Pediatric Pulmonology 45:481–486 (2010)

Accuracy of Three Transcutaneous Carbon Dioxide

Monitors in Critically Ill Children
Javier Urbano, MD,1,2* Verónica Cruzado, MD,2 Jesús López-Herce, MD, PhD,2
Jimena del Castillo, MD,2 José Marı́a Bellón, PhD,3 and Ángel Carrillo, MD, PhD2
Summary. Objectives: To study the accuracy of three devices for measuring transcutaneous CO2
tension in critically ill children. Methods: A prospective study comparing the values from three
transcutaneous CO2 monitors (SenTec1, TOSCA 5001, and TINA TCM31) with simultaneous
arterial CO2 (PaCO2) and end-tidal CO2 (EtCO2) values. Clinical data were collected from the
patients. Influence of core-skin temperature gradient and doses of inotropic drugs was evaluated.
Results: There were 62 samples from 41 critically ill children with ages between 2 and 192 months
(median, 18.5 months) and weights between 3.1 and 72 kg (median, 9 kg). The median PaCO2
was 42.5 mmHg (range, 28–85 mmHg). Transcutaneous CO2 (PtcCO2) values correlated
better with PaCO2 than with EtCO2. The correlation coefficient between PaCO2 and PtCO2
was 0.833 with the TINA TCM31 monitor, 0.931 with the SenTec1 monitor, and 0.765 with
the TOSCA 5001 monitor. The mean (SD) differences between the PaCO2 and PtcCO2 were
4.5 (3.7) mmHg, 4.3 (3.8) mmHg, and 5.6 (5.1) mmHg, respectively, with the three monitors, and
the differences between the PaCO2 and PtcCO2 were less than 7.5 mmHg in 77.7%, 81.2%, and
67.7% of the samples. Bland–Altman analysis showed a precision of
11.5 mmHg for TINA
TCM31 monitor,
10.6 mmHg for SenTec1 monitor, and
14.8 mmHg for TOSCA1 monitor. No
influence of core-skin temperature gradient and inotropic index on the differences between PaCO2
and PtcCO2 was observed. Conclusions: The three transcutaneous CO2 monitors have an
acceptable correlation with arterial CO2 tension and can be useful in critically ill children. Pediatr
Pulmonol. 2010; 45:481–486. ß 2010 Wiley-Liss, Inc.

Key words: transcutaneous carbon dioxide; ventilation; CO2; children.

INTRODUCTION Transcutaneous CO2 (PtcCO2) monitors are based on

arterialization of the capillary bed through the local
Arterial blood gas measurement is the gold standard for
application of heat and the use of Stow–Severinghaus
evaluating oxygenation and ventilation in critically ill
electrodes; they provide information on the transcuta-
patients. However, this is an invasive test that is not always
neous CO2 tension continuously and non-invasively.10,11
available in children and, furthermore, it carries certain
These monitors have been more extensively used in
risks such as infection, hemorrhage, anemia, and arterial
neonates as their skin is thinner and less keratinized, thus
vasospasm.1 In addition, arterial blood gases can only be
presenting a minimal resistance to gas diffusion;12 their
measured intermittently, and there is a delay between the
accuracy and reliability have been demonstrated in a
extraction of the sample and the availability of the results
in those pediatric intensive care units (PICU) that do not 1
Research Assistant Contract, Spanish Health Research Fund (Fondo de
have a gas analyzer in the unit. Investigaciones Sanitarias), Instituto de Salud Carlos III, Madrid, Spain.
Ideally, techniques should be available for the contin- 2
Pediatric Intensive Care Department, Hospital General Universitario
uous, non-invasive monitoring of oxygenation and Gregorio Marañón, Madrid, Spain.
ventilation. Pulse oximetry provides a rapid, reliable,
and non-invasive measurement of oxygen saturation,2 but 3
Preventive Medicine and Quality Management, Hospital General
there is no similar method for measuring carbon dioxide Universitario Gregorio Marañón, Madrid, Spain.
(CO2). Measurement of the end-tidal CO2 concentration
*Correspondence to: Javier Urbano, MD, Servicio de Cuidados Intensivos
(EtCO2) gives an approximate estimation of the partial Pediátricos, Hospital General Universitario Gregorio Marañón,
pressure of CO2 in arterial blood (PaCO2) in patients on C/Dr. Castelo 47, 28009 Madrid, Spain. E-mail: [email protected]
mechanical ventilation3 or spontaneous breathing.4 How-
ever, its reliability decreases with ventilation–perfusion Received 7 May 2009; Revised 1 December 2009; Accepted 8 December
mismatch.5,6 Thus, they are not useful in patients on high- 2009.
frequency7 or non-invasive ventilation8,9 without the DOI 10.1002/ppul.21203
need for the specialized equipment required for EtCO2 Published online 12 April 2010 in Wiley InterScience
monitoring. (www.interscience.wiley.com).

