Pulmonary embolism causes increased alveolar dead space (VDalv ) [1-4] 2 elimination and result in CO2 accumulation and retention in the central pulmonary and peripheral tissue compartments. However, during mild to moderate pulmonary embolism in patients, reflex hyperventilation usually maintains normal or even increased VA, resulting in normal or decreased arterial PCO2 (PaCO2 ) [3,5]
However, in the last 15 yr there have been at least five reports of severe pulmonary embolism documenting progressive hypercapnia refractory to maximal spontaneous reflex hyperventilation or even supramaximal VE during mechanical ventilation [6-10] [3,11] [5,12] [5] [7,12-14] [15] [11]
However, the functional recovery of gas exchange during emergency pulmonary revascularization also should be signaled by an increase in the volume of CO2 exhaled per breath (VCO2 ,br) [16,17] alv ). Conceivably, the resultant decrease in pulmonary vascular resistance could increase cardiac output (QT), venous return, and the transport of CO2 (that had accumulated in the large peripheral tissue compartment) to the lung [4] 2 , which reflects the CO2 content of the central pulmonary compartment, would depend on the balance between VCO2 ,br, QT, and the peripheral tissue CO2 content [18] 2 stores and pulmonary CO2 elimination to normal values would not be detected by end-tidal PCO2 (PETCO2 ), which depends only on PaCO (2 ) and VDalv and does not measure exhaled volume or the shape of the CO2 waveform [16,18] 2 kinetics in the immediate period after resolution of pulmonary embolism are not known; many studies focus only on the steady-state condition [11] 2 retention could help guide intraoperative management and assess the success of surgical embolectomy.
Accordingly, this canine model examined the nonsteady-state changes in CO2 kinetics after removal of pulmonary embolism. We tested the hypotheses that increased pulmonary CO2 elimination (VCO2 ,br) can detect pulmonary reperfusion after removal of a large pulmonary embolus and that PETCO2 does not detect the eventual return of pulmonary CO2 elimination to normal by reversibly occluding the right pulmonary artery (RPA) with a cloth snare in five anesthetized, mechanically ventilated dogs with open thorax. We chose mechanical RPA occlusion to simulate reversible, large pulmonary embolism, although this model may not encompass many vasoactive changes present in studies of smaller pulmonary emboli [19,20]
Methods
After animal research committee approval, we studied five adult mongrel dogs (22 +/- 1 kg). Anesthesia was induced with intravenous chloralose (125 mg/kg) and urethane (625 mg/kg) [21,22] 2 and adjusted tidal volume (VT = 22 +/- 2 mL/kg) and respiratory frequency (f = 14 +/- 1/min) to maintain PaCO2 near 40 mm Hg. Expired gas passed through a large three-way stopcock, to allow the mixed expired gas measurement of PCO2 (PECO2 , infrared capnometry) in a 5-L bag.
Femoral arterial and venous cannulae were inserted. Systemic arterial, pulmonary arterial (Ppa), and airway opening pressures were transduced, amplified, and recorded.
After median sternotomy, we opened both pleural spaces to atmosphere, and added 4 cm H2 O end-expiratory pressure to maintain physiologic transpulmonary pressure. We carefully dissected the anterior, superior mediastinum [22] -1 centered dot h-1 ) that was maintained throughout the protocol. Normal saline (5 mL centered dot kg-1 centered dot h-1 ) was infused throughout the experiment.
A calibrated electromagnetic flow probe (Model EP 450, 50 mm internal circumference; Carolina Medical Electronics, King, NC) was placed around the ascending aorta and connected to a conditioner amplifier (Model 501D-CME500) to continuously measure QT [18]
VCO2 ,br was calculated by VT centered dot PECO2 centered dot PB-1 , where VT was calculated by integration of exhaled flow (see below) and PB was the dry barometric pressure [22] 2 ,br was determined by computer calculation using digital measurements of exhaled flow and PCO2 [16] 2 (FCO2 ; Gambro Engstrom Eliza, Bromma, Sweden) were digitally sampled at the airway opening. FCO2 was corrected for transport delay (time to suction gas through the sampling line), and dynamic response (response time of the measuring chamber). Then, VCO2 ,br was computed by integrating the expression VE(t) centered dot FCO2 (t) centered dot dt for each breath, where dt was the sampling time interval (1/100 Hz). At the same time, exhaled VT (time interval of VE over expiration) and PETCO2 were determined by computer algorithm. Dynamic response and transport delay of the gas analyzer were measured by bench and in vivo methods, respectively [16,17] 2 ). During the calculation of VCO2 ,br, inspired VCO2 was measured and excluded to determine overall pulmonary CO2 elimination [23]
In a period of stabilization after surgery and before the experimental protocol, we expanded the lungs to total capacity (two to three sighs) and administered supplemental chloralose, urethane, and sodium bicarbonate as required.
