EDITORIAL |
Editorial II: Deadspace: invasive or not?
1 Edinburgh, UK
2 Cheadle, UK
* Corresponding author. E-mail: g.b.drummond{at}ed.ac.uk
The recent paper by Tang and colleagues1 brought a refreshing discipline to measurements of respiratory deadspace. With capnography being widely available and more sophisticated, respiratory deadspace is more relevant in anaesthesia and intensive care, and physiological deadspace may become an important clinical measurement. In a recent study of patients in the first day of onset of ARDS, increased physiological deadspace fraction was an independent and powerful predictor of mortality.2 The relative risk of death increased by 45% if deadspace was increased by 5%. This increased risk was greater than for other predictive features, such as a score of illness severity or respiratory compliance. However, oxygenation was also significantly worse in the non-survivors (more of which will be explained later).
Deadspace can be calculated invasively or non-invasively: the difference may be subtle but very important, particularly in lung disease. Non-invasive information is obtained from the carbon dioxide single breath test (SBT-CO2), in which expired carbon dioxide fraction is plotted against expired volume (Fig. 1). This plot yields three parts; a carbon dioxide-free first phase, which is gas from the large conducting airways; phase II represents the transition between gas from the airways and alveolar gas, and phase III represents alveolar gas. A point midway through phase II is taken as the position of the alveolar/fresh gas interface at the start of expiration. This point, usually determined by Fowler's classic ‘equal area’ method,3 defines the volume of the airway proximal to the diffusive boundary, known as the anatomical deadspace, although for obvious reasons the term ‘airway’ deadspace is preferable. An alternative method was used by Tang and colleagues, which was to plot the cumulative volume of carbon dioxide exhaled against the exhaled volume.4 This is the integral of the SBT-CO2. Although it adds little extra information, it does show nicely the gradual onset of carbon dioxide output as airway deadspace gives way to alveolar gas (Fig. 2), and can also be used to define the anatomical or airway deadspace. In intubated adults, this volume is 
85 ml. Since its volume is determined by a diffusive boundary, it is reduced by breath-holding or by an end-inspiratory pause, which allows more time for alveolar gas to mix by diffusion with the static gas in the airways.
|
|
The shape of phases II and III is determined by the degree of gas mixing within the lung, and in disease it may be difficult to determine where one ends and the other begins. Obstructive airways disease reduces ventilation in some units more than others, causing an increased scatter of ventilation/perfusion (V/Q) ratios and thus of the range of alveolar
values. The SBT-CO2 reflects the changing composition of the alveolar expirate as an upward slope of phase III, for which there are two commonly accepted explanations. The first is asynchronous emptying, in which low PCO2 (i.e., high V/Q) regions empty quickly, before less well ventilated, slower ones with lower V/Q ratios.5 This is a ‘between units’ cause of mismatching, or regional inhomogeneity. Regional differences within a unit, where the more distal alveolar sacs are less well ventilated and thus have a lower V/Q than the more proximal ones, produce ‘stratified inhomogeneity’, which is time-dependent. As with the airway deadspace, this mixing defect can be reduced by breath-holding or an end-inspiratory pause. A large alveolar deadspace can also be present despite an apparently normal phase III slope, as in pulmonary embolism and the right to left (R–L) intracardiac shunting of congenital cyanotic heart disease. In contrast, the intrapulmonary shunting of severe lung disease is usually associated with both a steep phase III slope and very large deadspace. In small airways obstructive disease, an increased slope is often combined with an indistinct transition from phase II to phase III which is easily recognized, even on the time plot. The slope of phase III is also caused by the continuing evolution of carbon dioxide from the mixed venous blood into a progressively emptying lung (and here the rate of expiration is crucial—the slower expiration of obstructive lung disease guarantees a steeper slope6).
SBT-CO2 is the standard tool for understanding carbon dioxide elimination and the deadspace concept. The area under the curve gives the volume of carbon dioxide in the breath, from which we calculate not only minute carbon dioxide production, a metabolic indicator, but also mixed expired FCO2, which would be the carbon dioxide concentration in the breath if the exhaled carbon dioxide were uniformly distributed throughout the exhalation. By assuming that the last part of the expirate represents the composition of alveolar gas, we can estimate the Bohr deadspace (Fig. 1). However in disease, this is an inappropriate assumption. Arterial blood sampling allows us to infer the average composition of alveolar gas, by calculating the FCO2 of a gas in equilibrium with arterial blood, and substituting this value into the deadspace equation (Fig. 1). This use of arterial PCO2 (
) as described by Enghoff,7 leads to the now standard Bohr–Enghoff equation. Riley and his co-workers8 described the rationale of using arterial blood as ‘a physiological integrator of the carbon dioxide pressures existing in all parts of the lung’. On SBT-CO2, a horizontal line representing the FCO2 of a gas in equilibrium with the arterial blood illustrates the physiological deadspace. The alveolar deadspace (VDalv) is the area between phase III and the blood FCO2 line. The alveolar and airway deadspaces together constitute the physiological deadspace VDphys.
