gas exchange

[Respiration][Circulation] Blood Flow to the Lung – General and Pulmonary Vascular Resistance

September 11, 2016 Cardiology, Critical Care, Hemodynamics, Mechanical Ventilation, Physiology and Pathophysiology, Pulmonary Medicine, Respirology No comments , , , , , , , , , , , ,

The lung receives blood flow via both the bronchial circulation and the pulmonary circulation. Bronchial blood flow constitutes a very small portion of the output of the left ventricle and supplies part of the tracheobronchial tree with systemic arterial blood. Pulmonary blood flow (PBF) constitutes the entire output of the right venticle and supplies the lung with the mixed venous blood draining all the tissues of the body.

There is about 250 to 300 mL of blood per square meter of body surface area in the pulmonary circulation. About 60 to 70 mL/m2 of this blood is located in the pulmonary capillaries.

Gas exchange starts to take place in smaller pulmonary arterial vessels, which are not truly capillaries by histologic standards. These arterial segments and successive capillaries may be thought of as functional pulmonary capillaries.

About 280 billion pulmonary capillaries supply approximately 300 million alveoli, resulting in a potential surface area for gas exchange estimated to be 50 to 100 m2.

Bronchial Circulation

The bronchial arteries arise variably, either directly from the aorta or from the intercostal arteries. They supply arterial blood to the tracheobronchial tree and to other structures of the lung down to the level of the terminal bronchioles. They also provide blood flow to the hilar lymph nodes, visceral pleura, pulmoonary arteries and veins, vagus, and esophagus. The bronchial circulation may be important in the "air-conditioning" of inspired air. The blood flow in the bronchial circulation constitutes about 2% of left ventricle output of the left ventricle. Blood pressure in the bronchial arteries is the same as that in other systemic arteries.

The venous drainage of the bronchial circulation is unusual. Although some of the bronchial venous blood enters the azygos and hemiazygos veins, a substantial portion of bronchial venous blood enters the pulmonary veins. Therefore, the bronchial venous blood entering the pulmonary venous blood is part of the normal anatomic right-to-left shunt. Histologists have also identified anastomoses, or connections, between some bronchial capillaries and pulmonary capillaries and between bronchial arteries and branches of the pulmonary artery. Thse connections probably play little role in a healthy person but may open in pathologic states, such as when either bronchial or PBF to a protion of lung is occluded. For example, if PBF to an area of the lung is blocked by a pulmonary embolus, bronchial blood flow to that area increases.

Pulmonary Circuation

The pulmonary vessels offer much less resistance to blood flow than do the systemic arterial tree. They are also much more distensible than systemic arterial vessels. These factors lead to much lower intravascular pressures than those found in the systemic arteries, which makes them more compressible. The pulmonary vessels are located in the thorax and are subject to alveolar and intrapleural pressures that can change greatly. Therefore, factors other than the tone of the pulmonary vascular smooth muscle may have profound effects on pulmonary vascular resistance (PVR). The transmural pressure difference across vessel walls is therefore a major determinant of PVR.

Because the right and left circulations are in series, the outputs of the right and left venticles must be approximately equal to each other over the long run. If the 2 outputs are the same and the measured pressure drops across the systemic circulation and the pulmonary circulation are bout 98 and 10 mm Hg, respectively, then the PVR must be about one tenth that of the systemic vascular resistance (SVR). Again, the low resistance to blood flow offered by the pulmonary circulation is due to the structural aspects of the pulmonary circulation.

The resistance is fairly evenly distributed among the pulmonary arteries, the pulmonary capillaries, and the pulmonary veins (from 15 mm Hg to 12 mm Hg, from 12 mm Hg to 8 mm Hg, from 8 mm Hg to 5 mm Hg, respectively). At rest, about one third of the resistance (PVR) is located in the pulmonary arteries, about one third is located in the pulmonary capillaries, and about one third is located in the pulmonary veins.


screen-shot-2016-09-11-at-3-44-33-pmThe relative small amounts of vascular smooth muscle, low intravascular pressures, and high distensibility of the pulmonary circulation lead to a much greater importance of extravascular effects ("passive factors") on PVR. Gravity, body position, lung volume, alveolar and intrapleural pressures, intravascular pressures, and right ventricular output all can have profound effects on PVR without any alteration in the tone of the pulmonary vascular smooth muscle.

