Measurement of respiratory function: gas exchange

PHYSIOLOGY

Measurement of respiratory function: gas exchange Iain Campbell James Waterhouse

Blood normally spends about 0.75 s in the alveolar capillary. Equilibration of carbon dioxide occurs in about 0.25 s and of oxygen in about 0.4 s. Therefore, equilibrium occurs rapidly and is normally complete when the blood has traversed only about half the capillary. Within the pulmonary capillary there is thus a large reserve for gas transfer. Although pulmonary blood flow may rise fivefold with exercise, gas transfer (at sea level) is still normally complete. The pulmonary circulation also has a large reserve. With increases in cardiac output, pulmonary capillaries have the capacity to expand (distension) and capillaries that are closed at rest or have no blood in them, open up (recruitment). The carbon monoxide transfer test is used in chest medicine as a noninvasive method of quantifying gas transfer. Carbon monoxide is more soluble in blood than carbon dioxide or oxygen. When it is present in inspired air, equilibrium is not reached by the time the blood has perfused the alveoli. No appreciable loss of diffusion gradient develops in the alveolus and the gas continues to move rapidly into the bloodstream, therefore its uptake is normally independent of blood flow, and thus limited by diffusion. Its rapid uptake from the alveolus makes it a sensitive test of gas transfer. It is considered principally as an index of diffusion, but probably is also affected by ventilation–perfusion abnormalities. The carbon monoxide transfer test is not widely used in anaesthetic practice. •

•

Measurement of ventilation (VA) and perfusion (Q) involves the inhalation or injection of radioactive isotopes, respectively. These methods are in common use in chest medicine to assess localized abnormalities. Multiple gas elimination is a research technique that allows ventilation–perfusion abnormalities to be quantified precisely. It involves injection of tracer gases of varying solubilities and allows distribution curves to be plotted for pulmonary blood

Iain Campbell is Consultant Anaesthetist at the University Hospitals of South Manchester NHS Trust and Visiting Professor of Human Physiology at Liverpool John Moores University. He qualified from Guy’s Hospital Medical School, London, and trained in anaesthesia in Zimbabwe, Southend, Montreal and Leeds. James Waterhouse is Professor of Biological Rhythms at the Research Institute for Sport and Exercise Sciences, Liverpool John Moores University. He qualified as an animal physiologist at the University of Oxford and gained his DPhil there. His research interests are circadian rhythms in humans, with particular reference to jetlag and assessing the clock-driven component of a measured circadian rhythm.

ANAESTHESIA AND INTENSIVE CARE MEDICINE 6:11

366

© 2005 The Medicine Publishing Company Ltd

PHYSIOLOGY •

•

flow and ventilation in relation to VA:Q ratios. Such techniques are not in routine use in anaesthetic practice and are beyond the scope of this article, which focuses on the common clinical techniques used to measure ventilation and gas exchange, and the derivation of parameters that can be calculated from them.

Inspired PO2 (PIO2) is calculated from barometric pressure (760 mm Hg at sea level) and the proportion of O2 in inspired air (FIO2; for atmospheric air 20.9%) allowing for humidification of the inspired air; the saturated vapour pressure of water at 37oC is 47 mm Hg. Thus: PI O2 = (760 – 47) × 20.9/100 = 149 mm Hg

Blood gases The ultimate test of how well ventilation is matched to perfusion is the measurement of blood gases. Inadequate ventilation (hypoventilation) is always accompanied by a raised partial pressure of carbon dioxide in arterial blood (PaCO2). Mismatch of ventilation to perfusion results in hypoxaemia. PaCO2 does not rise with ventilation–perfusion abnormalities unless ventilation is inadequate, because of depression of the respiratory centre (e.g. opioids) or pathology of the ventilatory apparatus (e.g. muscle paralysis, crush injury of the chest). • PaCO2 is a function of CO2 production (VCO2) and alveolar • ventilation (VA) and is given by the relationship: •

The decrease in PO2 from the inspired air to the alveolar air is due to dilution by the CO2 in the alveoli, but the body produces less CO2 than it uses O2. The ratio of CO2 produced to O2 consumed is given by the respiratory quotient or respiratory exchange ratio (R). This is normally 0.7–1. It is seldom measured and a value of 0.8 is normally assumed. Based on this, PAO2 can be calculated from the simplified form of the alveolar air equation: PAO2 = PIO2 – PACO2/R PACO2 can be taken as end-tidal CO2, though it is more usual to use PaCO2. In terms of the figures quoted above:

•

PaCO2 = VCO2/VA × K

PAO2 = 149 – 40/0.8 = 99 mm Hg

•

where K is a constant. Thus, halving VA doubles PaCO2.

