Share this article on:

The Arterial to End-Tidal Carbon Dioxide Gradient Increases with Uncorrected but Not with Temperature-Corrected PaCO2 Determination During Mild to Moderate Hypothermia

Sitzwohl, Christian MD; Kettner, Stephan C. MD; Reinprecht, Andrea MD; Dietrich, Wolfgang MD; Klimscha, Walter MD; Fridrich, Peter MD; Sladen, Robert N. MD; Illievich, Udo M. MD

doi: 10.1213/00000539-199805000-00043
General Articles

End-tidal carbon dioxide (PETCO2) monitoring is recommended as a basic standard of care and is helpful in adjusting mechanical ventilation. Gas solubility changes with temperature, which might affect the PaCO2 and thereby the gradient between Paco2 and PETCO2 (PA-ETCO2) under hypothermic conditions. We investigated whether the PA-ETCO2 changes during mild to moderate hypothermia (36[degree sign]C-32[degree sign]C) using PaCO2 measured at 37[degree sign]C (uncorrected PaCO2) and PaCO2 corrected to actual body temperature. We preoperatively investigated 19 patients. After anesthesia had been induced, controlled ventilation was established to maintain normocarbia using constant uncorrected PaCO (2) to adjust ventilation (alpha-stat acid-base regimen). Body core temperature was reduced without surgical intervention to 32[degree sign]C by surface cooling. Continuous PETCO2 was monitored with a mainstream PETCO2 module. The PA-ETCO (2) was calculated using the uncorrected and corrected PaCO2 values. During body temperature reduction from 36[degree sign]C to 32[degree sign]C, the gradient between PETCO2 and uncorrected PaCO2 increased 2.5-fold, from 4.1 +/- 3.7 to 10.4 +/- 3.8 mm Hg (P < 0.002). The PA-ETCO2 remained unchanged when the corrected Paco2 was used for the calculation. We conclude that when the alpha-stat acid-base regimen is used to adjust ventilation, the PA-ETCO2 calculated with the uncorrected PaCO2 increases and should be added to the differential diagnosis of widened PA-ETCO2. In contrast, when the corrected PaCO2 is used for the calculation of the PA-ETCO2, the PA-ETCO2 remains unaltered during hypothermia. Implications: We investigated the impact of induced hypothermia (36[degree sign]C-32[degree sign]C) on the gradient between PaCO2 and PETCO2 (PA-ETCO (2)). The PA-ETCO2 increased 2.5-fold when CO2 determinations were not temperature-corrected. Hypothermia should be added to the differential diagnosis of an increased PA-ETCO2 when the alpha-stat acid-base regimen is used.

(Anesth Analg 1998;86:1131-6)

Departments of (Sitzwohl, Kettner, Klimscha, Fridrich, Illievich) Anesthesiology and General Intensive Care and (Reinprecht, Dietrich) Neurosurgery, University of Vienna, Vienna, Austria; and (Sladen) Department of Anesthesiology, Division of Critical Care, Columbia-Presbyterian Medical Center, New York, New York.

Accepted for publication February 4, 1998.

Address correspondence and reprint requests to Christian Sitzwohl, MD, Department of Anesthesiology and General Intensive Care, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria.

Quantitative assessment of PETCO2 is simple and noninvasive, and it is recommended as a basic standard of care in anesthetic practice [1]. In anesthetized, supine, normothermic patients, a 4- to 6-mm Hg gradient of PaCO2 to PETCO2 (PA-ETCO2) is normal [2-5]. Given this small gradient, continuous monitoring of PETCO2 allows on-line adjustment of mechanical ventilation [6-8]. Additionally, PETCO2 correlates with pulmonary blood flow [9,10] and is therefore helpful in detecting changes in the ventilation-perfusion ratio of the lung.

Accidental and deliberate hypothermia are common in anesthetic practice. The use of therapeutic mild (36.5[degree sign]C-34[degree sign]C) and moderate (33.5[degree sign]C-28.5[degree sign]C) hypothermia is of interest in patients with head trauma [11] and in those with increased intracranial pressure [12,13].