Transmural Pressure On PVR

For distensible-compressible vessels, the transmural pressure difference is an important determinant of the vessel diameter. As the transmural pressure difference (which is equal to pressure inside minus pressure outside) increases, the vessel diameter increases and resistance falls; as the transmural pressure difference decreases, the vessel diameter decreases and the resistance increases. Negative transmural pressure differences lead to compression or even collapse of the vessel.

Lung Volume on PVR

Screen Shot 2016-09-06 at 12.48.46 PMTwo different groups of pulmonary vessels must be considered when the effects of changes in lung volume on PVR are analyzed: those vessels that are exposed to the mechanical influences of the alveoli and the larger vessels that are not – the alveolar and extraalevolar vessels.

As lung volume increases during a normal negative-pressure inspiration, the alveoli increase in volume. While he alveoli expand, the vessels found between them, mainly pulmonary capillaries, are elongated. As these vessels are stretched, their diameters decrease, just as stretching a rubber tube causes its diameter to narrow. Resistance to blood flow through the alveolar vessels increases as the alveoli expand because the alveolar vessels are longer (resistance is directly proportional to length) and because their radii are smaller (resistance is inversely proportional to radius to the fourth power). At high lung volumes, then, the resistance to blood flow offered by the alveolar vessels increases greatly; at low lung volumes, the resistance to blood flow offered by the alveolar vessels decreases.

One group of the extraalveolar vessels, the larger arteries and veins, is exposed to the intrapleural pressure. As lung volume is increased by making the intrapleural pressure more negative, the transmural pressure difference of the larger arteries and veins increase and they distend. Another factor tending to decrease the resistance to blood flow offered by the extraalveolar vessels at higher lung volumes is radial traction by the connective tissue and alveolar septa holding the larger vessels in place in the lung. Thus, at high lung volumes, the resistance to blood flow offered by the extraalveolar vessels decreases. During a forced expiration to low lung volumes, however, intrapleural pressure becomes very positive. Extraalveolar vessels are compressed, and as the alveoli decrease in size, they exert less radial traction on the extraalveolar vessels. The resistance to blood flow offered by the extraalveolar vessels increase greatly.

Because the alveolar and extraalveolar vessels may be thought of a 2 groups of resistances in series with each other, the resistances of the alveolar and extraalveolar vessels are additive at any lung volume. Thus, the effect of changes in lung volume on the total PVR gives the U-shape cruve. PVR is lowest near the functional residual capacity and increases at both high and low lung volumes.

There is another type of extraalveolar vessel called corner vessel, or extraalveolar capillary. Although these vessels are found between alveoli, their locations at junctions of alveolar septa give them different mechanical properties. Expansion of the alveoli during inspiration increases the wall tension of the alveolar septa, and the corner vessels are distended by increased radial traction, whereas the alveolar capillaries are compressed.

Also note that during mechanical positive-pressure ventilation, alveolar pressure (PA) and intrapleural pressure are positive during inspiration. In this case, and the resistance to blood flow offered by both alveolar and extraalveolar vessels increases during lung inflation. This is especially a problem during mechanical positive-pressure ventilation with positve end-expiratory pressure (PEEP). During PEEP, airway pressure (and thus alveolar pressure) is kept positive at end expiratory to help prevent atelectasis. In this situation, alveolar pressure and intrapleural pressure are positive during both inspiration and expriation. PVR is elevated in both alveolar and extraalveolar vessels throughout the respiratory cycle. In addition, because intrapleural pressure is always positive, the other intrathoracic blood vessels are subjected to decreased transmural pressure differences; the venae cavae, which have low intravascular pressure, are also compressed. If cardiovascular reflexes are unable to adjust to this situation, cardiac output may fall precipitously because of decreased venous return (for the reason see thread "Effects of Pressure Outside the Heart on Cardiac Output" at and high PVR.

Recruitment and Distention

During exercise, cardiac output can increase several-fold without a correspondingly great increase in MPAP. Although the MPAP does increase, the increase is only a few millimeters of mercury, even if cardiac output has doubled or tripled. Since the pressure drop across the pulmonary circulation is proportional to the cardiac output times the PVR, this must indicate a decrease in PVR.

