Tissue Oxygenation

The outcome of critical illness depends on the adequacy of oxygenation; therefore, accurate assessment and treatment of oxygenation disturbances is critical to optimal patient outcomes. A thorough evaluation of oxygenation is one of the most important components of the advanced practice nurse's assessment abilities.

Oxygenation is a physiologic process that is dependent upon the integration and coordination of multiple body systems including the pulmonary, cardiovascular, neurologic, hematologic, and metabolic systems. Adequacy of oxygenation depends on the integration of three physiologic components: pulmonary gas exchange, oxygen delivery, and oxygen consumption (Figure 1). Pulmonary gas exchange physiology involves the physiologic processes of ventilation, diffusion, and perfusion as oxygen is brought from the atmosphere to the pulmonary capillary bed. Delivery of oxygen via the bloodstream is dependent upon cardiac output and the content of oxygen in arterial blood. Use of oxygen for energy metabolism requires extraction and consumption of oxygen through aerobic metabolism. Each of these three physiologic processes must function properly to ensure adequate oxygenation. Each of these components may vary independently in response to pathophysiologic conditions and therapeutic interventions. Therefore, an assessment of oxygenation must include assessments of pulmonary gas exchange, oxygen delivery, and oxygen consumption. There are two goals in the assessment of oxygenation: (1) to determine overall adequacy of oxygenation and (2) to determine which element of oxygenation dysfunction should be manipulated to improve patient outcome.

Assessment of Pulmonary Gas Exchange

Pulmonary gas exchange involves the inspiration and delivery of oxygen from the external environment to the alveoli, diffusion across the alveolar-capillary membrane, and the combination of oxygen with hemoglobin (Hb) in the pulmonary capillaries. Pulmonary gas exchange is dependent upon the physiologic processes of alveolar ventilation, diffusion of gases across the alveolar-capillary membrane, pulmonary perfusion, and ventilation perfusion matching (V/Q) (see Figure 1). Therefore, an assessment of pulmonary gas exchange should ideally include an evaluation of these physiologic processes.

Assessment of Ventilation

Alveolar ventilation depends on respiratory rate and tidal volume. An average tidal volume of 600 mL results in an alveolar ventilation of 450 mL, with 150 mL required to overcome normal physiologic dead space of the conducting airways. At very low tidal volumes, the dead space alone may be ventilated even though the minute volume (rate times tidal volume) is normal due to a high respiratory rate.

Assessment of the adequacy of ventilation begins with a focused physical examination that can provide some of the most useful clues about ventilation. Lethargy and somnolence may indicate hypercapnea. Evaluation of rate, rhythm, and depth of respirations can provide an insight into the work of breathing and adequacy of alveolar ventilation which are important factors to analyze when considering initiation of mechanical ventilation, alteration of ventilator settings, weaning a patient from the ventilator, or advancing a patient's activity level. A normal respiratory rate and absence of the use of accessory muscles may reflect a manageable work of breathing. Conversely, tachypnea and the use of accessory muscles of inspiration (sternocleomastoid or external intercostals) or expiration (internal intercostals, abdominal muscles) may indicate the patient has insufficient muscle strength to maintain the work of breathing necessary to overcome increased resistance or decreased compliance. Resistance and compliance can be directly measured in patients receiving mechanical ventilation (See "Assessment of the Patient Receiving Mechanical Ventilation" in this issue).

Auscultation of the lungs can be used to evaluate the effectiveness of ventilation. Bronchial sounds may be indicative of extensive consolidation, and bronchovesicular sounds may indicate early pulmonary disease. Wheezing, produced by airflow through narrow airways, may be heard on expiration. Inspiratory wheezes may indicate bronchial constriction. Crackles may indicate fluid in the lung or weakened airway wall strength. Cardiac disease is more likely to produce excessive fluid type crackles that respond to fluid removal, while pulmonary disease is more likely to produce crackles that are nonresponsive to fluid removal.

The major route of elimination of carbon dioxide (CO₂) is through the lungs. Therefore, a quantitative assessment of ventilation is made through CO₂ evaluation. The higher the serum CO₂, the more significant the ventilatory failure. CO₂ can be measured through arterial blood gas analysis (PaCO₂) or end-tidal CO₂ monitoring (PetCO₂). PetCO₂ monitoring is a noninvasive measurement of exhaled CO₂. It ideally represents an assessment of alveolar ventilation as PetCO₂ should correlate with PaC0₂. In healthy individuals under normal conditions of V/Q matching, the difference between PetCO₂ and PaCO₂ is minimal (<5 mm Hg); as exhaled CO₂ is always slightly lower than PaCO₂. In these situations, PetCO₂ can be used as a substitute for PaCO₂.1-4

Several patient conditions inhibit the use of PetCO₂ to reflect PaCO₂. PetCO₂ does not accurately reflect PaCO₂ under conditions of respiratory failure, hemodynamic instability, or extremes in temperature. 5 For example, if perfusion to the lungs is decreased (cardiac arrest, pulmonary embolism), then less CO₂ is carried to the lungs from the tissues and PetCO₂ will be lower than PaCO₂. There must be adequate perfusion to the lungs to bring CO₂ to the alveolar capillary membrane so that gas exchange can occur and CO₂ is exhaled. Increased peripheral CO₂ production, such as that which occurs in a febrile state, can increase PetCO₂.

