Chapter 50: Acid–Base Management

Nearly all biochemical reactions in the body are dependent on maintenance of a physiological hydrogen ion concentration, and alterations beyond normal hydrogen ion concentration are associated with widespread organ dysfunction. This regulation—usually referred to as acid–base balance—is of prime importance in critical illness. Changes in ventilation and perfusion, as well as infusion of electrolyte-containing solutions, are common during anesthesia and can rapidly alter acid–base balance.

Our understanding of acid–base balance is evolving. In the past, we focused on the concentration of hydrogen ions [H⁺], CO₂ balance, and the base excess/deficit. We now understand that the strong ion difference (SID), PCO₂, and total weak acid concentration (A_TOT) best explain acid–base balance in physiological systems.

This chapter examines acid–base physiology and the perioperative care implications of common disturbances. Clinical measurements of blood gases and their interpretation are reviewed.

ACID–BASE CHEMISTRY

Hydrogen Ion Concentration & pH

In any aqueous solution, water molecules reversibly dissociate into hydrogen and hydroxide ions:

This process is described by the dissociation constant, K_W:

The concentration of water is omitted from the denominator of this expression because it does not vary appreciably and is already included in the constant. Therefore, given [H⁺] or [OH⁻], the concentration of the other ion can be readily calculated.

Example: If [H⁺] = 10^–8 nEq/L, then [OH⁻] = 10^–14 ÷ 10^–8 = 10^–6 nEq/L.

Arterial [H⁺] is normally 40 nEq/L, or 40 × 10^–9 mol/L. Hydrogen ion concentration is more commonly expressed as pH, which is defined as the negative logarithm (base 10) of [H⁺] (Figure 50–1). Normal arterial pH is therefore –log (40 × 10^–9) = 7.40. Hydrogen ion concentrations between 16 and 160 nEq/L (pH 6.8–7.8) are compatible with life.

Like most dissociation constants, K_W is affected by changes in temperature. Thus, the electroneutrality point for water occurs at a pH of 7.0 at 25°C, but at about a pH of 6.8 at 37°C; temperature-related changes may be important during hypothermia (see Chapter 22).

Because physiological fluids are complex aqueous solutions, SID, the PCO₂, and A_TOT are other factors that affect the dissociation of water into H⁺ and OH⁻.

Acids & Bases

An acid is usually defined as a chemical species that can act as a proton (H⁺) donor, whereas a base is a species that can act as a proton acceptor (Brönsted–Lowry definitions). In physiological solutions, it is probably better to use Arrhenius’ definitions: An acid is a compound that contains hydrogen and reacts with water to form hydrogen ions. A base is a compound that produces hydroxide ions in water. Using these definitions, the SID becomes important, as other ions in solutions (cations and anions) will affect the dissociation constant for water, and, therefore, the hydrogen ion concentration. A strong acid is a substance that readily and almost irreversibly gives up an H⁺ and increases [H⁺], whereas a strong base avidly binds H⁺ and decreases [H⁺]. In contrast, weak acids reversibly donate H⁺, whereas weak bases reversibly bind H⁺; both weak acids and bases tend to have less of an effect on [H⁺] (for a given concentration of the parent compound) than do strong acids and bases. Biological compounds are either weak acids or weak bases.

For a solution containing the weak acid HA, where

a dissociation constant, K, can be defined as follows:

The negative logarithmic form of the latter equation is called the Henderson–Hasselbalch equation:

From this equation, it is apparent that the pH of this solution is related to the ratio of the dissociated anion to the undissociated acid.

This approach works well with pure water: The concentration of [H⁺] must equal [OH⁻]. But physiological solutions are far more complex. Even in such a complex solution, the [H⁺] can be predicted using three variables: the SID, the PCO₂, and A_TOT.

Strong Ion Difference

The SID is the sum of all the strong, completely or almost completely dissociated cations (Na⁺, K⁺, Ca²⁺, Mg²⁺) minus the strong anions (Cl⁻, lactate⁻, etc; Figure 50–2). Although we can calculate the SID, because the laws of electroneutrality must be observed, if there is a SID, other unmeasured ions must be present. PCO₂ is an independent variable, assuming ventilation is ongoing. The conjugate base of HA is A⁻ and is composed mostly of phosphates and proteins that do not change independent of the other two variables. A⁻ plus AH is an independent variable because its value is not determined by any other variable. Note that [H⁺] is not a strong ion (water does not completely dissociate), but it can, does, and must change in response to any change in SID, PCO₂, or A_TOT to comply with the laws of electroneutrality and conservation of mass. Hydrogen ions are created or consumed based on changes in the dissociation of water.

