Blood pH is measured by exposing the specimen to a glass electrode (see Figure 11-28) or by measuring light absorbance with an optical pH indicator under anaerobic conditions. pH measurements are made at 37°C. The pH of arterial blood is related to the Paco₂ by the

Henderson-Hasselbalch equation:

pH=pK + log [HCO3−][CO2]

where:

pK=negative log of dissociation constant for carbonic acid (6.1)

HCO3−[]=molar concentration of serum bicarbonate

[CO₂]=molar concentration of CO₂

Paco₂ (measured directly by a CO₂ electrode or similar device) may be multiplied by 0.03 (the solubility coefficient for CO₂) to express Paco₂ in milliequivalents per liter (mEq/L). The equation then may be expressed as follows:

pH=pK + log [HCO3−][0.03(PaCO2)]

Specimen Collection for Blood Gases

Arterial samples are usually obtained from either the radial or brachial arteries via puncture or catheter in adults. Arterial specimens may also be drawn from the femoral, dorsalis pedis, posterior tibial, or umbilical arteries. The radial artery of the nondominant hand is usually the preferred site. A warmed capillary sample (arterialized) can also be obtained to approximate an arterial specimen.

PF Tip 6-1

Blood gas analyzers usually report blood gas values at 37°C. If the patient has an extreme body temperature (hypothermia or hyperthermia), correction may be useful. At low body temperatures, more gas dissolves in the blood, so Po₂ and Pco₂ show lower values when exposed to electrodes that measure their activity. At elevated temperatures, the opposite occurs; less gas dissolves and partial pressures appear higher. The pH changes in relation to Pco₂.

Before a radial artery puncture, adequacy of collateral circulation to the hand from the ulnar artery should be established, using the modified Allen’s test (Figure 6-1). The original Allen’s test was described by Dr. Edgar Allen, a vascular surgeon at the Mayo Clinic in the 1920s, to assist in the diagnosis of thromboangiitis obliterans (e.g., Raynaud’s disease). A positive test in his description denoted reduced circulation; however, using the modified technique, a “positive test” demonstrates adequate collateral circulation. In performing the modified Allen’s test, the technologist occludes both the radial and ulnar arteries by pressing down over the wrist. The patient is instructed to make a fist, then open the hand and relax the fingers. The palm of the hand is pale and bloodless because both arteries are occluded. The ulnar artery is released while the radial remains occluded. The hand should be reperfused rapidly (5–15 seconds) if the ulnar supply is adequate. If perfusion is inadequate, an alternative site should be used.

6-1   How To…

Perform a Modified Allen’s Test

1. Tasks common to all procedures.

a. Calibrate and prepare equipment.

b. Review order.

c. Introduce yourself and identify patient according to institutional policies.

d. Describe and demonstrate procedure.

2. Instruct the patient to tightly close his or her hand to form a fist. Pressure is then applied at the wrist, compressing and obstructing both the radial and ulnar arteries.

3. The hand is then opened (but not fully extended), revealing a blanched palm and fingers.

4. Release pressure from the ulnar artery only, while the palm and fingers, including the thumb, are observed.

5. These should become flushed within 15 seconds as the blood from the ulnar artery refills the empty capillary bed.

6. If the ulnar artery does not adequately supply the entire hand (a negative Allen’s test), the radial artery should not be used as a puncture site. An alternate artery should be selected.

Arterial puncture should not be performed through any type of lesion. Similarly, puncture distal to a surgical shunt (e.g., shunts used for dialysis) should be avoided. Infection or evidence of peripheral vascular disease should prompt selection of an alternative site. Some patients may be using anticoagulant drugs such as heparin, warfarin (Coumadin), or streptokinase. High doses of these drugs or a history of prolonged clotting times may be relative contraindications to arterial puncture. Box 6-1 lists some of the potential hazards associated with arterial puncture.

Box 6-1   Complications of Arterial Puncture for Blood Gas Testing

Pain and discomfort

Hematoma

Vasovagal response

Thrombosis and embolism

Infection or contamination

Inadvertent needle stick

Vascular trauma or occlusion

Arterial spasm

6-2   How To…

Perform an Arterial Puncture from the Radial Artery

1. Tasks common to all procedures.

a. Calibrate and prepare equipment.

b. Review order.

c. Introduce yourself and identify patient according to institutional policies.

d. Describe and demonstrate procedure.

2. Prepare necessary equipment: self-filling heparinized syringe, needle, and personal protection equipment (e.g., gloves and eyewear).

3. Assure the pre test conditions of the order are met (e.g. F_IO₂, liter flow, ventilator settings, etc.). If F_IO₂ has been modified wait 20-30 minutes prior to harvesting the sample. In the outpatient setting the patient should be in a stable condition for a minimum of 5 minutes or longer once the above criteria have been met…

4. Assess the collection site for adequate pulse, and perform a modified Allen’s test.

5. Hyperextension of the wrist may facilitate exposure of the artery; palpate for a pulse.

6. Prepare the puncture site; clean with a laboratory approved agent.

7. Make sure to instruct the subject to “breathe normally” and not to breath hold during the procedure.

8. Hold the syringe with the bevel up at a 30- to 45-degree angle, and advance toward the palpated artery until blood flow enters the syringe.

9. After obtaining the desired amount, place a dry gauze sponge over the puncture site and quickly remove the needle, applying pressure immediately and directly over the puncture site.

10. Immediately expel any air bubbles and mix thoroughly by rotating or inverting the sample numerous times.

11. Hold direct pressure to the site (2-5 minutes is typically adequate if the subject is not anticoagulated) until hemostasis occurs.

12. Report results and note comments related to sample quality.

Success in obtaining an arterial specimen by radial puncture requires careful positioning of the patient’s wrist. The hand should be hyperextended with good support under the wrist (Figure 6-2, arterial puncture). A topical anesthetic may be useful for some patients, but is usually unnecessary. Always perform the modified Allen’s test to ensure adequate collateral circulation before puncturing the artery. The person drawing the sample should be in a comfortable position (i.e., sitting) to maximize control of the needle during insertion. It is recommended that the person drawing the sample use standard precautions and personal protective equipment, including gloves and eyewear. A vented syringe or similar device allows the blood to “pulse” into the syringe, ensuring that the needle is in the lumen of the artery.

Drawing a sample from a catheter–infusion system for an arterial or mixed venous sample from a pulmonary artery (Swan-Ganz) catheter has its own nuances. Contamination of the sample with flush solution is a common problem. Withdrawing a small volume of blood into a “waste” syringe ensures that the sample is not diluted by flush solution in the catheter. Care should be taken to limit the volume of blood removed in this process. Significant blood loss can occur with repetitive measurements. Another problem associated specifically with mixed venous specimen collection is displacement of the Swan-Ganz catheter tip. If the catheter is advanced too far, it may “wedge” into a pulmonary arteriole. Specimens drawn from this position often reflect arterialized pulmonary capillary blood. Similarly, if the catheter tip is withdrawn or “loops back,” it may be in the right ventricle or atrium rather than in the pulmonary artery. Specimens obtained from this location may not represent true mixed venous blood.

Venous samples from peripheral veins are not useful for assessing oxygenation. Venous blood reflects only the metabolism of the area drained by that particular vein. Venous samples may be used for measurement of pH or blood lactate during exercise.

