janssens JP, Howarth-Frey C, Chevrolet JC, et al. Transcutaneous PCO2 to monitor noninvasive mechanical ventilation in adults. Chest 1998; 113: 768–73[Abstract/Free Full Text]

15 Sanders MH, Kern NE, Costantino JP, et al. Accuracy of end-tidal and transcutaneous PCO2 monitoring during sleep. Chest 1994; 106: 472–83[Abstract]

16 Evans EN, Ganeshalingam K, Ebden P. Changes in oxygen saturation and transcutaneous carbon dioxide and oxygen levels in patients undergoing fiberoptic bronchoscopy. Respir Med 1998; 92: 739–42[ISI][Medline]

17 Reid CW, Martineau RJ, Miller DR, et al. A comparison of transcutaneous, end-tidal and arterial measurements of carbon dioxide during general anesthesia. Can J Anaesth 1992; 39: 31–6[Abstract]

18 Phan CQ, Tremper KK, Lee SE, Barker SJ. Noninvasive monitoring of carbon dioxide: a comparison of the partial pressure of transcutaneous and end-tidal carbon dioxide with the partial pressure of arterial carbon dioxide. J Clin Monit 1987; 3: 149–54[ISI][Medline]