Acidosis and alkalosis—the underlying processes that contribute to the pH status—may be pathogenic or compensatory. The normal ranges for arterial pH and PaCO₂ are 7.38 to 7.44 and 35 to 45 mm Hg, respectively. A pH that exceeds 7.45 indicates alkalemia, generated by bicarbonate retention, hyperventilation relative to metabolic need, or both. A pH > 7.35 indicates acidemia, caused by metabolic or renal depletion of bicarbonate, hypoventilation relative to metabolic need, or both. Venous blood varies in composition depending on the delivery/consumption characteristics of the sampled tissue bed. Mixed-venous gases, representing admixture from tissues throughout the body, generally have a PCO₂ that is 4 to 8 mm Hg higher and a pH that is 0.05 to 0.10 units lower than arterial gases. Occasionally, these arteriovenous discrepancies can be much wider, and in such cases, the venous value may more accurately reflect the acid-base status of vital tissue beds.

The ABG Provides the pH and allows the clinician to determine the relative contributions. Because compensation is never complete, the dominant underlying mechanism—acidosis or alkalosis—is suggested by the pH. Does a blood gas demonstrating a PaCO₂ of 32 mm Hg and an HCO of 16 mEq/L indicate respiratory alkalosis with renal compensation, or metabolic acidosis with respiratory compensation? A major clue is provided by the pH—acidemia would suggest that the fundamental problem is metabolic. Failure of the PaCO₂ to fall to the expected level would suggest a superimposed problem with ventilatory drive or ventilatory pump. Classically, “Winters’ Formula” (following) predicts the respiratory compensation for an uncomplicated metabolic acidosis. Intensive care unit (ICU) clinicians must be alert for “triple” acid-base disorders in which two metabolic derangements with opposing influences on pH are in play. The anion gap calculation usually provides the key to appropriate interpretation.

Chronicity of the process can be judged by comparing the observed values of pH, bicarbonate, and base excess with those expected for acute hypercapnia or hypocapnia (see following). To place such information into proper perspective for diagnosis and management decisions, the clinician must take account of the clinical backdrop and examine the serum electrolytes and albumin for evidence of renal insufficiency, renal tubular dysfunction, and an anion gap that indicates the presence of such noncarbonic (metabolic) acids as lactate, ketoacids, and salicylate.

The anion gap is the difference between the serum sodium concentration and the sum of chloride and bicarbonate ions. Normally, the gap is 13 mEq or less—a reflection of the sulfates, phosphates,
and other unmeasured negatively charged ions that correspond to kidney-excreted “mineral” acids. Because of the anionic nature of serum proteins, the calculated gap should be increased approximately by 2 to 2.5 mEq/L for each g/dL of hypoalbuminemia. Another useful bedside computation is the “Δ/Δ,” the ratio of the anion gap to the bicarbonate gap. As lactic acidosis develops, the anion gap increases relatively more than the HCO falls because HCO has a wider volume of distribution. The ratio varies but averages approximately 1.5 in moderate lactic acidosis—less if the acuity is extreme and more as the severity worsens. Values less than 1 should prompt consideration of another source of H⁺ excess, whereas a value greater than 2 suggests a concurrent metabolic alkalosis.

As already noted, primary metabolic disturbances are incompletely compensated by changes in ventilation, and primary respiratory disturbances are partially offset by renal excretion or retention of bicarbonate. Knowledge of these expected compensations allows a judgment to be made regarding the nature and chronicity of the underlying processes.

A systematic approach to blood gas evaluation incorporates the elements of the foregoing discussion. The first priority is to verify the technical validity of the sample—errors in sampling, sample processing, analysis, and transcription occur commonly. (Is the sample characteristic of a venous rather than the intended arterial specimen? Are the pH, PaCO₂, and [HCO] internally consistent? Does the [HCO] correlate with the [HCO] directly analyzed from venous blood? Is the PaO₂ reported physically possible given the FiO₂ administered?)

