Chapter 27: Blood Gases

Blood gas and pH measurements allow evaluation of oxygenation, ventilation, and acid-base balance. Either arterial or mixed venous blood gases can be assessed. This chapter covers important aspects of blood gas assessment as it relates to the mechanically ventilated patient.

Partial Pressure of Oxygen

The normal range of Pao₂ is 80 to 100 mm Hg in healthy young persons breathing room air at sea level. Pao₂ decreases with age, altitude, and lung disease. Hypoxemia occurs when the lungs fail to adequately oxygenate arterial blood. Pao₂ is a reflection of lung function and not hypoxia per se. Hypoxia can occur without hypoxemia and vice versa. Causes of hypoxemia and hypoxia are listed in Table 27-1. The best Pao₂ in critically ill mechanically ventilated patients is unknown, but a target Pao₂ of 55 to 80 mm Hg (at sea level) is usually acceptable. Pao₂ must be balanced against the potentially toxic effects of Fio₂ and alveolar distending pressure. For mechanically ventilated patients with severe lung disease, permissive hypoxemia may be a desirable alternative to applying potentially injurious ventilator setting to normalize the Pao₂.

Oxygen Saturation

The relationship between Pao₂ and oxygen saturation of hemoglobin (Sao₂) is described by the oxyhemoglobin dissociation curve (Figure 27-1). This is a sigmoid relationship, with hemoglobin having a greater affinity for oxygen at a high Po₂ (eg, in the lungs, where the Po₂ is high) and a lower affinity for oxygen at a lower Po₂ (eg, in the tissues, where Po₂ is low). The affinity of hemoglobin for oxygen is also affected by the environment of the hemoglobin molecule, which can shift the curve to the left or to the right. Shifts of the curve to the right decrease the affinity of hemoglobin for oxygen (promote oxygen unloading), and shifts of the curve to the left increase the affinity of hemoglobin for oxygen (promote oxygen binding). Because of the variable relationship between hemoglobin saturation and Po₂, saturation cannot be precisely predicted from Po₂, and vice versa. To accurately evaluate oxygen saturation, CO-oximetry should be performed. CO-oximetry also allows measurement of total hemoglobin concentration, oxygen saturation, methemoglobin level, and carboxyhemoglobin level.

Oxygen Content and Oxygen Delivery

Oxygen content (Co₂) is a combination of dissolved oxygen and that combined with hemoglobin (Hb).

The amount of oxygen dissolved in plasma is small and related to the Po₂. The importance of Pao₂ is that it determines the Sao₂ and thus the amount of oxygen bound to hemoglobin. Note that a decrease in Pao₂ and Sao₂ may not result in a decrease in Co₂ if there is a concomitant increase in [Hb] (polycythemia).

Oxygen delivery is determined by cardiac output and oxygen content:

Note that oxygen delivery to tissues is determined by both Q̇c and Co₂. Thus, hypoxia can result from either a decrease in Q̇c or Co₂. Moreover, a decrease in oxygen delivery may not occur with a decrease in Co₂ if there is a concomitant increase in Q̇c.

Alveolar Po₂

Alveolar Po₂ (Pao₂) is a mathematically derived value using the alveolar gas equation:

where Fio₂ is the inspired O₂ fraction, Pb is barometric pressure, Ph₂o is water vapor pressure (47 mm Hg at 37°C), and R is the respiratory quotient (V̇co₂/V̇o₂). For calculation of Pao₂, R = 0.8 is commonly used. Note that the effect of R on Pao₂ depends on the Fio₂. For Fio₂ 0.60 or greater, the effect of R on Pao₂ becomes negligible. For a high Fio₂ 0.60 or greater, the alveolar gas equation thus becomes:

For Fio₂ less than 0.60, the alveolar Po₂ is estimated by:

Pressure-Based Indices

There are several oxygen-pressure-based indices. Each of these relates Pao₂ to either Pao₂ or the Fio₂. P(a–a)o₂ is calculated by subtracting the Pao₂ from the Pao₂. An increase in P(a–a)o₂ can result from V̇/Q̇ disturbances, shunt, or diffusion limitation. Changes in Paco₂ will not affect the P(a–a)o₂ because Paco₂ is included in the calculation of Pao₂. A problem with the use of the P(a–a)o₂ is that it changes as Fio₂ changes. The normal P(a–a)o₂ is 5 to 10 mm Hg breathing room air, but 30 to 60 mm Hg when breathing 100% O₂. This variability, when the Fio₂ is changed, limits its usefulness as an indicator of pulmonary function with Fio₂ changes and invalidates it as a predictor of the change in Pao₂ if the Fio₂ is changed. The P(a–a)o₂ is affected not only by the Fio₂ but also by the degree of intrapulmonary shunt and V̇/Q̇ mismatch. In critically ill patients, the P(a–a)o₂ does not correlate well with the degree of pulmonary shunt. The P(a–a)o₂ is also affected by changes in mixed venous oxygen content.