Tang and colleagues used a computer simulation in which V/Q distribution, VDalv, and R–L intrapulmonary shunt, could be adjusted in a tidally breathing cardio-respiratory model. They calculated airway deadspace by Fowler's method, by Bohr's method (calculated from ideal alveolar gas rather than end tidal FCO2 (
) as originally described), using the Bohr–Enghoff equation, and also using a graphical method based on the plot of cumulative expired carbon dioxide volume against expired volume, recently described by Koulouris and colleagues.9 This latter ‘non-invasive’ method essentially combines the impairment of carbon dioxide elimination by airway and alveolar deadspaces. Tang and co-workers' simulation shows that when VDalv is increased by V/Q mismatch as might occur in obstructive airways disease, VDphys is correctly measured by the Bohr–Enghoff equation and the original Bohr equation, but not by the Koulouris method.
It is not generally appreciated that R–L shunting, either intrapulmonary or intracardiac, can also affect the Bohr–Enghoff measure of deadspace. This is because the shunted mixed venous blood contains more carbon dioxide than pulmonary capillary blood, which loses some carbon dioxide as it equilibrates with the alveolar gas. Thus, shunted blood carries more carbon dioxide into the arterial blood, so that the is now greater than we would expect if the arterial blood were ‘integrating’ the blood from the lungs (Fig. 3). Thus, the assumption that the carbon dioxide in arterial blood is a measure of the alveolar 
becomes incorrect. This effect of shunt can be represented in SBT-CO2, where it raises the 
line, giving rise to an ‘apparent’ alveolar deadspace.10 A moment's experimentation with the classic shunt equation, rewritten for carbon dioxide, can demonstrate this phenomenon.
|
The apparent deadspace of R–L shunting is reflected by the calculated Bohr–Enghoff deadspace, but not by the Bohr or Koulouris deadspaces and clearly not by Fowler's airway deadspace.11 The true VDphys can, however, be estimated if shunt fraction is known, but this requires pulmonary artery catheterization in order to measure venous PCO2. Of the simpler clinically feasible methods, only the Bohr–Enghoff deadspace increases with increasing severity of V/Q mismatch. When alveolar PCO2 is increased by any mechanism,
calculated by Koulouris' method does not agree well with average alveolar PCO2.
To respiratory physiologists, these results should be intuitive, but confirmation by Tang and co-workers is welcome. Non-invasive respiratory measures only provide limited information, and the differences between them and invasive measures have caused much confusion. The misleading similarity between the plot of carbon dioxide against volume, and the more familiar plot of carbon dioxide against time, has induced some to confuse VDBohr, the deadspace measured by inserting end tidal FCO2 into the deadspace equation, with VDaw, the ‘Fowler’ deadspace. Thus any alveolar deadspace is included, and VDaw is overestimated.12 An equally misleading use of the end tidal value is to use this as a surrogate for , thus replacing the Bohr–Enghoff equation with the original Bohr equation, and underestimating the alveolar and physiological deadspaces.13
Tang's work also emphasizes the drawbacks of Cumming's alternative SBT-CO2 plot,4 also used by Koulouris, which presents the accumulated expired carbon dioxide volume, rather than FCO2, against expired carbon dioxide. However, this adds no new information to that available in the standard Fowler plot, as it is derived from the same observations. Delineation of VDaw, the first step necessary for any division of the physiological deadspace into its components VDaw and VDalv, is not improved. Further, the Cumming plot is less easy to grasp, and its pictorial representation of the alveolar deadspace is less helpful. Koulouris and colleagues attempted to extract more information out of Cumming's method than was there in the first place: it not possible to divine PaCO2 from it in patients with lung disease. Similar claims to those of Koulouris and colleagues have been made by others,14 based on limited experimental and correlational evidence.
Tang and colleagues also confirm the important effect of intrapulmonary shunting on gas exchange. By the 1980s, it was clear that the R–L shunting of cyanotic heart disease caused a large arterial-end tidal PCO2 difference. Some thought that intracardiac shunting caused poor perfusion of the lung,15 but this is not a necessary condition. Tang shows nicely that increasing intrapulmonary R–L shunting causes large arterial-end tidal PCO2 differences, confirming clinical observations.10 A clinically relevant observation is the strong relationship between impaired arterial oxygenation as indicated by pulse oximetry and the arterial-end tidal PCO2 difference.16
These facts may explain why the ‘new’ message, that increased deadspace in ARDS independently predicts outcome, may be none other than the ‘old’ message that arterial oxygenation is impaired by shunting. Indeed, concentrating on factors that could obstruct pulmonary capillaries or increase ventilation/perfusion ratios, in other words factors that could increase deadspace, as discussed by Nuckton and his colleagues,2 shows that these authors failed to consider how shunting can also impair carbon dioxide elimination. Impaired carbon dioxide elimination may well be a good marker of early lung damage, but perhaps only because it is less affected by variables such as PEEP (indeed, carbon dioxide elimination could be made worse by PEEP) and 
, which act to obscure the message from arterial hypoxaemia, than any intrinsic capacity to signal disturbed gas exchange. Disturbed gas exchange will impair carbon dioxide evolution and oxygen uptake in equal measure: and R–L shunting increases the measured deadspace.