ß 2010 Wiley-Liss, Inc.

482 Urbano et al.

number of studies.13–15 A good correlation between tropic index ¼ dose of dopamine þ dobutamine þ [100 
PtcCO2 and PaCO2 has been found in adults8,16,17 and epinephrine] þ [100  norepinephrine] þ [15  milrinone]
children.18–22 Although other authors have not been able [in mg/kg/min]). The arterial blood gas samples were
to confirm these findings.23–26 analyzed immediately after extraction using the GEM
The objectives of this study were to investigate the Premier 40001 blood gas analyzer (Instrumentation
accuracy of three devices for measuring transcutaneous Laboratory, Barcelona, Spain) located in the PICU. The
CO2 tension in critically ill children and to compare the pH, PaO2, and PaCO2 were recorded after correction for
results with those of EtCO2. The influence of the core-skin the core temperature of the patient. EtCO2 was measured
temperature gradient and of the use of vasoactive drugs on using laser-based molecular correlation spectroscopy
the PtcCO2 measurements was also analyzed. (MCS) as the infrared emission source (Microstream1,
Oridion, Jerusalem, Israel), in a multiparametric monitor
(EtCO2 module M3015A, Phillips Intelivue MP70 neo-
PATIENTS AND METHODS
natal1, Phillips Medizinsysteme, Boeblingen, Germany).
This was a prospective, comparative study performed in This system uses a sidestream aspirator with a flow rate of
children admitted to a PICU. The study was approved by 50 ml/min, connected to a pipe (Filterline1, Oridion)
the Institutional Review Board. placed between the endotracheal tube and the remainder
Patients of 1 month to 16 years of age who required of a pediatric respirator system.
invasive monitoring of the blood pressure were enrolled in The statistical analysis of the results was performed
the study after obtaining informed consent from the using the SPSS 15.0 program (SPSS, Inc., Chicago, IL).
parents or legal guardians. The possible influence of age and cyanotic heart disease
Three PtcCO2 monitors were compared: TINA on PaCO2/PtcCO2 differences was tested by regression
TCM31, (Radiometer, Copenhagen, Denmark), SenTec1 analysis. The absolute difference between the non-
(Therwil, Arlasheim, BL, Switzerland), and TOSCA 5001 invasive methods (EtCO2 and PtcCO2) and PaCO2 was
(Linde Medical Sensors AG, Basel, Switzerland). Only one analyzed using a two-tailed Wilcoxon’s signed rank test
of the three PtcCO2 monitors was used in each patient. The for pairs, Kruskal–Wallis and Mann–Whitney U-tests.
device was calibrated at the start of the experiment and Bias (mean difference), 95% limit of agreement (mean
every 8 hours thereafter. The sensor was maintained on the difference
1.96 standard deviation), and plots were
same site for a maximum 8 hr, changing from one ear lobe to calculated as proposed by Bland and Altman, to
the other in the case of the SenTec1 and TOSCA 5001 demonstrate the relationship of the PaCO2 and PtcCO2
monitors and from one area of the chest or abdomen to with the mean and to evaluate the differences between the
another in the case of the TINA TCM31 monitor. three measurement systems.27 The concordance between