Blood is usually collected in a syringe to which an anticoagulant (e.g., heparin) has been added, and sealed from the atmosphere immediately (Criteria for Acceptability 6-1). Blood gases are typically drawn using specially designed syringes containing a dry anticoagulant (i.e., a blood gas “kit”). Dry heparin is applied to the lumen of the needle and the interior of the syringe. A small heparin pellet is often placed in the syringe to provide additional anticoagulation. Care must be taken that heparin solution (sometimes used for flushing arterial catheters) does not contaminate the sample. Contamination by flush solution may affect Po₂ and Pco₂ values by dilution, but the buffering capacity of whole blood prevents large changes in pH.

Criteria For Acceptability 6-1   Blood Gases

1. Blood should be collected anaerobically. Syringe body and plunger should be tight-fitting. Commercially available blood gas kits should be used according to manufacturers’ specifications. Air bubbles should be expelled immediately.

2. The specimen must be adequately anticoagulated; sodium or lithium heparin is preferred. If liquid heparin is used, all excess should be expelled. Choice of anticoagulant should be determined by analyses to be performed (e.g., electrolytes).

3. A sample volume of 2–4 mL is recommended.

4. The specimen should be analyzed as soon as possible. If immediate analysis is not available, the specimen should be drawn in a glass syringe and stored in an ice-water slurry at 0°C and analyzed within 1 hour. Specimens drawn in plastic syringes should not be placed in ice water.

5. The specimen should be adequately identified, including patient name and/or number, date/time, ordering physician, and accession number. Information provided with the specimen should be the site from which it was obtained, F_IO₂ (if applicable), device (e.g., nasal cannula, Oxymizer pendant), continuous flow or pulsed via a conserving device (if applicable) and ventilator or assist device (e.g., BIPAP) settings (if applicable).

6. Analysis should be performed on an instrument that has been recently calibrated and whose function is documented by appropriate controls.

After the syringe has been capped (see Chapter 12 for safe handling of blood specimens), the sample should be thoroughly mixed by rolling or gently shaking. Mixing helps prevent the sample from clotting, whether dry or liquid heparin is used. Mixing is also important for analyzers that use spectrophotometric methods for measuring blood gases. Lithium heparin or a similar preparation should be used for specimens that will also be used for electrolyte analysis.

Air contamination of arterial or mixed venous blood specimens can seriously alter blood gas values. Room air at sea level has a Po₂ of approximately 150 mm Hg, and a Pco₂ near 0. If air bubbles are present in a blood gas specimen, equilibration of gases between the sample and air occurs (Table 6-2). Contamination commonly occurs during sampling when air is left in the syringe after the sample is collected. Small bubbles may also be introduced if the needle does not connect tightly to the syringe. Other sources of air contamination include poorly fitting plungers and failure to properly cap the syringe. Use of a vented syringe or one in which the pulse pressure of the blood displaces the plunger can help prevent air bubble contamination.

Table 6-2

Air Contamination of Blood Gas Samples

	In Vivo Values	Air Contamination*
pH	7.40	7.45
Pco2	40	30
Po2	95	110

^*Typical values that might occur when a blood gas specimen is exposed to air, either directly or by mixing with a solution that has been exposed to air (e.g., heparinized flush solution). The change in pH occurs because of the change in Pco₂.

Sample storage depends on the type of syringe used. If a glass syringe is used, the sample may be stored in ice-water slush if analysis is not done within a few minutes. Ice water reduces the metabolism of red and white blood cells in the sample. Specimens with O₂ tensions in the normal physiologic range (50–150 mm Hg) show minimal changes over 1–2 hours for glass syringes if kept in ice water. Changes in specimens held at room temperature are related to cellular metabolism in the blood, particularly in white blood cells and platelets. Specimens with Po₂ values above 150 mm Hg are most susceptible to alterations resulting from gas leakage or cell metabolism. When the Po₂ is 150 mm Hg or more, Hb is almost completely bound with O₂. In such cases, a small change in O₂ content results in a large change in Po₂.

If a plastic syringe is used, blood gas specimens should be analyzed within 30 minutes. Plastic syringes are not completely gas tight, so room air can contaminate the sample. The influx of O₂ may be counterbalanced by the consumption of O₂ if the sample is not iced. When a blood gas specimen is placed in an ice-water bath, the solubility of O₂ increases, as does the affinity of hemoglobin for oxygen. This lowers the partial pressure of O₂ in the sample and increases the gradient between the sample and the environment. This gradient exaggerates the leakage of O₂ into the specimen (as occurs with plastic syringes). When the sample is introduced into the analyzer at 37°C, solubility and Hb affinity return to their normal values, the oxygen that leaked in is released, and the Po₂ is falsely increased. Because of these phenomena, blood gas specimens in plastic syringes should not be placed in ice water. If specimens cannot be routinely analyzed within 30 minutes, glass syringes with ice-water storage may be preferable.

PF Tip 6-2

Blood gas specimens should be analyzed as soon as possible. Specimens drawn in plastic syringes should NOT be placed in ice water and should be analyzed within 30 minutes. Specimens that cannot be analyzed within 30 minutes should be drawn in a glass syringe and placed in an ice-water slush to retard cellular metabolism. Blood gas specimens drawn for shunt studies and specimens with elevated white blood cell counts should be analyzed immediately to minimize changes in Pao₂.

Capillary samples are useful in infants when arterial puncture is impractical. The area for collection (the heel is commonly chosen) should be heated by a warm compress and lanced. Blood is then allowed to fill the required volume of heparinized glass/plastic capillary tubes. Squeezing the tissue should be avoided because predominantly venous blood will be obtained. The capillary tubes should be carefully sealed to avoid air bubbles. Guidelines for quality control of blood gas analyzers and for the safe handling of blood specimens are included in Chapter 12.

Significance and Pathophysiology

See Interpretive Strategies 6-1.

pH

The pH of arterial blood in healthy adults averages 7.40 with a range of 7.35–7.45. Arterial pH below 7.35 constitutes acidemia. A pH above 7.45 constitutes alkalemia. Because of the logarithmic scale, a change of 0.3 pH units represents a twofold change in [H⁺] concentration. If the pH decreases from 7.40 to 7.10 with no change in Pco₂, the concentration of hydrogen ions has doubled. Conversely, if the concentration of [H⁺] is halved, the pH increases from 7.40 to 7.70, assuming the Pco₂ remains at 40 mm Hg. Changes of this magnitude represent marked abnormalities in the acid-base status of the blood and are almost always accompanied by clinical symptoms (e.g., cardiac arrhythmias) (Figure 6-3). If a metabolic acidemia is present, examination of the Anion Gap (AG) needs to be assessed. The anion gap is the difference between routinely measured cations (sodium) and anions (chloride and bicarbonate) in serum, plasma, or urine. The magnitude of the difference in serum is calculated to help identify the cause of a metabolic acidemia. The anion gap also reflects unmeasured cations (e.g., proteins, organic acids, phosphates) and anions (e.g., calcium, potassium, magnesium) (Table 6-3). The AG is expressed in units of mEg/L or mmol/L and is calculated by subtracting the serum concentrations of chloride and bicarbonate from the concentrations of sodium and potassium. However, in routine clinical practice, the potassium is frequently ignored, which leaves the following equation:

Table 6-3

Anion Gap

Metabolic Acidosis
Wide Gap	Normal Gap
“MUDPILES”	“HARDUP”
Methanol (Osm gap high)	Hyperventilation
Uremia	Acid infusion/carbonic anhydrase inhibitors/Addison’s disease
Diabetic/alcoholic ketoacidosis	Renal tubular acidosis (urine gap positive)
Paraldehyde	Diarrhea (urine gap normal; i.e., negative)
Isoniazid/Iron	Ureterosigmoidostomy
Lactic acidosis	Pancreatic fistula/drainage
Ethylene glycol (Osm gap high)
Salicylates/solvents

AG=[Na+]−[Cl-+HCO3−]AG=140−[105+24]AG=11 mEq/L

The normal range is 3 − 11 mEq/L; however, the reference range should be provided by the lab performing the tests.