However, hypothermia increases the solubility of CO2 and thereby decreases the partial pressure of CO2 for a given CO2 content of blood [14]. At body temperatures lower than 37[degree sign]C, the PaCO2 can be measured at 37[degree sign]C (uncorrected), or measured values can be corrected to actual body temperature. A letter to the editor described the PA-ETCO2 during body temperature reduction [15], but no study has quantified the effect of hypothermia on the PA-ETCO2 when the uncorrected PaCO2 is used to adjust controlled ventilation, which reflects the use of alpha-stat as an acid-base regimen. Changes in the gradient between uncorrected PaCO2 and PETCO2 could lead to inadequate adjustment of controlled ventilation in hypothermic patients managed with the alpha-stat regimen.

Therefore, we designed a study to investigate the effect of graded hypothermia (36[degree sign]C-32[degree sign]C) on the PA-ETCO2 calculated with the uncorrected PaCO2 when the alpha-stat acid-base regimen is used to adjust ventilation. Furthermore, we compared the effect of using uncorrected PaCO2 and corrected PaCO2 on the PA-ETCO2.

Back to Top | Article Outline


The study protocol was approved by our ethics committee, and written, informed consent was obtained. We investigated 19 ASA physical status I or II patients undergoing elective intracranial surgery with anticipated surgical duration longer than 4 h. Patients with pulmonary disease were excluded from the investigation.

During the 6 h preceding the study period, the patients received an IV infusion of 2 mL [center dot] kg-1 [center dot] h-1 isotonic sodium chloride solution. Four to five hours before the scheduled surgery, the patients arrived at the neurosurgical intensive care unit, and routine monitoring, including electrocardiogram, noninvasive blood pressure, and pulse oximetry (Component Monitoring System, Hewlett Packard, Palo Alto, CA), was applied.

Anesthesia was induced with bolus doses of 0.1 mg/kg midazolam, 2 mg/kg propofol, and 2-4 [micro sign]g/kg fentanyl. Orotracheal intubation was facilitated by the administration of 0.1 mg/kg vecuronium. Continuous positive pressure ventilation (Evita I, Draeger, Lubeck, Germany) was provided with a positive end-expiratory pressure of 5 cm H2 O. Basic ventilatory settings were a respiratory frequency of 12 breaths/min and a tidal volume of 8 mL/kg body weight, which was adjusted to maintain normocarbia based on the uncorrected PaCO2 (alpha-stat acid-base regimen). The fraction of inspired oxygen was 0.3 (air/oxygen). Anesthesia was maintained by a continuous infusion of 8-12 mg [center dot] kg-1 [center dot] h-1 propofol, 1.5-4 [micro sign] g [center dot] kg-1 [center dot] h-1 fentanyl, and 0.1-0.2 mg [center dot] kg-1 [center dot] h-1 vecuronium. After the induction of anesthesia, an arterial catheter was inserted into the radial artery. To maintain normovolemia, 2-4 mL [center dot] kg-1 [center dot] h-1 isotonic sodium chloride solution was infused during the study period.

To measure body core temperature, a temperature probe was inserted into the distal esophagus, and correct positioning was verified by chest radiograph. Mild to moderate hypothermia was induced preoperatively by surface cooling. The patients were covered with a forced-air cooling blanket (Polar Air, Augustine Medical Inc., Eden Prarie, MN) with the temperature set to 10[degree sign]C. Additionally, the patients were placed on a cooling water mattress (Blanketroll II; Sub-Zero, Cincinnati, OH) set to 10[degree sign]C. To avoid thermal damage to the skin, three blankets were placed between the cooling mattress and the patient. Body temperature was reduced to 32[degree sign]C after the induction of anesthesia and before surgical intervention. To avoid an afterdrop (i.e., unintentional body temperature reduction below target temperature), active cooling was stopped at a body temperature 32.5[degree sign]C-33[degree sign]C.

At 36[degree sign]C, 34[degree sign]C, and 32[degree sign]C, patients were hypo- and hyperventilated. Before hypo- and hyperventilation, arterial blood samples were obtained and, if necessary, the ventilator was adjusted to ensure a normocarbic uncorrected PaCO2 (35-40 mm Hg). At each temperature, the ventilation rate was decreased to 8 breaths/min to achieve hypercarbia (uncorrected PaCO2 40-44 mm Hg), followed by an increase to 16 breaths/min to achieve hypocarbia (uncorrected PaCO2 28-32 mm Hg). Based on the assessment of the PA-ETCO2 during normoventilation, PETCO2 was used as a guide to adjust the tidal volume to achieve the target uncorrected PaCO2. After an equilibration period of 5 min after the resetting of the ventilator, arterial blood samples were drawn. If the target uncorrected PaCO (2) was not reached, the ventilator was reset, and another equilibration period of 5 min was allowed to assure steady-state conditions. When the target uncorrected PaCO2 was achieved, PETCO2, arterial blood pressure, and body temperature were recorded simultaneously.