Isolated PetCO₂ values must be evaluated cautiously in critically ill patients. Multiple measurements of PetCO₂ or monitoring the PaCO₂-PetCO₂ gradient over time may provide important clues related to either improved or worsening patient clinical status. 5 A large PaCO₂-PaC0₂ gradient (>20 mm Hg) may reflect a large V/Q mismatch. A gradual improvement in the gradient (a narrowing of the gradient) may represent improved V/Q matching and ventilatory function.

Physiologic processes of the respiratory system can be evaluated by different pulmonary function tests (PFTs). PFTs evaluate lung mechanics by measuring the volume of air the patient is able to move in and out during ventilation and estimating several lung capacities (Table 1). Potential uses of PFTs in critically ill patients include: evaluation of known or suspected lung disease; evaluation of symptoms such as chronic cough, dyspnea, or chest tightness; monitoring the effects of exposure to pulmonary toxic drugs; risk stratification prior to surgery; and monitoring the effectiveness of therapeutic interventions. 6 Physiologic abnormalities that may be measured by PFTs include obstruction to airflow, restriction of lung size, and a decrease in transfer of gases across the alveolar-capillary membrane. PFTs can detect these abnormalities early in the course of a disease when the physical examination and radiographic studies still appear normal. 7

. Pulmonary Function Test Parameters

A major limitation of PFTs is that they require considerable patient effort and cooperation which limits their use in the intensive care unit (ICU) patient population. Inadequate patient effort can lead to misdiagnoses and inappropriate treatment. 8 However, in critically ill mechanically ventilated patients, it is important to monitor some of the components of PFTs including vital capacity, tidal volume, and minute ventilation (See "Assessment of the Patient Receiving Mechanical Ventilation" elsewhere in this issue). These parameters can help measure the effects of a disease process on ventilation. Decreased tidal volume results in alveolar hypoventilation and acute respiratory failure. Tidal volume and vital capacity monitor respiratory muscle strength. Therefore, as the patient experiences respiratory muscle fatigue, these values decrease. Increased minute ventilation is associated with increased work of breathing. Decreased minute ventilation is associated with hypoventilation.

Assessment of Intrapulmonary Shunt

Intrapulmonary shunt is the proportion of blood that flows past alveoli without participating in gas exchange. An elevated intra-pulmonary shunt indicates a large percent of venous blood has bypassed alveoli and entered arterial blood without being oxygenated. Increased intrapulmonary shunt can be attributed to: (1) diffusion impairment produced by a thickened alveolar capillary membrane, (2) ventilation to perfusion (V/Q) abnormalities, or (3) atelectasis, consolidation, and pulmonary edema. Normal intrapulmonary shunt is relatively low; <5% of the blood flow fails to make contact with functioning alveoli. A mild intrapulmonary shunt is 5 to 15%, a major intrapulmonary shunt is 15 to 30%, and a severe intrapulmonary shunt is greater than 30%. There are several methods to assess intrapulmonary shunt, each with advantages and limitations. The value of assessing intrapulmonary shunt is to identify alterations in pulmonary gas exchange, determine the degree to which the lung deviates from ideal as an oxygenator of blood, and to evaluate patient response to interventions.

PHYSIOLOGIC SHUNT.

The physiologic shunt equation (Qs/Qt) is identified as the gold standard for assessing impairment of pulmonary gas exchange in critically ill patients. 9 Qs/Qt can be measured by calculating the difference between the content of fully oxygenated pulmonary capillary (CcO₂) and arterial blood (CaO₂) divided by a difference between full CcO₂ blood and mixed venous blood (CvO₂) according to the formula in Table 2. It is the most accurate indicator of pulmonary gas exchange that is clinically available. 9,10 However, it is time consuming to obtain, complex to calculate, and costly.

Physiologic Shunt Equation

Data needed to calculate oxygen content in the Qs/Qt formula are available only from patients who have pulmonary and peripheral artery catheters in place. These data are obtained by drawing simultaneous mixed venous and arterial blood gases from these catheters. It is a complex formula to calculate, although most bedside monitoring systems in the ICU have the capability to calculate Qs/Qt once mixed venous and arterial blood gas data are available (Hb, arterial oxygen saturation [SaO₂], PaO₂, partial pressure of venous oxygen [PvO₂], and venous oxygen saturation [SvO₂]) and are entered into the bedside monitoring system.

OXYGEN TENSION DERIVED INDICIES.

Several indices to estimate Qs/Qt were developed because of the complexity of the classic shunt equation calculation, necessity for frequent sampling of arterial and mixed venous blood, and the need for a pulmonary artery catheter. 10-12 The indices are collectively referred to as "oxygen tension derived indices" of pulmonary gas exchange, and include: alveolar-arterial gradient, ratio of arterial-to-alveolar oxygen tension, and ratio of arterial oxygen tension to fraction of inspired oxygen (Table 3). All relate the driving pressure (fraction of inspired oxygen [FiO₂] or partial pressure of alveolar oxygen [PAO₂]) for diffusion into the pulmonary capillary blood, which is the mean determinant of PaO₂. There has been some debate on the use of PAO₂ versus FiO₂ in a hypoxemia index. Some argue that PAO₂ is more advantageous because it includes the effect of CO₂. 13 Others contend that this addition has little value, adds further assumptions, and may vary somewhat with permissive hypercapnea. 14

Oxygen Tension Based Indices of Pulmonary Gas Exchange

Alveolar-Arterial Gradient.