Like the effects of lung volume on PVR, this decrease appears to be passive – that is, it is not a result of changes in the tone of pulmonary vascular smooth muscle caused by neural mechanisms or humoral agents. In fact, a fall in PVR in response to increased blood flow or even an increase perfusion pressure can be demonstrated in a vascularly isolated perfused lung. There are two different mechanisms that can explain this decrease in PVR in response to elevated blood flow and perfusion pressure: recruitment and distention.


At resting cardiac outputs, not all the pulmonary capillaries are perfused. A substantial proportion of capillaries, perhaps as large as one half to two thirds, is probably not perfused because of hydrostatic effects. Others may be unperfused because they have a relatively high critical opening pressure. That is, these vessels, because of their high vascular smooth muscle tone or other factors such as positive alveolar pressure, require a higher perfusion pressure than that solely necessary to overcome hydrostatic forces. Under normal circumstances, it is not likely that the critical opening pressures for pulmonary blood vessels are very great because they have so little smooth muscle. Increased blood flow increases the MPAP, which opposes hydrostatic forces and exceeds the critical opening pressure in previously unopened vesels. This series of events opens new parallel pathways for blood flow, which lowers the PVR. This opening of new pathways is called recruitment. Note that decreasing the cardiac output or pulmonary artery pressure can result in a derecruitment of pulmonary capillaries.


As perfusion pressure increases, the transmural pressure gradient of the pulmonary blood vessels increases, causing distention of the vessels. This increases their radii and decreases their resistance to blood flow.

Control of Pulmonary Vascular Smooth Muscle

Pulmonary vascular smooth muscle is responsive to both neural and humoral influences. These produce "active" alterations in PVR, as opposed to those "passive" factors discussed in the previous section.

The pulmonary vasculature is innervated by both sympathetic and parasympathetic fibers of the autonomic nervous system. The innervation of pulmonary vessels is relatively sparse in comparsion with that of systemic vessels. There is relatively more innervation of the larger vessels and less of the smaller, more muscular vessels. There appears to be no innervation of vessels smaller than 30 um in diameter. There does not appear to be much innervation of intrapulmonary veins and venules.

The effects of stimulation of the sympathetic innervation of the pulmonary vasulature are somewhat controversial. Some investigators have demonstrated an increase in PVR with sympathetic stimulation of the innervation of the pulmonary vasculature, whereas others have shown only a decreased distensibility with no change in calculated PVR. Stimulation of the parasympathetic innervation of the pulmonary vessels generally causes vasodilation, although its physiologic function is not known.

The catecholamines epinephrine and norepinephrine both increase PVR when injected into the pulmonary circulation. Histamine, found in the lung in mast cells, is a pulmonary vasoconstrictor. Certain prostaglandins and related substances, such as PGF2alpha, PGE2, and thromboxane, are also pulmonary vasoconstrictors, as is endothelin, a 21-amino acid peptide synthesized by the vascular endothelium. Alveolar hypoxia and hypercapnia also cause pulmonary vasoconstriction. Acetylcholine, the beta-adrenergic agonist isoproterenol, nitric oxide (NO), and certain prostaglandins, such as PGE1, and PGI2 (prostacyclin), are pulmonary vasodilators.

Gravity's Impact on PVR

Determinations of the regional distribution of PBF (see discussion below) have shown that gravity is another important "passive" factor affecting local PVR and the relative perfusion of different regions of the lung (see discussion below). The interaction of the effects of gravity and extravascular pressures may have a profound influence on the relative perfusion of different areas of the lung.

The Regional Distribution of Pulmonary Blood Flow

Interaction of Gravity and Extravascular Pressure

Experiments done on excised, perfused, upright animal lungs have demonstrated the same gradient of increased perfusion per unit volume from the top of the lung to the bottom. When the experiments were done at low pump outputs so that the pulmonary artery pressure was low, the uppermost regions of the lung received no blood flow. Perfusion of the lung ceased at the point at which alveolar pressure (PA) was just equal to pulmonary arterial pressure (Pa). Above this point, there was no perfusion because alveolar pressure exceeded pulmonary artery pressure, and so the transmural pressure across capillary walls was negative. Below this point, perfusion per unit volume increased steadily with increased distance down the lung.