Although there is no best method for interpreting a technically valid report, one logical approach has the following steps:

Transcutaneous photometric oximetry is useful for monitoring patients with marginal or fluctuating oxygen exchange. For patients supported by mechanical ventilation, transcutaneous oximetry continuously measures SaO₂, enabling rapid adjustment of FiO₂, mean airway pressure, and positive end-expiratory pressure (PEEP) and warning of arterial desaturation during weaning, sleeping, or changes of body position. As a general rule, trends in oximetry values are of greater significance than the absolute value of saturation, at least over the clinical saturation range usually encountered.

Lightweight probes direct filtered light of several specific wavelengths onto the surface of the digit, nasal bridge, or earlobe. The relative absorption of these spectrophotometric beams as they pass through the tissue (which differs for O₂ saturated and desaturated blood) is converted into the appropriate saturation value by computer-stored algorithms. Phasic variations separate the incoming arterial component from venous and background absorption. Pulse oximetry probes do not require tissue heating because phasic changes in blood volume and optical density cue the instrument to the arterial component of the blood contained in the vascular bed. Most units also display pulse rate, and many display a simulated arterial waveform or some other visual indicator of pulse intensity. A tracing whose waveform baseline varies dramatically with ventilation strongly suggests variation of stroke output synchronous with the respiratory cycle—a condition typical for gas trapping and auto-PEEP (AP).

Oxygen saturation is not the only useful parameter that can be estimated noninvasively by spectrophotometry. Quite recently, widening of the optical wavelength spectrum and advanced signal processing algorithms have been developed to enable noninvasive monitoring of hemoglobin, carboxyhemoglobin, and methemoglobin concentrations. As the reliability of these measures is improved and validated for use in critical care, such technology promises to add considerably to early event detection.

The pulse rate should be “correlated” to the electrocardiogram (ECG) measured rate to assess signal quality. Good correlation does not ensure accuracy, but poor correlation of the heart rate displayed on the ECG and pulse oximeter calls the reported saturation value into question. With a good pulse signal, currently available instruments are quite accurate in their upper range (i.e., saturations > 80%) but become less reliable as the patient desaturates or perfusion deteriorates. Even when accurate, pulse oximetry presents a signal-averaged, and therefore delayed, report. Poor perfusion and/or
probe contact are the primary causes of an erroneous signal, and placement on a different digit, nose bridge, or earlobe may improve reliability.

Routine ABG reports of saturation are calculated from the measured PaO₂ and pH. A direct determination of arterial blood saturation (preferably by co-oximetry) is the most definitive check. Motion artifact often is an important problem for patients who are not immobilized. Because detection of the “arterial” segment of the cycle depends on small phasic changes in the tissue volume, large-amplitude vibrations of other kinds that are unassociated with arterial pulsation can confuse the sampling algorithm —especially when the frequencies of the rhythmic vibration approximate the patient’s own heart rate. When a patient has a rhythmic tremor (Parkinson disease, anxiety, agitation, seizures, essential tremor, shivering, etc.), tissue volumes can vary phasically in such a way as to invalidate the oximeter’s output, which trends toward the default value. In the absence of any detected discrimination between the “baseline” and “arterial” absorption differences, many devices default to a recorded display of 85% to 88%. Rarely, when arterial perfusion is poor and venous pulsations are vigorous, the recorded value may be erroneously depressed.

Anemia and jaundice do not routinely affect the accuracy of pulse oximetry. Pulse oximetry values tend to be misleadingly high in some black patients, but by no means in all. (The existing literature conflicts on this point.) Carboxyhemoglobin and methemoglobin can produce falsely high saturation values, and specific nail polishes (particularly blue, green, or black) interfere with light transmission and absorbance, as do certain blood-borne dyes, such as indocyanine green and methylene blue. These tend to artifactually reduce the O₂ saturation reported.