Acid-base disorders arising from lung disease are often related to Pco₂ and its transport as carbonic acid. If the pH is outside of its normal range (i.e., acidemia or alkalemia) but Pco₂ is within normal limits, the condition is termed nonrespiratory or metabolic (Table 6-4). The calculated HCO3− is a useful indicator of the relationship between pH and Pco₂. In the presence of acidemia (pH less than 7.35) and normal CO₂ (Pco₂ 35–45 mm Hg), HCO3− will be low and a metabolic acidosis is present. A Pco₂ of less than 35 mm Hg in the presence of acidosis suggests that ventilatory compensation for acidemia is occurring. The acid-base status would be considered partially compensated nonrespiratory (i.e., metabolic) acidosis. Complete compensation occurs if pH returns to the normal range. This happens when ventilation reduces Pco₂ in proportion to the HCO3−.

Table 6-4

Acid-Base Disorders

Status	pH	Pco2	HCO3−
Simple Disorders
Metabolic acidosis	Low	Normal	Low
Metabolic alkalosis	High	Normal	High
Respiratory acidosis	Low	High	Normal
Respiratory alkalosis	High	Low	Normal
Compensated Disorders
Compensated respiratory acidosis, or metabolic alkalosis	Normal*	High	High
Compensated metabolic acidosis, or respiratory alkalosis	Normal*	Low	Low
Combined Disorders
Metabolic/respiratory acidosis	Low	High	Low
Metabolic/respiratory alkalosis	High	Low	High

^*Compensation cannot return values to within normal limits in severe acid-base disturbances. In addition, a normal pH may result in instances of respiratory and metabolic disturbances that occur together but are not compensatory.

In the presence of alkalemia (pH above 7.45) and normal Pco₂ (35–45 mm Hg), calculated bicarbonate will be increased and metabolic alkalosis is present. If ventilatory compensation occurs, Pco₂ will be slightly increased. However, decreased ventilation is required so that the CO₂ can increase. Reduced minute ventilation may interfere with oxygenation. For this reason, Paco₂ seldom increases above 50 mm Hg to compensate for nonrespiratory (i.e., metabolic) alkalosis. Compensation may be incomplete if the alkalosis is severe.

Abnormal Pco₂ and HCO3− characterize combined respiratory and nonrespiratory acid-base disorders. In combined acidosis, Pco₂ is elevated (more than 45 mm Hg) and HCO3−is low (less than 22 mEq/L). In combined alkalosis, HCO3−is elevated (more than 26 mEq/L) and Pco₂ is low (less than 35 mm Hg). The pH is more severely deranged (i.e., much higher or lower) than if just one disorder were present.

Carbon Dioxide Tension

The arterial carbon dioxide tension (Paco₂) in a healthy adult is approximately 40 mm Hg; it may range from 35–45 mm Hg. The Pco₂ of venous or mixed venous blood is seldom used clinically. Body temperature affects Paco₂, as described in Table 6-1.

Paco₂ is inversely proportional to alveolar ventilation (_A) (see Chapter 5). When _A decreases, CO₂ may not be removed by the lungs as fast as it is produced. This causes Paco₂ to increase. The pH falls (i.e., [H⁺] increases) as CO₂ is hydrated to form carbonic acid:

CO2+ H2O↔H2CO3↔H++ HCO3−

Increasing levels of CO₂ in the blood drive this reaction to the right. The patient has respiratory acidosis resulting from hypoventilation. Conversely, when alveolar ventilation removes CO₂ more rapidly than it is produced, Paco₂ decreases. The pH increases (i.e., [H⁺] falls) as the patient becomes alkalotic. This condition is called hyperventilation, or respiratory alkalosis.

If dead space increases, high-minute ventilation (_E) may be required to adequately ventilate alveoli and keep Paco₂ within normal limits. Respiratory dead space occurs because some lung units are ventilated but not perfused by pulmonary capillary blood. Pulmonary embolization is an example of dead space–producing condition. Emboli may block pulmonary arterioles, causing ventilation of the affected lung units to be “wasted.” To maintain normal Paco₂, total ventilation must be increased to compensate for wasted ventilation. Paco₂ may be normal, or decreased, even though significant pulmonary disease is present.

Patients who have disorders such as lobar pneumonia may increase their _E to provide more alveolar ventilation of functional lung units. This mechanism compensates for lung units that do not participate in gas exchange. Hypoxemia is a common cause of hyperventilation (i.e., respiratory alkalosis). Hyperventilation may be seen in patients with asthma, emphysema, bronchitis, or foreign body obstruction. Anxiety, stress or central nervous system disorders may also cause hyperventilation. Hyperventilation may also be an intentional or unintentional result of mechanical ventilation.

Increased Paco₂ (i.e., hypercapnia) commonly occurs in patients who have advanced obstructive or restrictive disease. These individuals are characterized by markedly abnormal ventilation-perfusion (/) patterns. They are unable to maintain adequate alveolar ventilation. Not all patients with advanced pulmonary disease retain CO₂. Those who do become hypercapnic often have a low ventilatory response to CO₂ (see Chapter 5). Their response to the increased work of breathing caused by obstruction or restriction is to allow CO₂ to increase rather than increase ventilation. When respiratory acidosis results from increased Pco₂, renal compensation usually occurs (see Table 6-4).

Increased Paco₂ may also be seen in patients who hypoventilate as a result of central nervous system or neuromuscular disorders. Whether CO₂ retention is the result of lung disease, central nervous system dysfunction, or neuromuscular disease, pH is usually maintained close to normal. The kidneys retain and produce bicarbonate (HCO3−) to match the increased Paco₂. This response may completely compensate for a mildly elevated Paco₂. However, it can seldom produce normal pH when the Paco₂ is greater than 65 mm Hg. If the disorder causing the increased Paco₂ is acute (e.g., foreign body aspiration), little or no renal compensation may be observed.

Some degree of hypoxemia is always present in patients who retain CO₂ while breathing air. As alveolar CO₂ increases, alveolar O₂ decreases. If the cause of hypercapnia is either obstructive or restrictive lung disease, hypoxemia may be severe because of /abnormalities. Because O₂ therapy is common in these patients, changes in Pco₂ while breathing supplementary O₂ must be carefully monitored. Some patients with chronic hypoxemia have a decreased ventilatory response to CO₂. O₂ administered to these patients may reduce their hypoxic stimulus to ventilation. As a result, Paco₂ may increase further. O₂ therapy is usually titrated to maintain Paco₂ values less than 60 mm Hg without hypercapnia and acidosis.