The alveolar-arterial gradient (A-a) gradient was developed based on the relationship between alveolar and arterial oxygen tension: PAO₂ equals PaO₂ when ventilation and perfusion are perfectly matched. The gradient describes the overall efficiency of oxygen uptake from alveolar gas to pulmonary capillary blood. In healthy conditions, there is generally a small difference between PAO₂ and PaO₂ because PAO₂ is approximately 100 mmHg and PaO₂ is about 95 mmHg. As the gradient between PAO₂ and PaO₂ increases, intrapulmonary shunt increases.

Table 4 illustrates the use of the A-a gradient in different clinical situations. Patient A reflects a normal healthy person breathing room air. The A-a gradient is 1.7, which represents a near-perfect match between ventilation and perfusion. Patient B reflects respiratory failure in a patient breathing room air. In this example, the A-a gradient is 31, which is a large gradient and, therefore, a large ventilation perfusion disturbance.

Clinical Application of Oxygen Tension Derived Indices

Several studies have compared the A-a gradient and Qs/Qt and have demonstrated the two measures moderately correlate in evaluating ventilation perfusion imbalance in critically ill patients (r = 0.58-0.68). 15-18 However, one study reported that 51% of the A-a gradients did not accurately reflect Qs/Qt 16; another found that in 25% of the measurements, Qs/Qt and A-a gradient varied in opposite directions. 15 Thus, while it appears A-a gradient is useful in situations where the patient is breathing room air, its use in critically ill patients is limited because it cannot differentiate the severity of different clinical situations. 15-19 Patients C and D in Table 4 illustrate this point. Patient C is receiving FiO₂^1.0 and is in respiratory failure; a large A-a gradient is noted. Patient D is receiving FiO₂ 0.4, but has a normal PaCO₂ and PaO₂, yet the A-a gradient is still large. Another limitation of the A-a gradient is that varying FiO₂ concentrations affect the measurement. 20,21 Patients E and F in Table 4 reflect that with varying FiO₂, A-a gradient increases markedly. Patient E is receiving FiO₂ 0.7 with a PaCO₂ 40 mmHg and a PaO₂ of 70 mmHg that results in an A-a gradient of 379. Patient F is receiving an FiO₂ 0.9 with PaCO₂ 40 mmHg and PaO₂ 200 mmHg and the A-a gradient is still high at 392. At increasing FiO₂, A-a gradient increases markedly, causing this index to lose clinical utility in critically ill patients. Its clinical usefulness is limited to patients with a Qs/Qt <15% and FiO₂ >0.5. 12

Arterial-Alveolar Oxygen Tension Ratio.

The arterial alveolar oxygen tension ration (a/A ratio) was developed based on the relationship between alveolar and arterial oxygen tension. This index uses the same variables as the A-a gradient: PAO₂ (PaCO₂, FiO₂) and PaO₂. As PaO₂ decreases relative to PAO₂, the ratio decreases and intrapulmonary shunt increases. A normal value is >=0.75 and a ratio <0.75 indicates pulmonary dysfunction due to ventilation perfusion abnormality, shunt, or a diffusion limitation. 22Table 4 illustrates the clinical use of a-A ratio. Patient A reflects a normal healthy person breathing room air with an a-A ratio of 0.98. This represents a near-perfect match between ventilation and perfusion. Patient B reflects respiratory failure in a patient breathing room air. Here, the a-A ratio is 0.61, which reflects a worsening intrapulmonary shunt.

There are conflicting data on the accuracy with which this index reflects Qs/Qt. Cane and colleagues examined the relationship of Qs/Qt and a-A ratio in a heterogeneous group of 75 critically ill patients (50 medical and 25 surgical ICU patients) and reported a high correlation between Qs/Qt and a-A ratio (r = -0.78). 18 Rasanen and colleagues examined the relationship between Qs/Qt and the a-A ratio in 17 critically ill patients with respiratory failure, but reported a poor correlation (r = 0.47). 12 Like A-a gradient, a-A ratio appears to be vulnerable to changes in peripheral circulation and oxygen therapy as can be seen by analyzing the data in Table 4. Patient C in respiratory failure on FiO₂ 0.40 has a low a-A ratio that reflects a significant intrapulmonary shunt. Patient D on FiO₂ 1.0 has a normal PaCO₂ and PaO₂, yet the a-A ratio is only 0.15. Again, as was noted with the A-a gradient, the a-A ratio can also not differentiate the severity of two clinical situations as noted in Patients E and F in Table 4.

PaO₂/FiO₂.

The PaO₂/FiO₂ ratio was in-troduced in an attempt to overcome the limitations of A-a gradient and a-A ratio and enable the evaluation of PaO₂ at varying FiO₂.23 A normal ratio is 300 to 500 and a value <250 reflects a clinically significant impairment of pulmonary gas exchange. 23Table 4 illustrates the clinical use of PaO₂/FiO₂. Patient A reflects a normal healthy person breathing room air with a normal PaO₂/FiO₂. Patient B, in respiratory failure, appears to have an intrapulmonary shunt as reflected by a PaO₂/FiO₂ of 238.