screen-shot-2016-09-12-at-1-51-11-pmThus, under circumstances in which alveolar pressure is greater than pulmonary artery pressure in the upper parts of the lung, no blood flow occurs in that region, and the region is referred to as being in zone 1, as shown in Fingure 4-9. Any zone 1, then, is ventilated but not perfused. It is alveolar dead space. Fortunately, during normal, quiet breathing in a person with a normal cardiac output, pulmonary artery pressure, even in the uppermost regions of the lung, is greater than alveolar pressure, and so there is no zone 1. Some experiments have also demonstrated perfusion of the corner vessels under zone 1 conditions.

The lower portion of the lung in Figure 4-9 is said to be in zone 3. In this region, the pulmonary artery pressure and the pulmonary vein pressure (Pv) are both greater than alveolar pressure. The driving pressure for blood flow through the lung in this region is simply pulmonary artery pressure minus pulmonary vein pressure. Note that this driving pressure stays constant as one moves further down the lung in zone 3 because the hydrostatic pressure effects are the same for both the arteries and the veins.

The middle portion of the lung in Figure 4-9 is in zone 2. In zone 2, pulmonary artery pressure is greater than alveolar pressure, and so blood flow does occur. However, because alveolar pressure is greater than pulmonary vein pressure, the effective driving pressure for blood flow is pulmonary artery pressure minus alveolar pressure in zone 2. Notice that in zone 2 the increase in blood flow per distance down the lung is greater than it is in zone 3. This because the upstream driving pressure, the pulmonary artery pressure, increases according to the hydrostatic pressure increase, but the effective downstream pressure, alveolar pressure, is constant throughout the lung at any instant.

It is important to realize that the boundaries between the zones are dependent on physiologic conditions – they are not fixed anatomic landmarks. Alveolar pressure changes during the course of each breath. During eupneic breathing these changes are only a few centimeters of water, but they may be much greater during speech, exercise, and other conditions. A patient on a positive-pressure ventilator with PEEP may have substantial amounts of zone 1 because alveolar pressure is always high. Similarly, after a hemorrhage or during general anesthesia, PBF and pulmonary artery pressure are low and zone 1 conditions are also likely. During exercise, cardiac output and pulmonary artery pressure increase and any existing zone 1 should be recruited to zone 2. The boundary between zones 2 and 3 will move upward as well. Pulmonary artery pressure is highly pulsatile, and so the borders between the zones probably even move up a bit with each contraction of the right ventricle.

Changes in lung volume also affect the regional distribution of PBF and will therefore affect the boundaries between zones. Finally, changes in body position alter the orientation of the zones with respect to the anatomic locations in the lung, but the same relationships exist with respect to gravity and alveolar pressure.

Hypoxic Pulmonary Vasoconstriction

Alveolar hypoxia or atelectasis causes an active vasoconstriction in the pulmonary circulation. The site of vascular smooth muscle constriction appears to be in the arterial (precapillary) vessels very close to the alveoli.

The mechanism of hypoxic pulmonary vasoconstriction is not completely understood. The response occurs locally, that is, only in the area of the alveolar hypoxia. Connections to the central nervous system are not necessary: An isolated, excised lung, perfused with blood by a mechanical pump with a constant output, exhibits an increased perfusion pressure when ventilated with hypoxic gas mixtures. This indicates that the increase in PVR can occur without the influence of extrinsic nerves. Thus, it is not surprising that hypoxic pulmonary vasoconstriction persists in human patients who had received heart-lung transplants. Hypoxia may cause the release of a vasoactive substance from the pulmonary parenchyma or mast cells in the area. Histamine, serotonin, catecholamines, and prostaglandins have all been suggested as the mediator substance, but none appears to completely mimic the response. Decreased release of a vasodilator such as nitric oxide may also be involved in hypoxic pulmonary vasoconstriction. Possibly several mediators act together. More recent studies have strongly indicated that hypoxia acts directly on pulmonary vascular smooth muscle to produce hypoxic pulmonary vasoconstriction.