PETCO2 was continuously monitored by a mainstream infrared carbon dioxide light absorption device (M1016A CMS, Hewlett Packard) that was calibrated immediately before use and inserted between the endotracheal tube and the breathing circuit.

Arterial blood samples were drawn from the arterial catheter into a preheparinized syringe (Aspitator, Marquest, Englewood, CO) and analyzed for pH, oxygen, and PaCO (2) at 37[degree sign]C (Radiometer ABL330, Radiometer, Copenhagen, Norway). Temperature correction of Paco2 was performed off-line according to the formula of Severinghaus (reviewed in [16]).

Data are presented as means +/- SD and were analyzed using the Friedman statistic and Wilcoxon's signed rank test (StatView 4.51, Abacus Inc., Berkeley, CA) to test the significance of a change in the PA-ETCO2 with body temperature reduction. Additional regression analysis was performed to determine the association between PaCO2 and PETCO2. Differences were considered significant at P < 0.05.

Back to Top | Article Outline


Nineteen patients (7 male, 12 female; age 36 +/- 11 yr, weight 71 +/- 16 kg) were included in this study. The mean time required to achieve a body temperature of 32[degree sign]C was 171 +/- 63 min. Because of continuous cooling and the time required for equilibration after resetting the ventilator, not all comparisons were performed at the exact target temperature (Table 1). The 164 comparisons of PaCO (2) with PETCO2 were performed using both uncorrected and corrected PaCO2 (Table 1).

Table 1

Table 1

PETCO2 decreased significantly with all modes of ventilation during body temperature reduction while PaCO2 was kept constant using the alpha-stat regimen to adjust ventilation (Figure 1). The PA-ETCO2 was not influenced by the mode of ventilation. Therefore, data for normoventilation, hyperventilation, and hypoventilation at the different temperatures were pooled. Calculating the PA-ETCO (2) with uncorrected PaCO2 resulted in an increase of PA-ETCO2 in all patients during body temperature reduction. In contrast, PA-ETCO2 did not change when corrected PaCO2 was used for the calculation (Figure 2). Of note, the inter-individual variability of PA-ETCO2 was large, whether the uncorrected or the corrected PaCO2 was used for calculation (Table 2).

Figure 1

Figure 1

Figure 2

Figure 2

Table 2

Table 2

PaCO2 correlated with PETCO2 at all temperatures, whether uncorrected or corrected PaCO2 was used (Figure 3 and Figure 4). However, the correlation was more strongly statistically significant with the corrected than with the uncorrected PaCO2 (P = 0.024).

Figure 3

Figure 3

Figure 4

Figure 4

Mean arterial blood pressure increased when body temperature was decreased from 36[degree sign]C to 32[degree sign]C (Table 1). There was no correlation between mean arterial pressures and PETCO2 (R = 0.25, R2 = 0.06).

Back to Top | Article Outline


The physiologic basis for the PA-ETCO2 is the inhomogenous ventilation-perfusion ratio throughout the lung. The partial pressure of CO2 in exhaled gas from perfused alveoli closely approximates the PaCO2. Gas from ventilated alveoli without perfusion contains no CO2. Because end-expiratory gas is a mixture of air from both well and poorly perfused alveoli, PETCO2 is usually less than PaCO2. According to Nunn and Hill [5], the PA-ETCO2 is 4.6 +/- 2.5 mm Hg in healthy, anesthetized patients. This finding was confirmed by numerous investigators under various conditions, such as during ventilatory weaning [6], in critically ill and mechanically ventilated children [7], and during laparoscopic surgery [17]. Our study demonstrates a similar PA-ETCO2 at normothermia.