PaO₂ /FiO₂ and Qs/Qt are moderately to highly correlated (r -0.51 to -0.90). 10,17,18,24 Covelli and colleagues found that a PaO₂/ FiO₂ <200 correlated with a Qs/Qt >20% in critically ill patients with acute respiratory distress syndrome (ARDS). 17 In a retrospective analysis of previously published data of 16 patients with ARDS, PaO₂/FiO₂ in patients with moderate shunts (<30%) varied considerably with alteration in FiO₂. 14 When the use of the ratio was restricted to FiO₂ values >=0.5 and PaO₂ values <=100, there was little variation in the ratio for a given patient. With low values of true shunt or with substantial perfusion of alveolar units with low ventilation/perfusion ratios, PaO₂ increased to >100 and diminished the value of PaO₂/FiO₂ as an index of hypoxemia.

The American-European consensus conference definition of acute lung injury and ARDS is partially based on the PaO₂/FiO₂. A PaO₂/FiO₂ of less than 300 defines acute lung injury and PaO₂/FiO₂ <200 defines ARDS. 25 PaO₂/FiO₂ has been shown to be higher in patients with ARDS who survive 26 and significantly lower in nonsurvivors. 27 However, several studies 26,28 and a meta-analysis 29 suggest that it is an inconsistent predictor of outcome in patients with ARDS. Doyle and colleagues found no difference in mortality between patients with a PaO₂/FiO₂ ratio of 150 to 300 and those whose ratio was <150. 28 The Prostaglandin E1 Study Group found that an improvement in the PaO₂/FiO₂ ratio after day 1 of conventional therapy predicted a favorable prognosis in their control patients. 26 This improvement in oxygenation among survivors was maintained over a 7-day period. Thus, monitoring trends over time may provide more useful information than any single measurement. Despite these reports, PaO₂/ FiO₂ remains the most important clinical physiologic variable used in the diagnosis and assessment of ARDS. 30-32 In a survey of 448 ICU medical directors in the United States, respondents considered the PaO₂/ FiO₂ ratio to be the physiologic variable most important to determine the respiratory status of a patient with ARDS. 31

Assessment of Oxygen Delivery

The second component in the assessment of oxygenation is oxygen delivery (DO₂) which involves the process of transporting oxygen to cells. The major function of the cardiovascular system is to transport oxygen from the lungs to the tissues at a rate to meet cellular oxygen demands. The concept of a failure of DO₂ to meet metabolic demand requirements commonly defines shock. Under normal resting conditions, DO₂ is more than adequate to meet tissue oxygen requirements for aerobic metabolism.

DO₂ is dependent on cardiac output (CO) and the oxygen content in arterial blood (CaO₂) as expressed in the Fick Formula (Table 5). The overall transport of oxygen to cells is dependent on the quantity of blood being pumped (CO) and the quality of the blood (CaO₂). DO₂ is approximately 1000 mL/min and when indexed to body surface area, is approximately 600 mL/min/m². DO₂ may be compromised by anemia, oxygen desaturation, and a low CO, either singularly, or in combination.

Fick Formula for Oxygen Delivery

Individual cells and organs vary in their sensitivity to impaired DO₂. Cardiomyocytes, neurons, and renal tubular cells are particularly sensitive to an acute reduction in DO₂. The kidneys and liver can tolerate 15 to 20 minutes of hypoxia, skeletal muscle 60 to 90 minutes; however, in contrast, hair and nails can continue to grow for several days after death. 33 This variation in tissue tolerance to impaired DO₂ has important clinical implications. Maintenance of DO₂ to the most hypoxia-sensitive organs is crucial. Measurement of DO₂ to individual organs is an important goal for the future, but is not currently possible in the clinical setting. At present, only near infrared spectroscopy and gastric tonometry are used clinically to detect organ hypoxia.

Physical Assessment of DO₂

Direct physical assessment of oxygen delivery is difficult for a variety of reasons including that oxygen is a colorless, odorless gas. Indirect assessments of DO₂ can be made using parameters such as level of consciousness, skin color, capillary refill, and skin temperature. A diminished level of consciousness, confusion, or agitation may be manifestations of hypoxemia. Capillary refill as a marker of DO₂ is controversial. It may be useful as a marker of hypovolemia and poor myocardial function during early resuscitation in children 34; however, in elderly patients it does not correlate with objective measures of hypovolemia. 35 The poor sensitivity of capillary refill to blood loss (6% sensitivity and 93% specificity) 36 leads to the conclusion that capillary refill has no proven diagnostic value in the adult. 37 In a more recent investigation of the usefulness of the physical examination of subjective extremity skin temperature in a population of critically ill adult patients with a variety of disease processes, cool distal extremities correlated with other markers of hypoperfusion (base deficit, high lactate levels, and low mixed venous oxygen saturation). 38