Measurement stages consisted of the following: 60-s digital sampling of VE and FCO2 , mixed expired gas collection, arterial and mixed venous blood samples, aortic QT, systemic arterial Ppa, and airway opening pressures. Then, the snare was tightened to completely occlude the RPA. After waiting for 70 min to approach the steady-state condition, we conducted baseline measurements (occluded RPA) [24] 2 began and the RPA snare was abruptly released to restore blood flow to the right lung. Measurement sequences were repeated 1, 2, 5, 10, 25, 45, and 70 min after RPA reperfusion. The position of the snare around the right pulmonary artery was confirmed at postmortem examination. Arterial and mixed venous PO2 , PCO2 , and pH were measured [25]
The physiologic dead space fraction was calculated by the Enghoff modification [26] [27] phy/VT = (PaCO2 - PECO2 )/PaCO2 [28] [29] ana ) from analysis of the PCO2-VT plot Figure 1 [18,28] ana was given by segment FE. The average alveolar expired PCO2 (PAECO2 ) was the average PCO2 plateau value (midpoint of line BC) [28] 2 was the end PCO2 of that PCO2-VT plot. The alveolar dead space fraction (VDalv/VT alv ) was calculated by (PaCO2 - PACO2 )/PaCO (2 ), where PACO2 was estimated by PETCO2 [31] alv = VDalv/VT alv centered dot (VT - VDana ) [28]
Figure 1: CO
2 expirogram (exhaled PCO
2 versus tidal volume) during baseline mechanical ventilation with right pulmonary artery (RPA) occlusion (Panel A) and after RPA reperfusion (Panel B) in Dog 5. The volume of CO
2 exhaled per breath (VCO
2 ,br) was the area under the data curve. By setting trapezoid BCDE equal to VCO
2 ,br, the Fowler anatomical dead space (VD
ana ) was calculated (segment FE). Thus, area Z represented the effects of VD
ana . Average alveolar expired PCO
2 (PAECO
2 ) was the midpoint of line BC. Arterial PCO
2 (PaCO
2 ) was 68 mm Hg in Panel A and 44 mm Hg in Panel B. Area Y represented the effect of alveolar dead space (VD
alv ). Then, total physiological dead space (VD
phy ) was the sum of area Y and area Z
[30] . (See text for details.)
We calculated the pulmonary shunt fraction of QT by Qs/QT = (Cc prime O2 - CaO2 )/(Cc prime O2 - CVO2 ), where Cc prime O2 , CaO (2 ), and CVO2 were the O2 contents of endcapillary, arterial, and mixed venous blood, respectively [18] 2 content (CO2 ) was determined by (Hb centered dot SO2 centered dot 1.34)+(PO2 centered dot 0.003), where Hb was the hemoglobin concentration, SO2 was the calculated oxyhemoglobin fraction, 1.34 was the hemoglobin O2 binding capacity (mL O2/g Hb), and 0.003 was solubility of O2 (mL O2 centered dot mm Hg-1 centered dot dL blood (-1 )). For end-capillary blood, alveolar PO2 was calculated by PB centered dot FIO2-PACO 2 , where PB was the dry barometric pressure, FIO2 (fraction of inspired oxygen) was unity (O2 ventilation), and alveolar PCO2 (PACO (2 )) was assumed equal to PaCO2 .
For each variable, we used repeated-measures analysis of variance to test for differences among measurement stages. If the analysis of variance F-statistic was significant (P < 0.05), a multiple comparison test (Student-Newman-Keuls) detected the differing stages. Selected variables were analyzed with the t-test. Values are reported as mean +/- SD.
