Chapter 5, cont… (Page 3)

Oxygen transfer

Clinical problem 8
A 54­year­old man is admitted to the hospital following a stroke that has left him partially paralyzed and comatose. He is severely hypoxemic. His chest x­ray shows pneumonia in the right lung and a clear left lung. He has been intubated and artificially ventilated, and an arterial line is inserted to allow monitoring of his blood gases. After several adjustments are made, he is found to require 70% inspired oxygen.
To help prevent bed sores, the patient is turned every 2 hours. During a period in which the ventilator settings are not changed, the following blood gas values are obtained:
1 PM Patient lying on his right side: pH 7.45; PaCO2 30 mm Hg; PaO2 76 mm Hg
2 PM Patient lying on his left side: pH 7.43; PaCO2 32 mm Hg; PaO2 123 mm Hg
How do you explain these blood gas values?

West and his colleagues have done studies to plot the distribution of ventilation and perfusion in healthy and abnormal lungs, and their figures are useful for visualizing V/Q distribution (see Fig. 5-9, which shows V/Q distribution in two subjects, one age 22 and the other age 44). Most of the ventilation and perfusion goes to lung units with V/Q ratios around 1. Contrast this distribution to the abnormal curves in Fig. 5­10.

Figure 5-8

Fig. 5­8. Effect of gravity on regional ventilation m the lung. A, A coiled spring suspended vertically; the greatest separation between coils is at the top of the spring, the least separation is at the bottom. This is the same situation with alveoli in the upright lung. B, The pressure­volume curve of the lung demonstrating the relative position of apical and basilar alveoli. (Lung recoil pressure, transpulmonary pressure or Ppl [see Fig. 3­3].) C, Apical and basilar alveoli in the upright lung. Dotted circle, the end of inspiration; solid circle, end of expiration. For a given change in lung recoil pressure, the basilar alveoli have a larger change in volume and therefore a larger ventilation. (From Gibson, G.J.: Clinical tests of respiratory function, New York, 1984, Raven Press.)

Fig. 5­10 shows the V/Q curves obtained from two patients with chronic obstructive pulmonary disease. The curve in Fig. 5­10, A, is from a patient with pulmonary emphysema. His V/Q problem is a result of unperfused or underperfused alveoli and a resulting large amount of alveolar dead space. His pulmonary blood flow is matched by adequate alveolar ventilation, so his blood gas readings are normal or nearly so. However, so much extra air has to be brought in to fill the increased alveolar dead space that, although hypoxemia is not present, dyspnea is marked. Clinically this type of patient is often called a "pink puffer" because he has plenty of oxygen (is therefore not cyanotic), yet he has to labor to breathe. Typical blood gas values might be a PaO2 of 79 mm Hg, PaCO2 of 38 mm Hg, and pH of 7.43, with a respiratory rate of 30/min.

The curve in Fig. 5­10, B, is from a patient with chronic bronchitis who is hypoxemic, cyanotic, and perhaps in right­sided heart failure­­the so-called "blue bloater." He is bloated (often manifested by leg edema and ascites) because of failure of the right side of the heart. Note that a large amount of perfusion goes to areas that are underventilated. As a result, relatively unoxygenated blood passes through the lungs and mixes with the blood in the left side of the heart, thus severely depressing PaO2. Typical blood gas readings from such a patient are: PaO2 of 45 mm Hg, PaCO2 of 48 mm Hg, and pH of 7.36. (For reasons that are poorly understood, blue bloaters tend to be much less dyspneic at rest than are emphysema patients .)

Although V/Q imbalance can explain most cases of hypoxemia, exactly how the clinical disorder causes the imbalance is not always clear.

Figure 5-9

Fig. 5­9. Distribution of ventilation to perfusion in healthy people. A, V/Q distribution for a 22 year­old man; B, V/Q distribution for a 44­year­old man. In the older man 10% of the blood flow is to regions with V/Q ratios less than 0.1. (Reprinted from West, J.B., and Wagner, P.D.: Bioengineering aspects of the lung, p. 405, N.Y., 1977, by courtesy of Marcel Dekker, Inc.)