Conjugate Pairs & Buffers

As previously discussed, when the weak acid HA is in solution, HA can act as an acid by donating an H⁺, and A⁻ can act as a base by taking up H⁺. A⁻ is therefore often referred to as the conjugate base of HA. A similar concept can be applied for weak bases. Consider the weak base B, where

BH⁺ is therefore the conjugate acid of B.

A buffer is a solution that contains a weak acid and its conjugate base or a weak base and its conjugate acid (conjugate pairs). Buffers minimize any change in [H⁺] by readily accepting or giving up hydrogen ions. It is readily apparent that buffers are most efficient in minimizing changes in the [H⁺] of a solution (ie, [A⁻] = [HA]) when pH = pK. Moreover, the conjugate pair must be present in significant quantities in solution to act as an effective buffer.

CLINICAL DISORDERS

A clear understanding of acid–base disorders and compensatory physiological responses requires precise terminology (Table 50–1). The suffix “-osis” is used here to denote any pathological process that alters arterial pH. Thus, any disorder that tends to reduce pH to a less than normal value is an acidosis, whereas one tending to increase pH is termed an alkalosis. If the disorder primarily affects [HCO₃⁻], it is termed metabolic. If the disorder primarily affects PaCO₂, it is termed respiratory. Secondary compensatory responses (discussed in the next section) should be referred to as just that and not as an “-osis.” For example, one might refer to a metabolic acidosis with respiratory compensation.

When only one pathological process occurs by itself, the acid–base disorder is considered to be simple (Figure 50–3). The presence of two or more primary processes indicates a mixed acid–base disorder.

The suffix “-emia” is used to denote the net effect of all primary processes and compensatory physiological responses (described next) on arterial blood pH. Because arterial blood pH is normally between 7.35 and 7.45 in adults, the term acidemia signifies a pH less than 7.35, whereas alkalemia signifies a pH greater than 7.45.

Physiological responses to changes in [H⁺] are characterized by three phases: (1) immediate chemical buffering, (2) respiratory compensation (whenever possible), and (3) a slower, but more effective, renal compensatory response that may nearly normalize arterial pH even if the underlying pathological process remains present.

BODY BUFFERS

Physiologically important buffers in humans include bicarbonate (H₂CO₃/HCO₃⁻), hemoglobin (HbH/Hb⁻), other intracellular proteins (PrH/Pr⁻), phosphates (H₂PO₄⁻/HPO₄^2–), and ammonia (NH₃/NH₄⁺). The effectiveness of these buffers in the various fluid compartments is related to their concentration. Bicarbonate is the most important buffer in the extracellular fluid compartment. Hemoglobin, though restricted inside red blood cells, also functions as an important buffer in blood. Other proteins probably play a major role in buffering the intracellular fluid compartment. Phosphate and ammonium ions are important urinary buffers.

Buffering of the extracellular compartment can also be accomplished by the exchange of extracellular H⁺ for Na⁺ and Ca²⁺ ions from bone and by the exchange of extracellular H⁺ for intracellular K⁺. Acid loads can demineralize bone and release alkaline compounds (CaCO₃ and CaHPO₄). Alkaline loads (NaHCO₃) increase the deposition of carbonate in bone.

Buffering by plasma bicarbonate is almost immediate, whereas that due to interstitial bicarbonate requires 15 to 20 min. In contrast, buffering by intracellular proteins and bone is slower (2–4 h). Up to 50% to 60% of acid loads may ultimately be buffered by bone and intracellular buffers.

The Bicarbonate Buffer

Although in the strictest sense, the bicarbonate buffer consists of H₂CO₃ and HCO₃⁻, CO₂ tension (PCO₂) may be substituted for H₂CO₃ because:

This hydration of CO₂ is catalyzed by carbonic anhydrase. If adjustments are made in the dissociation constant for the bicarbonate buffer and if the solubility coefficient for CO₂ (0.03 mEq/L) is taken into consideration, the Henderson–Hasselbalch equation for bicarbonate can be written as follows:

where pK′ = 6.1.

Note that its pK′ is well removed from the normal arterial pH of 7.40, which means that bicarbonate would not be expected to be an efficient extracellular buffer (see earlier discussion). The bicarbonate system is, however, important for two reasons: (1) Bicarbonate (HCO₃⁻) is present in relatively high concentrations in extracellular fluid, and (2) more importantly, PaCO₂ and plasma [HCO₃⁻] are closely regulated by the lungs and the kidneys, respectively. The ability of these two organs to alter the [HCO₃⁻]/PaCO₂ ratio allows them to exert important influences on arterial pH.