PAO2 can be influenced by raising or lowering the FIO2 or by changing ventilation, which changes PaCO2 and thus alters the PAO2 in line with the alveolar air equation. O2 uptake from the alveoli may also have an effect, particularly in patients with respiratory impairment and a raised oxygen consumption caused by shivering, pain, anxiety or fever. The difference between PAO2 and PaO2 (the alveolar–arterial oxygen difference or gradient) is due mainly to ventilation– perfusion mismatch, specifically alveoli over-perfused relative to • • • • their ventilation (i.e. VA:Q < 0.8) ranging from alveoli with a VA:Q ratio only slightly less than 0.8 to alveoli that are perfused and not • • ventilated at all (i.e. VA:Q = ∞; atelectasis, lobar collapse, pneu• • monia). In health (in the upright position), the alveoli with VA:Q ratios less than 0.8 occur in the lower part of the lung. There is a • • gradation of VA:Q ratios from about 3.3 at the top (over-ventilated and under-perfused) to 0.63 at the base (under-ventilated and • • over-perfused). This distribution of VA:Q ratios leads to a spread of PAO2 values, with those at the top of the lung being high (about 130 mm Hg) and those at the base being low (about 80 mm Hg). With such a spread of PAO2 values how can one justify using a single figure? The figure given by the alveolar air equation is conventionally taken to be an average value for PAO2 over the whole lung. However, the figure is heavily weighted by basal alveoli, which is where most of the ventilation (and perfusion) takes place. It is also known as the ideal PAO2. The principal factor affecting ventilation and perfusion in the lung is gravity in relation to the weight of the lung and the low pressures in the pulmonary circulation. All of the above arguments about the distribution of ventilation and perfusion have applied to the conscious individual in the upright position. When the • • individual lies down gravity still applies, so the areas of high VA:Q • • ratios move to the anterior aspect of the chest and of low VA:Q to the posterior. Airway closure then takes place posteriorly rather than at the base.

The oxygen cascade Figure 1 illustrates how the partial pressure of oxygen (PO2) falls as gas is moved from the inspired air to the tissues. Inspired air is humidified in the upper airways and gas exchange with the alveoli is a function of ventilation. From the alveoli, oxygen moves into the arteries. The difference between PO2 in the alveoli (PAO2) and in the arteries (PaO2) is a function of ventilation–perfusion relationships and, in some chest diseases, diffusion across the alveolar capillary membrane. From the arteries, gas exchange with the peripheral tissues occurs in the capillaries. O2 and CO2 diffuse through the tissues. Most oxygen is used, and most CO2 produced, in the mitochondria where PO2 is very low (1–4 mm Hg).

Oxygen cascade

Partial pressure of oxygen (mm Hg)

150

100

Inspired

Alveoli

Arteries

Capillaries 50

Veins

0

Mitochondria

1

ANAESTHESIA AND INTENSIVE CARE MEDICINE 6:11

367

© 2005 The Medicine Publishing Company Ltd

PHYSIOLOGY

and an alveolar–arterial difference of 23 mm Hg. In other words, the relative over-ventilation of the 80 and 120 mm Hg alveoli has little effect on arterial saturation and does not compensate for the under-ventilation of the 40 mm Hg alveolus which, because of its position on the curve, has a very marked effect. This is even greater in practice because blood flow through the base of the lung is normally much greater than through the apex.

Measurement of gas exchange The causes of hypoxaemia are broadly classified as impairment of diffusion (abnormalities of the pulmonary capillary membrane, pulmonary embolus, exercise at altitude), ventilation–perfusion abnormalities (specifically over-perfusion of under-ventilated alveoli) and ‘shunt’, the most extreme form of ventilation–perfusion • • mismatch (VA:Q = ∞). In normal circumstances, most arterial hypoxaemia is due to ventilation–perfusion mismatch.