Other clinical indices used in the assessment of DO₂ (such as heart rate, skin temperature, and urine output) are unreliable and slow to change, particularly in compensatory shock states, and abnormal values may only occur in the late stages of severe DO₂ impairment. 39 Even though urine output is used as an index of kidney perfusion (based on the presumed sensitivity of the kidney to intravascular volume and pressure changes), it is now widely accepted that even in the face of adequate urine output, other crucial tissue beds may be underperfused. 40 In addition, polyuria may be observed in the course of some altered perfusion states including sepsis. 41 Using these traditional clinical signs is more problematic in the aging patient population because of chronic disease states and the use of concurrent medications. Examples include the inability of some patients using cardioactive medications to mount an appropriate tachycardia from volume loss, hypertensive patients presenting with normal blood pressure as a manifestation of hypovolemia, and patients with limited cardiac reserve in shock states from relatively minor blood loss or whose cardiac function deteriorates as the result of pain and stress. 42

The presence of peripheral edema should alert the clinician about the potential for impaired DO₂ to individual cells. Tissue edema due to increased vascular permeability or excessive intravascular volume expansion may result in impaired oxygen diffusion from capillary blood to individual cells, particularly in clinical situations associated with arterial hypoxemia. In these situations, avoiding tissue edema may improve DO₂ to cells. 33

Assessment of Cardiac Output

Cardiac output (CO) is a product of stroke volume and heart rate. The role of hemodynamic monitoring is to evaluate the three components of stroke volume (preload, afterload, and contractibility) and optimize components to ensure adequate CO (See "Hemodynamic Assessment" elsewhere in this issue).

The two most commonly used techniques to measure CO are bolus thermodilution and continuous CO. The choice of method depends on the patient and the clinical situation. A detailed description of these and other techniques to measure CO is beyond the scope of this article, but has been reviewed elsewhere. 43 ICU clinicians must have a thorough understanding of the limitations, bias, precision, and risks of the method used to determine CO to avoid treatment decisions based on spurious data. An even greater challenge lies in "understanding the number" once it's generated. Tibby and Murdock suggest an approach that may assist clinicians in interpreting CO data. 44 They recommend CO should ideally be interpreted from four aspects: (1) a quantitative element, (2) a qualitative element, (3) a temporal element, and (4) as part of a global assessment of metabolic well-being. In other words, adequacy of CO should be determined by answering the following question: "Is the CO ('x' L/min.) adequate for this patient at this time?" Integration of CO into a global metabolic assessment requires an appreciation of the contribution of CO to DO₂ and an understanding of the balance between DO₂ and oxygen consumption (VO₂). This global assessment can be made by considering the following questions: (1) Is the DO₂ adequate to meet metabolic needs of the patient (both globally and regionally)? (2) Is DO₂ occurring with adequate perfusion pressure? (3) Is the patient able to use oxygen delivered? (4) If the answer is "no" to any of the above, why is this so? 44

Measurement of Oxygen Content in Arterial Blood

As Table 5 demonstrates, many measurements are required to calculate DO₂. If all measurements had zero error, the Fick method would be accurate and reproducible. However, all clinical measurements, particularly in critically ill patients, have associated systematic measurement errors (bias) and random measurement errors (precision). In the Fick equation these measurement errors are not simply added in the equation, but they are multiplied in the final equation. Even with modest individual measurement errors, large overall errors occur in the calculation of DO₂.

CaO₂ is calculated according to the formula in Table 5. Normal CaO₂ is 20 mL O₂/dL. PaO₂ contributes minimally to overall DO₂ and is frequently omitted from the calculation. Global DO₂ depends more on SaO₂ than PaO₂. Therefore, there is little extra benefit from increasing PaO₂ above 90 mmHg due to the shape of the oxyhemoglobin dissociation curve when over 90% of Hb is already saturated with oxygen. Increasing Hb through blood transfusion may seem to be an appropriate intervention to optimize CaO₂. However, blood viscosity increases with Hb >=10 g/dL 33 and can impair blood flow. Recent evidence suggests that traditionally accepted Hb concentrations for critically ill patients of 10 g/dL may be too high since an impaired outcome was observed if Hb was maintained between 7 to 9 g/dL (with the exception of patients with coronary artery disease in whom a level of 10 g/dL may be appropriate). 45

Pulse oximetry is used for continuous measurement of SaO₂ and when measured this way it is designated as SpO₂. Because desaturation is detected earlier by pulse oximetry than by clinical observation, the use of pulse oximetry is recommended for any patient at risk for hypoxemia. 46 Pulse oximetry has become a standard monitoring device in the ICU upon which therapeutic interventions are frequently made. Its uses include detection of hypoxemia, 47 reduction in the frequency of arterial blood gases, 48 titration of FiO₂, 49 and weaning from mechanical ventilation. 50 A full review of research related to pulse oximetry can be found in research-based clinical protocols on pulse oximetry. 46,51

Comparison of pulse oximetry with direct CO-oximeter measurements is reported in terms of bias (mean difference between two techniques) and precision (standard deviation between two techniques). In healthy volunteers, oximeters have a bias of <2% and a precision of <3% when Sa0₂ >=90%. 52,53 Comparable results have been reported in critically ill patients with good arterial perfusion. 54,55 However, decreased bias and precision have been reported in hemodynamically compromised patients and patients with hypoxemia, in whom accurate and reliable monitoring is of major importance.