Arterial blood gas tests were only performed when the different measurements was analyzed using the
considered necessary for the clinical control patient; no intraclass correlation coefficient (ICC).28 Linear regres-
tests were performed exclusively for the purposes of the sion model was used to determine the influence of the
study. To avoid bias, the number of PtcCO2 measurements temperature gradient and inotropic index on the accuracy
was limited to a maximum of four per patient on of the measurements. An analysis was performed of the
different days during the admission. The samples were percentage of measurements with a difference less than or
drawn after the monitors had been functioning for equal to 7.5 mmHg (1 kPa) and 4.5 mmHg; these limits
a minimum of 20 min (at the end of the ‘‘QuickStart’’ were chosen based on previous studies.20,24,26 Chi-square
rapid arterialization function in the case of the TOSCA1 and Fishers’ tests were used to compare these proportions.
monitor, with the sensor at 448C) in order to ensure Statistical significance was taken as a P-value less than
arterialization of the skin. The working temperature of the 0.05.
sensors was maintained at 428C for TOSCA1 and
SenTec1, and 43.58C for TINA TCM31, in accordance
RESULTS
with manufacturer’s recommendations. A number of
parameters measured by the PtcCO2 monitors—heart A total of 62 samples were analyzed from 46 patients
rate, transcutaneous oxygen saturation, PtcCO2, sensor (61% male) with ages between 2 and 192 months (median,
temperature, perfusion index, and oxygen tension (only 18.5 months) and weights between 3.1 and 72 kg (median,
available on the TINA TCM31 monitor)—were recorded 9 kg). The diagnoses of the patients are shown in Table 1;
at the moment of drawing the arterial sample. In addition, 29.3% presented cyanotic heart disease and 65.9%
the demographic and diagnostic data of the patients required mechanical ventilation. The median gradient
were recorded, as well as the heart rate, blood pressure, between the core and skin temperatures was 5.58C (range,
core temperature (sensor in the urinary catheter), skin 0.8–13.08C). Continuous perfusions of vasoactive drugs
temperature (sensor on the sole of the foot), parameters were administered to 31 (75.6%) patients. The median
of mechanical ventilation, EtCO2, inotropic index (ino- inotropic index was 15.5 (range, 0–35.5). There were 14
Pediatric Pulmonology
 Transcutaneous CO2 in Critically Ill Children 483
TABLE 1— Diagnoses of the Patients The mean absolute difference (expressed as mean and
Diagnosis Number of patients %
95% confidence interval) between PaCO2 and EtCO2 (8.2
[6–10.5] mmHg) was greater than that found between the
Cardiac surgery 21 51.2 PaCO2 and PtcCO2 (4.9 [3.9–6.1] mmHg) (P ¼ 0.045).
Heart failure 7 17.1 The percentage of samples with PaCO2 –PtcCO2
Respiratory insufficiency 5 12.2
Sepsis 2 4.9 differences less than or equal to 7.5 mmHg (1 kPa) and
Other 6 14.6 to 4.5 mmHg are shown in Table 3. There were no
significant differences among the three PtcCO2 monitors
Other: multiple injury (n ¼ 1), postoperative neurosurgery (n ¼ 1), (P ¼ 0.613 and 0.492, respectively). A larger proportion of