Hypothermia increases the solubility of CO2 and the binding of CO (2) to hemoglobin, thereby decreasing the PaCO2 for a given blood CO2 content [14]. At body temperatures lower than 37[degree sign]C, the PaCO2 can be measured at 37[degree sign]C (uncorrected PaCO2), or measured values can be corrected to actual body temperature (corrected PaCO2). The difference in PA-ETCO2 calculated with the corrected or the uncorrected PaCO2 during hypothermia has been previously described in one patient in a letter to the editor [15]. The author concludes that PETCO2 can predict the temperature-corrected, but not the uncorrected, PaCO2, which is similar to our findings. However, the author does not state whether the patient was managed following the pH-stat or alpha-stat regimen, whether measurements were made under steady-state conditions, or whether mainstream or sidestream technology was used for PETCO2 measurements. The author recommends that PETCO2 be decreased approximately 4.5% per degree Celsius decrease in body temperature when the alpha-stat regimen is used. Our findings in 19 patients suggest a lesser adjustment of PETCO2 (approximately 3.6% per degree Celsius decrease of body temperature) to obtain a normocarbic uncorrected PaCO2.

In this study, we quantified the impact of the use of uncorrected PaCO (2) on the PA-ETCO2 during mild to moderate hypothermia with the alpha-stat regimen. Our data reveal that a 2[degree sign]C body temperature reduction results in an approximately 100% increase in the PA-ETCO2 during the alpha-stat regimen and when the uncorrected PaCO2 is used for calculation. A further 2[degree sign]C body temperature reduction increases the PA-ETCO2 by an additional 50%.

It could certainly be argued that the PA-ETCO2 calculated with the corrected PaCO2 reflects the physically true PA-ETCO2, and that the use of the uncorrected PaCO2 for the calculation creates an artifactual PA-ETCO (2). However, the alpha-stat regimen has strong proponents for its use during hypothermia, whether during cardiopulmonary bypass or neurosurgery [18,19]. We therefore consider the PA-ETCO2 calculated with uncorrected PaCO2 clinically relevant. It is the PaCO2 used for the alpha-stat management, and the resulting PA-ETCO2 is seen at the bedside of the patient.

The stability of the PA-ETCO2 during anesthesia is controversial. Some studies have found a stable PA-ETCO2 throughout the maintenance of anesthesia [20] in children undergoing heart surgery with acyanotic-shunting and mixing congenital heart lesions [21] or during laparoscopic surgery [17]. In contrast, other studies have suggested that the PA-ETCO2 is highly variable and that PETCO2 does not provide a stable reflection of PaCO2 in patients with head trauma [22], during craniotomies [23], and in patients with multisystem trauma [24]. The latter studies are notable for having used the alpha-stat regimen for acid-base management and the uncorrected PaCO2 for the calculation of the PA-ETCO2. Our data showed that the correlation of PETCO2 with the uncorrected PaCO (2) is weaker than that with the corrected PaCO2. This suggests that the variability in PA-ETCO2 and the poor correlation between PaCO2 and PETCO2 found in these studies could have been the result of fluctuations in temperature.

The PA-ETCO2 can be used as an index of alveolar dead space, which is increased in conditions such as pulmonary embolism and atelectasis [25]. Our study demonstrates that during hypothermia, the use of uncorrected PaCO2 for the calculation of the PA-ETCO2 must be added to the differential diagnosis of increased PA-ETCO2.

A potential limitation of our study is the change in mean arterial blood pressure during the reduction of body temperature, which might indicate a change in cardiac output. However, the poor correlation between mean arterial blood pressure and PETCO2 (R2=0.06) indicates that the changes in blood pressure did not affect the PETCO2.

In conclusion, our study demonstrates that the use of the uncorrected PaCO2 for the calculation of the PA-ETCO2 during the alpha-stat acid-base regimen and body temperature reduction results in an increased PA-ETCO2. In contrast, PA-ETCO2 did not change during induced hypothermia when the corrected PaCO2 was used for the calculation. Furthermore, the correlation of PETCO2 with the corrected PaCO2 was stronger than that with the uncorrected PaCO (2). Thus, PETCO2 reflects the corrected PaCO2 more adequately than the uncorrected PaCO2 using the alpha-stat acid-base regimen and body temperature reduction. If PETCO2 is used for correcting ventilation in the hypothermic patient managed with the alpha-stat regimen, the increasing uncorrected PA-ETCO2 must be considered. PETCO2 should be decreased approximately 3.6% per degree Celsius decrease of body temperature to prevent erroneously low uncorrected PaCO2 values. This is of clinical importance in situations in which precise control of CO2 is necessary, such as in patients with increased intracranial pressure or those undergoing neurosurgical procedures.

We thank Mr. Fritz Buschacher for all his help in the laboratory and Dr. Schmetterer for his help with the CO2 physics.