Physiologic Function of Hypoxic Pulmonary Vasoconstriction

The function of hypoxic pulmonary vasoconstriction in localized hypoxia is fairly obvious. If an area of the lung becomes hypoxic because of airway obstruction or if localized atelectasis occurs, any mixed venous blood flowing to that area will undergo little or no gas exchange and will mix with blood draining well-ventilated areas of the lung as it enters the left atrium. This mixing will lower the overall arterial PO2 (PaO2) and may even increase the arterial PCO2 (PaCO2). The hypoxic pulmomary vasoconstriction diverts mixed venous blood flow away from poorly ventilated areas of the lung by locally increasing vascular resistance. Therefore, mixed venous blood is sent to better-ventilated areas of the lung. The problem with hypoxic pulmonary vasoconstriction is that it is not a very strong response because there is so little smooth muscle in the pulmonary vasculature. Very high pulmonary artery pressures can interfere with hypoxic pulmonary vasoconstriction, as can other physiologic disturbances, such as alkalosis.

In hypoxia of the whole lung, such as might be encountered at high altitude or in hypoventilation, hypoxic pulmonary vasoconstriction occurs throughout the lung. Even this may be useful in increasing gas exchange because greatly increasing the pulmonary artery pressure recruits many previously unperfused pulmonary capillaries. This increases the surface area available for gas difusion and improves the matching of ventilation and perfusion. On the other hand, such a whole-lung hypoxic pulmonary vasoconstriction greatly increases the workload on the right venticle, and the high pulmonary artery pressure may overwhelm hypoxic pulmonary vasoconstriction in some parts of the lung, increase the capillary hydrostatic pressure in those vessels, and lead to pulmonary edema.

[Respiratory] Pulmonary Gas Exchange and Its Measurement

September 9, 2016 Cardiology, Clinical Skills, Critical Care, Physiology and Pathophysiology, Pulmonary Medicine No comments , , , , , , , , , , , , , , , , ,

Pulmonary Gas Exchange

The efficiency of gas exchange in the lungs is determined by the balance between alveolar ventilation and pulmonary capillary blood flow. This balance is commonly expressed as the ventilation-perfusion (V/Q) ratio. The influence of V/Q ratios on pulmonary gas exchange can be described using an alveolar-capillary unit (V/Q = 1; V/Q >1; and V/Q <1).

A V/Q ratio above 1.0 describes the condition where ventilation is excessive relative to pulmonary capillary blood flow. The excess ventilation, known as dead space ventilation, does not participate in gas exchange with the blood. Dead space ventilation includes anatomic dead space, which is the gas in the large conducting airways that does not come in contact with capillary blood, and physiologic dead space, which is alveolar gas that does not equilibrate fully with capillary blood. In normal subjects, dead space ventilation (VD) accounts for 20% to 30% of the total ventilation (VT).

Dead space ventilation increases in the following situations:

1.When the alveolar-capillary interface is destroyed; e.g., emphysema

2.When blood flow is reduced; i.e., low cardiac output

3.When alveoli are overdistended; e.g., during positive-pressure ventilation

An increase in VD/VT above 0.3 results in both hypoxemia (decreased arterial PO2) and hypercapnia (increased arterial PCO2), which is analogous to what would happen if you held your breath. The hypercapnia usually appears when the VD/VT is above 0.5.

A V/Q ratio below 1.0 occurs when pulmonary capillary blood flow is excessive relative to ventilation.The excess blood flow, known as intrapulmonary shunt, does not participate in pulmonary gas exchange. There are two types of intrapulmonary shunt. True shunt indicates the total absence of gas exchange between capillary blood and alveolar gas (V/Q = 0), and is equivalent to an anatomic shunt between the right and left sides of the heart. Venous admixture represents the capillary flow that does not equilibrate completely with alveolar gas (0 <V/Q <1). As the venous admixture increases, the V/Q ratio decreases until it becomes a true shunt (V/Q = 0).

The fraction of the cardiac output that represents intrapulmonary shunt is known as the shunt fraction. In normal subjects, intrapulmonary shunt flow (Qs) represents less than 10% of the total cardiac output (Qt), so the shunt fraction (Qs/Qt) is less than 10%.