A simplified and more practical derivation of the Henderson–Hasselbalch equation for the bicarbonate buffer is as follows:

This equation is very useful clinically because pH can be readily converted to [H⁺] (Table 50–2). Note that below 7.40, [H⁺] increases 1.25 nEq/L for each 0.01 decrease in pH; above 7.40, [H⁺] decreases 0.8 nEq/L for each 0.01 increase in pH.

Example: If arterial pH = 7.28 and PaCO₂ = 24 mm Hg, what should the plasma [HCO₃⁻] be?

Therefore,

It should be emphasized that the bicarbonate buffer is effective against metabolic but not respiratory acid–base disturbances. If 3 mEq/L of a strong nonvolatile acid, such as HCl, is added to extracellular fluid, the following reaction takes place:

Note that HCO₃⁻ reacts with H⁺ to produce CO₂. Moreover, the CO₂ generated is normally eliminated by the lungs such that PaCO₂ does not change. Consequently, [H⁺] = 24 × 40 ÷ 21 = 45.7 nEq/L, and pH = 7.34. Furthermore, the decrease in [HCO₃⁻] reflects the amount of nonvolatile acid added.

In contrast, an increase in CO₂ tension (volatile acid) has a minimal effect on [HCO₃]. If, for example, PaCO₂ increases from 40 to 80 mm Hg, the dissolved CO₂ increases only from 1.2 mEq/L to 2.2 mEq/L. Moreover, the equilibrium constant for the hydration of CO₂ is such that an increase of this magnitude minimally drives the reaction to the left:

If the valid assumption is made that [HCO₃⁻] does not appreciably change, then

[H⁺] therefore increases by 40 nEq/L, and because HCO₃⁻ is produced in a 1:1 ratio with H⁺, [HCO₃⁻] also increases by 40 nEq/L. Thus, extracellular [HCO₃⁻] increases negligibly, from 24 mEq/L to 24.000040 mEq/L. Therefore, the bicarbonate buffer is not effective against increases in PaCO₂, and changes in [HCO₃⁻] do not reflect the severity of a respiratory acidosis.

Hemoglobin as a Buffer

Hemoglobin is rich in histidine, which is an effective buffer from pH 5.7 to 7.7 (pK_a 6.8). Hemoglobin is the most important noncarbonic buffer in intravascular fluid. Simplistically, hemoglobin may be thought of as existing in red blood cells in equilibrium as a weak acid (HHb) and a potassium salt (KHb). In contrast to the bicarbonate buffer, hemoglobin is capable of buffering both carbonic (CO₂) and noncarbonic (nonvolatile) acids:

RESPIRATORY COMPENSATION

Changes in alveolar ventilation responsible for the respiratory compensation of PaCO₂ are mediated by chemoreceptors within the brainstem and the carotid and aortic bodies (see Chapter 23). These receptors respond to changes in cerebrospinal spinal fluid pH. Minute ventilation increases 1 to 4 L/min for every (acute) 1 mm Hg increase in PaCO₂. In fact, the lungs are responsible for eliminating the approximately 15 mEq of CO₂ produced every day as a byproduct of carbohydrate and fat metabolism. Respiratory compensatory responses are also important in defending against marked changes in pH during metabolic disturbances.

Respiratory Compensation During Metabolic Acidosis

Decreases in arterial blood pH stimulate medullary respiratory centers. The resulting increase in alveolar ventilation lowers PaCO₂ and tends to restore arterial pH toward normal. The respiratory response to lower the PaCO₂ occurs rapidly but may not reach a predictably steady state until 12 to 24 h; pH is never completely restored to normal. PaCO₂ normally decreases 1 to 1.5 mm Hg below 40 mm Hg for every 1 mEq/L decrease in plasma [HCO₃⁻].

Respiratory Compensation During Metabolic Alkalosis

Increases in arterial blood pH depress respiratory centers. The resulting alveolar hypoventilation tends to elevate PaCO₂ and restore arterial pH toward normal. The respiratory response to metabolic alkalosis is generally less predictable than the respiratory response to metabolic acidosis. Hypoxemia, as a result of progressive hypoventilation, eventually activates oxygen-sensitive chemoreceptors; the latter stimulates ventilation and limits the compensatory respiratory response. Consequently, PaCO₂ usually does not increase above 55 mm Hg in response to metabolic alkalosis. As a general rule, PaCO₂ can be expected to increase 0.25 to 1 mm Hg for each 1 mEq/L increase in [HCO₃⁻].