Shunt True shunt refers to blood that enters the arterial system without passing through ventilated lung such as occurs with pneumonia, (consolidation) or pulmonary collapse. There is a small amount of shunt in the normal lung; blood supplying the bronchi (bronchial arteries) drains into the pulmonary veins and a small amount of coronary venous blood drains directly into the cavity of the left ventricle through the Thebesian veins. The addition of this desaturated blood depresses PaO2. The same occurs in individuals with a defect between the right and left sides of the heart. In clinical practice, true shunt is mainly seen in pneumonia with inflammatory consolidation of the alveoli or with the alveolar flooding of pulmonary oedema. If all hypoxaemia were caused by true shunt it would be possible to quantify the shunt from measurements and calculation of the oxygen content of arterial and mixed venous blood and a knowledge of PAO2. Figure 3 shows a shunt with a proportion of • • the cardiac output (Qt) passing through ventilated lung (Qc) and • a proportion through the shunt (Qs):

Alveolar–arterial difference The difference between PAO2 and PaO2 can be used as an indicator of ventilation–perfusion mismatch. In the young healthy individual, the alveolar–arterial difference is less than 10 mm Hg, but it increases with age and with any condition that impairs gas exchange. It exists because of the shape of the haemoglobin dissociation curve. Consider three alveoli of equal size and equal perfusion in the upper, mid and lower zones with PO2 of 120, 80 and 40 mm Hg (i.e. in an individual with some ventilation–perfusion mismatch; Figure 2). The average PAO2 is 80 mm Hg. The blood leaving these three alveoli has partial pressures corresponding to the PAO2, but there is little difference in the oxygen content between the 120 and 80 mm Hg alveoli because they both lie on the flat part of the oxygen–haemoglobin dissociation curve and the blood from them will be 98% and 95% saturated, respectively. However, the blood leaving the 40 mm Hg alveolus lies on the steeper part of the dissociation curve and its saturation is only 75%. Assuming equal blood flow through each of the three alveoli gives an overall arterial saturation of 89%, corresponding to a PaO2 of 57 mm Hg

•

•

•

Qt = Qc + Qs

Shunt

Origin of the alveolar–arterial gradient Upper zone Mid zone

Alveolar gas

Lower zone Cc'O2

Saturation of blood P AO 2 120 mm Hg

Qc

Qc

98%

CaO2

P AO 2 80 mm Hg

95%

89%

Qt

= 57 mm Hg P AO 2 40 mm Hg

Qt

CvO2

75%

Degree of venous admixture can be quantified by assuming that it is all due to true shunt and calculating the amount of shunt that would be required to produce that degree of hypoxaemia. End pulmonary capillary oxygen content (Cc'O2) is calculated assuming c'PO2 to be the same as partial pressure of oxygen in alveoli (PAO2) Qt, cardiac output; Qc, blood flow through ventilated lung; Qs, shunt. Oxygen content of arterial blood (CaO2) and mixed venous blood (CvO2)

Alveoli Lung

Mean partial pressure of oxygen in alveoli (PAO2) is 80 mm Hg Alveolar–arterial gradient = 80 – 57 = 23 mm Hg 3

2

ANAESTHESIA AND INTENSIVE CARE MEDICINE 6:11

Qs

368

© 2005 The Medicine Publishing Company Ltd

PHYSIOLOGY

Oxygen content (C) for each component can be calculated from the haemoglobin concentration and the PO2 of each component, _ such as mixed venous blood (CvO2), arterial blood (CaO2) and end pulmonary capillary blood (Cc'O2), the pressure of which is assumed to equal PAO2. Thus: •

•

Where VD is physiological dead space, VT is tidal volume, and PÉCO2 (the partial pressure of mixed expired CO2) is calculated from barometric pressure and the fractional concentration of CO2 in the mixed expired air. Mixed expired air is a mixture of all the air breathed out over a period and includes both the anatomical dead space and the alveolar gas. It is collected either into a spirometer or reservoir (Douglas) bag for analysis or passed through a mixing box and sampled. The important feature is that it includes all the dead space air and all the air that has taken part in gas exchange. Mixed expired CO2 concentration is usually about 2–4% CO2 depending on the amount of dead space. It must be distinguished from endtidal CO2 which normally reflects PaCO2 and is about 5–5.5% CO2. Rearranging the Bohr equation:

•

Qt × CaO2 = Qc × Cc'O2 + Qs × CvO2 Rearranging • Qs Cc'O2 – CaO2 _ • = Qt Cc'O2 – CvO2 _

gives shunt as a percentage of cardiac output. Calculation of CvO2 necessitates the measurement of mixed venous oxygen pressure using a pulmonary artery catheter. This also allows the measurement of cardiac output and a figure for shunt could be obtained in terms of volume/minute.