Figure 5-10 A Figure 5-10 B

Fig. 5­10. Distribution of ventilation to perfusion in patients with chronic obstructive pulmonary disease. A, V/Q distribution for a patient with emphysema. B, V/Q distribution for a patient with chronic bronchitis. (From Wagner, P.D., et al.: J. Clin. Invest. 59:203­206, 1977.)

Clinical problem 9
A patient with severe chest pain is admitted to the hospital's coronary care unit. A few rales are heard in his chest, and an arterial blood gas is obtained while he is receiving 2 L/min nasal oxygen with the following results: PaO2 77 mm Hg, PaCO2 36 mm Hg, and pH 7.45. A chest x­ray is normal.

For the first 24 hours he is stable. He then develops severe pain on the right side of his chest. A repeat chest x­ray reveals slightly elevated lung diaphragms but no infiltrates or other abnormality. A repeat blood gas while on 2 L/min nasal oxygen shows a PaO2 of 45 mm Hg, PaCO2 of 28 mm Hg, and pH of 7.51. A perfusion lung scan reveals markedly decreased perfusion in the right lower lung field.

How do you explain these blood gas values''


Shunt is an overused word in pulmonary physiology and often means different things to different people.

In its simplest definition, a shunt occurs whenever one thing bypasses another. In the lungs, a shunt can be thought of as an extreme form of venous admixture, i.e., a mixing of totally unventilated, unoxygenated blood with ventilated, oxygenated blood.*

Venous admixture can occur in one of three situations, only two of which are traditionally called "shunt."

1. An anatomic shunt occurs when blood bypasses the lungs through an anatomic channel, such as from the right to the left ventricle through a ventricular septal defect or from a branch of the pulmonary artery directly to a pulmonary vein.

2. A physiologic shunt occurs when a portion of the cardiac output goes through the regular pulmonary vasculature without coming into any contact with alveolar air. There is no abnormal connection between the blood vessels; rather, there is a severe redistribution of pulmonary blood flow. Physiologic shunting is often seen in conditions such as pulmonary edema, pneumonia, and lobar atelectasis.

3. Low ventilation­perfusion ratios occur when there is relatively more blood in the pulmonary capillary than can be fully oxygenated by the alveolar air. Although blood flow is to some extent redistributed, the blood is still exposed to some alveolar air. Low V/Q ratios account for most cases of hypoxemia seen clinically.

In terms of its effect on oxygenation, a physiologic shunt is not different from an anatomic shunt. In both, some unoxygenated blood bypasses the alveoli and mixes with oxygenated blood. Although both types of shunt represent venous admixture, they differ in one important aspect from venous admixture that occurs from low V/Q ratios. Since shunted blood contacts no air, increasing the fraction of inspired oxygen (FIO2) will not improve oxygenation (except by adding more dissolved oxygen to the normally oxygenated blood). In contrast, oxygenation of the blood from low V/Q areas will definitely be improved by increasing FIO2 because blood in low V/Q units is in contact with some air. Increasing the FIO2 should eventually denitrogenate the alveolar air in the low V/Q units and completely oxygenate the blood that serves these units; 100% oxygen should accomplish this exchange completely.

Administration of 100% oxygen was recommended in the past to determine whether hypoxemia was from low V/Q areas or from a shunt. It is now known that 100% oxygen can cause shunting by converting areas of low V/Q to 0 V/Q. This conversion happens when the pure oxygen in the poorly ventilated alveoli is fully absorbed by the capillary blood and the alveoli collapse. Well­ventilated alveoli (normal or high V/Q) are anatomically larger, and their collapse is less likely. Even if 100% oxygen gave an accurate measure of the percent shunt, the calculation would not ordinarily affect therapy. (Shunts and their calculation are discussed further in Chapter 11.)


