Volumetric Capnography for Monitoring Lung Function
during Mechanical Ventilation
F. Suarez-Sipmann, G. Tusman, and S. H. Bæhm
z Introduction
Capnography has become standard of care in monitoring respiratory function during anesthesia [1] and together with pulse oximetry has contributed to a major improvement in safety and reduction in morbidity over the last three decades [2].
Carbon dioxide is an end product of the body's metabolism and is continuously
produced in the cells. In a normal person, about 280 l of CO2 is produced every
day and after its transport by the systemic and pulmonary circulation is eliminated
by the lungs via tidal ventilation. The amount of CO2 reaching the alveoli depends
on several factors including the rate of production, the equilibrium between the tissue stores, the venous return, cardiac output and pulmonary perfusion and finally
alveolar ventilation.
In intensive care medicine, in contrast to anesthesiology, capnography has
gained only limited acceptance as a monitoring tool. However, a better understanding of the pathophysiology of CO2 kinetics as well as the introduction of new measurement techniques have increased the interest and the potential of capnography.
In addition, the introduction of lung protective ventilation strategies [3, 4] in the
clinical management of patients with acute lung injury (ALI) requires us to revisit
our views of mechanical ventilation as a rather passive and supportive process and
start to consider it instead as a highly dynamic and therapeutic intervention. This
has increased the demand for improved bedside respiratory monitoring in order to
facilitate its adequate and safe implementation.
Breath by breath volumetric capnography represents a very attractive monitoring
option that provides the clinician with information not only about the amount of
CO2 eliminated but also about its elimination process, thus adding valuable information about the lung's physiological condition. In this chapter, we will review some of
the theoretical and physiological principles behind volumetric capnography. Finally
we will discuss some clinical applications that can be derived from a systematic use
of this methodology especially in the context of lung protective ventilation strategies.
z Time and Volume Capnography
The first capnographic measurement to be introduced and currently the most
widely used is the time based capnogram that is obtained by plotting exhaled CO2
against time. This measure provides continuous monitoring of end tidal CO2 (PETCO2) and, more importantly, changes in the shape of its graphical display assist in
Volumetric Capnography for Monitoring Lung Function during Mechanical Ventilation
Fig. 1. The volume capnogram is obtained by integrating the CO2 and the volume signals. The resulting
graph, the single breath test of CO2, can be divided into three main phases. Phase I: proximal airway
gas, Phase II: expiratory upstroke, and Phase III: alveolar plateau. The slopes of phases II (SII) and III (SIII),
the angle between them (a) and the area under the curve define its shape. The area under the curve
represents the volume of CO2 expired in a single breath (VTCO2, br)
detecting a number of clinically relevant problems during mechanical ventilation
such as: esophageal or bronchial intubation, circuit disconnections, spontaneous
breathing, ventilator malfunctions, etc.
The synchronous measurement of both the CO2 and the flow/volume signals
measured at the airway opening allowed changes in CO2 in the volume domain to
be studied in real time, this way obtaining the volume capnogram also called expirogram or single breath test of CO2 (SBT-CO2). Figure 1 shows a normal volume
capnogram with its components and phases.
The volume capnogram provides all the features of the time based capnogram,
supplementing it, however, with important physiologic information related to the
dynamics of CO2 exhalation and the ability to analyze the sequence of tidal ventilation and dead spaces (VD).
z The Normal Capnogram: Definitions of Phases and Derived Variables
z Phase I begins with the start of expiration and ends when the concentration of
CO2 increases beyond 0.1% from baseline. The volume of gas in phase I comprises airway VD (VDaw) and represents part of the gas in the proximal airway.
z Phase II, or expiratory upstroke, starts at the end of phase I and ends at the intersection of the predictive slope lines of phases II and III. The midpoint of
phase II (50% of the slope) is the limit between VDaw and alveolar gas, and represents the `interface' where gas transport by convection changes into transport
by diffusion within the lung acini. Thus, phase II contains part of both VDaw
459
460
F. Suarez-Sipmann et al.
z
z
z
z
z
and alveolar gas. Phase II is highly influenced by the time constant of emptying
acini.