Oxygen Tension

The Pao₂ of healthy young adults at sea level ranges from 85–100 mm Hg and decreases slightly with age. Breathing room air at sea level results in an inspired Po₂ of approximately 150 mm Hg:

PIO2=FIO2(PB−47) =0.21(760−47) =0.21(713) =149.7

where:

F_IO₂= fractional concentration of inspired O₂

P_B= barometric pressure

47= partial pressure of water vapor at 37°C

The partial pressure of O₂ in alveolar gas is usually close to 100 mm Hg and can be calculated using the alveolar air equation:

PAO2=(FIO2×  [PB−47])−PaCO2 (FIO2+1−FIO2R)

where:

Paco₂= arterial CO₂ tension (presumed equal to alveolar CO₂ tension)

R= respiratory exchange ratio (co₂/o₂)

Substituting 40 mm Hg for Paco₂, and 0.8 for R:

PAO2=(0.21  ×   [760−47])−40  (0.21+1−0.210.8)

=(0.21  ×  713)−40(1.1975)

=149.7−47.9

=101.8

In healthy lungs with good gas exchange, arterial oxygen tension can approach the value of the alveolar Po₂. The difference between the alveolar and arterial oxygen tensions is described as the alveolar-arterial gradient, or A-a gradient. This gradient is usually less than 20 mm Hg in healthy individuals breathing air at sea level.

Hyperventilation may increase Pao₂ as high as 120 mm Hg in a patient with normal lung function (see the previous alveolar air equation). Healthy subjects breathing 100% O₂ may exhibit Pao₂ values higher than 600 mm Hg. The alveolar Po₂ (PAo₂) for a particular inspired O₂ fraction can be calculated, as previously described. Decreased Pao₂ can result from hypoventilation, diffusion abnormalities, /imbalances, shunt, and inadequate atmospheric O₂ (high altitude) (see Table 6-4).

PF Tip 6-3

In patients breathing air at sea level, the sum of Pao₂ and Paco₂ is usually less than 150 mm Hg. If the sum is greater than 150 mm Hg, the patient was breathing supplementary O₂, or the blood gas values are erroneous. For patients breathing air at sea level, a quick estimate of alveolar Po₂ can be derived from the equation, 150− (Paco₂/0.8) or 150−(Paco₂*1.25).

Table 6-1 lists changes that occur in Pao₂ because of body temperatures above and below normal (37°C). Changes in Po₂ and Pco₂ reflect solubility of the gas. The partial pressure of each gas is a measure of its activity. Hypothermia (low body temperature) is accompanied by decreased partial pressure as more gas dissolves. Hyperthermia (elevated body temperature) causes elevated gas tensions as gas comes out of solution. All blood gas analyzers perform analyses at 37°C, and most allow temperature corrections to be made. Although blood gas tensions vary with temperature, the clinical significance of correcting measurements is unclear. Blood gas values should be reported at 37°C. Care must be taken to ensure that blood gas analyzers are maintained at 37°C. Measurements made at other temperatures can significantly alter results.

Po₂ is the pressure of O₂ dissolved in blood. The amount of Hb and whether it is capable of binding O₂ has only a minimal effect on Po₂. Hypoxemia (decreased O₂ content of the blood) may occur even if Pao₂ is normal or elevated by breathing O₂. Hypoxemia commonly results from inadequate or abnormal Hb. Many automated blood gas analyzers calculate oxygen saturation (SaO₂). Saturation may be calculated from the Pao₂ and pH, assuming a normal oxygen-hemoglobin reaction occurs (Figure 6-4). Calculated SaO₂ may differ significantly from true saturation measured by a spectrophotometer (see the section on oxygen saturation) because a number of factors can affect the relationship of the binding capacity of hemoglobin (see Figure 6-4)

An example of the discrepancy between calculated and measured saturation is the patient with elevated carboxyhemoglobin (COHb), resulting from smoking or smoke inhalation. The patient’s Pao₂ may be normal while O₂ saturation is markedly decreased. Calculating saturation from Po₂, in this case, overestimates the O₂ content of the blood. Measured SaO₂ is preferred to a calculated value.

The severity of impaired oxygenation is indicated by the Pao₂ at rest. Pao₂ is a good index of the lungs’ ability to match pulmonary capillary blood flow with adequate ventilation. If ventilation matches perfusion, pulmonary capillary blood leaves the lungs with a Po₂ close to that of the alveoli. If ventilation is adequate, pulmonary capillary blood is almost completely saturated. If either of these conditions is not met (i.e., poor ventilation or /mismatching), pulmonary capillary blood has reduced O₂ content. Pao₂ is reduced in proportion to the number of lung units contributing blood with low O₂ content. Lung units with good /cannot compensate for their poorly functioning counterparts because pulmonary capillary blood leaving them is already almost fully oxygenated. O₂ binding to Hb is almost complete when the Pao₂ is greater than 60 mm Hg (approximately 90% saturation). As Pao₂ decreases from 60 to 40 mm Hg, saturation decreases from 90% to 75%, with increasing symptoms of hypoxia (e.g., mental confusion, shortness of breath).

Delivery of O₂ to the tissues, however, depends on Hb concentration and cardiac output as well as adequate gas transfer in the lungs. Arterial oxygen content (Cao₂ in mL/dL) is defined as follows:

CaO2=(1.34  ×  Hb  ×  SaO2)  +  (PaO2×  0.0031)

where:

1.34= O₂ binding capacity of Hb, mL/g

Hb= hemoglobin concentration, g/dL

SaO₂= arterial oxygen saturation as a fraction

Pao₂= arterial oxygen tension, mm Hg

0.0031= solubility coefficient for O₂ at 37°C

Because most O₂ transported is bound to Hb, there must be an adequate supply (12–15 g/dL) of functional Hb. Adequate cardiac output (4–5 L/min) is necessary to deliver the oxygenated arterial blood to the tissues. Signs and symptoms of hypoxia may be present despite adequate Pao₂ because of severe anemia and/or reduced cardiac function.

The mixed venous oxygen tension (Pv¯o₂) in healthy patients at rest ranges from approximately 37–43, with an average of 40 mm Hg (see Figure 6-4). Mixed venous O₂ content (Cv¯o₂) averages 15 mL/dL. In healthy individuals, arterial oxygen content (Cao₂) averages 20 mL/dL. The resulting content difference, or C(a-v¯)o₂, is thus 5 mL/dL (or 5 vol%). Although Pao₂ varies with the inspired O₂ fraction and matching of /, Pv¯o₂ changes in response to alterations in cardiac output and O₂ consumption. If cardiac output increases while oxygen consumption (o₂) remains constant, the C(a-v¯)o₂ decreases. Conversely, if cardiac output decreases with no change in O₂ consumption, C(a-v¯)o₂ increases. Increased cardiac output sometimes occurs in response to pulmonary shunting. This allows mixed venous oxygen content to increase, reducing the deleterious effect of the shunt. Critically ill patients often have low Pv¯o₂ values and increased C(a-v¯)o₂ as a result of poor cardiovascular performance. Alterations in Pv¯o₂ often occur even if Pao₂ is normal. Pv¯o₂ values less than 28 mm Hg in critically ill patients, accompanied by C(a-v¯)o₂ greater than 6 vol%, suggest marked cardiovascular decompensation.