RENAL COMPENSATION

The ability of the kidneys to control the amount of HCO₃⁻ reabsorbed from filtered tubular fluid, form new HCO₃⁻, and eliminate H⁺ in the form of titratable acids and ammonium ions (see Chapter 30) allows them to exert a major influence on pH during both metabolic and respiratory acid–base disturbances. The kidneys are responsible for eliminating the approximately 1 mEq/kg per day of sulfuric acid, phosphoric acid, uric acid, and incompletely oxidized organic acids that are normally produced by the metabolism of dietary and endogenous proteins, nucleoproteins, and organic phosphates (from phosphoproteins and phospholipids). Incomplete metabolisms of fatty acids and glucose produces keto acids and lactic acid. Endogenous bases are produced during the metabolism of some anionic amino acids (eg, glutamate and aspartate) and other organic compounds (eg, citrate, acetate, and lactate), but the quantity is insufficient to offset the endogenous acid production.

Renal Compensation During Acidosis

The renal response to acidemia is three-fold: (1) increased reabsorption of the filtered HCO₃⁻, (2) increased excretion of titratable acids, and (3) increased production of ammonia. Although these mechanisms are probably activated immediately, their effects are generally not appreciable for 12 to 24 h and may not be maximal for up to 5 days.

A. Increased Reabsorption of HCO₃⁻

Bicarbonate reabsorption is shown in Figure 50–4. CO₂ within renal tubular cells combines with water in the presence of carbonic anhydrase. The carbonic acid (H₂CO₃) formed rapidly dissociates into H⁺ and HCO₃⁻. Bicarbonate ion then enters the bloodstream while the H⁺ is secreted into the renal tubule, where it reacts with filtered HCO₃⁻ to form H₂CO₃. Carbonic anhydrase associated with the luminal brush border catalyzes the dissociation of H₂CO₃ into CO₂ and H₂O. The CO₂ thus formed can diffuse back into the renal tubular cell to replace the CO₂ originally consumed. The proximal tubules normally reabsorb 80% to 90% of the filtered bicarbonate load along with sodium, whereas the distal tubules are responsible for the remaining 10% to 20%. Unlike the proximal H⁺ pump, the H⁺ pump in the distal tubule is not necessarily linked to sodium reabsorption and is capable of generating steep H⁺ gradients between tubular fluid and tubular cells. Urinary pH can decrease to as low as 4.4 (compared with a pH of 7.40 in plasma).

B. Increased Excretion of Titratable Acids

After all of the HCO₃⁻ in tubular fluid is reclaimed, the H⁺ secreted into the tubular lumen can combine with HPO₄^2– to form H₂PO₄⁻ (Figure 50–5); the latter is not readily reabsorbed because of its charge and is therefore eliminated in urine. The net result is that H⁺ is excreted from the body as H₂PO₄⁻, and the HCO₃⁻ that is generated in the process can enter the bloodstream. With a pK of 6.8, the H₂PO₄⁻/HPO₄^2– pair is normally an ideal urinary buffer. When urinary pH approaches 4.4, however, all of the phosphate reaching the distal tubule is in the H₂PO₄⁻ form; HPO₄^2– ions are no longer available for eliminating H⁺.

C. Increased Formation of Ammonia

After complete reabsorption of HCO₃⁻ and consumption of the phosphate buffer, the NH₃/NH₄⁺ pair becomes the most important urinary buffer (Figure 50–6). Deamination of glutamine within the mitochondria of proximal tubular cells is the principal source of NH₃ production in the kidneys. Acidemia markedly increases renal NH₃ production. The ammonia formed is then able to passively cross the cell’s luminal membrane, enter the tubular fluid, and react with H⁺ to form NH₄⁺. Unlike NH₃, NH₄⁺ does not readily penetrate the luminal membrane and is therefore trapped within the tubules. Thus, excretion of NH₄⁺ in urine effectively eliminates H⁺.