Phase III, or the alveolar plateau, begins at the intersection of the predictive
slopes lines of phase II and III and terminates at the end of expiration. This volume represents the gas inside the alveoli in contact with pulmonary capillary
blood and is considered as the efficient part of the tidal volume (VT).
The area under the curve represents the volume of CO2 expired in a single
breath measured by flow integration, and represents the portion of alveolar gas
that is in contact with the pulmonary capillary blood.
The slope of phase II is derived from linear regression using data points collected commonly between 25-75% of phase II, and expressed as fraction/litre.
Similar to the volume of phase II, the slope represents the spread of acini expiratory times. If all acini are empty at the same time ventilation is more homogeneous and the slope increases.
The slope of phase III is derived from linear regression using data points collected between 25±75% of phase III, and expressed as fraction/litre. The slopes
of phases II and III of individual breaths can be normalized by dividing the
slope value by the corresponding mean alveolar fraction of CO2 (expressed in
%). The phase III slope is related to the ventilation/perfusion relationship (V/Q);
when the V/Q ratio is more homogeneous, the phase III slope decreases whereas
it increases when the V/Q ratio is more heterogeneous.
Angle II±III or angle alpha is defined as the angle defined by the intersections
of the slopes of phase II and III.
Finally, the slopes of phases II and III, their intercepts, angle alpha, the area under
the curve, the volumes of phases I, II and III are the variables that can be determined non-invasively without having to use additional arterial blood gas determinations. They are representatives of the `shape' of volume capnography and change
with changing pulmonary states and diseases (Fig. 1).
z Measurement of Dead Space and Efficiency of Ventilation
VD is defined as `wasted ventilation', i.e., the part of the VT that does not reach
perfused alveoli and therefore does not participate in gas exchange [5±7]. At the alveolar level it constitutes one extreme of the V/Q relationship (V/Q = ?), while
shunt represents the opposite phenomenon (V/Q = 0). Shunt can be considered a
mirror concept of VD. Therefore, it can be called `wasted perfusion'.
The expired CO2-volume curve has facilitated the bedside assessment of VD as a
measurement of the efficiency of ventilation that could previously be determined in
the pulmonary function laboratory [5]. When adding the value of the arterial partial pressure of CO2 (PaCO2) to this curve, a complete VD analysis can be performed at the bedside delivering the following parameters (Fig. 2):
z Instrumental VD (VDinst) consists of any additional VD beyond the `Y' piece
causing a re-breathing of CO2 with each breath.
z VDaw, also known as anatomical or serial dead space, is the volume of gas in the
upper and main airways down to the boundary between convective and diffusive
gas transport at the bronchiolar level.
z Alveolar VD (VDalv) represents the alveolar gas within alveoli that are not perfused (West's zone I condition).
Volumetric Capnography for Monitoring Lung Function during Mechanical Ventilation
Fig. 2. Dead space analysis. Introducing the value of PaCO2 allows for the calculation of dead space and
its fractions. Physiological dead space is the total tidal ventilation dead space, i.e., the sum of airway
(VDaw) and alveolar (VDalv) dead space. Fowler's equal area method uses a vertical line across the mid of
phase II slope determining the areas p and q of equal size. The intersection of this line with phase II represents the limit between convective and diffusive gas transport within the lungs. PAECO2: mean alveolar
expiratory concentration of CO2 ± the value at the mid point of phase III must be taken to avoid an underestimation of VD/VT measurement
z Physiological VD (VDphys) comprises the total tidal VD, thus the sum of VDaw
and VDalv.
z Physiological dead space-to-tidal volume ratio (VD/VT) represents the relationship between the inefficient and efficient part of the VT.
z VTalv is the portion of VT distal to the bronchial interface, in alveolar gas. This
volume is constituted by the sum of alveolar CO2 volume plus VDalv. VTalv is derived as VT±VDalv by Fowler's method.
z VDalv-to-VTalv ratio represents the relationship between the inefficient and efficient portion of the VTalv.
z Bohr's dead space is defined as the sum of VDaw and a portion of the VDalv.
z Arterial to end-tidal difference of CO2 (Pa-ETCO2) is a clinically useful index
that represents the magnitude of VDalv.