Back to Top | Article Outline


1. American Society of Anesthesiologists. Standards for basic anesthetic monitoring. In: Directory for Members. Washington, DC: ASA, 1997:394.
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.
3. Phan CQ, Tremper KK, Lee SE, et al. 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.
4. Weinger MB, Brimm JE. End-tidal carbon dioxide as a measure of arterial carbon dioxide during intermittent mandatory ventilation. J Clin Monit 1987;3:73-9.
5. Nunn J, Hill D. Respiratory deadspace and arterial to end tidal CO (2) tension difference in anesthetized man. J Appl Physiol 1960;15:383-9.
6. Healey CJ, Fedullo AJ, Swinburne AJ, et al. Comparison of noninvasive measurements of carbon dioxide tension during withdrawal from mechanical ventilation. Crit Care Med 1987;15:764-8.
7. Sivan Y, Eldadah MK, Cheah TE, et al. Estimation of arterial carbon dioxide by end-tidal and transcutaneous Pco2 measurements in ventilated children. Pediatr Pulmonol 1992;12:153-7.
8. Sharma SK, McGuire GP, Cruise CJ. Stability of the arterial to end-tidal carbon dioxide difference during anaesthesia for prolonged neurosurgical procedures. Can J Anaesth 1995;42:498-503.
9. Gudipati CV, Weil MH, Bisera J, et al. Expired carbon dioxide: a noninvasive monitor of cardiopulmonary resuscitation. Circulation 1988;77:234-9.
10. Weil MH, Bisera J, Trevino RT, et al. Cardiac output and end-tidal carbon dioxide. Crit Care Med 1985;13:907-11.
11. Marion DW, Penrod LE, Kelsey SF, et al. Treatment of traumatic brain injury with moderate hypothermia. N Engl J Med 1997;336:540-6.
12. Shiozaki T, Sugimoto H, Taneda M, et al. Effect of mild hypothermia on uncontrollable intracranial hypertension after severe head injury. J Neurosurg 1993;79:363-8.
13. Marion DW, Obrist WD, Carlier PM, et al. The use of moderate therapeutic hypothermia for patients with severe head injuries: a preliminary report. J Neurosurg 1993;79:354-62.
14. Kofstad J. Blood gases and hypothermia: some theoretical and practical considerations. Scand J Clin Invest 1996;56:21-6.
15. Baraka A. Does end-tidal PCO2 predict the temperature-uncorrected or corrected PaCO2 during hypothermia? [letter]. Anesth Analg 1995;80:208.
16. Baraka AS, Baroody MA, Haroun ST, et al. Effect of alpha-stat versus pH-stat strategy on oxyhemoglobin dissociation and whole-body oxygen consumption during hypothermic cardiopulmonary bypass. Anesth Analg 1992;74:32-7.
17. Nyarwaya JB, Mazoit JX, Samii K. Are pulse oximetry and end-tidal carbon dioxide tension monitoring reliable during laparoscopic surgery? Anaesthesia 1994;49:775-8.
18. Patel RL, Turtle MR, Chambers DJ, et al. Alpha-stat acid-base regulation during cardiopulmonary bypass improves neuropsychologic outcome in patients undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg 1996;111:1267-79.
19. Murkin JM. The role of CPB management in neurobehavioral outcomes after cardiac surgery. Ann Thorac Surg 1995;59:1308-11.
20. Whitesell R, Asiddao C, Gollmann D, et al. Relationship between arterial and peak expired carbon dioxide pressure during anesthesia and factors influencing the difference. Anesth Analg 1981;66:508-12.
21. Lazzell VA, Burrows FA, McNulty SE, et al. Stability of the intraoperative arterial to end-tidal carbon dioxide partial pressure difference in children with congenital heart disease. Can J Anaesth 1991;38:859-65.
22. Kerr ME, Zempsky J, Sereika S, et al. Relationship between arterial carbon dioxide and end-tidal carbon dioxide in mechanically ventilated adults with severe head trauma. Crit Care Med 1996;24:785-90.
23. Russell GB, Graybeal JM. The arterial to end-tidal carbon dioxide difference in neurosurgical patients during craniotomy. Anesth Analg 1995;81:806-10.
24. Russell GB, Greybeal JM. Reliability of the arterial to end-tidal carbon dioxide gradient in mechanically ventilated patients with multisystem trauma. J Trauma 1994;36:317-22.
25. Hatle L, Rosketh R. The arterial to end-expiratory carbon dioxide tension gradient in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974;66:352-7.
© 1998 International Anesthesia Research Society