In low perfusion states with decreased CO, the bias and precision reach unacceptable limits of agreement (>4%). 56-58 Accuracy of pulse oximeters appears to deteriorate when SaO₂ falls below 80%. In healthy individuals under hypoxemic conditions, bias of pulse oximetry ranges from -;15% to 13%, while the precision ranges from 1 to 16%. 52,59-61 In a study of 54 mechanically ventilated patients with SaO₂ >90%, the bias and precision of pulse oximetry was 1.7% and 1.2%, respectively, but when the SaO₂ was <90%, the bias and precision of pulse oximetry was 5.1 and 2.7%, respectively. 49 Elevated carboxyhemoglobin or methemoglobin levels cause inaccurate oximetry readings. 62-64 In patients with sickle cell anemia in acute crisis, the mean bias of pulse oximetry was 4.5%. 65 More recently, Sequin and colleagues 66 found that SpO₂ consistently overestimated SaO₂ and concluded that a minimum threshold SpO₂ value of 96% is more reliable to ensure an SaO₂ value >=90%.

Pulse oximeters have several technical limitations that may lead to inaccurate readings. Accuracy decreases during states of diminished blood flow through the point of attachment. 67-69 as a result of vasoactive drugs, hypotension, or hypothermia. Accuracy of finger probes is generally found to be better than performance at other sites. 70 Because the earlobe is the least vasoactive site and is least susceptible to signal loss, it may show faster response and greater accuracy during periods of vasoconstriction and hypotension. 46 Blue, green, or black nail polish causes inaccurate SpO₂ readings. 71 Motion of the probe continues to be a significant source of error and false alarms. 72-74 Nearly one-half of all false alarms in the ICU have been attributed to SpO₂ signals. 75

Many clinicians fail to appreciate the physical and technical limitations of pulse oximetry. 76 The pulse oximeter remains a valuable tool in the care of critically ill patients, but an awareness of its limitations is an important component of enhancing the quality of care. The major challenge facing pulse oximetry is whether this technology can be incorporated into diagnostic and management algorithms that can improve the efficiency of clinical management in the ICU. 77

Assessment of Oxygen Consumption and Oxygen Extraction

The third component of oxygenation is oxygen consumption (VO₂); the process by which cells use oxygen to generate energy. Ingested substrates are converted in the Kreb's Cycle to form adenosine triphosphate (ATP) through aerobic metabolism and results in the creation of 38 molecules of ATP. As a "back-up" mechanism, cells can still produce ATP with a limited oxygen supply through the process of anaerobic metabolism where carbohydrates are broken down to generate two ATP molecules. However, in addition to ATP, the byproducts pyruvate and lactate are produced. The more time spent in anaerobic metabolism, the more lactate produced. Normal VO₂ is approximately 250 mL/min. When indexed to body surface area, VO₂ is approximately 130 mL/min/m².

Direct measures of cellular VO₂ are in various developmental phases but are not yet available in the clinical setting. There are no physical assessment parameters that can be used to evaluate VO₂. Current clinical techniques for measuring VO₂ focus on global measurements that reflect the balance between DO₂ and VO₂.

Assessment of Oxygen Consumption

Indirect calorimetry is the gold standard for the determination of VO₂78 and is the recommended method to measure VO₂ in critically ill adults. 79 It uses a direct measure of VO₂ and CO₂ production. Although the use of indirect calorimetry is considered to be the most accurate method, it is time consuming, involves the use of expensive and specialized equipment, and requires trained personnel to perform it. 80 These limitations prevent its widespread acceptance as a method to measure VO₂ in critically ill patients.

VO₂ can also be measured using the Reverse Fick equation according to the formula in Table 6. Data needed to calculate VO₂ using this equation are available from patients who have pulmonary artery and peripheral arterial catheters in place to allow for obtaining simultaneous mixed venous and arterial blood gases. Like the Qs/Qt, it is time consuming to obtain, complex to calculate, and costly. Most bedside monitoring systems in the ICU have the capability to calculate it once mixed venous and arterial blood gas data are available (Hb, PaO₂, SaO₂, PvO₂, SvO₂) and entered into the bedside monitoring system.

Reverse Fick Equation to Measure VO

There are several methodological concerns about the accuracy and precision of the reverse Fick equation. It may underestimate whole body VO₂ because it does not include the VO₂ of the bronchial and thebesian circulations. 81,82 Pulmonary VO₂ is a physiologic variable of importance especially in pulmonary inflammatory conditions such as acute lung injury and pneumonia because it may be substantially increased. 83 The Fick method is episodic and may not reflect actual trends in VO₂. 84 Investigators have demonstrated there can be large, apparently spontaneous, changes in VO₂ in critically ill patients. 85,86

Studies that have compared simultaneous measurements using indirect calorimetry have reported that indirect calorimetry measurements are 8 to 27% higher than measurements made using the Fick method. 87-89 A more recent study reported a bias of 41 mL/min/m², precision of 3.95 mL/min/m², and a 95% confidence interval of 20 to 63 mL/min/m²; which they considered to be far too wide and concluded that these two methods should not be considered to be interchangeable. 90

Assessment of Oxygen Extraction

With a normal resting DO₂ of approximately 1000 mL/min and VO₂ of 250 mL/min, resting oxygen extraction is approximately 25%. The amount of oxygen consumed as a fraction of oxygen delivered defines the oxygen extraction ratio (O₂ER). The O₂ER estimates the balance between oxygen delivery and oxygen consumption and is calculated as O₂ER = VO₂/DO₂. As VO₂ increases or DO₂ decreases the O₂ER rises to maintain aerobic metabolism. An O₂ER greater than 0.35 implies an excessively high extraction of oxygen to meet metabolic demands and is often associated with shock states. 9 Tissue hypoxia, either because DO₂ is inadequate or cells do not extract and use oxygen normally, may be a contributor to organ failure in critical illness. 83

One of the most important clinical questions about oxygenation that clinicians must evaluate is whether DO₂ is adequate to meet the metabolic needs of the patient. There are no direct methods currently clinically available to assist clinicians in making this assessment. Instead, clinicians must rely on global and regional parameters. These parameters indirectly assess the balance between DO₂ and VO₂.