Results
The significant increase in VDalv/VT alv from 31% to 54% required PVCO2 and PaCO2 to increase to steady values of 63 and 55 mm Hg, respectively, in order to maintain normal pulmonary CO2 elimination (VCO2 ,br) Table 1 [24]
Table 1: Effect of 70 min of Right Pulmonary Artery (RPA) Occlusion on Selected Cardiopulmonary Variables in Five Dogsa
(Figure 2 ) displays the changes during the first 23 breaths after RPA reperfusion. Compared to baseline (8.9 +/- 2.7 mL), average VCO2 ,br (Panel A, solid symbols) increased to 11.6 +/- 3.6 mL by 12 breaths after RPA reperfusion began and remained steady through the 2-min measurement stage Figure 3 A. Thereafter, VCO2 ,br steadily decreased and approached the baseline value by 70 min of RPA reperfusion (9.6 +/- 2.6 mL). Compared to baseline (24.9 +/- 4.6 mm Hg), PETCO2 increased to 33.2 +/- 5.2 mm Hg by 1 min after RPA reperfusion Figure 2 B. Then, PETCO2 began to slowly decrease but was still significantly greater than baseline at 70 min of RPA reperfusion (27.6 +/- 3.5 mm Hg, open triangle, Figure 3 B). From the baseline value of 62.8 +/- 6.5 mm Hg, RPA reperfusion resulted in a steady decrease in PVCO2 (Figure 3 B, open circle) to the 70 min value of 47.6 +/- 4.4 mm Hg. During the same period, RPA reperfusion decreased PaCO2 (open square) from 55.1 +/- 8.1 to 39.1 +/- 2.9 mm Hg.
Figure 2: In five mechanically ventilated dogs, breath-by-breath effects of right pulmonary artery (RPA) reperfusion on CO2 volume exhaled per breath (VCO2 ,br), end-tidal PCO2 (PETCO2 ), and cardiac output (QT). Periodic measured values of arterial (PaCO2 ) and mixed venous (PVCO2 ) PCO2 are shown. RPA reperfusion began after Breath 0. Individual and average values (open and solid symbols) are shown.
Figure 3: In five mechanically ventilated dogs, effect of 70 min of right pulmonary artery (RPA) reperfusion on CO2 volume exhaled per breath (VCO2 ,br) (Panel A); mixed venous (PVCO2 , open circle), arterial (PaCO2 , open square), and end-tidal (PETCO2 , open triangle) PCO2 (Panel B); and physiologic dead space (VDphy , open circle) and alveolar dead space fraction (VDalv/VT alv , open square) (Panel C). *Significant difference (P < 0.05) from baseline measurement (Time 0). dagger All stages during RPA reperfusion were different (P < 0.05) from baseline.
From a baseline value of 379 +/- 38 mL, VDphy (Figure 3 C, open circle) decreased to 327 +/- 45 mL by 1 min of RPA reperfusion, due to the decrease in VDalv/VT alv (open square) from 54.1% +/- 10.3% to 32.1% +/- 12.3%. Then, these dead space variables remained steady through 70 min of RPA reperfusion. In the CO2 expirogram for Dog 5 Figure 1 , the decrease in VDalv was represented by the decrease in area Y from baseline (RPA occlusion, Panel A) to RPA reperfusion (Panel B). Figure 1 A also shows that VDana at baseline (segment FE, 148 mL) was changed little by RPA reperfusion Figure 1 B. For the five dogs, average VD (ana ) at baseline (176 +/- 28 mL) was not significantly different at 70 min of RPA reperfusion (165 +/- 24 mL).
There was a transient trend for QT to increase for about 10 breaths after RPA reperfusion began Figure 2 E. However, compared to the baseline value of 4.7 +/- 0.9 L/min, QT did not significantly change during any stage of RPA reperfusion Table 2 . Ppa decreased from 31 mm Hg at baseline to 22 mm Hg at 1 min after RPA reperfusion Table 2 . This decrease in Ppa was sustained to the 70-min stage (20 mm Hg).
Table 2: Changes in Selected Cardiopulmonary Variables During 70 min of Right Pulmonary Artery (RPA) Reperfusion in Five Mechanically Ventilated, Anesthetized Dogs
Discussion
In this canine study, resolution of pulmonary thromboembolism was modeled by unclamping the RPA. We present the first breath-by-breath data that demonstrate how reperfusion of the RPA immediately increased VCO2 ,br by 30% and PETCO2 by 33%, as the decrease in VDalv improved VA. By 70 minutes of RPA reperfusion, VCO2 ,br had decreased back to baseline as body CO2 retention resolved. However, even by 70 minutes of RPA reperfusion, PETCO2 was significantly higher than the baseline value.
