Introduction

Dead space is that part of the tidal volume that does not participate in gas exchange. Although the concept of pulmonary dead space was introduced more than a hundred years ago, current knowledge and technical advances have only recently lead to the adoption of dead space measurement as a potentially useful bedside clinical tool.

Concept of dead space

The homogeneity between ventilation and perfusion determines normal gas exchange. The concept of dead space accounts for those lung areas that are ventilated but not perfused. The volume of dead space (Vd) reflects the sum of two separate components of lung volume: 1) the nose, pharynx, and conduction airways do not contribute to gas exchange and are often referred to as anatomical Vd or herein as airway Vd (Vdaw); 2) well-ventilated alveoli but receiving minimal blood flow comprise the alveolar Vd (Vdalv). Mechanical ventilation, if present, adds additional Vd as part of the ventilator equipment (endotracheal tubes, humidification devices, and connectors). This instrumental dead space is considered to be part of the Vdaw. Physiologic dead space (Vdphys) is comprised of Vdaw (instrumental and anatomic dead space) and Vdalv and it is usually reported in mechanical ventilation as the portion of tidal volume (Vt) or minute ventilation that does not participate in gas exchange [1, 2].

A device that measures partial pressures (PCO2) or fractions (FCO2) of CO2 during the breathing cycle is called a capnograph. The equation to transform FCO2 into PCO2 is PCO2 = FCO2 multiplied by the difference between barometric pressure minus water-vapour pressure. Time-based capnography expresses the CO2 signal as a function of time and from this plot mean expiratory (Douglas bag method) or end-expiratory (end-tidal) CO2 values can be obtained. The integration of the volume signal using an accurate flow sensor (pneumotachograph) and CO2 signal (with a very fast CO2 sensor) is known as volumetric capnography. Combined with the measurement of arterial PCO2 (PaCO2) it provides a precise quantification of the ratio of Vdphys to Vt. The three phases of a volumetric capnogram are shown in Fig. 1 and Fig. 2. The combination of airflow and mainstream capnography monitoring allows calculation of breath by breath CO2 production and pulmonary dead space. Therefore, the use of volumetric capnography is clinically more profitable than time-based capnography.

Fig. 1
figure 1

A Single-breath expiratory volumetric capnogram recorded in a healthy patient receiving controlled mechanical ventilation. Dead-space components are shown graphically and equations are depicted and explained in the text. Phase I is the CO2 free volume which corresponds to Vdaw. Phase II represents the transition between airway and progressive emptying of alveoli. Phase III represents alveolar gas. PaCO2 is arterial PCO2; PetCO2 is end-tidal PCO2. Drawings adapted from [2]; B Single-breath expiratory carbon dioxide volume (VCO2) plotted as a function of exhaled tidal volume. The alternative method to measure airway dead space (Vdaw) described by Langley et al. [3] is graphically shown in a healthy patient receiving controlled mechanical ventilation

Full size image
Fig. 2
figure 2

A Single-breath expiratory volumetric capnogram recorded in a chronic obstructive pulmonary disease patient receiving controlled mechanical ventilation. The three phases of the volumetric capnogram are depicted. The transition from phase II to III is less evident due to heterogeneity of ventilation and perfusion ratios. Dead-space components are shown graphically and equations are depicted and explained in the text. PaCO2 is arterial PCO2; PetCO2 is end-tidal PCO2. Drawings adapted from [2]; B Single-breath expiratory carbon dioxide volume (VCO2) plotted as a function of exhaled tidal volume. The alternative method to measure airway dead space (Vdaw) described by Langley et al. [3] is graphically shown in a chronic obstructive pulmonary disease patient receiving controlled mechanical ventilation

Full size image

Measurement of dead space using CO2 as a tracer gas

Bohr originally defined Vd/Vt [2] as: Vd/Vt = (FACO2–FECO2)/FACO2, where FACO2 and FECO2 are fractions of CO2 in alveolar gas and in mixed expired gas, respectively. End-tidal CO2 is used to approximate FACO2, assuming end-tidal and alveolar CO2 fractions are identical. Physiologic dead space calculated from the Enghoff modification of the Bohr equation uses PaCO2 with the assumption that PaCO2 is similar to alveolar PCO2 [2], such that: Vdphys/Vt = (PaCO2–PECO2)/PaCO2, where PECO2 is the partial pressure of CO2 in mixed expired gas and is equal to the mean expired CO2 fraction multiplied by the difference between the atmospheric pressure and the water-vapour pressure. Since Vdphys/Vt measures the fraction of each tidal breath that is wasted on both Vdalv and Vdaw, the Vdaw must be subtracted from Vdphys/Vt to obtain the Vdalv/Vt. Vdphys/Vt is the most commonly and commercially (volumetric capnographs) formula used to estimate pulmonary dead space at the bedside.