postoperative scoliosis (n ¼ 1), postoperative abdominal surgery
(n ¼ 1), postoperative airway surgery (n ¼ 1), acute respiratory distress
samples showed differences of less than 4.5 or 7.5 mmHg
syndrome (n ¼ 1). between PaCO2 and PtcCO2 than between PaCO2 and
EtCO2 (Table 4).
Minimal changes in the slopes of the linear regression
models of PaCO2 and PtcCO2 were observed when
(34%) patients receiving epinephrine dose equal or greater including temperature gradient and inotropic index (data
than 0.1 mg/kg/min or dopamine dose equal or greater than not shown). The tolerance of the PtcCO2 sensors was
10 mg/kg/min (6 in the TINA1 TCM3 group, 4 in the good. Slight reddening of the skin was observed in two
TOSCA1 500 group, and 4 in the SenTec1 group). patients, though this was resolved completely in less than
Table 2 shows the characteristics of the patients grouped 12 hr.
according to the monitor used. Regression analysis
showed no influence of age or cyanotic heart disease on
DISCUSSION
PaCO2/PtcCO2 differences in each group.
The mean absolute difference (expressed as mean and This is the first study that has compared three systems
95% confidence interval) between PaCO2 and PtcCO2 was for the measurement of PtcCO2 in children. Our results
4.5 mmHg (2.8–6.2) for TINA1 TCM3 group, 5.6 mmHg show that the limits of agreement are over the level defined
(3.8–7.6) for TOSCA1 500 group, and 4.3 mmHg (2.4– a priori as clinically useful (maximum 7.5 mmHg),24,26
6.2) for SenTec1 group without significant difference and that the concordance was similar with the three
(P ¼ 0.330). The PaCO2, PtcCO2 values, ICC, bias, and monitors used. These results are in agreement with other
95% limits of agreement for each monitor are listed in studies in critically ill adults,24–26 and newborns23 with
Table 3. Figures 1–3 show the Bland–Altman plots of wide limits of agreement. No studies reporting these limits
PaCO2 and PtcCO2 for the three PtcCO2 monitors. of agreement have been found in children.
The ICC between PaCO2 and PtcCO2 was 0.833 Although the agreement between the PaCO2 and
(0.612–0.934) for TINA1 TCM3 group, 0.765 (0.557– PtcCO2 was poorer than that observed in other
0.883) for TOSCA1 500 group, and 0.931 (0.814–0.975) studies,8,16–18,20,22 a high percentage of values were
for SenTec1 group. In the overall group ICC was 0.847 within the limits of clinical reliability considered in other
(0.759–0.905) for PaCO2 –PtcCO2 and 0.649 (0.051– studies as acceptable for taking clinical decisions.18,24,26
0.857) for PaCO2 –EtCO2. In the samples with a PaCO2 This could therefore be sufficient for clinical management
greater than 45 mmHg, the ICC between PaCO2 and in the majority of patients. However, PtcCO2 cannot
PtcCO2 was 0.759 and between PaCO2 and EtCO2 was replace arterial blood gases in patients with severe lung
0.463. In patients with cyanotic heart disease, the ICC disease, raised intracranial pressure, or pulmonary hyper-
between PaCO2 and PtcCO2 was 0.816, and between tension, where greater accuracy is needed. In this sample,
PaCO2 and EtCO2 was 0.317. the percentage of patients with a difference less than