Figure 2 represents the volume capnogram and its VD subdivisions. VDaw is commonly calculated by Fowler's equal area method, where a vertical line across the
mid of phase II slope determines the areas p and q of equal size. This line represents
the limit between convective and diffusive gas transport within the lungs and is called
the `transition zone' by some authors [6±8]. This border is dynamic, not fixed and
changes with different physiological events, diseases or ventilatory settings.
Bohr was the first to describe the measurement of VD/VT using a Douglas bag:
VD =VT ˆ FACO2
FECO2 =FACO2
where FACO2 is the alveolar gas and FECO2 is the mixed CO2 concentration in this
expired gas. Later, Enghoff introduced a modification of Bohr's formula that simpli-
461
462
F. Suarez-Sipmann et al.
fied the measurement of VD/VT in patients [9, 10]. He replaced the alveolar CO2
concentration (or partial pressure) by the PaCO2 assuming that the partial pressure
of CO2 at the arterial side truly integrated the gas exchange function of each single
alveolus within the lung:
VD =VT ˆ PaCO2
ETCO2 =PaCO2
Since the slope of phase III is almost always positive, the mean alveolar expiratory
concentration of CO2 (PAECO2), i.e., the concentration of CO2 at the mid point of
this phase should replace the end-tidal concentration in order to avoid a systematic
underestimation of VD/VT [7, 11].
VD =VT ˆ PaCO2
PAECO2 =PaCO2
VDphys is calculated as:
VD phys ˆ VD =VT VT
VDalv is derived simply by subtracting VDaw from VDphys.
z Theoretical Principles Related to the Kinetics of CO2
The kinetics of exhaled gases are useful indicators of gas transport within the lung
in healthy individuals and in patients with respiratory diseases [12±14].
Gas transport within the lungs occurs by two main mechanisms: convection and
diffusion [8, 12, 15]. Convection, at the main airways, is the bulk flow created by
the energy stored within the respiratory system at end-inspiration and the inertial
forces created by the diffusive transport. Diffusion, which occurs at the alveolar
level, is the passive movement of molecules following a concentration or partial
pressure gradient that is given by Fick's law:
J ˆ Dmol A Dc=Dx
where J is the instantaneous flux of CO2, Dmol represents the gas-phase molecular diffusivity of CO2 in air, A is the area of gas exchange, Dc the venous-alveolar gas concentration gradient for CO2 and Dx is the width of the alveolar-capillary membrane.
Diffusion also promotes the intra-acinar gas transport of CO2 from the alveolus into
the airways. For this movement, A is the cross-sectional area of the small airways and
Dc the differential CO2 partial pressure between the alveoli and the interface of convective-diffusive transport and Dx the distance between alveoli and the above interface. During normal physiology and in most of the pathological conditions, Dmol,
Dc, and Dx are constant so that the area becomes the main variable affecting diffusion. A reduction of the area for gas exchange is observed in atelectatic lungs where
the number of alveoli is reduced [16]. Here, diffusion is negatively affected and can be
considered as an increment in the resistance to the removal of CO2 from the blood.
Similarly, any reduction in the cross-sectional area of the small airways increases
the resistance to intrapulmonary gas transport by diffusion as observed in patients
with chronic obstructive pulmonary disease (COPD) [17].
Volumetric Capnography for Monitoring Lung Function during Mechanical Ventilation
The slopes of phase II and III are closely related to the mechanism of transport
of expiratory gases but their true genesis is still a matter of debate. The phase II
slope represents the transition between alveolar and airway gas transport and depends on the spread of transit times of lung units with different time constants [6,
8, 12, 15, 18, 19]. An increase in the cross-sectional area of the bronchial tree in
the lung periphery decreases the linear velocity of the bulk flow until a point where
the two transport mechanisms within the lungs, convection and diffusion, are of
equal magnitude.