Many practitioners do not fully understand the oxyhemoglobin dissociation relationship (Fig. 5-1) or the value and limitations of transmission oximetry. Over the clinically relevant range, the oxyhemoglobin dissociation curve is highly nonlinear, so a drop of a few percentage points in SaO₂ over the 95% to 100% interval reflects a much larger change in PaO₂ than does a similar decrement over the 80% to 85% interval. Pulse oximeters record the relative absorption of light by oxyhemoglobin and deoxyhemoglobin. Therefore, for a fixed value of viable hemoglobin, the saturation parallels its relative O₂ content, but a high saturation guarantees neither its total O₂ content nor the adequacy of tissue O₂ delivery. For example, a patient may have a “full” SaO₂ after inhaling a high concentration of carbon monoxide, and yet directly measuring arterial oxygen content per deciliter of blood (by co-oximetry) may demonstrate profound arterial O₂ depletion. Moreover, a patient in circulatory shock may maintain a perfectly normal SaO₂ despite serious O₂ privation. Cyanide blocks the uptake of oxygen by the tissues, so O₂ consumption is low even as arterial and mixed-venous saturations are normal or increased. Arterial oxygen saturation also bears no direct relationship to the adequacy of ventilation; a patient breathing a high-inspired concentration of oxygen will maintain a nearly normal SaO₂ for extended periods in the face of a full respiratory arrest.

Other gas-measuring techniques (e.g., transcutaneous and transconjunctival measurements of O₂ and CO₂) have been used widely in neonatology to monitor tissue gas tensions, but traditional monitors have been generally less helpful for adults. These transcutaneous techniques require frequent calibration, excellent skin and electrode preparation to ensure gas transfer to the skin surface, and regular site changes to avoid burning the warmed patch of skin they monitor. More importantly, they are profoundly affected by inadequacy of perfusion
and therefore track arterial gas tensions unreliably during many critical illnesses. Alert patients tolerate conjunctival probes poorly. Newer tissue oxygen sensors appear to hold considerably more promise.

Several methods now in active development have been used largely in a research setting but show considerable promise for clinical monitoring of microcirculatory function during circulatory failure. These include CO₂ measurements for sublingual, buccal, and subcutaneous microcirculatory CO₂ levels, as well as absorbance, reflectance, and near infrared spectroscopy (NIRS) for measuring microcirculatory hemoglobin saturation. NIRS can probe to considerable depth and has already found clinical application in the assessment of cerebrocortical viability. Orthogonal polarization spectral (OPS) imaging and sidestream dark field technology allow microscopic visualization of the deeper lying microcirculation and the flow of red blood cells in the microcirculation. Sublingual capnography combined with OPS imaging has been used to investigate the relationship between the microcirculation and metabolic status during resuscitation. Combinations of these technologies, which look at different functional compartments of regional microcirculations, can integratively probe the distributive alterations of oxygen transport during sepsis, septic shock, and therapy that are not provided by conventional monitoring of systemic hemodynamic and oxygen-derived variables.

Controversy has surrounded the concept of supply dependency of O₂ consumption for patients having sustained trauma, massive surgery, or sepsis. Failure to provide sufficient O₂ delivery may result in anaerobiosis, multisystem organ failure, and an adverse or fatal outcome. Moreover, it generally is agreed that prognosis in these conditions is somewhat better for critically ill patients in whom higher O₂ delivery is manifest. By inference, it has been suggested that in these settings, supranormal O₂ delivery is needed to satisfy the O₂ demands of certain vital organs. Prompt and vigorous resuscitation must be carried out, as patients who do not spontaneously generate sufficient O₂ delivery or who cannot extract O₂ effectively have a worse prognosis than other patients undergoing the same stress who do. Yet, it is highly questionable whether attempts to sustain O₂ delivery at supranormal values are well advised. Some data even suggest potential harm. Specific subgroups of surgical patients could, in fact, benefit; there are no tightly controlled data available to settle this question in either direction. Patients having sustained massive trauma or extensive surgery may represent a fundamentally different physiologic problem and respond more favorably than patients with medical crises such as sepsis. It now seems clear that targeting supranormal values for oxygen delivery confers no routine benefit for patients in the latter category. Without better evidence, therefore, maximizing [V with dot above]O₂ cannot be accepted as the primary target variable for circulatory support.