TABLE 2— Patient Characteristics

Parameter Overall population TINA TCM31 SenTec1 TOSCA 5001

Number of patients (number of samples) 41 (62) 11 (18) 11 (16) 19 (28)

Sex: male (%) 61.0 63.6 66.7 58.8
Age (months) 24 (2–192) 72 (5–192) 13 (2–156) 12 (2–166)
Weight (kg) 12 (3.1–72) 18 (5–72) 7.7 (3.1–34) 11 (4–65)
Cyanotic heart disease (%) 29.3 36.4 17.7 32.1
Mechanical ventilation (%) 65.9 81.8 66.7 59.2
Inotropic index 15.5 (0–36) 21 (0–35.5) 14.5 (0–28.5) 15.5 (0–36)
Temperature gradient (8C) 5.5 (0.8–13) 5.3 (2.4–11) 4.3 (1.3–8.2) 6.5 (0.8–13)

Results are expressed as median (range). Temperature gradient: difference between core and skin temperatures.

Pediatric Pulmonology
484 Urbano et al.
TABLE 3— Comparison between PaCO2 and PtcCO2

Parameter Overall population TINA TCM31 TOSCA 5001 SenTec1

PaCO2(mmHg) 44.0 (28–85) 48.7 (36–79) 47.7 (36–51) 40 (36–59)

PtcCO2 (mmHg) 42.1 (29–90) 48.5 (36–69) 41.5 (35–59) 41.3 (34–69)
Bias 0.4 0.9 1.5 2.1
95% limits of agreement
13.0
11.5
14.8
10.6
Diff. 7.5 mmHg 75% 77.7% 67.7% 81.2%1
Diff. 4.5 mmHg 57.7% 61.1% 50% 62.5%2

PaCO2, arterial partial pressure of CO2; PtcCO2, transcutaneous partial pressure of CO2 (results expressed as median and range); bias, mean
difference between PaCO2 and PtcCO2 (mmHg); 95% limits of agreement, 95% limit of agreement (
1.96 standard deviation of the mean difference
between PaCO2 and PtcCO2, in mmHg); Diff. 7.5 mmHg, percentage of samples with a difference less than or equal to 7.5 mmHg between PaCO2
and PtcCO2; Diff. 4.5 mmHg, percentage of samples with a difference less than or equal to 4.5 mmHg between PaCO2 and PtcCO2.
1
P ¼ 0.613.
2
P ¼ 0.492.

4.5 mmHg between PaCO2 and PtcCO2 was low when between PaCO2 and PtcCO2, as it has also been found
measured with a transcutaneous device, although it was by other authors.26,31,32 In the same way, inotrope-induced
even lower when measured with end tidal capnography. vasoconstriction could be expected to reduce the accuracy
Several factors may have influenced in these results. of transcutaneous monitoring. Poorer correlation between
Low cardiac output and hypothermia can be excluded PaCO2 and PtcCO2 in patients receiving dopamine at
because the patients in this study remained normothermic 20 mg/kg/min or epinephrine at 0.3 mg/kg/min has been
and hemodynamically stable. Cutaneous perfusion is a reported.21 In this study the inotropic index did not
determining factor in the accuracy of transcutaneous significantly affect PtcCO2 measurements, as observed by
monitoring, as some studies have demonstrated that.29,30 other authors.18,26 However, it must be realized that the
The application of heat to the skin surface increases blood inotropic index includes both vasoconstrictor and vaso-
flow, and capillary blood becomes arterialized, leading to dilator drugs and no patients were receiving doses as high
a closer correlation of the CO2 and oxygen values with as described before.21
those of arterial blood. In addition, heat alters the stratum Palmisano and Severinghaus32 found that the accuracy
corneum of the skin, facilitating diffusion of arterial of transcutaneous monitoring systems decreased in hyper-
gases.10–12 Despite the skin perfusion was not specifically capnic states; however, the majority of authors did not
assessed (i.e., capillary refill time), temperature gradient detect these variations,18,20,21,24 nor were they observed in
did not produce any deterioration in the correlation the present study.

Fig. 1. Bland–Altman plot of the arterial PCO2 –transcutaneous Fig. 2. Bland–Altman plot of the arterial PCO2 –transcutaneous
PCO2 for the TINA TCM31 monitor. The mean difference (‘‘bias’’) PCO2 for the TOSCA 5001 monitor. The mean difference (‘‘bias’’)
is represented as a continuous line, and 95% limits of agreement is represented as a continuous line, and 95% limits of agreement
are represented as dotted lines. are represented as dotted lines.