The Pao₂/Pao₂ is calculated by dividing the Pao₂ by Pao₂. Unlike the P(a–a)o₂, the Pao₂/Pao₂ remains relatively stable with Fio₂ changes. A Pao₂/Pao₂ less than 0.75 indicates pulmonary dysfunction due to V̇/Q̇ abnormality, shunt, or diffusion abnormality. The Pao₂/Pao₂ is most stable when it is less than 0.55, when the Fio₂ is greater than 0.30, and when the Pao₂ is less than 100 mm Hg. The Pao₂/Pao₂ is more useful than the P(a–a)o₂ for comparing the pulmonary function of patients on different Fio₂ and for following a patient's pulmonary function as Fio₂ is changed.

The Pao₂/Fio₂ is easier to calculate than P(a–a)o₂ and Pao₂/Pao₂ because it does not require calculation of Pao₂. The Pao₂/Fio₂ is used in the classification of the acute respiratory distress syndrome (ARDS). A Pao₂/Fio₂ of 100 mm Hg or less is consistent with severe ARDS, Pao₂/Fio₂ greater than 100 but less than 200 indicates moderate ARDS, and Pao₂/Fio₂ greater than 200 but less than or equal to 300 indicates mild ARDS, when patients are receiving 5 cm H₂O or greater positive end-expiratory pressure (PEEP).

The oxygenation index (OI) relates Pao₂, Fio₂, and mean airway pressure (P̄aw):

Although not commonly used in adults, this index is used to classify respiratory failure in infants and children.

Pulmonary Shunt

Shunting is the portion of the cardiac output that moves from the right side of the heart to the left side of the heart without participating in gas exchange. Shunt is calculated from the oxygen content of pulmonary end-capillary (Cc′o₂), arterial (Cao₂) and mixed venous (Cv̄o₂) blood:

where Q̇_S is shunted cardiac output, Q̇_T is total cardiac output, Cc′o₂ is pulmonary end-capillary oxygen content, Cao₂ is arterial oxygen content, and Cv̄o₂ is mixed venous oxygen content.

The arterial oxygen content (Cao₂) is calculated from arterial blood gas values, and mixed venous oxygen content (Cv̄o₂) is calculated from pulmonary artery blood gas values. Cc′o₂ is calculated based on the assumption that pulmonary end-capillary Po₂ is equal to the alveolar Po₂. When Pao₂ is greater than 150 mm Hg, it is assumed that the end-capillary blood is 100% saturated with oxygen. When a pulmonary artery catheter is not in place to sample mixed venous blood, shunt can be estimated from the equation:

The 3.5 vol% can replace Cao₂ – Cv̄o₂ if there is cardiovascular stability and body temperature is normal. The Cc′o₂ – Cao₂ can be replaced by (Pao₂ – Pao₂) × 0.003 in settings where it can be assumed that the Sao₂ is 100%. When the patient has a high Pao₂ (>150 mm Hg), the modified shunt equation can be used:

Oxygen Delivery and Oxygen Consumption

Oxygen delivery (Do₂) is the volume of oxygen delivered to the tissues each minute and is calculated as:

Normal Do₂ is 1000 mL/min. Of this, the tissues normally extract 250 mL/min (V̇o₂), and 750 mL is returned to the lungs. V̇o₂ can be calculated using the Fick equation:

Oxygen extraction is the oxygen consumption divided by the oxygen delivery.

Partial Pressure of Carbon dioxide

The adequacy of alveolar ventilation is usually assessed by the arterial partial pressure of carbon dioxide (Paco₂) due to the relationship between Paco₂, V̇a, and V̇co₂:

Thus, Paco₂ is an indication of the body's ability to sustain V̇_A adequate for V̇co₂. V̇co₂ is determined by metabolic rate and is normally about 200 mL/min. An increase in V̇co₂ requires a higher minute ventilation (V̇_E). Dead space ventilation also affects the relationship between V̇_E and Paco₂; minute ventilation must increase to maintain the same Paco₂ in the presence of increased dead space. Clinical causes of hypoventilation (increased Paco₂) and hyperventilation (decreased Paco₂) are listed in Table 27-2. Although a goal of mechanical ventilation in the past has been to normalize Paco₂, an elevated Paco₂ (permissive hypercapnia) may be more desirable than the high alveolar distending pressure required to normalize the Paco₂.