Intrapulmonary shunt fraction is increased in the following situations:

1. When the small airways are occluded; e.g., asthma

2. When the alveoli are filled with fluid; e.g., pulmonary edema, pneumonia

3. When the alveoli collapse; e.g., atelectasis

4. When capillary flow is excessive; e.g., in nonembolized regions of the lung in pulmonary embolism

The influence of shunt fraction on aterial blood gas includes: The PaO2 falls progressively as shunt fraction increases, but the PaCO2 remains constant until the shunt fraction exceeds 50% (4). The PaCO2 is often below normal in patients with increased intrapulmonary shunt as a result of hyperventilation triggered by the disease process or by the accompanying hypoxemia.

The shunt fraction also determines the influence of inhaled oxygen on the arterial PO2. As intrapulmonary shunt increases from 10% to 50%, an increase in fractional concentration of inspired oxygen (FiO2) produces less of an increment in the arterial PO2. When the shunt fraction exceeds 50%, the arterial PO2 is independent of changes in FiO2, and the condition behaves like a true (anatomic) shunt. This means that, in conditions associated with a high shunt fraction (e.g., acute respiratory distress syndrome), the FiO2 can often be lowered to non-toxic levels (FiO2 below 60%) without further compromising arterial oxygenation. This can be valuable maneuver for preventing pulmonary oxygen toxicity.

Measures of Gas Exchange


The calculation of dead space ventilation (VD/VT) is based on the difference between the PCO2 in exhaled gas and end-capillary (arterial) blood. In the normal lung, the capillary blood equilibrates fully with alveolar gas, and the exhaled PCO2 (PECO2) is equivalent to the arterial PCO2 (PaCO2). As dead space ventilation (VD/VT) increases, the PECO2 decreases relative to the PaCO2.

VD/VT =  (Paco2 – PEco2)/Paco2

Thus, when the PECO2 decreases relative to the PaCO2, the calculated VD/VT rises. The PECO2 is measured in a random sample of expired gas (mean exhaled PCO2), and is not measured at the end of expiration (end-tidal PCO2).

Intrapulmonary Shunt Fraction

The intrapulmonary shunt fraction (Qs/Qt) is derived by the relationship between the O2 content in arterial blood (CaO2), mixed venous blood (CvO2), and pulmonary capillary blood (CcO2).

Qs/Qt = (CcCO2 – CaCO2) / (CcCO2 – CvCO2)

The problem with this formula is the inability to measure the pulmonary capillary O2 content (CcO2) directly. As a result, pure oxygen breathing (to produce 100% oxyhemoglobin saturation in pulmonary capillary blood) is recommended for the shunt calculation. However, in this situation, Qs/Qt measures only true shunt.

The A-a PO2 Gradient

The PO2 difference between alveolar gas and arterial blood (PAO2 – PaO2) is an indirect measure of ventilation-perfusion abnormalities. The PAO2 – PaO2 (A-a PO2) gradient is determined with the alveolar gas equation shown below.

PAO2 = PiO2 – (PaCO2/RQ)

This equation defines the relationship between the PO2 in alveolar gas (PAO2), the PO2 in inhaled gas (PiO2), the PCO2 in arterial blood (PaCO2), and the respiratory quotient (RQ). The RQ defines the relative rates of exchange of O2 and CO2 across the alveolar-capillary interface: i.e., RQ = VCO2/VO2. The PiO2 is determined using the fractional concentration of inspired oxygen (FiO2), the barometric pressure (PB), and the partial pressure of water vapor (PH2O) in humidified gas:

PiO2 = FiO2 x (PB – PH2O)

So if we combine the above two equations, the A-a PO2 gradient can be calculated as follows:

A-a PO2 = FiO2 x (PB – PH2O) – (PaCO2/RQ) – PaO2

In a healthy subject breathing room air at sea level, FiO2 = 0.21, PB = 760 mm Hg, PH2O = 47 mm Hg, PaO2 = 90 mm Hg, PaCO2 = 40 mm Hg, and RQ =0.8:

A-a PO2 = 0.21 x (760 – 47) – 40 / 0.8 – 90 = 10 mm Hg [Equation 20.6]

screen-shot-2016-09-08-at-9-57-59-pmThis represents an idealized rather than normal A-a PO2 gradient, because the A-a PO2 gradient varies with age and with the concentration of inspired oxygen.