To explain the changes in CO2 storage and transport after RPA reperfusion, we invoke a hydraulic model of CO2 kinetics in the body [18,32] Figure 4 . A large peripheral tissue and venous blood CO2 compartment is shown on the left and a small central pulmonary and arterial blood compartment is shown on the right. CO2 is generated in the peripheral tissue compartment (VCO2 ,ti), transported by QT to the central compartment, and eliminated through pulmonary ventilation (VCO (2 ),br). The height and width of a compartment reflect its CO2 concentration (PCO2 ) and size, respectively. Then, CO2 will flow from higher to lower levels, as liquid flows downward by gravity. The central pulmonary compartment is further subdivided (left-to-right) into pulmonary shunt, normal lung, and VDalv . Pulmonary shunt (Qs) consists of zero VA/Q units that are perfused but devoid of ventilation. Hence, the height of the Qs region is equal to PVCO2 . VDalv are lung regions with high VA/Q ratios, depicted by the ventilated lung region on the right that receives minimal perfusion. The remaining normal lung has approximately equal values of ventilation and perfusion.
Figure 4: Scheme of mammalian CO
2 kinetics
[18,32] . CO
2 is produced (VCO
2 ,ti) in the large peripheral tissue compartment (left), transported by cardiac output (QT) to the small central pulmonary compartment (right), and eliminated through pulmonary ventilation (VCO
2 ,br). PaCO
2 and PVCO
2 = arterial and mixed venous PCO
2 ; VE = minute ventilation; Qs = pulmonary shunt; VD
alv and VD
ana = alveolar and anatomical dead spaces; VA/Q = ventilation-to-perfusion ratio; FRC = functional residual capacity. (See text for details.)
Major pulmonary thromboembolism was modeled by clamping and occluding the RPA for 70 minutes [24] alv , ventilated lung units now deprived of pulmonary perfusion, resulted in decreased VA at constant minute ventilation (Figure 4 , right). Consequently, VCO2 ,br decreased, CO2 retention occurred, and PCO2 in the peripheral tissues (PVCO2 ) and the central alveolar compartment (PaCO2 ) increased (increased compartment heights in Figure 4 ). By 70 minutes of RPA occlusion, steady state was approached as alveolar CO2 concentration (PaCO2 ) increased enough to restore VCO2 ,br to the initial value of tissue VCO2 .
To simulate surgical relief of pulmonary thromboembolism, the RPA was unclamped to immediately reperfuse and convert VDalv to normally perfused and ventilated lung units (Figure 4 , center of pulmonary compartment). The resultant sudden increase in VA caused the abrupt 30% increase in VCO2 ,br Figure 2 A and Figure 3 A and the 33% increase in PETCO2 (Figure 2 B and Figure 3 B, open triangle). Then, the increase in VCO2 ,br steadily decreased the CO2 concentration Figure 3 B in the central alveolar compartment (PaCO2 , Figure 4 , right) and subsequently in the peripheral tissue compartment (PVCO2 , Figure 4 , left). By 70 min of RPA reperfusion, VCO2 ,br had returned to the baseline value, indicating that steady state CO2 kinetics had been reestablished.
The reduction of CO2 in the peripheral compartment resulted solely from the increased CO2 elimination from the central alveolar compartment (VCO (2 ),br), since QT did not seem to significantly increase after RPA reperfusion in the five dogs. After RPA unclamping, significant reduction of pulmonary vascular resistance was signaled by the decrease in Ppa by 9 mm Hg Table 2 . Normally, the right pulmonary artery carries 56% of the total pulmonary blood flow in the canine lung [33] [34]
The increase in VCO2 ,br was an immediate and noninvasive monitor of reperfusion of the right lung in this model of experimental pulmonary embolism. We therefore speculate that increased VCO2 ,br would signal gas exchange recovery during surgical or transvenous pulmonary thromboembolectomy in patients, although other factors (such as hypotension, presence of anticoagulation, etc.) that might affect VDalv must be considered [4] 2 ,br to the initial value signaled the return to steady-state CO2 kinetics, including the decrease of body compartment stores of CO2 (PaCO2 and PVCO2 ) back to normal values. However, in a patient suffering a large clinical pulmonary embolus, the period of pulmonary vasculature occlusion could be longer than 70 minutes. Because most CO2 is stored in the peripheral tissues [32] 2 kinetic variables would be longer than the observations in this experimental model.