To judge the efficiency of pulmonary gas exchange, mean alveolar oxygen tension (PAO₂) must first be computed. The ideal PAO₂ is obtained from the modified alveolar gas equation:

PAO₂ = PIO₂ − (PaCO₂/R) + [(PaCO₂ × FiO₂ × (1 − R)/R)]

Here R is the respiratory exchange ratio and PIO₂ is the inspired oxygen tension adjusted for FiO₂ and water vapor pressure at body temperature (47 mm Hg at 37°C).

PIO₂ = (barometric pressure − 47) × FiO₂

Under steady-state conditions, R normally varies from approximately 0.7 to 1.0, depending on the mix of metabolic fuels. When the same patient is
monitored over time, R generally is assumed to be 0.8 or neglected entirely. Under most clinical conditions, the alveolar gas equation can be simplified to:

PAO₂ = PIO₂ − (1.25 × PaCO₂)

For example, at sea level with a normally ventilated patient breathing room air:

PAO₂ = 0.21 × (760 − 47) − 1.25 × (PaCO₂) = 150 − (1.25 × 40) [all equal to] 100 mm Hg

The difference between alveolar and arterial oxygen tensions, P(A-a)O₂, takes account of alveolar CO₂ tension and therefore eliminates hypercapnia from consideration as the sole cause of hypoxemia. However, although useful, a single value of P(A-a)O₂ does not characterize the efficiency of gas exchange across all FiO₂s—even in normal subjects. The P(A-a)O₂ in a young normal subject ranges from approximately 10 mm Hg (on room air) to approximately 100 mm Hg (on an FiO₂ of 1.0). (Breathing room air, the upper limit of normal approximates age/4+ 4 mm Hg.) Moreover, PAO₂ changes nonlinearly with respect to FiO₂ as the extent of [V with dot above]/ mismatch increases. When the [V with dot above]/ abnormality is severe and abnormally distributed among gas exchanging units, the PAO₂ may vary little with FiO₂ until high fractions of inspired oxygen are given (Fig. 5-2). Finally, the P(A-a)O₂ may be influenced by the fluctuations in venous oxygen content.

Under normal circumstances, fluctuations in mixedvenous O₂ saturation (SvO₂) do not contribute significantly to hypoxemia. However, as ventilation/perfusion inequality or shunting develops, the O₂ content of mixed-venous blood (CvO₂) exerts an increasingly important effect on SaO₂. Measuring SvO₂ with a fiberoptic Swan-Ganz catheter enable venous admixture () to be computed with relative ease. In the steady state:

= (CAO₂ − CaO₂)/(CAO₂ − C[V with bar above]O₂)

where the oxygen content of alveolar capillary blood (CAO₂), arterial blood (CaO₂), or mixedvenous blood (C[V with bar above]O₂), expressed in mL of O₂ per 100 mL of blood, equal the sum:

[0.003 × PO₂] + [0.0138 × (SO₂ ×Hgb)]

(In the latter equation, PO₂ [mm Hg] and SO₂ [%] refer to the oxygen tension and saturation of blood at the respective sites. Hemoglobin [Hgb] is expressed in gm/dL.) Like P(A-a)O₂, is also influenced by variations in [V with dot above]/ mismatching and by fluctuations in SvO₂ and FiO₂. If is abnormally high but all alveoli are patent, calculated admixture will diminish toward the normal physiologic value (approx. 5%) as FiO₂ increases. Conversely, if the abnormality results from blood bypassing the patent alveoli through intrapulmonary communications or through an intracardiac defect, there will be no change in as FiO₂ increases (“true” shunt).