Global Parameters of DO₂/VO₂ Balance

The relationship between whole body DO₂ and VO₂ is illustrated in Figure 2. (It should be noted that individual organ systems have their own DO₂/VO₂ relationship). As noted in Figure 2, VO₂ remains relatively constant over a wide range of DO₂ because tissues can extract more oxygen when needed. When this occurs, venous oxygen saturation decreases. However, when DO₂ reaches a critical threshold, tissue extraction of oxygen cannot be further increased to meet VO₂. It is at this point that VO₂ becomes directly dependent on DO₂ (DO_2[Crit]) and cells convert to anaerobic metabolism. This is manifested by an increase in lactate, a more significant base deficit, a decrease in mixed venous oxygen saturation (SvO₂), and an increase in the O₂ ER. The DO_2(Crit) in humans has not been widely determined, but may be in the range of 180 to 330 mL/min/m².91,92

Global indices of DO

SERUM LACTATE.

As noted in Figure 2, a critical reduction in DO₂ results in anaerobic metabolism, which, in turn, produces a significant elevation in serum lactate levels. Serum lactate levels are considered to be the clinical gold standard as a marker of inadequate cellular oxygenation. 93-95 Normal serum lactate levels are <2 mMol/L. The magnitude and duration of elevated lactate levels maybe predictors of mortality and morbidity in some critically ill patient populations. In critically ill trauma, surgical, and burn patient populations, lactate normalization within 24 hours of admission is associated with increased survival. 96-100 Initial and highest lactate levels appear to be higher in nonsurvivors than survivors, and significantly higher in patients with multiple organ failure than without organ failure. 101 However, in a recent study 102 blood lactate levels were not affected by worsening congestive heart failure severity in status I cardiac transplant candidates.

Certain limitations and cautions must be considered when interpreting lactate levels. Serum lactate concentrations represent the balance of many complex factors that influence its production and clearance including cancer, acute alcohol intoxication, cocaine, and grand mal seizure. 103-105 Exogenous catecholamine infusions can stimulate anaerobic metabolism in skeletal muscle resulting in increased production of lactate in the absence of tissue hypoxia. 106 Use of large-volume lactated Ringer's solution (28 mMol/L) may result in transiently elevated serum lactate levels because of the lag in lactate catabolism, particularly if hepatic and renal perfusion remain poor. 107 Because the kidneys contribute to lactate removal, lactate can accumulate in the absence of tissue hypoxia in situations of impaired renal function. 108 Evidence suggests that the lung can release lactate in the presence of acute lung injury. 109 Despite these limitations, sequential use of lactate levels can establish and evaluate trends in cellular oxygenation. Any increase in lactate is a cause for concern and the etiology should be aggressively pursued.

BASE DEFICIT.

Base deficit is defined as the amount of base (mMol) required to titrate 1 L of arterial blood to a pH of 7.40 with the sample fully saturated with oxygen at 37[degrees]C and partial pressure of CO₂ at 40 mmHg. It is calculated from an arterial blood gas (ABG) analysis and is usually reported with the ABG results. A normal base deficit is -;3 mMol to + 3mMol. Positive values reflect metabolic alkalosis and negative values reflect metabolic acidosis. Base deficit can result from an accumulation of lactate associated with anaerobic metabolism and has been classified in the trauma patient population as mild (-2 to -5 mMol), moderate (-6 to -14 mMol), or severe (> -15 mMol). 110

Base deficit appears to be a sensitive measure of the degree and duration of inadequate DO₂ in the critically ill trauma patient population. Rutherford and colleagues 111 conducted a retrospective study of 3791 trauma patients and reported that a base deficit of -15 mMol within 24 hours postin-jury was a significant marker of mortality in patients <55 years of age. However, in patients ages 55 and older, a base deficit of -8mMol was a significant marker of mortality. The use of a lower threshold of base deficit in elderly trauma patients is recommended because elderly patients with significant injuries and mortality risk may not manifest a base deficit out of the normal range. 112 A more recent study found that trauma patients who had persistently high base deficit also had lower VO₂ (126 +/- 40 mL/min/m² versus 156 +/- 30 mL/min/²) than with patients with a low base deficit. 113 The investigators concluded that a persistently high arterial base deficit is associated with altered oxygen extraction and an increased risk of multiple organ failure and mortality. For these reasons, many trauma centers support the use of a normal base deficit as an appropriate endpoint of adequate DO₂94

The use of base deficit in other patient populations is unclear. In samples of critically ill surgical patients 98 and burn patients, 100 initial base deficit was a poor predictor of mortality and did not correlate with lactate levels. However, in a mixed sample of medical and surgical ICU patients, Smith and colleagues 114 found that a base deficit more negative than -4 mMol/L and a lactate >1.5 mMol/L led to a sensitivity of 80.3% and a specificity of 58.7% for mortality.