Tang and co-workers also confirmed that Bohr's deadspace, obtained from substituting in the deadspace equation, has another of the drawbacks of a ‘non-invasive’ measure. The Bohr deadspace does not reveal the parallel deadspace of pulmonary embolism, where the SBT-CO2 has a flat phase III and a very large a-A gradient, a feature recognized since the experiments of Severinghaus and Stupfel.13
Although measurements of SBT-CO2, and physiological and alveolar deadspace are not yet widely available at the bedside, clinical measurement of the respiratory deadspace is beginning to establish itself. The study of Nuckton and colleagues2 suggests that deadspace measurement may be a robust index of gas exchange impairment, and a good prognostic index, whatever the mechanism may be.
There are several other uses for SBT-CO2. In the diagnosis of pulmonary embolism, the shape of SBT-CO2 combined with a large alveolar deadspace is almost pathognomonic.17 Bohr's deadspace to tidal volume ratio, and the slope of the phase III can distinguish patients with asthma from patients with emphysema who have the same degree of airways obstruction.18
Given the abstract nature of the deadspace concept it is perhaps not surprising that from time to time blind alleys will be entered, and ‘will-o’-the-wisps' pursued. Analysing exhaled carbon dioxide alone is unduly optimistic: Tang and colleagues have done valuable service by showing that arterial blood is generally necessary to fully understand impaired carbon dioxide elimination, and even with an arterial sample, we can be misled by unwarranted assumptions.
References
1 Tang Y, Turner MJ, Baker AB. Effects of alveolar dead-space, shunt and V/Q distribution on respiratory dead-space measurement. Br J Anaesth 2005; 95: 538–48
2 Nuckton TJ, Alonso JA, Kallet RH, et al. Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N Eng J Med 2002; 346: 1281–6
3 Fowler WS. Lung function studies II. The respiratory deadspace. Am J Physiol 2005; 154: 405–16
4 Cumming G. Carbon-dioxide dead space. Bull Physiopathol Respir 1974; 10: 607
5 Otis AB, Mckerrow CB, Bartlett RA, et al. Mechanical factors in distribution of pulmonary ventilation. J Appl Physiol 1956; 8: 427–43
6 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: 77–88
7 Enghoff H. Volumen inefficax. Upsala Lakareforen Forh 1938; 44: 191
8 Riley RL, Lilienthal JL Jr, Proemmel DD, Franke RE. On the determination of the physiologically effective pressures of oxygen and carbon dioxide in alveolar air. J Appl Physiol 1946; 10: 335–41
9 Koulouris NG, Latsi P, Dimitroulis J, Jordanoglou B, Gaga M, Jordanoglou J. Noninvasive measurement of mean alveolar carbon dioxide tension and Bohr's dead space during tidal breathing. Eur Respir J 2001; 17: 1167–74
10 Fletcher R. Relationship between alveolar deadspace and arterial oxygenation in children with congenital cardiac disease. Br J Anaesth 1989; 62: 168–76
11 Fowler WS. Lung function studies II. The respiratory deadspace. Am J Physiol 1948; 154: 405–16
12 Fletcher R. Airway deadspace, end-tidal CO2, and Christian Bohr. Acta Anaesthesiol Scand 1984; 28: 408–11[Web of Science][Medline]
13 Fletcher R. Deadspace, invasive and non-invasive. Br J Anaesth 1985; 57: 245–9
14 Romero PV, Lucangelo U, Aguilar JL, Fernandez R, Blanch L. Physiologically based indices of volumetric capnography in patients receiving mechanical ventilation. Eur Respir J 1997; 10: 1309–15[Abstract]
15 Lindahl SG, Olsson AK. Congenital heart malformations and ventilatory efficiency in children. Effects of lung perfusion during halothane anaesthesia and spontaneous breathing. Br J Anaesth 1987; 59: 410–8
16 Fletcher R. The relationship between the arterial to end-tidal PCO2 difference and hemoglobin saturation in patients with congenital heart-disease. Anesthesiology 1991; 75: 210–6[CrossRef][Web of Science][Medline]
17 Olsson K, Jonson B, Olsson CG, Wollmer P. Diagnosis of pulmonary embolism by measurement of alveolar dead space. J Intern Med 1998; 244: 199–207[CrossRef][Web of Science][Medline]
18 Kars AH, Bogaard JM, Stijnen T, deVries J, Verbraak AFM, Hilvering C. Dead space and slope indices from the expiratory carbon dioxide tension–volume curve. Eur Respir J 1997; 10: 1829–36[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. D. Macnaughton New ventilators for the ICU--usefulness of lung performance reporting Br. J. Anaesth., July 1, 2006; 97(1): 57 - 63. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||