The time course of embolus resolution and pulmonary arterial reperfusion would be slower during thrombolytic therapy. Consequently, the increase in VCO2 ,br and decrease in central (PaCO2 ) and peripheral (PVCO2 ) contents of CO2 would be less abrupt but still signal functional recovery of pulmonary gas exchange. Furthermore, during medical therapy of pulmonary embolism in spontaneously breathing patients, reflex increases in minute ventilation will partly counter the decreases in VA caused by increased VDalv . Then, examination of VCO2 ,br normalized for VT (VCO2 ,br/VT), which is proportional to PECO2/br , could help demonstrate the amount of ventilation required to effect a given VCO2 ,br. In fact, the difference between PaCO2 and PECO2 normalized for PaCO2 , (PaCO2 - PECO2 )/PaCO2 , is the physiological dead space fraction of VT.
There are other examples of pulmonary artery occlusion that can occur during thoracic surgery. During surgery for hemoptysis, a main branch of the pulmonary arterial tree may be temporarily occluded to locate and gain control of the bleeding vessel. During esophageal surgery, the mediastinum and lungs may be shifted laterally, compressing and occluding a major pulmonary artery branch. Functional return of CO (2 ) elimination after relief of these temporary occlusions of the pulmonary vasculature would be signaled by abrupt increases in VCO2 ,br and PETCO2 , followed by slower reductions in PaCO2 and PVCO2 .
PETCO2 is variably used to estimate either PaCO2 or VCO2 ,br [18] 2 can estimate PaCO2 , the alveolar dead space fraction equation (VDalv/VT alv = (PaCO2 - PETCO2 )/PaCO2 ) may be rearranged to show the relationship between PETCO2 and PaCO2 [4] 2 = PaCO2 centered dot (1 - VDalv/VT alv ). Immediately after RPA reperfusion, the conversion of VDalv to perfused and ventilated alveolar regions resulted in the abrupt increase in PETCO2 , even though PaCO2 was progressively decreasing. PETCO2 only began to parallel the decrease in PaCO2 (constant PaCO2 - PETCO2 gradient) by 5 to 10 minutes of RPA reperfusion, because VDalv remained constant through the duration of RPA reperfusion. The final value of PETCO (2 ) at 70 minutes of RPA reperfusion was the balance of decreased VDalv and decreased PaCO2 .
On the other hand, PETCO2 appropriately monitored the initial increase in VCO2 ,br at the onset of RPA reperfusion after experimental pulmonary embolism, as VDalv was converted to perfused, ventilated lung units and pulmonary CO2 output increased. However, as VCO2 ,br approached its baseline value (i.e., tissue VCO2 ) to signal the return of steady state, PETCO2 was still significantly higher than its initial value because the decrease in VDalv outweighed the decrease in PaCO2 during 70 minutes of RPA reperfusion. Because it does not measure exhaled volume nor the shape of the exhaled CO2 waveform [16,18] 2 cannot consistently measure VCO2 ,br under the circumstances of this study.
We conclude, in this canine model of resolution of pulmonary thromboembolism, that reperfusion of the RPA immediately increased VCO2 ,br by 30% and PETCO2 by 33%, as the decrease in VDalv improved VA. By 70 minutes of RPA reperfusion, VCO2 ,br had decreased back to baseline as retention of CO2 in the body resolved (decreasing PaCO2 and PVCO2 ). However, even by 70 minutes of RPA reperfusion, PETCO2 was significantly higher than baseline. PETCO2 did not accurately approximate PaCO2 (due to the effect of VDalv ), nor did it reliably estimate VCO2 ,br (because PETCO2 does not measure exhaled volume).
REFERENCES
1. Severinghaus JW, Stupfel M. Alveolar dead space as an index of distribution of blood flow in pulmonary capillaries. J Appl Physiol 1957;10:335-48.
2. Eriksson L, Wollmer P, Olsson C-G, et al. Diagnosis of pulmonary embolism based upon alveolar dead space analysis. Chest 1989;96:357-62.
3. Elliott CG. Pulmonary physiology during pulmonary embolism. Chest 1992;101:163S-171S.
4. Isserles SA, Breen PH. Can changes in end-tidal PCO
2 measure changes in cardiac output? Anesth Analg 1991;73:808-14.
5. Dehring DJ, Arens JF. Pulmonary thromboembolism: disease recognition and patient management. Anesthesiology 1990;73:146-64.
6. Haynes JB, Iseman MD. Massive pulmonary embolism. Rocky Mountain Med J 1979;77:135-7.
7. Greenfield LJ, Jay SJ, Moeller D. Immediate management of massive pulmonary embolism. Chest 1982;82:775-80.
8. Goldberg SK, Lipschutz JB, Fein AM, Lippmann ML. Hypercapnia complicating massive pulmonary embolism. Crit Care Med 1984;12:686-8.