Renal Compensation During Alkalosis

The tremendous amount of HCO₃⁻ normally filtered and subsequently reabsorbed allows the kidneys to rapidly excrete large amounts of bicarbonate, if necessary (see Chapter 49). As a result, the kidneys are highly effective in protecting against metabolic alkalosis, which therefore generally occurs only in association with concomitant sodium deficiency or mineralocorticoid excess. Sodium depletion decreases extracellular fluid volume and enhances Na⁺ reabsorption in the proximal tubule. To maintain neutrality, the Na⁺ ion is brought across with a Cl⁻ ion. As Cl⁻ ions decrease in number (<10 mEq/L of urine), HCO₃⁻ must be reabsorbed. Increased H⁺ secretion in exchange for augmented Na⁺ reabsorption favors HCO₃⁻ formation with metabolic alkalosis. Similarly, increased mineralocorticoid activity augments aldosterone-mediated Na⁺ reabsorption in exchange for H⁺ secretion in the distal tubules. The resulting increase in HCO₃⁻ formation can initiate or propagate metabolic alkalosis. Metabolic alkalosis is commonly associated with increased mineralocorticoid activity, even in the absence of sodium and chloride depletion.

Base Excess

Base excess is defined as the amount of acid or base (expressed in mEq/L) that must be added for blood pH to return to 7.40 and PaCO₂ to return to 40 mm Hg at full O₂ saturation and 37°C. Moreover, it adjusts for noncarbonic buffering in the blood. Simplistically, base excess represents the metabolic component of an acid–base disturbance. A positive value indicates metabolic alkalosis, whereas a negative value reveals metabolic acidosis. Base excess is usually derived from a nomogram and requires measurement of hemoglobin concentration.

PHYSIOLOGICAL EFFECTS OF ACIDEMIA

Biochemical reactions are very sensitive to changes in [H⁺]. [H⁺] is strictly regulated (36–43 nmol/L), as H⁺ ions have high charge densities and “large” electric fields that can affect the strength of hydrogen bonds that are present on most physiological molecules. The overall effects of acidemia represent the balance between the direct biochemical effects of H+ and the effects of acidemia-induced sympathoadrenal activation. With severe acidosis (pH <7.20), direct myocardial and smooth muscle depression reduces cardiac contractility and peripheral vascular resistance, resulting in progressive hypotension. Severe acidosis can lead to tissue hypoxia, despite a rightward shift in hemoglobin affinity for oxygen. Both cardiac and vascular smooth muscle become less responsive to endogenous and exogenous catecholamines, and the ventricular fibrillation threshold is decreased. The movement of K⁺ out of cells in exchange for increased extracellular H⁺ results in hyperkalemia that is also potentially lethal. Plasma [K⁺] increases approximately 0.6 mEq/L for each 0.10 decrease in pH.

Central nervous system depression is more prominent with respiratory acidosis than with metabolic acidosis. This effect is often termed CO₂ narcosis. Unlike CO₂, H⁺ ions do not readily penetrate the blood–brain barrier.

RESPIRATORY ACIDOSIS

Respiratory acidosis is defined as a primary increase in PaCO₂. This increase drives the reaction

to the right, leading to an increase in [H⁺] and a decrease in arterial pH. For the reasons described above, [HCO₃⁻] is minimally affected.

PaCO₂ represents the balance between CO₂ production and CO₂ elimination:

CO₂ is a byproduct of fat and carbohydrate metabolism—and muscle activity, body temperature, and thyroid hormone activity can all have major influences on CO₂ production. Because CO₂ production does not appreciably vary under most circumstances, respiratory acidosis is usually the result of alveolar hypoventilation (Table 50–3). In patients with a limited capacity to increase alveolar ventilation, however, increased CO₂ production can precipitate respiratory acidosis.

Acute Respiratory Acidosis

The compensatory response to acute (6–12 h) elevations in PaCO₂ is limited. Buffering is primarily provided by hemoglobin and the exchange of extracellular H⁺ for Na⁺ and K⁺ from bone and the intracellular fluid compartment (see earlier discussion). The renal response to retain more bicarbonate is acutely very limited. As a result, plasma [HCO₃⁻] increases only about 1 mEq/L for each 10 mm Hg increase in PaCO₂ above 40 mm Hg.

Chronic Respiratory Acidosis

Renal compensation in respiratory acidosis is appreciable only after 12 to 24 h and may not be maximal until 3 to 5 days have elapsed. During chronic respiratory acidosis, plasma [HCO₃⁻] increases approximately 4 mEq/L for each 10 mm Hg increase in PaCO₂ above 40 mm Hg.

