Chapter 4, cont… (Page 3)

PCO2 and alveolar ventilation


The normal ratio of dead space to tidal volume (VD/VT or VD/VE, which is the same thing) is approximately 150 ml VD/500 ml VT, or 0.3. Normal VD/VT ranges from approximately 0.28 to 0.33.

As pointed out previously, VD/VT can be elevated from either a reduction of VT or an actual increase in VD. Either cause of increased VD/VT can cause a decrease in alveolar volume (VA) and hence in alveolar ventilation (VA).

When the cause of increased VD/VT is lung disease, ventilatory adaptations will try to keep VA and PaCO2 normal (Table 4­1). These adjustments of course require an intact central nervous system and intact chest bellows. When ventilatory adaptations fail, VA will fall and PaCO2 will rise.

On occasion it is useful to quantitate the VD/VT. This can be done using the Bohr dead space equation:

(Eqn 4-11)

where PeCO2 is the mean expired carbon dioxide pressure, which is obtained from an expired air sample that is collected over a few minutes time. Normal PeCO2 is approximately 28 mm Hg. Thus 40--28/40 = 0.30.

Clinical problem 11
During an attempt to wean a patient from the use of artificial ventilation, her PaCO2 and PeCO2 are measured with the following results: PaCO2, 56 mm Hg; PeCO2, 26 mm Hg. What is the VD/VT? Apart from any other factors, does this ratio indicate the patient can be removed from the ventilator?


The dangers from an elevated PaCO2 are usually not from the excess carbon dioxide per se. In fact carbon dioxide is a respiratory stimulant until very high PaCO2 values are reached (90 mm Hg or greater), at which time carbon dioxide may depress breathing. In addition to indicating a state of respiratory failure, there are three distinct dangers associated with an elevated PaCO2.

Low PaO2 from high PaCO2. For a constant fraction of inspired oxygen (FIO2), as the PaCO2 rises, the alveolar oxygen pressure (PAO2) falls, roughly on a mm Hg­for­mm Hg basis. This change can be appreciated from the relationship of PaCO2 to PAO2 in the alveolar air equation (discussed extensively in Chapter 5); in addition a fall in PAO2, results in a corresponding drop in PaO2. Although this cause of hypoxemia can often be corrected by judicious use of supplemental oxygen, there are situations in which a further rise in PaCO2 will cause PaO2 to fall to a dangerous level.

Low pH from high PaCO2. A rise in PaCO2 will lead to a fall in pH; this can be seen in the relationship of PaCO2 to pH (Henderson­Hasselbalch equation; see Fig. 4­5 and Chapter 7). Acidemia is a potential trigger of cardiac arrhythmias. Although the critical level for hydrogen ion concentration varies in each situation, an arterial pH below 7.30 that is not improving should be considered potentially life­threatening.

Decreased ventilatory reserve. Finally, a high PaCO2 represents a precarious situation in terms of ventilatory reserve. Small changes in alveolar ventilation (VA) that would be inconsequential in a healthy individual can be disastrous in someone retaining carbon dioxide. The reason why can be seen in the hyperbolic relationship when VA is plotted against PaCO2 (Fig. 4­3). For example, when carbon dioxide production (VCO2) is 200 ml/ min, a 500 ml/min decrease in VA (which may occur during central nervous system depression) will increase PaCO2 only 5 mm Hg, from a normal baseline PaCO2 of 40 mm Hg (VA = 4.3 L/min) to 45 mm Hg (VA = 3.8 L/min). When the baseline PaCO2 is 60 mm Hg (VA = 2.9 L/min), the same 500 ml/min decrease will elevate PaCO2 by 12.5 mm Hg to 72.5 mm Hg (VA = 2.4 L/min). The resultant changes in PaO2 and pH will also be amplified.

Clinical problem 12
For the following two patients, an initial VA and PaCO2 are given. Assuming constant VCO2 of 200 ml/min, what is the new PaCO2, after the described change in VA?

Patient A:

VA = 6 PaCO2 = 29 VA decreases by 1 L/min

L/min mm Hg because of administra­

tion of anesthesia

Patient B:

VA = 3 PaCO2 = 57.5 VA decreases by 1 L/min

L/min mm Hg because of pulmonary edema


Because hyperventilation represents an excess of alveolar ventilation (VA), intubation to increase VA is never needed if the PaCO2 is low. (Patients with low PaCO2 may need artificial ventilation to help correct other problems, e.g. hypoxemia or severe alkalosis.)

Ventilatory assistance is used only for patients with elevated PaCO2 or in those unusual situations where PaCO2 is in the normal range but the patient is in imminent danger of life­threatening hypoventilation. Ventilatory assistance usually requires tracheal intubation and connection to an artificial respirator. Alternate methods of directly augmenting VA, such as by using a tight­fitting face mask and intermittent positive pressure breathing, are almost never satisfactory in adults needing ventilatory assistance.