Dead Space and Alveolar Ventilation

Dead space is that portion of the minute ventilation that does not participate in gas exchange. It consists of anatomic dead space and alveolar dead space. Dead space is calculated using the Bohr equation:

In the Bohr Equation, V_D/V_T is the fraction of the total ventilation that is dead space and Pēco₂ is the partial pressure of CO₂ in mixed expired gas. Normal V_D/V_T is 0.2 to 0.4. Causes of increased V_D/V_T include pulmonary embolism, positive pressure ventilation, pulmonary hypoperfusion, low tidal volume, and alveolar overdistention. The traditional method to determine Pēco₂ uses mixed exhaled gas collected for 5 to 15 minutes (Figure 27-2). An arterial blood sample for Paco₂ is obtained during this time. However, many current-generation mechanical ventilators have a constant bias flow through the circuit, which complicates the collection of mixed exhaled gas to calculate V_D/V_T. In this case, Pēco₂ can be calculated from V̇co₂ and V̇_E using volumetric capnography.

Because dead space determinations require a leak-free system, they cannot be measured in patients with a bronchopleural fistula.

V_D/V_T correlates with mortality in patients with ARDS; high V_D/V_T is associated with higher mortality. V_D/V_T can also be used to determine the balance between alveolar recruitment and overdistention with titration of PEEP in patients with ARDS. The best level of PEEP is associated with the lowest V_D/V_T, with a PEEP too low and a PEEP too high associated with a higher V_D/V_T.

From the exhaled CO₂ and V̇_E, alveolar ventilation (V̇_A) can be calculated:

V̇_A can also be calculated from V_D/V_T as:

Acid-base balance is explained by the Henderson-Hasselbalch equation:

Metabolic acid-base disturbances are those that affect the numerator of the Henderson-Hasselbalch equation, and respiratory acid-base disturbances are those things that affect the denominator. The arterial pH is normal (7.40) when the ratio [HCO₃^–]/(0.03 × Pco₂) is 20:1. The metabolic component of acid-base interpretation is usually given as the [HCO₃^–]. The metabolic component can also be expressed as base excess (BE):

In other words, [HCO₃^–] less than 24mmol/L corresponds with a negative BE, and [HCO₃^–] greater than 24mmol/L corresponds with a positive BE. An algorithm for classification of acid-base disturbances is shown in Figure 27-3. Clinical causes of metabolic acid-base disturbances are listed in Table 27-3, and the expected degree of compensation for acid-base disturbances is shown in Table 27-4.

Anion Gap and Osmol Gap

The anion gap (AG) is useful to differentiate causes of metabolic acidosis. Metabolic acidosis can be associated with a normal AG (hyperchloremic acidosis) or with an increased AG (normochloremic acidosis). The AG is calculated as:

The normal AG is 8 to 12 mmol/L. Causes of metabolic acidosis with an increased AG include lactic acidosis, diabetic ketoacidosis, and azotemic (renal) acidosis. Causes of metabolic acidosis with a normal AG include loss of bicarbonate from the gastrointestinal tract (eg, diarrhea), acetazolamide therapy, or excessive chloride administration (eg, HCl, NH₄Cl). The traditionally defined AG does not take into account the large changes in plasma albumin concentration often seen in critically ill patients. Unless a correction is used, an increased AG may go unrecognized. This has led to the concept of albumin-corrected AG. AG is reduced approximately 2.5 mmol/L for every 1 g/dL fall in albumin:

The osmol gap is the difference between the measured osmolality of the plasma and that calculated as:

where osmol is osmolality and BUN is the blood urea nitrogen. If the measured osmolality is more than 10 above that calculated, there may be unmeasured osmotically active particles present, whose metabolites may be organic acids. Metabolic acidosis with an osmol gap is consistent with the presence of the toxins methanol and ethylene glycol.

Strong Ion Difference

The strong ion difference (SID) is a method of evaluating acid-base disturbances based on Stewart's approach to acid-base chemistry. Using Stewart's approach, the only variables that affect pH are the Pco₂, SID, and the concentration of unmeasured strong ions. SID is calculated as:

Alternatively, SID is calculated as:

The normal value for SID is 40 mmol/L. Classification of primary acid-base disturbances using SID is shown in Table 27-5. Metabolic acidosis is associated with a decreased SID and metabolic alkalosis is associated with an increased SID.