VD = VT ×

The calculation of VD thus entails measuring VT, PaCO2 and PECO2 all at the same time. This can be done by collecting expired air over a timed period and measuring its volume, its CO2 content and PaCO2 during the period of collection. VT can be calculated from the timed expired volume and the respiratory rate. It is important to maintain a steady state throughout; in a conscious individual this can be difficult when a mouthpiece and the sampling of arterial blood both tend to induce hyperventilation. With an intubated unconscious patient undergoing mechanical ventilation it is relatively easy and some ventilators calculate physiological dead space automatically using end-tidal CO2 as an approximation of PaCO2. As discussed earlier, in the diseased lung this can cause problems because the assumption that end-tidal CO2 approximates arterial CO2 may not be valid.

Ventilation–perfusion mismatch Hypoxaemia is seldom caused by true shunt; much of it arises from ventilation–perfusion mismatch. To measure this requires the sophisticated laboratory-based techniques referred to earlier, which are not readily available. A method of quantifying impairment of oxygenation in clinical practice is to assume that all the hypoxaemia, including ventilation–perfusion mismatch is caused by true shunt and to calculate the amount of shunt that would be required to produce that degree of hypoxaemia. The lung is considered to consist of: • alveoli that are ventilated but not perfused (dead space) • alveoli that are perfused but not ventilated (shunt) • alveoli that are perfused and ideally ventilated (i.e. blood leaving these alveoli (end pulmonary capillary blood) has the same PO2 as ideal alveolar air). In terms of ventilation–perfusion mismatch, the over-perfused under-ventilated alveoli at the base of the lung contribute to ‘shunt’ and hypoxaemia. This is often referred to as venous admixture. Similarly the over-ventilated under-perfused alveoli at the top of the lung contribute to ‘dead space’. A feature of hypoxaemia caused by true shunt is that increasing the FIO2 makes little difference to the severity of the hypoxaemia whereas hypoxaemia caused by ventilation–perfusion mismatch can be abolished by increasing the inspired concentration of oxygen.

Alveolar–arterial gradients and mixed expired PCO2: in practice, shunt can be measured only if a pulmonary artery line is in place, but the alveolar–arterial PO2 difference in breathing air is a reasonable reflection of the extent of ventilation–perfusion mismatch. Because of the shape of the haemoglobin-dissociation curve the relationship at higher FIO2 is more complex. From the Bohr equation it is self-evident that the mixed expired CO2 concentration has an inverse relationship with the size of the dead space. 

Dead space is gas that takes part in ventilation but not in gas exchange. It is made up of anatomical dead space and alveolar dead space, both of which constitute physiological dead space. Physiological and anatomical dead space can both be measured and alveolar dead space can therefore be calculated. Anatomical dead space is measured using Fowler’s technique. This is a research tool not commonly used in clinical practice. The method is based on the single-breath nitrogen test (see Anaesthesia and Intensive Care Medicine 6:10: 356). Physiological dead space can be measured in clinical practice and is calculated from the Bohr equation:

SI units are not routinely used in respiratory physiology and we have chosen to follow convention in these articles. For comparison 1 atmosphere = 100 kPa = 760 mm Hg = 1000 cm H2O

FURTHER READING Hedenstierna G. Respiratory measurement. London: BMJ Books, 2000. Lumb A B. Nunn’s applied respiratory physiology. Oxford: ButterworthHeinemann, 2000.

VD PaCO2 – PÉCO2 = VT PaCO2

ANAESTHESIA AND INTENSIVE CARE MEDICINE 6:11

PaCO2 – PÉCO2 PaCO2

369

© 2005 The Medicine Publishing Company Ltd