Limitations of base deficit exist and clinicians should consider this in the interpretation of these data. Administration of sodium bicarbonate, hypothermia, and hypocapnea can affect base deficit. Certain conditions that result in a metabolic acidosis, unrelated to lactic acidosis, can produce a base deficit that does not reflect DO₂/VO₂ balance. These include hyperchloremic acidosis as a result of infusions of large volumes of normal saline, 115 preexisting renal failure, acute ingestion of certain substances (ie, alcohol, cocaine, aspirin), conditions associated with chronic CO₂ retention (emphysema), and diabetic ketoacidosis. 95,116

MIXED VENOUS OXYGEN SATURATION.

Mixed venous oxygen saturation (SvO₂) represents a global balance between DO₂ and VO₂. Normal SvO₂ is 60 to 75%. If SvO₂ decreases, one of two problems exist: (1) DO₂ has decreased, or (2) VO₂ has increased. To determine which of these situations is occurring, an assessment of DO₂ must be made. If the components of DO₂ have not changed, then the source for an increase in VO₂ must be investigated.

In order to circumvent the need for a pulmonary artery catheter, oxygen saturation of central venous blood (ScvO₂) obtained from the lower superior vena cava or right atrium has been advocated as a surrogate. Studies evaluating the impact of this measure on the management of critically ill trauma patients have varied and its use is the subject of ongoing debate. 122 A prospective study of 40 critically ill trauma patients found that ScvO₂ did not correlate with base deficit or lactate concentration. 123 Others have demonstrated that ScvO₂ parallels elevations in lactate and that it can be of value in detecting tissue hypoxia not confirmed by vital signs in patients who are critically ill. 124,125

Very few studies have examined the agreement between ScvO₂ and SvO₂ and the results are conflicting. Ladakis and colleagues 126 examined the mean ScvO₂ and SvO₂ values in 61 mechanically ventilated patients and reported the mean values were similar (mean SvO₂ 68.2% and ScvO₂ 69.4%); however, comparing mean values between methods does not answer if the methods agree. They did report similar precision between the two methods (SvO₂ 1.2%; ScvO₂ 1.1%). In a more recent study in 32 critically ill medical surgical patients, ScvO₂ averaged 7 +/- 4% higher than SvO₂127 More studies are needed to more fully understand the implications and limitations of the use of ScvO₂ in the management of critically ill patients.

MIXED VENOUS OXYGEN SATURATION: LACTATE RATIO.

Preliminary data are available to indicate that using SvO₂ (or ScvO₂) lactate ratio may be valuable in recognizing DO₂/VO₂ imbalances. 128 For example, when a high SvO₂ or ScvO₂ is coupled with an elevated lactate level, then inadequate cellular DO₂ may be present. More information is needed to determine if this ratio is clinically relevant and meaningful.

Organ Specific Monitoring of DO₂/VO₂

Global indices of DO₂/VO₂ balance lack the sensitivity necessary to be early warning signals, particularly in shock states when maldistribution of circulating volume occurs and not all organs become hypoxic all at once. The optimal tissue bed to monitor for DO₂ impairment is unclear. Access to various tissues to make monitoring useful is an obvious consideration.

GASTRIC TONOMETRY.

Splanchnic DO₂ has been quantified using gastric tonometry which is based on the knowledge that when DO₂ to the stomach decreases, anaerobic metabolism in gastric cells produces the byproducts of excess hydrogen ions, lactate, and CO₂. As gastric CO₂ levels increase, hypoperfusion may be present. Currently available methods of monitoring regional CO₂ include gastric tonometry and sublingual capnography. An extensive review of these methods is beyond the scope of this article, but can be found elsewhere. 129,130 Gastric tonometry is useful, particularly when used to monitor the difference between gastric and arterial CO₂. In normal physiologic states, this difference is usually <10 mmHg. A widening gap indicates compromised blood flow to the splanchnic bed and a gap exceeding 20 mmHg requires aggressive treatment. 129,131 Carbon dioxide is measurable in other tissue beds and using gastric tonometry principles, it has been extrapolated to other parts of the gastrointestinal system. Sublingual capnography was developed to overcome some of the limitations in gastric tonometry. Sublingual capnography is noninvasive and portable, and data are available within minutes. It correlates well with gastric PCO₂. 132 A full review of this technology can be found elsewhere. 130

Summary

Oxygenation is dependent upon three physiologic processes: pulmonary gas exchange, oxygen delivery, and oxygen consumption. An accurate and thorough assessment of oxygenation must include an evaluation of each of these three processes. Diagnostic tools available to ICU clinicians to monitor pulmonary gas exchange, oxygen delivery, and oxygen consumption presented in this review are summarized in Figure 3. It is imperative that ICU clinicians recognize and appreciate the implications and limitations of these diagnostic tools to avoid making treatment decisions based on spurious inaccurate data. Accurate assessment and treatment of disturbances in oxygenation are crucial to optimize patient outcomes.

Clinical measures of pulmonary gas exchange, oxygen delivery, and oxygen consumption. Adapted with permission from Taylor CR, Weibel ER. Design of the mammalian respiratory system. Respiration Physiology. 1981;41(p2). Copyright 1981 by Elsevier Science. CO

References