To assess mixed venous blood gases (Pv̄o₂ and Pv̄co₂), blood is obtained from the distal port of the pulmonary artery catheter. Because pulmonary artery catheters are no longer commonly used, blood is drawn from a central venous catheter as a proxy for mixed venous blood. There is a reasonable relationship between central venous blood gases and mixed venous blood gases if the distal tip of the catheter is in the right atrium. Peripheral venous blood gases, however, have little utility for assessment of cardiopulmonary function; they reflect the metabolic conditions of the local tissue and cannot be used to assess respiratory function.

Mixed Venous Po₂

Normal mixed venous Po₂ (Pv̄o₂) is 40 mm Hg and is a global indication of the level of tissue oxygenation. However, normal or supranormal values of Pv̄o₂ can coexist with severe tissue hypoxia caused primarily by arterial admixture, septicemia, hemorrhagic shock, congestive heart failure, and some febrile states. Further, Pv̄o₂ reveals little about the oxygenation status of individual tissue beds. Factors affecting Pv̄o₂ can be illustrated from rearrangement of the Fick equation:

Cv̄o₂ (and its components Pv̄o₂ and Sv̄o₂) is decreased with decreases in Cao₂ (ie, Pao₂, So₂, or Hb), decreases in Q̇, or increases in V̇o₂. Note that an increase in V̇o₂ with a proportional increase in Q̇ does not affect Cv̄o₂ (eg, exercise). Also note that breathing 100% oxygen by persons with normal lung function does not affect Cv̄o₂ because increasing the Pao₂ affects Cao₂ very little (ie, oxygen is very insoluble in blood and the hemoglobin is nearly 100% saturated when breathing room air). In patients with abnormal lung function, a decrease in Pv̄o₂ may produce a decrease in Pao₂.

Mixed Venous and Central Venous Oxygen Saturation

Mixed venous oxygen saturation (Sv̄o₂) can be determined from a blood sample obtained from the distal port of the pulmonary artery catheter or by an oximeter that monitors Sv̄o₂ continuously using a system incorporated into the pulmonary artery catheter. The oximeter reflects light from red blood cells near the pulmonary artery catheter, and Sv̄o₂ is determined as the ratio of transmitted and reflected light. Central venous oxygen (Scvo₂) saturation can be measured when a pulmonary artery catheter is not present. When the central venous catheter tip is 15 cm away from the inlet of the right atrium, Scvo₂ overestimates Sv̄o₂ by 8%, but when the tip of the catheter is in the right atrium, Scvo₂ overestimates Sv̄o₂ by only 1%. Sv̄o₂ monitoring is used during early goal-directed therapy for sepsis treatment, with a Scvo₂ goal 70% or greater.

Mixed Venous Pco₂

Mixed venous Pco₂ (Pv̄co₂) is a global indication of tissue Pco₂. Normal Pv̄co₂ is 45 mm Hg, which is only slightly greater than Paco₂. Under conditions of low perfusion (eg, cardiac arrest), there can be a great disparity between Paco₂ and Pv̄co₂. Under these conditions, a respiratory acidosis can be present at the tissue level and in the venous circulation, concurrent with a respiratory alkalosis in the arterial circulation. Paco₂ is determined by V̇_A, whereas Pv̄co₂ is determined by perfusion (Figure 27-4).

There is an increasing interest in monitoring brain tissue oxygenation directly. In traumatic brain injury, brain tissue Po₂ (Pbto₂) can be monitored directly with a thin, metallic electrode that measures Po₂ in a small area of brain tissue. Normal values for Pbto₂ are 25 to 30 mm Hg. Higher mortality occurs with increasing duration of Pbto₂ less than 15 mm Hg, but it is unknown whether interventions to increase Pbto₂, such as increasing Fio₂ to increase the Pao₂, result in better outcomes.

Blood gases and pH are measured at 37°C (normal body temperature). If the patient's temperature is abnormal, the in vivo blood gas and pH values will differ from those measured and reported by the blood gas laboratory. The use of temperature-adjusted values for blood gases and pH is controversial. Although normal values are known for euthermia, normal values during hypothermia and hyperthermia are unknown. The acid-base changes that occur with hypothermia and hyperthermia may be homeostatic. The treatment of acid-base disturbances should be guided by the unadjusted values (ie, those measured at 37°C). Temperature adjustment of blood gases and pH is useful to follow changes in these values with changes in body temperature. Temperature adjusted blood gas values are used during therapeutic hypothermia. Temperature adjusted values should also be used to compare blood gases to exhaled gas values (eg, end-tidal Pco₂) and to assess lung function by comparing Pao₂ and Pao₂. Temperature adjustment allows the clinician to differentiate temperature-related changes from pathophysiologic changes.