Several pragmatic approaches have been taken to simplify bedside assessment of O₂ exchange efficiency. The first is to quantitate P(A-a)O₂ during the administration of pure O₂. After a suitable wash-in time (5 to 15 min, depending on the severity of the disease), pure shunt accounts for the entire P(A-a)O₂. Furthermore, if hemoglobin is fully saturated with O₂, dividing the P(A-a)O₂ by 20 approximates the shunt percentage (at FiO₂ = 1.0). As pure O₂ replaces alveolar nitrogen, some patent but poorly ventilated units may collapse— the process of “absorption atelectasis.” Moreover, because shunt percentage is affected by changes in CO and mixed-venous O₂ saturation, these simplified measures may give a misleading impression of changes within the lung itself. Whatever its shortcomings, determining the shunt fraction is worthwhile because it alerts the clinician to consider nonparenchymal causes of hypoxemia (e.g., arteriovenous malformation, intracardiac right-to-left shunting). Furthermore, because PaO₂ shows little response to variations in FiO₂ at true shunt fractions greater than 25%, the clinician may be encouraged to reduce toxic and marginally effective concentrations of oxygen.

The PaO₂/FiO₂ (or “P/F”) ratio is a convenient and widely used bedside index of oxygen exchange that attempts to adjust for fluctuating FiO_2. Although simple to calculate, this ratio is affected by changes in SvO₂ and does not remain equally sensitive across the entire range of FiO₂—especially when shunt is the major cause for admixture. Another easily calculated index of oxygen exchange properties, the PaO₂/PAO₂ (or “a/A”) ratio, offers similar advantages and disadvantages as FiO₂ is varied. Like the P/F ratio, it is a useful bedside index that does not require blood sampling from the central circulation but loses reliability in proportion to the degree of shunting. Furthermore, in common with all measures that calculate an “ideal” PAO₂, even the a/A ratio can be misleading when fluctuations occur in the primary determinants of SvO₂ (hemoglobin and the balance between oxygen consumption and delivery).

None of the indices discussed thus far account for changes in the functional status of the lung that result from alterations in PEEP, AP, or other techniques for adjusting average lung volume (e.g., inverse ratio ventilation, lateral positioning, or prone positioning). If the objective is to categorize the severity of disease or to track the true O₂ exchanging status of the lung in the face of such interventions, the P/F ratio falls short. The oxygenation index, PaO₂/(FiO₂ × mean P_AW), that takes the effects of PEEP and inspiratory time fraction into account has gained widespread popularity in neonatal and pediatric practice but has yet to catch hold in adult critical care. Although preferable, this index, too, is imperfect; mean airway pressure and FiO₂ bear complex and alinear relationships to PaO₂ when considered across their entire ranges.

The physiologic dead space (V_D) refers to the “wasted” portion of the tidal breath that fails to participate in CO₂ exchange. A breath can fail to accomplish CO₂ elimination either because fresh (CO₂-free) gas is not brought to the alveoli or because fresh gas fails to contact systemic venous blood. Thus, tidal ventilation is wasted whenever CO₂-laden gas is recycled to the alveoli with the next tidal breath. Alternatively, a portion of the tidal volume is wasted if fresh gas distributes to inadequately perfused alveoli, so CO₂-poor gas is exhausted during exhalation (Fig. 5-4). If this concept is understood, then it becomes clear why V_D cannot be considered accurately as a composite of physical volumes. Nonetheless, wasted ventilation traditionally is characterized as the sum of the “anatomic” (or “series”) dead space and the “alveolar” dead space. Because the airways fill with CO₂-containing alveolar gas at the end of the tidal breath, the physical volume of the airways corresponds rather closely to their contribution to
wasted ventilation (the “anatomic” dead space)— provided mixed alveolar gas is similar in composition to the gas within a well-perfused alveolus. This is almost true for a quietly breathing normal subject, in whom the alveolar dead space (poorly perfused alveolar volume) is negligible. When the parenchyma is well aerated and well perfused, the anatomic dead space is relatively fixed at approximately 1 mL/lb of body weight. Quite the opposite is true for patients with most lung diseases, in whom alveolar dead space predominates. Here, the lung is composed of well and poorly perfused units, so the mixed alveolar gas within the airways at end-exhalation has a CO₂ concentration lower than that of pulmonary arterial blood. Although V_D may increase dramatically, the contribution of stale airway gas to V_D is much less important because less amount of airway CO₂ is recycled to the alveoli.