Patients who have severe obstructive or restrictive diseases may have decreased Pao₂ at rest, occasionally as low as 40 mm Hg. Mild pulmonary disease may show little decrease in Pao₂ if hyperventilation is present. Pao₂ may also be relatively normal if the disease process affects ventilation and perfusion similarly. In patients with emphysema, destruction of alveolar septa may eliminate pulmonary capillaries as well, resulting in poor ventilation and equally poor blood flow. These patients may have severe airway obstruction, markedly reduced Dlco, but only a small decrease in resting Pao₂. Patients with chronic bronchitis or asthma, particularly during acute exacerbations, may have moderate or severe resting hypoxemia because of /abnormalities.

During exercise in patients with obstructive disease, Pao₂ often decreases commensurate with the extent of the disease. Pao₂ during exercise is correlated with the patient’s Dlco and FEV₁, but wide variability exists. Patients with markedly decreased Dlco may show low Pao₂ values at rest that decrease during exercise. Studies have demonstrated sensitivity and specificity in the range of 75%–90% of exercise desaturation if the Dlco is 50%–60% percent of predicted. The degree of arterial desaturation cannot be predicted from static pulmonary function measurements.

Pao₂ may be decreased for nonpulmonary reasons such as anatomic shunts (intracardiac) or hypoventilation because of neuromuscular disease. Tissue hypoxia can occur because of inadequate or nonfunctional Hb, or because of poor cardiac output. Pao₂ should be correlated with spirometry (FEV₁), Dlco, ventilation (_E, tidal volume, dead space), and lung volumes (VC, RV, TLC) to distinguish pulmonary from nonpulmonary causes of inadequate oxygenation.

Technique

O₂ saturation of Hb (O₂Hb, or SaO₂ when referring to arterial saturation) may be measured in several ways. In the first method, O₂ saturation is measured using a spectrophotometer (see Chapter 11). Whole blood is hemolyzed and the absorbances of the various components measured. The total Hb, O₂Hb, COHb, and MetHb are usually reported. Hemoximetry is usually performed in conjunction with blood gas analysis, using a single specimen of blood.

An indirect method may be used to measure mixed venous oxygen saturation (Sv¯o₂). Saturation may be measured by a reflective spectrophotometer in a pulmonary artery catheter (Swan-Ganz catheter). In this case, absorbances are measured without hemolyzing the blood. A special catheter that includes fiberoptic bundles is used to perform in vivo measurements.

Blood specimens for hemoximetry should be prepared, as described for arterial blood gas specimens. Guidelines for quality control of blood gas analysis are included in Chapter 12. Measurement of percent saturation allows the calculation of the O₂ content of either arterial or mixed venous blood (Cao₂ and Cv¯o₂, respectively) (see section, Oxygen Tension).

Significance and Pathophysiology

See Interpretive Strategies 6-2. SaO₂ for a healthy young adult with a Pao₂ of 95 mm Hg is approximately 97%. The O₂Hb dissociation curve is relatively flat when the Pao₂ is above 60 mm Hg (i.e., SaO₂ is 90% or more). Saturation changes only slightly when there is a marked change in Pao₂ at partial pressures above 60 mm Hg (see Figure 6-4). Therefore, Pao₂ is a more sensitive indicator of oxygenation in lungs that do not have gross abnormalities. At Pao₂ values of approximately 150 mm Hg, Hb becomes completely saturated (SaO₂ is 100%). At Pao₂ values above 150 mm Hg, further increases in O₂ content are caused by increased dissolved oxygen. Alterations in /patterns in the lungs can be monitored by allowing the patient to breathe 100% O₂ and by measuring the changes in dissolved oxygen. In practice, this is accomplished by using the clinical shunt equation (see section, Shunt Calculation).

When Pao₂ falls below 60 mm Hg, SaO₂ decreases rapidly. Small changes in Pao₂ result in large changes in saturation. As SaO₂ falls below 90%, O₂ content decreases rapidly. At saturations less than 85% (i.e., Pao₂ <55 mm Hg), symptoms of hypoxemia increase and supplementary O₂ may be indicated.

The ability of Hb to bind O₂ may be measured by the P₅₀. P₅₀ specifies the partial pressure at which Hb is 50% saturated (see Figure 6-4). The P₅₀ of normal adult Hb is approximately 27 mm Hg. P₅₀ may be determined by equilibrating blood with several gases at low oxygen tensions. An O₂Hb dissociation curve is then constructed to estimate the partial pressure at which Hb is 50% saturated. A second method of estimating P₅₀ compares measured SaO₂ (using a spectrophotometer) with the expected saturation. Calculated saturations presume a P₅₀ of 26–27 mm Hg, but it may differ significantly, depending on the types of Hb and interfering substances present.

Healthy individuals have small amounts of Hb that cannot carry O₂. COHb is present in blood from metabolism and from environmental exposure to carbon monoxide (CO) gas. Normal COHb, expressed as a percentage, ranges from 0.5%–2% of the total Hb. CO comes from smoking (cigarettes, cigars, and pipes), smoke inhalation, improperly vented furnaces, automobile emissions, and other sources of combustion. In smokers, levels may increase from 3%–15%, depending on recent smoking history. Smoke inhalation or CO poisoning from other sources also results in elevated COHb levels, sometimes as high as 50%. CO combines rapidly with Hb. Exposures of short duration can cause a high level of COHb if high concentrations of CO are present. Because O₂Hb saturation decreases as COHb increases, COHb levels greater than 15% almost always result in hypoxemia. High levels of COHb are rapidly fatal because of the profound hypoxemia that occurs.

COHb absorbs light at wavelengths similar to O₂Hb. When COHb is elevated, arterial blood appears bright red. Cyanosis, which appears when there is an increased concentration of reduced Hb, is absent. In spite of elevated levels of COHb, Pao₂ may be close to normal limits. Blood gas analysis that includes calculated saturation will give erroneously high O₂ saturations. For this reason, O₂Hb and COHb should be measured by hemoximetry whenever possible.

COHb interferes with O₂ transport in two ways: it binds competitively to Hb, and it shifts the O₂Hb curve to the left (see Figure 6-4). Increased COHb causes reduced O₂Hb with a decrease in O₂ content. The left shift of the dissociation curve causes O₂ to be bound more tightly to Hb. The combination of these two effects can seriously reduce O₂ delivery to the tissues. COHb concentrations in blood begin to decrease once the source of CO has been removed. Removal of CO from the blood depends on the minute ventilation. Breathing air may require several hours to reduce even moderate levels to normal. Breathing 100% O₂ speeds the washout of CO. High concentrations of O₂ (often including hyperbaric therapy) are indicated whenever dangerously high levels of COHb are encountered.

Methemoglobin (MetHb) forms when iron atoms of the Hb molecule are oxidized from Fe⁺⁺ to Fe⁺⁺⁺. The normal MetHb level is less than 1.5% of the total Hb in adults. High levels of MetHb can result from ingestion of or exposure to strong oxidizing agents. Methemoglobinemia has also been linked to high doses of medications such as benzocaine, dapsone, nitroprusside and inhaled nitric oxide. Like COHb, MetHb reduces O₂ carrying capacity of the blood by reducing the available Hb and shifting the O₂Hb dissociation curve to the left (see Figure 6-4).