The characteristic upward slope of phase III is explained by a number of mechanisms [6, 9, 20]: stratified inhomogeneity, continuous evolution of CO2 from the
blood and diffusive Pendelluft or sequential emptying of units with different CO2
concentrations.
Changes in the acinar structure, that is, in pulmonary 3D-morphology, as in emphysema, bronchospasm, atelectasis, airway overdistension or embolism, directly affect gas kinetics. Therefore, any change in gas kinetics is reflected in a changing
shape of the volume capnogram and in VD [7, 11, 13, 14, 16, 17, 21]. These changes
in the volume capnogram have been described for asthmatic [13, 14] and emphysema patients [17].
During mechanical ventilation, the acinar morphology is affected by different
factors such as lung volume history, surfactant function, ventilator settings and
pathologic lung condition. These changes can be dynamic, breath-by-breath and reversible (e.g., atelectasis, airway collapse, edema, etc.) or fixed and irreversible
(e.g., fibrotic phase of acute respiratory distress syndrome [ARDS]).
Important factors influencing the phase III slope are the size of the VT and the
magnitude of pulmonary perfusion [19]. Schwartz et al. observed that pulmonary
perfusion mainly affects the area under the curve. We have recently described that
progressive increases in the amount of lung perfusion cause parallel increases in
the phase III slope provided that all other ventilator settings are kept constant. On
the other hand, this dependency of volume capnography on both VT and perfusion
can make distinguishing between changes brought about by changing ventilatory
conditions and those due to hemodynamic alterations impossible. One proposed
solution is to cancel out these effects by normalizing each slope by the mean exhaled alveolar concentration of CO2 [22, 23]. In a study performed on patients during the weaning phase of cardiopulmonary bypass during cardiac surgery, we
showed that this normalization eliminated the effects of a wide range of changes on
the amount of pulmonary perfusion (e.g., cardiac output). Due to the revival of interest in capnography for monitoring purposes the clinical value of non-normalized
and normalized phase III slopeI and the use of other volume capnography-derived
variables must be further investigated.
z The Role of Volume Capnography for Monitoring Recruitment
and Lung Protective Ventilation
The development of atelectasis is a constant phenomenon in ventilated patients
with and without pulmonary disease. In addition to the impairment of gas exchange [24], lung collapse has been recognized as a direct cause of post-operative
complications [25] and, together with lung overdistension, as a major contributing
factor to the development of ventilation-induced lung injury in patients suffering
463
464
F. Suarez-Sipmann et al.
from respiratory insufficiency [26]. Recently, ventilation strategies aimed at protecting the lung from such damage by re-expanding the collapsed lung using recruitment maneuvers, stabilizing the lung with high levels of positive end-expiratory
pressure (PEEP) and by limiting inspiratory overdistension, have been shown to reduce mortality of ventilated patients with severe respiratory insufficiency [3, 4].
Atelectasis causes a loss of functional units due to a decrease in alveolar, capillary and airway cross-sectional area. During lung collapse and recruitment, gas
mixing and transport are affected causing a detectable change in the shape and
variables of the volume capnogram and VD fractions. Different lung conditions follow the same qualitative behavior although the magnitude of these changes is different. At similar lung volume history and ventilatory settings in a particular patient, the morphometric `state' of the acini is the main determinant of gas mixing,
gas exchange through the alveolar-capillary membrane and gas emptying. These
phenomena, together, are responsible for the changes observed in the volume capnography curve. Figure 3 shows the changes in the shape of a volume capnogram
due to recruitment. Table 1 lists the expected changes in volume capnogram variables after an effective lung recruitment.
In a recent study in anesthetized patients with perioperative atelectasis we
showed that treating patients with a recruitment maneuver, in which inspiratory
pressure was set to 40 cmH2O [16] and PEEP to a level safely above the lung's closing pressure, led to a normalization of the shape of the volume capnogram and reduced VD ventilation as compared to patients ventilated with the same level of
PEEP but without a previous recruitment maneuver. There was a significant increase in the phase II slope and the area under the curve and a reduction in phase
III slope and VD/VT. These changes were accompanied by an improvement in oxygenation, end-expiratory lung volume and lung compliance [16].