As shown in Table 20.1, the normal A-a PO2 gradient rises steadily with advacing age. Assuming that most adult patients in an ICU are 40 years of age or older, the normal A-a PO2 gradient in an adult ICU patient can be as high as 25 mm Hg when the patient is breathing room air. However, few ICU patients breathe room air, and the A-a PO2 gradient is increased further when oxygen is added to inhaled gas.

screen-shot-2016-09-09-at-12-11-52-pmThe influence of inspired oxygen on the A-a PO2 gradient is shown in Figure 20.4. The A-a PO2 gradient increases from 15 to 60 mm Hg as the FiO2 increases from 21% (room air) to 100%. According to this relationship, the normal A-a PO2 gradient increase 5 to 7 mm Hg for every 10% increase in FiO2.This effect is presumably caused by the loss of regional hypoxic vasoconstriction in the lungs. Hypoxic vasoconstriction in poorly ventilated lung regions diverts blood to more adequtely ventilated regions, and this helps to preserve the normal V/Q balance. Loss of regional hypoxic vasoconstriction during supplemental O2 breathing maintains blood flow in poorly ventilated lung regions, and this increases intrapulmonary shunt fraction and increases the A-a PO2 gradient.

Positive-pressure mechanical ventilation elevates the pressure in the airways above the ambient barometric pressure. Therefore, when determining the A-a PO2 gradient in a ventilator-dependent patient, the mean airway pressure should be added to the barometric pressure. In the example presented in equation 20.6, a mean airway pressure of 30 cm H2O would increase the A-a PO2 gradient from 10 to 16 mm Hg (a 60% increase). Thus, neglecting the contribution of positive airway pressure during mechanical ventilation will underestimate the degree of abnormal gas exchange.

The a/A PO2 Ratio

Unlike the A-a PO2 gradient, the a/A PO2 ratio is relatively unaffected by the FiO2. This is demonstrated in Figure 20.4. The independence of the a/A PO2 ratio in relation to the FiO2 is explained by the equation below.

a/A PO2 = 1 – (A-a PO2) / PAO[Equation 20.7]

Because the alveolar PO2 is in both the numerator and denominator of the equation, the influence of FiO2 on the PAO2 is eliminated. Thus, the a/A PO2 ratio is a mathematical manipulation that eliminates the influence of FiO2 on the A-a PO2 gradient. The normal a/A PO2 ratio is 0.74 to 0.77 when breathing room air, and 0.80 to 0.82 when breathing 100% oxygen.

The PaO2/FiO2 Ratio

screen-shot-2016-09-09-at-1-17-24-pmThe PaO2/FiO2 ratio is used as an indirect estimate of shunt fraction. The left correlations have been reported. The major limitation of the PaO2/FiO2 is the inability to estimate the FiO2 accurately when supplemental O2 is delivered through nasal prongs or "open" face masks.

The Diagnostic Evaluation of Hypoxemia

Hypoxemia can be defined as an arterial PO2 below what is expected for a patient's age, as defined in Table 20.1. However, hypoxemia usually doesn't raise red flags until the arterial PO2 falls below 60 mm Hg (or the arterial O2 saturation falls below 90%). The causes of hypoxemia can be separated into 3 categories based on the physiological process involved, including: 1) Hypoventilation; 2) V/Q mismatch; and 3) DO2/VO2 imbalance.


Alveolar hypoventilation causes both hypoxemia and hypercapnia, similar to breath-holding. There is no V/Q imbalance in the lungs, so the A-a PO2 gradient is not elevated (both the PAO2 and PaO2 decrease in parallel). Common causes of alveolar hypoventilation in ICU are listed in Table 20.4.screen-shot-2016-10-10-at-8-36-23-pm

Most cases of respiratory muscle weakness in the ICU are the result of an idiopathic polyneuropathy and myopathy that is specific to ICU patients, particualrly those with sepsis, prolonged mechnaical ventilation, and prolonged neuromuscular paralysis. The standard method of evaluating respiratory muscle strength is to measure the maximum inspiratory pressure (PImax), which is the maximum pressure recorded during a maximum inspiratory effort against a closed valve. The normal PImax varies with age and gender, but most healthy adults can generate a negative PImax of at least 80 cm H2O. A PImax that does not exceed -25 cm H2O is considered evidence of respiratory muscle failure.