Figure 4-3

Fig. 4­3. PaCO2 vs. alveolar ventilation (VA). The relationship is shown for carbon dioxide production rates of 200 ml/min and 300 ml/min. A decrease in alveolar ventilation (VA) in the hypercapnic patient will result in a greater rise in PaCO2 than will the same VA change when PaCO2 is low or normal. Also, an increase in carbon dioxide production when VA is fixed will result in an increase in PaCO2.

Given an elevated PaCO2, a good clinical rule is never intubate for hypercapnia alone. Different durations of hypercapnia, different buffering capacities, oxygen levels, and a host of other variables influence the need for ventilatory assistance. In the presence of hypercapnia, intubation and artificial ventilation are indicated only if one or more of the following are also present and judged to be life threatening:

  1. Decreased mental status, not improving and potentially worsening
  2. Increased fatigue, not improving and potentially worsening
  3. Low pH (usually less than 7.30), not improving and potentially worsening
  4. Low arterial oxygen pressure (PaO2) that cannot otherwise be improved except by lowering the PaCO2
  5. Secretions or mucus that is threatening upper airways patency

Obviously, a great deal of clinical judgment must enter into the decision to intubate a patient. Once a patient is intubated, PaCO2 can only be followed by specific measurement for the reasons given previously. Even though the tidal volume and the respiratory rate are set by the ventilator and minute ventilation is therefore known, VA and carbon dioxide production remain unknown. Furthermore, in the nonstable patient, one cannot assume a constant dead space ventilation (VD) since changing tidal volume, respiratory rate, and the time course of acute parenchymal disease (changing V/Q relationships) may influence VD throughout a patient's course.

Weaning a patient from the ventilator also involves careful observation of his arterial blood gases and, on occasion, measurements of lung mechanics (see Chapter 10). It is impossible to give specific guidelines about how often to obtain blood gas analysis; this decision has to be individualized for each patient's course. A single blood gas measurement following intubation may not reflect a stable PaCO2; generally at least two or more blood gas measurements should be obtained in the first few hours after intubation to help assure that the patient has reached a ventilatory steady state.


Because carbon dioxide is never diffusion limited, alveolar carbon dioxide pressure (PACO2) is assumed equal to arterial carbon dioxide pressure (PaCO2). In theory, measurement of PACO2 could substitute for PaCO2, although in practice this is not always the case.

Figure 4­4A shows a normal tracing of partial pressure of carbon dioxide (PCO2) measured during a single expired tidal volume with an infrared carbon dioxide analyzer. The first part of

the expired air is the same as the last part that was inspired on the previous breath; (it is dead­space air from the upper airways and will contain almost no carbon dioxide). Gradually, air from some of the alveoli begins to join this dead­space air, and the PCO2 rises. By the very end of exhalation all the deadspace air has left the lungs, and the last few milliliters of air are from the alveoli only. This tracing shows that the end­tidal PCO2 (PetCO2) is approximately 38 mm Hg, which indicates a normal PaCO2.

Figure 4-4

Fig. 4­4. A, Carbon dioxide measurement during a single expired breath. In this example from a healthy patient, the end­tidal point reflects alveolar, and hence arterial, partial pressure of carbon dioxide. B, Continuous monitoring of end­tidal carbon dioxide (PetCO2). This patient has severe chronic obstructive pulmonary disease. Some variation is seen during quiet breathing, but average PetCO2 is approximately 50 mm/Hg. PaCO2 measured at the same time was 74 mm Hg.

PetCO2 can be measured on a continuous basis (Fig. 4­4, B), but the measurement has limitations. One has to assure that the carbon dioxide cannula, which delivers the expired air to the carbon dioxide analyzer, is not contaminated with room air. This is not so much of a problem with intubated patients for whom the cannula is inserted in the ventilator's expiratory circuit as it is in other patients.

Perhaps the major pitfall is the difficulty of obtaining true PACO2 in patients with severe lung disease. In such cases PetCO2 may not reflect alveolar and arterial PCO2 because of severe ventilation/perfusion imbalance and a resulting large increase in physiologic dead space (see Chapter 5). In the example shown in Fig. 4­4. B, from a patient with severe chronic obstructive pulmonary disease, the PetCO2 averaged approximately 50 mm Hg, but PaCO2 was 74 mm Hg, resulting in a PaCO2­PetCO2 difference of 24 mm Hg. In this situation the diseased alveoli do not empty evenly, and the end tidal sample still reflects considerable dead space air.