The saturation of mixed venous blood (Sv¯o₂) in healthy patients averages 75% at a Pv¯o₂ of 40 mm Hg. Healthy patients have a content difference, C(a-v¯)o₂, of 5 vol%. Arterial blood, in adults with normal levels of Hb, typically carries approximately 20 vol% O₂, and mixed venous blood carries 15 vol% O₂. Pulmonary diseases that cause arterial hypoxemia may reduce Sv¯o₂ if oxygen uptake and cardiac output remain constant. Cardiac output often increases to combat arterial hypoxemia caused by intrapulmonary shunting. Increased cardiac output increases O₂ delivery to the tissues. This results in a reduced extraction of O₂ from the blood. Mixed venous blood then returns to the lungs with normal or even increased O₂ saturation. When this blood is shunted, it has a higher O₂ content, thereby reducing the shunt effect. With or without arterial hypoxemia, Sv¯o₂ decreases if cardiac output is compromised.

Sv¯o₂ is useful in assessing cardiac function in the critical care setting and during exercise, but does require the placement of a pulmonary artery catheter. Patients who have good cardiovascular reserves maintain a mixed venous saturation of 70%–75%. Patients whose Sv¯o₂ values are in the 60%–70% range have a limited ability to deliver more O₂ to the tissues. Sv¯o₂ values less than 60% usually indicate cardiovascular decompensation and may be associated with tissue hypoxemia. The indwelling reflective spectrophotometer (see Chapter 11) allows continuous monitoring of this important parameter. Sv¯o₂ also decreases during exercise. Despite increased cardiac output, O₂ extraction by the exercising muscles reduces the content of blood returning to the lungs.

Estimation of SaO₂ by most pulse oximeters (SpO₂) is based on the absorption of light at two wavelengths. When only two wavelengths are analyzed, only two species of Hb can be detected. Absorption in the red and near-infrared portions of the visible spectrum allows measurement of the oxyhemoglobin and reduced Hb, providing an estimate of the oxygen saturation of available Hb (see section, Pulse Oximetry).

Technique

Pulse oximeters (see Chapter 11) measure the light absorption of a mixture of two forms of Hb: O₂Hb and reduced Hb (rHb). The relative absorptions at 660 nm (red) and 940 nm of light (near infrared) can be used to calculate the combination of the two Hb forms. Absorption at two wavelengths provides an estimate of the saturation of available Hb (see the section on oxygen saturation). Most pulse oximeters use a stored calibration curve to estimate oxygen saturation.

Pulse oximetry may be used in any setting in which a noninvasive measure of oxygenation status is sufficient. This includes monitoring of O₂ therapy and ventilator management. Pulse oximetry is commonly used during diagnostic procedures such as bronchoscopy, sleep studies, or stress testing. It is also used for monitoring during patient transport or rehabilitation. Pulse oximetry may be used for continuous monitoring with inclusion of appropriate alarms to detect desaturation. Many pulse oximeter systems use memory (RAM) to record SpO₂ and heart rate for extended periods. This type of recorded monitoring allows pulse oximetry to be used for overnight studies of nocturnal desaturation. Alternatively, pulse oximetry can be used for discrete measurements or “spot checks.”

Most pulse oximeters use a sensor that attaches to the finger, toe (nail bed), or ear. The choice of attachment site should be dictated by the type of measurement being made. The ear site may be preferred in patients undergoing exercise testing or in whom arm movement precludes use of the finger site. Both finger and ear sites presume pulsatile blood flow. Most pulse oximeters adjust light output to compensate for tissue density or pigmentation. In some patients, impaired perfusion to one site determines which site is preferred. Low perfusion or poor vascularity can cause the oximeter to be unable to detect pulsatile blood flow. Rubbing or warming of the site often improves local blood flow and may be indicated to obtain reliable data. Some pulse oximeters use reflective sensors that detect light reflected from bone underlying the tissue bed. These sensors are usually placed on the patient’s forehead and may function when finger or ear sensors do not.

Some pulse oximeters display a representation of the pulse waveform derived from the absorption measurements (see Chapter 11). Such waveforms may be helpful in selecting an appropriate site or troubleshooting questionable SpO₂ values. Most pulse oximeters report heart rate (HR), which is also detected from pulsatile blood flow at the sensor site. Comparison of oximeter HR with palpated pulse or with an electrocardiograph (ECG) signal can assist with selection of an appropriate site. Inability to obtain a valid HR reading or acceptable pulse waveform suggests that SpO₂ values should be interpreted cautiously (Criteria for Acceptability 6-2).

A number of factors limit the validity of SpO₂ measurements (Box 6-2). To validate pulse oximetry readings, direct measurement of arterial saturation is required. Simultaneous measurement of SpO₂ and SaO₂ can be used to confirm pulse oximeter readings. Pulse oximetry used during exercise testing has been shown to produce variable results. Hemoximetry may be necessary to validate pulse oximetry readings at peak exercise.

COHb absorbs light at wavelengths similar to oxyhemoglobin. Most pulse oximeters sense COHb as O₂Hb and overestimate the O₂ saturation; however, pulse oximeters that measure O₂Hb and COHb are available. MetHb increases absorption at each of the two wavelengths used by pulse oximeters. This causes the ratio of the two Hb forms to approach 1, which is usually represented as a saturation of 85% (see Chapter 11). Other interfering substances may cause the pulse oximeter to underestimate saturation.

Significance and Pathophysiology

See Interpretive Strategies 6-3. Arterial oxygen saturation estimated by pulse oximetry (SpO₂) should approximate that measured by hemoximetry (SaO₂) in healthy nonsmoking adults. Most pulse oximeters are capable of accuracy of ±2% of the actual saturation when SaO₂ is above 90%. For SaO₂ values of 85%–90%, accuracy may be slightly less. For very low saturations (i.e., less than 80%), pulse oximeter accuracy is less of an issue because the clinical implications are the same.

Pulse oximetry is most useful when it has been shown to correlate with blood oximetry in a patient in a known circumstance. When this is the case, pulse oximetry can be used for noninvasive monitoring, either continuously or by taking discrete measurements. Uses include monitoring of O₂ therapy, ventilatory support, pulmonary or cardiac rehabilitation, bronchoscopy, surgical procedures, sleep studies, and cardiopulmonary exercise testing. In each of these applications, careful attention must be paid to minimizing known interfering agents or substances (see Box 6-2).

Because of its limitations, pulse oximetry should be used cautiously when assessing oxygen need. This is particularly true when using pulse oximetry to detect exercise desaturation. Pulse oximetry may not accurately reflect SaO₂ during exertion. For this reason, SpO₂, even if demonstrated to correlate with SaO₂ at rest, may yield false-positive or false-negative results during exercise. Blood gas analysis with hemoximetry should be used to resolve discrepancies between the SpO₂ reading and the patient’s clinical presentation.

Pulse oximetry may not be appropriate in all situations. To evaluate hyperoxemia (Pao₂ greater than 100–150 mm Hg) or acid-base status in a patient, blood gas analysis is required. Measurement of O₂ delivery, which depends on Hb concentration, cannot be adequately assessed by pulse oximetry alone.