Pediatric Pulmonology
 Transcutaneous CO2 in Critically Ill Children 485
34,35
value. However, its accuracy varies in patients with
abnormalities of lung function5,20 or congenital cyanotic
heart disease.6 In our study, there was a greater difference
and poorer correlation between the PaCO2 and EtCO2 than
between the PaCO2 and PtcCO2, as has been reported by
other authors.18,20,36 In addition, there was no correlation
between EtCO2 and PaCO2 in patients with cyanotic heart
disease.
The advantages of PtcCO2 over EtCO2, apart from its
greater accuracy, are the possibility for use in patients on
high-frequency oscillatory ventilation,19 non-invasive
ventilation,8 and high-flow through nasal cannulae oxygen
therapy systems. However, PtcCO2 monitoring systems
require regular calibration and represent a greater work-
load for the nursing staff.18 They may also be affected by
edema and tissue hypoperfusion.1
Its non-invasive character and continuous readings are
Fig. 3. Bland–Altman plot of the arterial PCO2 –transcutaneous
PCO2 for the SenTec1 monitor. The mean difference (‘‘bias’’) is
its main advantages compared to arterial blood gases.
represented as a continuous line, and 95% limits of agreement PtcCO2 cannot replace arterial blood gases in patients with
are represented as dotted lines. severe lung disease, intracranial or pulmonary hyper-
tension, but it can be very useful to detect early changes in
ventilation, facilitate clinical management, and reduce the
number of arterial blood samples. This technique can also
The agreement in PtcCO2 measurements among the be used to monitor non-invasive home ventilation.37
three monitors used was similar, with slightly better The main limitations of our study were that the sample
accuracy of SenTec1 monitor. size was relatively small, which could affect the level of
EtCO2 monitoring is useful in patients on mechanical significance and power of the study and that, for technical
ventilation. Capnography is used to confirm airway and ethical reasons, simultaneous or randomized compar-
patency and pulmonary ventilation. It provides early ison of the three monitors in all patients was not
warning of airway obstruction and respiratory compro- performed. Despite the sample is small, this is the first
mise in intubated and non-intubated patients. Capnog- study comparing three transcutaneous monitors in a
raphy may aid the management of patients during pediatric population.
transport and other situations requiring strict control of In conclusion, the three PtcCO2 measurement systems
ventilation, such as raised intracranial pressure or analyzed do not provide an acceptably reliable estimation
pulmonary hypertension.33 Also, measurement of EtCO2 of the PaCO2 in critically ill children. However, a high
during cardiopulmonary resuscitation is a signal of return percentage of values were within the limits of clinical
of spontaneous circulation and may have a predictive reliability considered acceptable. In our sample, the
measurement is not affected by the core-skin temperature
gradient or by the use of inotropic or vasoactive drugs.
The choice of non-invasive CO2 monitoring systems
TABLE 4— Comparison Between PaCO2, PtcCO2, and
EtCO2 must be made on an individual basis, evaluating the
possible correlation with arterial blood gases in each
PaCO2 –EtCO2 PaCO2 –PtcCO2 patient. PtcCO2 monitoring is particularly indicated in
Diff. 7.5 mmHg 50.3% 75%1 children with lung disease or cyanotic heart disease,
Diff. 4.5 mmHg 35.3% 57.7%2 situations in which the EtCO2 is not so reliable, and in
Bias (mmHg) 8.3 0.4 patients who are not intubated or in whom respiratory
Precision (mmHg)
14.5
13.0 techniques are being used in which it is not possible to
PaCO2, arterial partial pressure of CO2; PtcCO2, transcutaneous partial measure the EtCO2.
pressure of CO2; EtCO2, end-tidal concentration of CO2; Diff.
7.5 mmHg, percentage of samples with a difference less than or equal
to 7.5 mmHg; Diff. 4.5 mmHg, percentage of samples with a ACKNOWLEDGMENTS
difference less than or equal to 4.5 mmHg; bias, mean difference; The authors thank the nursing staff of the Intensive Care
precision, 95% limit of agreement (mean difference
1.96 standard
deviation). Department of the Gregorio Marañón University General
1
P ¼ 0.016. Hospital for their collaboration in the performance of this
2
P ¼ 0.011. study, and the families of the patients.
Pediatric Pulmonology
486 Urbano et al.