In a recent study it was shown that the increase in dead space was an independent predictor of mortality in patients with early ARDS [27]. Bedside monitoring
using volume capnography has shown that VD is not a static value but is strongly
Fig. 3. Differences in the shape of
the volume capnogram in an atelectatic and in a recruited lung.
After recruitment the phase II slope
increases, becoming steeper, and
the phase III slope decreases
Volumetric Capnography for Monitoring Lung Function during Mechanical Ventilation
Table 1. Changes in the volume capnogram after an effective lung recruitment technique
Improved gas exchange
Improved intra-acinar gas transport
; Phase III slope
; Phase III slope
; AngleII±III
: Phase II slope
: Area under the curve (VTCO2, br)
; VolIII
; Dead-space
; Pa-ETCO2
VTCO2, br: Volume of CO2 expired per breath; Pa-ETCO2: arterial to end-tidal CO2 gradient; angle II±III: alpha
angle between phase II and III: VolIII: volume of phase III
influenced by ventilatory interventions such as recruitments and PEEP titration
[16, 28].
Figure 4 shows changes in the VD fractions in an experimental model of ALI
where after an incremental PEEP trial, a recruitment maneuver was performed that
was then followed by a decremental PEEP titration maintaining the other ventilatory settings unchanged. At each PEEP level a CT scan was performed to assess
lung condition.
During the incremental PEEP steps, progressive recruitment resulting in a decrease in alveolar and physiologic dead space, was observed. During the decremental PEEP trial VD fractions decreased initially until CT confirmed the beginning of
collapse. At this point VD fractions reached their minimum values before increasing
again finally reaching pre-recruitment values at the lowest PEEP values. In contrast,
VDaw changed almost linearly with the changes in PEEP. This example shows how
volume capnography could aid in implementing a lung protective ventilation strategy by identifying the settings at which VD is minimal.
Fig. 4. Changes in dead space fractions during incremental PEEP steps and decremental PEEP steps after
a lung recruitment maneuver while maintaining all other ventilator parameters unchanged. Triangles:
physiological deadspace (ml); Squares: airway dead space in ml; Diamonds: Alveolar deadspace in ml;
Crosses: Physiological dead space-to-tidal volume ratio (VD/VT ) in %
465
466
F. Suarez-Sipmann et al.
z Conclusion
Volume capnography is a promising non-invasive, inexpensive, breath by breath
measurement that has the potential to become an indispensable bedside monitoring
tool that can guide the therapeutic process of mechanically ventilated critically ill
patients. The better understanding of the pathophysiology of the acutely injured
lung on the one hand and the increasing knowledge about the kinetics of CO2-exchange on the other has created a growing interest in this technology. Volume capnography provides information about the changes in the lung's condition and
might therefore allow for improved monitoring of complex ventilatory interventions
such as lung recruitment and PEEP titration. In addition, the bedside assessment
of VD helps to identify patients at risk with uneven modes of ventilation and to
evaluate their response to different ventilatory strategies. Further studies will help
define the true role of volumetric capnography in clinical decision making.