9. Padmanabhan K, Dhar SR. Irreversible hypercapnia secondary to embolic occlusion of the pulmonary artery. Chest 1984;86:927-8.
10. Bouchama A, Curley W, Al-Dossary S, Elguindi A. Refractory hypercapnia complicating massive pulmonary embolism. Am Rev Respir Dis 1988;138:466-8.
11. Prediletto R, Paoletti P, Fornai E, et al. Natural course of treated pulmonary embolism. Chest 1990;97:554-61.
12. Goldhaber SZ, Morpurgo M. Diagnosis, treatment, and prevention of pulmonary embolism. JAMA 1992;268:1727-33.
13. Greenfield LJ, Peyton MD, Brown PP, Elkins RC. Transvenous management of pulmonary embolic disease. Ann Surg 1974;180:461-8.
14. Greenfield LJ, Proctor MC, Williams DM, Wakefield TW. Longterm experience with transvenous catheter pulmonary embolectomy. J Vasc Surgery 1993;18:450-8.
15. Schmitz-Rode T, Gunther RW, Pfeffer JG, et al. Acute massive pulmonary embolism: use of a rotatable pigtail catheter for diagnosis and fragmentation therapy. Radiology 1995;197:157-62.
16. Breen PH, Isserles SA, Harrison BA, Roizen MF. Simple, computer measurement of pulmonary VCO
2 per breath. J Appl Physiol 1992;72:2029-35.
17. Breen PH, Mazumdar B, Skinner SC. Capnometer transport delay: measurement and clinical implications. Anesth Analg 1994;78:584-6.
18. Breen PH, Mazumdar B. How does positive end-expiratory pressure decrease CO
2 elimination from the lung? Respir Physiol 1996;103:233-42.
19. Delcroix M, Melot C, Vachiery J-L, et al. Effects of embolus size on hemodynamics and gas exchange in canine embolic pulmonary hypertension. J Apply Physiol 1990;69:2254-61.
20. Duranceau A, Jamieson GG, Jones RH, et al. An experimental model of pulmonary embolism. J Surg Res 1979;26:33-44.
21. Breen PH, Becker L, Ruygrok P, et al. Canine bronchoconstriction, gas trapping, and hypoxia with methacholine. J Appl Physiol 1987;63:262-9.
22. Breen PH, Isserles SA. Right atrial bypass model in the dog. Am J Vet Res 1995;56;208-14.
23. Breen PH, Serina ER, Barker SJ. Measurement of pulmonary CO
2 elimination must exclude inspired CO
2 measured at the capnometer sampling site. J Clin Monit 1996;12:1-6.
24. Breen PH, Mazumdar B, Skinner SC. How does experimental pulmonary embolism decrease CO
2 elimination? Respir Physiol. In press.
25. Thomas LJ. Algorithms for selected acid-base and blood gas calculations. J Appl Physiol 1972;33:154-8.
26. Enghoff H. Volumen inefficax. Bemerkungen zur Frage des schadlichen Raumes. Uppsala Lakarefoeren Forh 1938;44:191-218.
27. Bohr C. Ueber die Lungenathmung. Skand Arch Physiol 1891;2:236-68.
28. Breen PH, Mazumdar B, Skinner SC. Comparison of end-tidal PCO
2 and average alveolar expired PCO
2 during positive end-expiratory pressure. Anesth Analg 1996;82:368-73.
29. Fowler WS. Lung function studies. II. The respiratory dead space. Am J Physiol 1948;154:405-16.
30. Fletcher R, Jonson B, Cumming G, Brew J. The concept of deadspace with special reference to the single breath test for carbon dioxide. Br J Anaesth 1981;53:77-87.
31. Nunn JF, Hill DW. Respiratory dead space and arterial to end-tidal CO
2 tension difference in anesthetized man. J Appl Physiol 1960;15:383-9.
32. Cherniack NS, Longobardo GS. Oxygen and carbon dioxide gas stores of the body. Physiol Rev 1970;50:196-243.
33. Ebert PA, Allgood RJ, Jones HW III, Sabiston DC Jr. Hemodynamics during pulmonary artery occlusion. Surgery 1967;62:18-24.
34. McIntyre KM, Sasahara AA. The hemodynamic response to pulmonary embolism in patients without prior cardiopulmonary disease. Am J Cardiol 1971;28:288-94.