REFERENCES 19. Berkenbosch JW, Tobias JD. Transcutaneous carbon dioxide

monitoring during high-frequency oscillatory ventilation in
1. Tobias JD. Transcutaneous carbon dioxide monitoring in infants infants and children. Crit Care Med 2002;30:1024–1027.
and children. Paediatr Anaesth 2009;19:434–444. 20. Tobias JD, Meyer DJ. Noninvasive monitoring of carbon dioxide
2. Shrake K, Blonshine S, Brown R, Crapo R, Martineau K, Ruppell during respiratory failure in toddlers and infants: end-tidal
G, Wagner J. AARC (American Association for Respiratory Care) versus transcutaneous carbon dioxide. Anesth Analg 1997;85:
clinical practice guideline: pulse oximetry. Respir Care 55–58.
1991;36:1406–1409. 21. Tobias JD, Wilson WR, Meyer DJ. Transcutaneous monitoring of
3. McDonald MJ, Montgomery VL, Cerrito PB, Parrish CJ, Boland carbon dioxide tension after cardiothoracic surgery in infants and
KA, Sullivan JE. Comparison of end-tidal CO2 and PaCO2 in children. Anesth Analg 1999;88:531–534.
children receiving mechanical ventilation. Pediatr Crit Care Med 22. Tschupp A, Fanconi SA. Combined ear sensor for pulse oximetry
2002;3:244–249. and carbon dioxide tension monitoring: accuracy in critically ill
4. Tobias JD, Flanagan JF, Wheeler TJ, Garrett JS, Burney C. children. Anesth Analg 2003;96:82–84.
Noninvasive monitoring of end-tidal CO2 via nasal cannulas in 23. Lacerenza S, De Caroli MP, Fusco FP, La Torre G, Chiaradia G,
spontaneously breathing children during the perioperative period. Romagnoli C. An evaluation of a new combined SpO2/PtcCO2
Crit Care Med 1994;22:1805–1808. sensor in very low birth weight infants. Anesth Analg 2008;107:
5. Yamakana MK, Sue DY. Comparison of arterial–end tidal pCO2 125–129.
difference and dead space/tidal volume ratio in respiratory failure. 24. Bendjelid K, Schütz N, Stotz M, Gerard I, Suter PM, Romand JA.
Chest 1987;92:832–835. Transcutaneous PCO2 monitoring in critically ill adults: clinical
6. Short JA, Paris ST, Booker PD, Fletcher R. Arterial to end tidal evaluation of a new sensor. Crit Care Med 2005;33:2203–2206.
carbon dioxide tension difference in children with congenital 25. Rodrı́guez P, Lellouche F, Aboab J, Brunn Buisson C, Brochard L.
heart disease. Br J Anesth 2001;86:349–353. Transcutaneous arterial carbon dioxide pressure monitoring in
7. Nishimura M, Imanaka H, Tashiro C, Taenaka N, Yoshiya I. critically ill adult patients. Intensive Care Med 2006;32:309–312.
Capnometry during high-frequency oscillatory ventilation. Chest 26. Bolliger D, Steiner LA, Kasper J, Aziz OA, Filipovic M, Seeberge
1987;92:832–835. MD. The accuracy of non-invasive carbon dioxide monitoring: a
8. Janssens JP, Howarth-Frey C, Chevrolet JC, Abajo B, Rochat T. clinical evaluation of two transcutaneous systems. Anaesthesia
Transcutaneous pCO2 to monitor non-invasive mechanical 2007;62:394–399.
ventilation in adults: assessment of a new transcutaneous pCO2 27. Bland JM, Altman DG. Statistical methods for assessing agree-
device. Chest 1998;113:769–773. ment between two methods of clinical measurement. Lancet
9. Sanders MH, Kern NB, Costantino JP, Stiller RA, Strollo PJ, Jr., 1986;1:307–310.
Studnicki KA, Coates JA, Richards TJ. Accuracy of end-tidal and 28. Prieto L, Lamarca R, Casado A. Assessment of the reliability of