PaO2. It has been stated that V/Q imbalance is the most common cause of hypoxemia, a result of lung units with low V/Q ratios. The mechanism of hypoxemia can now be examined more closely. In Fig. 5­11, arterial partial pressure of oxygen (PaO2) is plotted against the arterial oxygen content; this is the oxygen dissociation curve for a hemoglobin content of 15 grams%. The shape of the curve is the same as when PaO2 is plotted against the percent oxygen saturation of hemoglobin (see Chapter 6). Note that the curve is nearly flat in the range of physiologic PaO2 values (above 70 mm Hg) and falls steeply below 60 mm Hg. Points representing oxygen contents from three separate alveolar­capillary units are also shown. These units have V/Q ratios of 0.1, 1.0, and 10.0. Note that the decrement in capillary oxygenation caused by the low V/Q unit is not compensated for by the high V/Q unit.

Units with low V/Q ratios have low alveolar PO2 values. Blood perfusing these units has a low endcapillary PO2 PCO2) and hence a low end­capillary oxygen content. If there were a range of V/Q units from 1.0 (normal) down to 0, the result could only be hypoxemia since low oxygen contents would be mixing with normal oxygen contents. In fact, V/Q imbalance implies that at least some units are overventilated (high V/Q ratios) while others are underventilated (low V/Q ratios). The shape of the oxygen dissociation curve (Fig. 5­11) shows that high V/Q ratios will not balance out the low V/Q units.

The final PaO2 is determined not by an average of oxygen partial pressures but by an average of the oxygen contents. This may seem confusing at first, especially since the point was already made that PaO2 determines oxygen saturation and that oxygen saturation is a determinant of oxygen content. However, the oxygen dissociation curve shows that when aliquots of unequal oxygen content mix, the resulting PO2 is not an average of the mixing PO2's, but instead is an average of the mixing oxygen contents. Partial pressures of gases do not average out when equal aliquots of blood mix. It is the gas contents (in this case, oxygen and carbon dioxide) that mix and average out.

Figure 5-11

Fig. 5­11. Oxygen dissociation curve: PaO2 vs. oxygen content. Oxygen content from alveolarcapillary units with V/Q ratios of 0.1, 1, and 10 are, respectively, 16, 19.5, and 20.0 ml O2/100 ml blood. Lines are drawn for each content to its point on the dissociation curve. The average oxygen content, 18.5 ml 02/100 ml, is represented by a circle on the dissociation curve. Note that the arterial oxygen content after all the blood is mixed (18.5 ml O,/100 ml) is lower than the oxygen content from the normal unit (19.5 ml O,/100 ml).

Think of it this way. Alveolar PO2 determines PO2 and percent oxygen saturation in the pulmonary capillary; the percent oxygen saturation and the hemoglobin content determine the oxygen content (this is discussed further in Chapter 6). When a range of capillary oxygen contents mix, the average oxygen content will determine PaO2. Careful study of the dissociation curve will help clarify this important point.

Hyperventilation of some units does not add enough oxygen to balance out the low oxygen con tent from the hypoventilated units. Hyperventilation will increase PaO2 in high V/Q units, but above a PaO2 of 70 mm Hg, there is not much increase in the oxygen content of blood perfusing these units. The result is a final oxygen content determined mainly by the low V/Q areas and a resulting PaO2 that is lower than would be predicted by averaging the PO2 values from each pulmonary capillary.

Figure 5-12

Fig. 5­12. Gas exchange effects from progressive V/Q imbalance. Ventilation­perfusion inequality and overall gas exchange in computer model of the lung. For this model, oxygen uptake and carbon dioxide output were kept constant. As the amount of V/Q inequality increases (represented by a log scale on the abscissa), PaO2 decreases and Paco, increases; this change is because V/Q inequality causes increases in both venous admixture and alveolar dead space. (From Murray, J.F.: The normal lung, Philadelphia, 1976, W.B. Saunders Co.; modified from West, J.B.: Respir. Physiol. 7:88­110, 1969.