References
1. ASA House of Delegates (2005) Standards for basic anesthetic monitoring. Available at:
www.asahq.org/publicationsAndServices/standards/02.pdf Accessed October 2005
2. Eichhorn JH (1989) Prevention of intraoperative anesthesia accidents and related severe injury through safety monitoring. Anesthesiology 70:572±577
3. Amato MB, Barbas CS, Medeiros DM, et al (1998) Effect of a protective ventilation strategy
on mortality in the acute respiratory distress syndrome. N Engl J Med 338:347±354
4. The Acute Respiratory Distress Network (2000) Ventilation with lower tidal volumes as
compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 342:1301±1308
5. Fowler WS (1948) Lung function studies. II. The respiratory dead space. Am J Physiol 154:
405±416
6. Fletcher R, Jonson B (1981) The concept of deadspace with special reference to the single
breath test for carbon dioxide. Br J Anaesth 53:77±88
7. Fletcher R, Jonson B (1984) Deadspace and the single breath test for carbon dioxide during
anesthesia and artificial ventilation. Br J Anaesth 56:109±119
8. Engel LA (1982) Gas mixing within the acinus of the lung. J Appl Physiol 54:609±618
9. Enghoff H (1938) Volume inefficax. Bemerkungen zur Frage des schådlichen Raumes. Upsala Låkaref Færh 44:191±218
10. Nunn JF, Holmdahl MH (1990) Henrik Enghoff and the volume inefficax. Acta Anesthesiol
Scand 34:24±28
11. Breen PH, Mazumdar B, Skinner SC (1996) Comparison of end-tidal PCO2 and average alveolar expired PCO2 during positive end-expiratory pressure. Anesth Analg 82:368±373
12. Crawford ABH, Makowska M, Paiva M, Engel LA (1985) Convection- and diffusion-dependent ventilation misdistribution in normal subjects. J Appl Physiol 59:838±846
13. You B, Peslin R, Duvivier C, Vu VD, Grilliat JP (1994) Expiratory capnography in asthma:
evaluation of various shape indices. Eur Respir J 7:318±323
14. Kars AH, Bogaard JM, Stijnen T, de Vries J, Verbraak AF, Hilvering C (1997) Deadspace and
slope indices from the expiratory carbon dioxide-tension volume curve. Eur Respir J
10:1829±1836
15. Verbank S, Paiva M (1990) Model simulations of gas mixing and ventilation distribution in
the human lung. J Appl Physiol 69:2269±2279
16. Tusman G, Bæhm SH, Suarez-Sipmann F, Turchetto E (2004) Dead space analysis before
and after lung recruitment. Can J Anesth 51:723±727
17. Schwardt JD, Neufeld GR, Baumgardner JE, Scherer PW (1994) Noninvasive recovery of acinar anatomic information from CO2 expirograms. Ann Biomed Eng 22:293±306
18. Dutrieue B, Vanholsbeeck F, Verbank S, Paiva M (2000) A human acinar structure for simulation of realistic alveolar plateau slopes. J Appl Physiol 89:1859±1867
Volumetric Capnography for Monitoring Lung Function during Mechanical Ventilation
19. Schwardt JF, Gobran SR, Neufeld GR, Aukburg SJ, Scherer PW (1991) Sensitivity of CO2
washout to changes in acinar structure in a single-path model of lung airways. Ann Biomed
Eng 19:679±697
20. Hofbrand BI (1966) The expiratory capnogram: a measure of ventilation-perfusion inequalities. Thorax 21:518±524
21. Folkow B, Pappenheimer IR (1955) Components of the respiratory dead space and their variation with pressure breathing and with bronchoactive drugs. J Appl Physiol 8: 102±116
22. Ream RS, Schreiner MS, Neff JD, et al (1995) Volumetric capnography in children: influence
of growth on the alveolar plateau slope. Anesthesiology 82:64±73
23. Tusman G, Areta M, Climente C, et al (2005) Effect of pulmonary perfusion on the slopes
of single breath test of CO2. J Appl Physiol 99:650±655
24. Tokics L, Hedenstierna G, Svenson L, et al (1996) V/Q distribution and correlation to atelectasis in anesthetized paralyzed humans. J Appl Physiol 81:1822±1833
25. Magnusson L, Spahn DR (2003) New concepts of atelectasis during general anaesthesia. Br J
Anaesth 91:61±72
26. Dreyfuss D, Saumon G (1998) Ventilator-induced lung injury: lessons from experimental
studies. Am J Respir Crit Care Med 157:294±323
27. Nuckton T, Alonso JA, Kallet RH, et al (2002) Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N Engl J Med 346:1281±1286
28. Tusman G, Bæhm SH, Suarez-Sipmann F, Maisch S (2004) Lung recruitment improves efficiency of ventilation and gas exchange during one lung ventilation anesthesia. Anesth Analg
98:1604±1609
467