A PaCO2­PetCO2 difference does not obviate the value of the end­tidal measurement for physiologic monitoring; a rise in PetCO2 still suggests a rise in PaCO2, but one cannot equate the measured PetCO2 with PaCO2. For physiologic monitoring of critically ill patients, one or two comparisons should be made of PetCO2 with PaCO2 before following the PetCO2 trend.

The absolute value of the PaCO2­PetCO2 difference has also been advocated for diagnostic purposes, especially in acute pulmonary embolism where the value is often much higher than in chronic lung conditions. The pulmonary embolus creates extra dead space by blocking perfusion to a group of alveoli. However, because of a lack of specificity, this measurement is not widely used in clinical practice.


Any discussion of gas exchange should begin with PaCO2 since it is the only blood gas value that provides information on ventilation, oxygenation, and acid­base balance. Fig. 4­5 shows the relationship of PaCO2 to alveolar ventilation (the PaCO2 equation), alveolar partial pressure of oxygen (the alveolar­air equation; see Chapter 5), and pH (Henderson­Hasselbalch equation; see Chapter 7).

Figure 4-5

Fig. 4­5. Arterial carbon dioxide pressure (PaCO2) in ventilation, oxygenation, and acid­base equations. A rise in PaCO2 indicates diminished VA in relation to VCO2, and will result in a fall in PaO2 and pH. See Chapters 5 and 7.


For gas exchange to occur fresh air must be brought into the alveoli. Alveolar ventilation (VA) is defined as the amount of fresh air that enters the alveoli and takes part in gas exchange; it is the difference between total or minute ventilation (VE) and the amount of air that does not take part in gas exchange--the dead­space ventilation (VD).

VA is inversely related to the partial pressure of carbon dioxide in arterial blood (PaCO2) and is directly related to metabolic carbon dioxide production (VCO2). When VA rises proportionately higher than VCO2, PaCO2 is reduced, a condition known as hyperventilation; conversely, a level of VA proportionately lower than normal will raise PaCO2 (hypoventilation).

By employing a constant (0.863) to equate the different units for PaCO2, VA, and VCO2, the three variables can be related thus: PaCO2 = VCO2 x 0.863/VA. Normally, VA will rise to match any increase in VCO2. During mild to moderate exercise, both VA and VCO2 increase proportionately, so that PaCO2 stays the same; the exercising person neither hyperventilates nor hypoventilates.

Based on this equation, it follows that hypercapnia is always caused by a level of VA that is inadequate for VCO2. Furthermore, since VA = VE ­ VD, all cases of hypercapnia can be seen as caused by a reduced or inadequate VE, or an elevated VD (or a combination of the two). Hypercapnia caused by drug overdose, for example, can be explained by a reduction in VE. Hypercapnia in chronic obstructive lung disease can be explained by an elevation of VD.

The most common cause of elevated VD is ventilation­ perfusion imbalance. Dead­space ventilation can also be elevated in states of rapid shallow breathing, in which a larger­than­normal proportion of each tidal volume goes to satisfy anatomic dead space.

PaCO2 is a key blood gas measurement. Not only does it help assess adequacy of VA, but it is also a component of the alveolar air equation and the Henderson­Hasselbalch equation.


State whether each of the following are true or false.

1. To estimate PaCO2 at the bedside, one can start with a measurement of 40 mm Hg, then subtract 2 mm Hg for every breath above 10/min.

2. PaCO2 is inversely related to alveolar ventilation.

3. PaCO2 is directly related to level of carbon dioxide production.

4. PaCO2 is always low if the alveolar ventilation is twice the resting level.

5. Normally, one can voluntarily hyperventilate to lower PaCO2 more than 10 mmHg.

6. Normally, one can voluntarily hypoventilate to raise PaCO2 more than 10 mmHg.

7. Dead­space ventilation can rise solely from a change in the pattern of breathing, i.e., without a change in the lung architecture.

8. Most of the blood carbon dioxide is carried in the form of bicarbonate.

9. To calculate the ratio of dead space to tidal volume using the Bohr equation, one need measure only tidal volume and PaCO2.

10. As PaCO2 goes up, alveolar PO2 goes down.


Mithoefer, J.C., Bossman, O.G., Thibeault, D.W., et al.: The clinical estimation of alveolar ventilation, Am Rev Resp Dis 98:868, 1968.

Suggested readings

Goldring, R.M., Heinemann, H.O., and Turino, G.M.: Regulation of alveolar ventilation in respiratory failure, Am. I. Med. Sci. 269:160, 1975.

Javahen, S., Blum, J., and Kazemi, H.: Pattern of breathing and carbon dioxide rentention in chronic obstructive lung disease, Am. J. Med. 71:228, 1981.

Thomas, H.M.: Ventilation and PCO2: make the distinction, Chest 79:617, 1981. See also General References (Physiology) in Appendix G.

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