Technique

Continuous monitoring of expired CO₂ is performed by sampling gas from the proximal airway. This gas may be pumped to an infrared analyzer (see Chapter 11) or to a mass spectrometer. An alternative method inserts a “mainstream” sample window directly into the expired gas stream. The analyzer signal is then passed to either a recorder or a computer. CO₂ waveforms may be displayed either individually (Figure 6-5) or as a series of deflections to form a trend plot. Petco₂ may be read from the peaks of the waveforms. It can also be obtained by a simple peak detector and displayed digitally. Continuous CO₂ monitoring is commonly used in patients with artificial airways in the critical care setting. Paco₂ can be measured at intervals to establish a gradient with Petco₂. Respiratory rate may be determined from the frequency of the CO₂ waveforms. The change in CO₂ concentration during a single expiration may be analyzed to detect ventilation-perfusion abnormalities (see Figure 6-5).

Technical problems involved in capnography include the necessity of accurate calibration and management of the gas sampling system (Criteria for Acceptability 6-3). Calibration using known gases (preferably two gas concentrations) is required if the system will be used to monitor Paco₂. Many systems use ambient air, containing minimal CO₂, and a 5% CO₂ mixture for calibration (see Chapter 12). Condensation of water in sample tubing, connectors, or the sample chamber can affect accuracy. Some infrared analyzers may be affected if sample flow changes after calibration. Saturation of a desiccator column, if used, can also lead to inaccurate readings. Long sample lines or low sample flows can cause damping of the CO₂ waveform, invalidating analysis of the shape of the expired gas curve.

Significance and Pathophysiology

See Interpretive Strategies 6-4. In healthy patients, CO₂ rises to a plateau as alveolar gas is expired (Figure 6-5). If all lung units empty CO₂ evenly, the plateau appears flat. However, even healthy lungs have ventilation and blood flow imbalances. Healthy lung units empty CO₂ at varying rates. Alveolar CO₂ concentration increases slightly as the exhalation continues. Petco₂ theoretically should not exceed Paco₂. In healthy patients at rest, the Petco₂ is usually close to the arterial value. During maximal exercise Petco₂ may exceed Paco₂. When ventilation and perfusion become grossly mismatched (e.g., in severe obstruction), end-tidal CO₂ may exceed the Paco₂ as well. Paco₂ reflects gas-exchange characteristics of the entire lung. Thus, Petco₂ may differ significantly if some lung units are poorly ventilated.

Continuous CO₂ analysis provides useful data for monitoring critically ill patients, particularly those requiring ventilatory support. Petco₂ measurements allow trending of changes in Paco₂ provided there is little or no change in the shape of the CO₂ waveform (indicating /abnormalities). When a reference blood gas sample is obtained, Petco₂ can be used as a continuous, noninvasive monitor. Respiratory rate can be measured from the frequency of expired CO₂ waveforms. Marked changes in breathing rate (e.g., hyperpnea or apnea) can be detected quickly. Analysis of the individual CO₂ waveforms, along with Paco₂, may help identify abrupt changes in dead space. This can be useful in detecting pulmonary embolization or reduced cardiac output.

Problems related to ventilatory support devices can also be detected. Disconnection or leaks in breathing circuits can be quickly recognized by the loss of the CO₂ signal. Increased mechanical dead space (i.e., gas rebreathed in the ventilator circuit) can be identified by a baseline CO₂ concentration greater than zero. Irregularities in the CO₂ waveform often signal that the patient is “out of phase” with the ventilator.

The shape of the expired CO₂ curve is determined by ventilation-perfusion matching. Only lung units that are ventilated and perfused contribute CO₂ to expired gas. In patients without lung disease, the CO₂ waveform shows a flat initial segment of anatomic dead space gas containing little or no CO₂. This phase is followed by a rapid increase in CO₂ concentration, reflecting a mixture of dead space and alveolar gas. Finally, an “alveolar” plateau occurs in which gas composition changes only slightly. This slight change is caused by different emptying rates of various lung units. The absolute concentration of CO₂ at the alveolar plateau depends on factors such as minute ventilation and CO₂ production. Dead space–producing disease (e.g., pulmonary embolization or marked decrease in cardiac output) may show a profound decrease in expired CO₂ concentration. Patients who have pulmonary disease, especially obstruction, show poorly delineated phases of the CO₂ washout curve. The alveolar plateau may actually be a continuous slope throughout expiration, causing the measurement of Petco₂ to be misleading.

Technique

Two methods for measuring the shunt fraction are available. The first uses O₂ content differences between arterial and mixed venous blood. The first method is called the physiologic shunt equation:

Equation 1.

Q·sQ·t=CcO2–CaO2CcO2–Cv¯O2

where:

CcO₂= O₂ content of end-capillary blood, estimated from saturation associated with calculated P_Ao₂

CaO₂= arterial O₂ content, measured from an arterial sample

Cv¯o₂= mixed venous O₂ content, measured from a sample obtained from a pulmonary artery catheter

The term in the denominator of this equation reflects potential arterialization of mixed venous blood. The term in the numerator reflects the actual arterialization.

The second method is more commonly used in clinical practice. The patient breathes 100% O₂ in a unidirectional valved system until the Hb is completely saturated (see Figure 6-6). Twenty minutes of O₂ breathing is usually sufficient. Percent shunt is then calculated from differences in dissolved O₂. The second method is called the clinical shunt equation:

Equation 2.

Q·sQ·t=(PAO2−PaO2)×0.0031C(a−v¯)O2+[(PAO2−PaO2)×0.0031]

where:

P_Ao₂= alveolar O₂ tension

Pao₂= arterial O₂ tension

C(a–v¯)o₂= arteriovenous O₂ content difference

0.0031= solubility factor for O₂ at 37°C

When the patient is breathing 100% O₂, P_Ao₂ can be estimated from an abbreviated form of the alveolar gas equation as follows:

PAO2=PB−PH2O−(PaCO20.8)

where:

P_B= barometric pressure

PH₂O= partial pressure of water vapor at body temperature (47 mm Hg)

Paco₂= arterial CO₂ tension (as an estimate of alveolar Pco₂)

0.8= normal (assumed) respiratory exchange ratio

This calculation of alveolar Po₂ assumes that the inspired gas is 100% O₂. If the F_IO₂ is measured and found to be less than 1, the standard alveolar air equation can be used (see Evolve http://evolve.elsevier.com/Mottram/Ruppel/).

The clinical shunt calculation is accurate only when Hb is completely saturated. This normally requires a Pao₂ greater than 150 mm Hg. A saturation of 100% is usually easily accomplished if O₂ is breathed long enough. If breathing 100% O₂ does not increase the Pao₂ high enough to completely saturate Hb, the content difference method (physiologic method) should be used. With either method, shunt fraction may be multiplied by 100 and reported as a percentage (e.g., 0.20 ratio × 100 equals a 20% shunt).