The decrease in PaO2 and the increase in venous admixture from progressive V/Q imbalance are shown in Fig. 5­12. This figure represents a computer lung model in which cardiac output, minute ventilation, and oxygen uptake are kept constant, and the effects of increasing V/Q imbalance on gas exchange are analyzed.

PaCO2. For many years it was taught that V/Q imbalance does not lead to elevated PaCO2 (hypercapnia) and that a patient had to hypoventilate to have carbon dioxide retention. No mention was made of the possible role of V/Q imbalance in carbon dioxide retention (West, 1971). In fact, "hypoventilation" and "elevated PaCO2" mean the same thing: decreased alveolar ventilation relative to carbon dioxide production, which can result from either decreased minute ventilation (VE) or increased dead space ventilation (VD) (see Chapter 4).

The most common cause of increased VD is V/Q imbalance. That V/Q imbalance can increase PaCO2 may not be intuitively obvious. Computer simulation (Fig. 5­12) also shows a progressive increase in alveolar dead space and in PaCO2 as V/Q imbalance decreases and when other factors are held constant.

Clinically, V/Q imbalance has a different effect on PaO2 and PaCO2 than that predicted in the computer model. The reason is that ''other factors" are almost never kept constant in a real patient. The variable effect of V/Q imbalance is discussed in the last section of this chapter.


Low mixed venous oxygen was earlier listed as one of the nonrespiratory factors that may lower arterial oxygen pressure PaO2) (Table 5­1). Discussion of this mechanism was purposely delayed until the concept of shunt and venous admixture was covered. It is only in the presence of shunting and/or low V/Q ratios that low mixed venous oxygen will depress the PaO2.

Factors that affect mixed venous oxygen content (CVO2) include cardiac output, hemoglobin content, and oxygen consumption. When the cardiac output or the hemoglobin content is reduced, the amount of oxygen delivered to the systemic capillaries will be reduced; if the body's oxygen consumption stays the same, it follows that the oxygen content of the blood returning to the right side of the heart will be reduced. This reduction will also be the case when the amount of oxygen delivered to the tissues is fixed and total oxygen consumption increases.

If there is no venous admixture, mixed venous blood will be fully oxygenated on one pass through the lungs. In the presence of venous admixture (low V/Q ratios or shunt), some of the mixed venous blood will not be oxygenated but will pass through the lungs and will join with the oxygenated blood in the pulmonary veins. The resulting oxygen content will be determined by the relative contribution of the amount of shunted blood that is poor in oxygen and the amount of nonshunted blood that is rich in oxygen. Given a certain amount of venous admixture, the lower the CVO2 (and hence the mixed venous oxygen pressure [PVO2]) the more will the arterial oxygen content (and PaO2) be reduced.

Reduction in PaO2 from low CVO2 is shown in Table 5­3, where the effect of anemia on PaO2 is calculated. In this example, the following are held constant: percent venous admixture, inspired oxygen pressure, PaCO2, cardiac output, and oxygen uptake. The result is a progressive fall in PaO2 as the hemoglobin is reduced­­solely because the anemia leads to reduced arterial oxygen delivery. Since oxygen uptake is unchanged, the amount of venous oxygen transport is reduced and hence CVO2 (and PVO2) is reduced. Low CVO2 depresses the oxygen content (and hence PaO2) of the normally oxygenated blood. If there were no venous admixture, pulmonary venous blood would be completely oxygenated on one pass through the lungs and hemoglobin content would not affect PaO2.

A real patient with reduced mixed venous oxygen resulting from anemia would try to compensate, mainly by increasing cardiac output and hyperventilating; this compensation would either prevent or ameliorate the fall in PaO2. Effects of reduced CVO2 are discussed further in Chapter 6.

[Table of Contents][Previous Chapter][Chapter 5, Page 1][Chapter 5, Page 2][Chapter 5, Page 4][Next Chapter]