Additional methods mostly used in research to calculate all the Vd components are shown in Fig. 1A and Fig. 2A. Fowler [1] introduced a procedure for measuring Vdaw based on the geometric method of equivalent areas (p = q), obtained by crossing the back extrapolation of phase III of the expired CO2 concentration over time with a vertical line traced so as to have equal p and q areas. Airway dead space is then measured from the beginning of expiration to the point where the vertical line crosses the volume axis [1]. By tracing a line parallel to the volume axis and equal to the PaCO2, it is possible to determine the readings from areas y and z, which respectively represent the values of alveolar and airway dead space. Referring these values to the Vt, it is possible to single out several Vd components [2]:

$$ \begin{array}{*{20}l} {{{\text{Vd}}_{{{\text{phys}}}} /{\text{Vt}}} \hfill} & { = \hfill} & {{({\text{Y}} + {\text{Z}})/({\text{X}} + {\text{Y}} + {\text{Z}})} \hfill} \\ {{{\text{Vd}}_{{{\text{alv}}}} /{\text{Vt}}} \hfill} & { = \hfill} & {{{\text{Y}}/({\text{X}} + {\text{Y}} + {\text{Z}})} \hfill} \\ {{{\text{Vd}}_{{{\text{aw}}}} /{\text{Vt}}} \hfill} & { = \hfill} & {{{\text{Z}}/({\text{X}} + {\text{Y}} + {\text{Z}})} \hfill} \\ \end{array} $$

An alternative method to measure airway dead space introduced by Langley et al. [3] is based on determination of the VCO2 value, which corresponds to the area inscribed within the CO2 versus volume curve (indicated in Fig. 1A and Fig. 2A as X area). Figure 1B and Fig. 2B are examples of Vdaw calculation using the Langley et al. [3] method. Briefly, VCO2 is plotted versus expired breath volume. Thereafter, Vdaw can be calculated from the value obtained on the volume axis by back extrapolation from the first linear part of the VCO2 versus volume curve.

Although these indexes are clinically useful, they are always bound to visual criteria for the definition of phase III of the expired capnogram. Often, the geometric analysis establishing the separation between the phase II and phase III is hardly seen and the rate of CO2 raising of the phase III is nonlinear in patients with lung inhomogeneities (Fig. 2A).

Utility of dead space in different clinical scenarios

The CO2 tension difference between pulmonary capillary blood and alveolar gas is usually small in normal subjects and end-tidal PCO2 is close to alveolar and arterial PCO2. Physiologic dead space is the primary determinant of the difference between arterial and end-tidal PCO2 (ΔPCO2) in patients with a normal cardio-respiratory system. Patients with cardiopulmonary diseases have altered ventilation to perfusion (VA/QT) ratios producing abnormalities of Vd, as well as in intrapulmonary shunt, and the latter may also affect the ΔPCO2. A ΔPCO2 beyond 5 mmHg is attributed to abnormalities in Vdphys/Vt and/or by an increase in venous admixture (the fraction of the cardiac output that passes through the lungs without taking oxygen) or both. The increase in Vdphys/Vt seen in normal patients when anaesthetised may be attributed to muscle paralysis, which causes a reduction of functional residual capacity and alters the normal distribution of ventilation and perfusion across the lung [2, 4, 5, 6].

Ventilation to regions having little or no blood flow (low alveolar PCO2) affects pulmonary dead space. In patients with airflow obstruction, inhomogeneities in ventilation are responsible for the increase in Vd. Shunt increase VDphys/Vt as the mixed venous PCO2 from shunted blood elevates the PaCO2, increasing VDphys/Vt by the fraction that PaCO2 exceeds the nonshunted pulmonary capillary PCO2 [7]. Vdalv is increased by shock states, systemic and pulmonary hypotension, obstruction of pulmonary vessels (massive pulmonary embolus and microthrombosis), even in the absence of a subsequent decrease in ventilation and low cardiac output. Vdaw is increased by lung overdistension and additional ventilatory apparatus dead space. Endotracheal tubes, heat and moisture exchangers, and other common connectors may increase ventilator dead space and induce hypercapnia during low Vt or low minute ventilation. Vdaw calculations include the ventilator dead space. Because the anatomic dead space remains relatively constant as Vt is reduced, very low Vt is associated with a high Vd/Vt ratio [1, 2, 7, 8, 9].

Positive end-expiratory pressure (PEEP) is used to increase lung volume and to improve oxygenation in patients with acute lung injury. Vdalv is large in acute lung injury and does not vary systematically with PEEP. However, when the effect of PEEP is to recruit collapsed lung units resulting in an improvement of oxygenation, Vdalv may decrease, and alveolar recruitment is associated with decreased arterial minus end-tidal CO2 difference [4, 5, 6]. Conversely, PEEP-induced overdistension may increase Vdalv and widen this difference [7].

In patients with sudden pulmonary vascular occlusion due to pulmonary embolism, the resultant high VA/QT mismatch produces an increase in Vdalv. The association of a normal D-dimer assay result plus a normal Vdalv is a highly sensitive screening test to rule out the diagnosis of pulmonary embolism [9].

Dead space and outcome prediction

Characteristic features of acute lung injury are alveolar and capillary endothelial cell injuries that result in alterations of pulmonary microcirculation. Consequently, adequate pulmonary ventilation and blood flow across the lungs are compromised and Vdphys/Vt increases. A high dead-space fraction represents an impaired ability to excrete CO2 due to any kind of VA/QT mismatch [7]. Nuckton et al. [10] demonstrated that a high Vdphys/Vt was independently associated with an increased risk of death in patients diagnosed with acute respiratory distress syndrome.

Conclusions

The advanced technology combination of airway flow monitoring and mainstream capnography allows breath-by-breath bedside calculation of pulmonary Vd and CO2 elimination. For these reasons, the use of volumetric capnography is clinically more useful than time capnography. Measurement of dead-space fraction early in the course of acute respiratory failure may provide clinicians with important physiologic and prognostic information. Further studies are warranted to assess whether the continuous measurement of different derived capnographic indices is useful for risk identification and stratification, and to track the effect of a therapeutic intervention during the course of disease in critically ill patients.