Several technical considerations should be noted regarding shunt measurement (Criteria for Acceptability 6-4). Using O₂ content differences (equation 1) requires a pulmonary artery catheter to obtain mixed venous O₂ content. Using dissolved O₂ differences (equation 2) also relies on measured C(a-v¯)o₂. If placement of a pulmonary artery catheter is not practical, the clinical shunt equation may be used with an assumed a-v¯ content difference or a presentation of the calculated data with several a-v¯ content differences (Table 6-5) (see the discussion on Significance and Pathophysiology in this section). The estimated technique assumes the subject is breathing 100% oxygen, so a closed circuit with directional value is required. This can be accomplished with a demand valve driven with a 50 psig 100% oxygen gas source or reservoir bag (e.g., Douglas bag) filled with 100% oxygen connected to a directional value. The subject needs to be monitored for leaks around the mouth or noseclips. The sample should be collected in a glass syringe or analyzed immediately to reduce the decrease in Pao₂ secondary to environmental exposure (How To Box 6-3).

Table 6-5

Shunt Study
Male, Age 48, Barometer: 719, Time: 10:00

Sample No.	Position	Activity	Pao2 (mm Hg)	Paco2 (mm Hg)	P(a-a)o2 (mm Hg)	C(a-)o2 (cc/100 mL)	sp/tot (%)
3	Supine	Rest	228.0	38-6	405-4	4	23.9
						5	20.1
						6	17.3
						8	13.6
4	Stand	Rest	355.0	36	281.0	4	17.9
						5	14.8
						6	12.7
						8	9.8

(Note: Shunt fraction (sp/tot) is estimated from arterial blood gas while breathing at 100% oxygen. Mixed venous blood is not sampled (sp/tot) is reported assuming 4, 5, 6, and an 8 CcO₂/100-mL difference between arterials and mixed venous blood. Clinical evaluation should dictate which value is appropriate.)

The physiologic shunt equation may be used for patients on any known F_IO₂. The clinical shunt measurement requires O₂ breathing for 20 minutes or longer. Prolonged O₂ breathing may be contraindicated in patients whose respiration is driven by hypoxemia. Breathing 100% O₂ washes nitrogen (N₂) out of the lungs. Washout of N₂ combined with O₂ uptake by perfusing blood flow can reduce the size of alveoli to their critical limit. In poorly ventilated lung units, this may cause alveolar collapse. The effect of this “nitrogen shunting” may be shunt values that are falsely high because some shunting was induced by the test itself.

Significance and Pathophysiology

See Interpretive Strategies 6-5. In healthy individuals, approximately 5% of the cardiac output is shunted past the pulmonary system. An increased pulmonary shunt fraction indicates that some lung units have little ventilation in relation to their blood flow. These patterns may be found in both obstructive and restrictive diseases. However, even in severe obstruction or restriction, blood flow to areas of poor ventilation may be reduced by the lesions themselves. In emphysema, destruction of the alveolar septa obliterates pulmonary capillaries. As the terminal airways lose their support, they also have reduced blood flow. In poorly ventilated lung units, vasoconstriction of pulmonary arterioles redirects blood flow away from the affected area. In these cases, there may be minimal shunting, even though severe ventilatory impairment exists.

Intrapulmonary shunting is common in acute disease patterns such as atelectasis or foreign body aspiration. Diseases such as pneumonia or adult (in some references this has been broaden to “acute”) respiratory distress syndrome (ARDS) often result in a “shunt-like effect.” This is caused by reduced ventilation in relation to blood flow in many lung units. Foreign body aspiration may cause shunting by blocking an airway and depriving all distal lung units of ventilation. Blood flow to the affected units cannot participate in gas exchange. The degree of shunting is directly related to the number of lung units with /ratios close to zero.

Intrapulmonary shunting may also occur when the pulmonary vascular system is involved, as is the case in hepatopulmonary syndrome. Patients who have liver disease with portal hypertension commonly have arteriovenous malformations that allow significant volumes of blood to pass through the lungs without participating in gas exchange. These malformations are prominent at the lung bases and result in increased shunting when the patient is upright.

Intracardiac shunting occurs when defects in the atrial or ventricular septa allow mixed venous blood to pass from the right side of the heart to the left side of the heart without traversing the pulmonary capillaries. In the ventricles, shunting usually requires an elevated right-heart pressure to overcome the normal pressure difference between the systemic and pulmonary systems. Atrial septal defects, such as patent foramen ovale (PFO), allow blood to pass from the right atrium to the left atrium during systole.

Very large shunt values (greater than 30%) suggest that a significant volume of blood is moving from the right side of the heart to the left side of the heart without participating in gas exchange. Further testing may be required to determine whether the shunt is occurring in the lungs or in the heart. Echocardiography with contrast media or cardiac catheterization may be necessary to identify intracardiac shunts.

The accuracy of clinical shunt measurement (i.e., dissolved O₂ differences) depends on the accuracy of Po₂ determinations. In small shunts, Hb becomes 100% saturated. The difference between alveolar and arterial Po₂ values results simply from the amount of O₂ dissolved. The difference between the actual content of dissolved oxygen and the content that can potentially dissolve is the basis for the calculation. Measurements of Po₂ used for shunt calculations (200–600 mm Hg) are much higher than the normal physiologic range. Additional calibration and quality control of the Po₂ electrode may be necessary to provide accurate values for oxygen tension. Blood gases drawn for shunt studies should be analyzed immediately to avoid large changes in Pao₂ values caused by cell metabolism in the specimen and the effect of diffusion across a plastic syringe if used. Glass syringes should be used if any delay in sampling greater than a few minutes is expected.

Table 6-6

Causes of Hypoxia and Hypoxemia

Type	Cause
Hypoxemic hypoxia	Altitude
	Hypoventilation
	Shunt
	V./Q. mismatching
	Diffusion
Circulatory hypoxia	Hypovolemia
	Reduced cardiac output (impaired LV function, outflow tract abnormalities, etc.)
Hemic hypoxia	Anemia, dyshemoglobinemias (COHb, MetHb, etc.)
Histotoxic hypoxia	Cyanide poisoning
Demand hypoxia	Seizures, exercise

The calculated shunt fraction also depends on the O₂ content difference between arterial and mixed venous blood. C(a-v¯)o₂ is a component of the denominator in the clinical shunt equation. The a-v¯content difference is determined not only by the lungs but also by cardiac output and the perfusion status of the tissues. Ideally, the value used in the equation should be measured rather than estimated. Arterial content can be determined easily from a sample taken from a peripheral artery. However, mixed venous content can only be measured accurately from a pulmonary artery sample. In patients who do not have a pulmonary artery catheter in place, an estimated value must be used. C(a-v¯)o₂ values from 4.5–5.0 vol% are reasonable content differences in patients who have good cardiac output and perfusion status. Values of 3.5 vol% may be more realistic in patients who are critically ill.

In some instances, a-v¯ content difference cannot be reliably estimated, or Hb cannot be maximally saturated by breathing 100% oxygen. In such cases, the alveolar-arterial oxygen gradient P(a−a)o₂ may be useful as an index of ventilation to blood flow. In healthy patients breathing 100% O₂ at sea level, arterial Pao₂ should increase to approximately 600 mm Hg. s/t does not directly provide absolute values for s, but if the cardiac output (t) is known, s can be determined simply. Measurement of the shunt fraction is sometimes performed in conjunction with the determination of the V_D/V_T ratio (see Chapter 5) to assess both types of gas exchange abnormalities together.