transcutaneous PCO2 monitoring during sleep. Chest 1994;106: clinical findings: the intraclass correlation coefficient. Med Clin
472–483. (Barc) 1998;110:142–145.
10. Severinghaus JW. The current status of transcutaneous blood gas 29. Hazinski TA, Severinghaus JW. Transcutaneous analysis of
analysis and monitoring. Blood Gas News 1998;7:4. arterial PCO2. Med Instrum 1982;16:150–153.
11. Severinghaus JW, Bradley AF. Electrodes for blood pO2 and 30. Sivan Y, Eldadah MK, Cheah TE, Newth TJ. Estimation of
pCO2 determination. J Appl Physiol 1958;13:515–520. arterial carbon dioxide by end-tidal and transcutaneous PCO2
12. Franklin ML. Transcutaneous measurement of partial pressure of measurements in ventilated children. Pediatr Pulmonol 1992;12:
oxygen and carbon dioxide. Respir Care Clin North Am 1995;1: 153–157.
119–131. 31. Carter BG, Wiwczaruk M, Hochmann M, Osborne A, Henning R.
13. Löfgren O, Andersson D. Simultaneous transcutaneous carbon Performance of transcutaneous PCO2 and pulse oximetry
dioxide and transcutaneous oxygen monitoring in neonatal monitors in newborns and infants after cardiac surgery. Anaesth
intensive care. J Perinat Med 1983;11:51–56. Intensive Care 2001;29:260–265.
14. Binder N, Atherton H, Thorkelsson T, Hoath SB. Measurement of 32. Palmisano BW, Severinghaus J. Transcutaneous PCO2 and PO2: a
transcutaneous carbon dioxide in low birthweight infants during multicenter study of accuracy. J Clin Monit 1990;6:189–195.
the first two weeks of life. Am J Perinatol 1994;11:237–241. 33. Anderson CT, Breen PH. Carbon dioxide kinetics and capnog-
15. Bernet-Buettiker V, Ugarte MJ, Frey B, Hug MI, Baenziger O, raphy during critical care. Crit Care 2000;4:207–215.
Weiss M. Evaluation of a new combined transcutaneous measure- 34. Falk JL, Rackow EC, Weil MH. End-tidal carbon dioxide
ment of PCO2/pulse oximetry oxygen saturation ear sensor in concentration during cardiopulmonary resuscitation. N Engl J
newborn patients. Pediatrics 2005;115:e64–e68. Med 1988;318:607–611.
16. Senn O, Clarenbach CF, Kaplan V, Maggiorini M, Bloch KE. 35. Asplin BR, White RD. Prognostic value of end-tidal carbon
Monitoring carbon dioxide tension and arterial oxygen saturation dioxide pressures during out-of-hospital cardiac arrest. Ann
by a single ear lobe sensor in patients with critical illness or severe Emerg Med 1995;25:756–761.
apnea. Chest 2005;128:1291–1296. 36. Nosovitch MA, Johnson JO, Tobias JD. Noninvasive intra-
17. Herrejón A, Inchaurraga I, Palop J, Ponce S, Peris R, Terrádez M, operative monitoring of carbon dioxide in children: end-tidal
Blanquer R. Usefulness of transcutaneous carbon dioxide versus transcutaneous techniques. Paediatr Anaesth 2002;12:48–
pressure monitoring to measure blood gases in adults hospitalized 52.
for respiratory disease. Arch Bronconeumol 2006;42:225–229. 37. Paiva R, Krivec U, Aubertin G, Cohen E, Clément A, Fauroux B.
18. Berkenbosch JW, Lam J, Burd RS, Tobias JD. Non invasive Carbon dioxide monitoring during long-term noninvasive respi-
monitoring of carbon dioxide during mechanical ventilation in ratory support in children. Intensive Care Med 2009;35:1068–
older children: end tidal versus transcutaneous techniques. Anesth 1074.
Analg 2001;92:427–431.

Pediatric Pulmonology