V/Q Mismatch

Most cases of hypoxemia are the result of a V/Q mismatch in the lungs. Virtually any lung disease can be included in this category, but the common ones encountered in the ICU are pneumonia, inflammatory lung injury (acute respiratory distress syndrome), obstructive lung disease, hydrostatic pulmonary edema, and pulmonary embolism. The A-a PO2 gradient is almost always elevated in these conditions, but the elevation can be minimal in patients with severe airways obstruction (which behaves like hypoventilation).

DO2/VO2 Imbalance

A decrease in systemic O2 delivery (DO2) is usually accompanied by an increase in O2 extraction from capillary blood, and this serves to maintain a constant rate of O2 uptake (VO2) into the tissues. The increased O2 extraction from capillary blood results in a decrease in the PO2 of venous blood, and this can have an deleterious effect on arterial oxygenation, as explained below.

The O2 in arterial blood represents the sum of the O2 in mixed venous blood and the O2 added from alveolar gas. When gas exchange is normal, the PO2 in alveolar gas is the major determinant of the arterial PO2. However, when gas exchange is impaired (that is the V/Q mismatch), the contribution of the alveolar PO2 declines and the contribution of the mixed venous PO2 rises. The greater the impairment in gas exchange, the greater the contribution of the mixed venous PO2 to the arterial PO2 (if there is no gas exchange in the lungs, the mixed venous PO2 would be the sole determinant of the arterial PO2). The relationship between O2 delivery (DO2), O2 uptake (VO2), and the mixed venous PO2 (PvO2) can be stated as follows:

PvO2 = k x (DO2/VO2)

k is a proportionality constant. Thus, any condition that reduces DO2 (e.g., low cardiac output, anemia) or inceases VO2 (e.g., hypermetabolism) can decrease the PvO2 and aggravate the hypoxemia caused by abnormal gas exchange in the lungs.

Spurious Hypoxemia

Spurious hypoxemia is a rarely reported phenomenon that is characterized by hypoxemia in an arterial blood sample without corresponding hypoxemia in circulating blood (as measured by pulse oximetry). This phenomenon seems to occur only in patients wtih hematologic malignancies who have marked leukocytosis (WBC >100,000) or thromboctosis (platelet count >1,000,000). The reduced PO2 in the blood sample has been attributed to O2 consumption by activated leukocytes in the sample, a phenomenon that has been called leukocyte larceny. This does not explain why marked thrombocytosis can also produce spurious hypoxemia because platelet are not oxygen-guzzlers like activated leukocytes. Regardless of the mechanism, there is no accepted method of preventing spurious hypoxemia, so you should be aware of the phenomenon and the value of pulse oximetry for validating in vitro PO2 measurements.

Let's Make the Diagnostic Evaluation

The evaluation of hypoxemia can proceed according to the flow diagram in Figure 20.6. This approach uses three measures: A-a PO2 gradient, mixed venous PO2 (PvO2), and maximum inspiratory pressure. The PO2 in superior vena cava blood (central venous PO2) can be used as the mixed venous PO2 when there is no indwelling pulmonary artery catheter.

screen-shot-2016-10-11-at-9-53-45-amThe first step in the approach involves a determination of the A-a PO2 gradient. After correcting for age and FiO2, the A-a PO2 gradient can be interpreted as follows:

1) Normal A-a PO2 gradient indicates hypoventilation rather than a cardiopulmonary disorder. In this situation, the most likely problems are drug-induced respiratory depression and neuromuscular weakness. The latter condition can be uncovered by measuring the maximum inspiratory pressure (PImax), which is described earlier.

2) Increased A-a PO2 gradient indicates a V/Q abnormality (cardiopulmonary disorder) and a possible superimposed DO2/VO2 imbalance (e.g., a decrease in cardiac output). The mixed venous (or central venous) PO2 will help to identify a DO2/VO2 imbalance. a) If the venous PO2 is 40 mm Hg or higher, the problem is solely a V/Q mismatch in the lungs; b) if the venous PO2 is below 40 mm Hg, there is a DO2/VO2 imbalance adding to the hypoxemia created by a V/Q mismatch in the lungs. The source of this imbalance is either a decreased DO2 (from anemia or a low cardiac output) or an increased VO2 (from hypermetabolism).