Treatment of Respiratory Acidosis

Respiratory acidosis is treated by reversing the imbalance between CO₂ production and alveolar ventilation. In most instances, this is accomplished by increasing alveolar ventilation. Measures aimed at reducing CO₂ production are useful only in specific instances (eg, dantrolene for malignant hyperthermia, muscle paralysis for status epilepticus, antithyroid medication for thyroid storm, or reduced caloric intake in patients receiving excessive enteral or parenteral nutrition). Potential temporizing measures aimed at improving alveolar ventilation (in addition to controlled mechanical ventilation) include bronchodilation, reversal of narcosis, or improving lung compliance via diuresis. Severe acidosis (pH <7.20), CO₂ narcosis, and respiratory muscle fatigue are indications for mechanical ventilation. An increased inspired oxygen concentration is also usually necessary, as coexistent hypoxemia is common. Intravenous NaHCO₃ is rarely necessary, unless pH is less than 7.10 and HCO₃⁻ is less than 15 mEq/L. Sodium bicarbonate therapy will transiently increase PaCO₂:

Buffers that do not produce CO₂, such as Carbicarb or tromethamine (THAM), are theoretically attractive alternatives; however, there is almost no evidence showing that they have greater efficacy than bicarbonate. Carbicarb^TM is a mixture of 0.3 M sodium bicarbonate and 0.3 M sodium carbonate; buffering by this mixture mainly produces sodium bicarbonate instead of CO₂. Tromethamine has the added advantage of lacking sodium and may be a more effective intracellular buffer.

Patients with chronic respiratory acidosis require special consideration. When such patients develop acute ventilatory failure, the goal of therapy should be to return PaCO₂ to the patient’s “normal” baseline. Normalizing the patient’s PaCO₂ to 40 mm Hg will produces the equivalent of a respiratory alkalosis (see later discussion). Oxygen therapy must also be carefully controlled, because the respiratory drive in these patients may be dependent on hypoxemia, not PaCO₂. “Normalization” of PaCO₂ or relative hyperoxia can precipitate severe hypoventilation.

METABOLIC ACIDOSIS

Metabolic acidosis is defined as a primary decrease in [HCO₃⁻]. Pathological processes can initiate metabolic acidosis by one of three mechanisms: (1) consumption of HCO₃⁻ by a strong nonvolatile acid, (2) renal or gastrointestinal wasting of bicarbonate, or (3) rapid dilution of the extracellular fluid compartment with a bicarbonate-free fluid.

A fall in plasma [HCO₃⁻] without a proportionate reduction in PaCO₂ decreases arterial pH. The pulmonary compensatory response in a simple metabolic acidosis (see preceding discussion) characteristically does not reduce PaCO₂ to a level that completely normalizes pH but nevertheless can produce marked hyperventilation (Kussmaul’s respiration).

Table 50–4 lists disorders that can cause metabolic acidosis. Note that differential diagnosis of metabolic acidosis may be facilitated by calculation of the anion gap.

The Anion Gap

The anion gap in plasma is most commonly defined as the difference between the major measured cations and the major measured anions:

Or

Some practitioners also include plasma K⁺ in the calculation. Using normal values,

In reality, an anion gap cannot exist because electroneutrality must be maintained in the body; the sum of all anions must equal the sum of all cations. Therefore,

“Unmeasured cations” include K⁺, Ca²⁺, and Mg²⁺, whereas “unmeasured anions” include all organic anions (including plasma proteins), phosphates, and sulfates. Plasma albumin normally accounts for the largest fraction of the anion gap (approximately 11 mEq/L). The anion gap decreases by 2.5 mEq/L for every 1 g/dL reduction in plasma albumin concentration. Any process that increases “unmeasured anions” or decreases “unmeasured cations” will increase the anion gap. Conversely, any process that decreases “unmeasured anions” or increases “unmeasured cations” will decrease the anion gap.

Mild elevations of plasma anion gap up to 20 mEq/L may not be helpful diagnostically during acidosis, but values greater than 30 mEq/L usually indicate the presence of a high anion gap acidosis. Metabolic alkalosis can also produce a high anion gap because of extracellular volume depletion, an increased charge on albumin, and a compensatory increase in lactate production. A low plasma anion gap may be encountered with hypoalbuminemia, bromide or lithium intoxication, and multiple myeloma.

High Anion Gap Metabolic Acidosis

Metabolic acidosis with an increased anion gap is characterized by an increase in relatively strong nonvolatile acids. These acids dissociate into H⁺ and their respective anions; the H⁺ consumes HCO₃⁻ to produce CO₂, whereas their anions (conjugate bases) accumulate and take the place of HCO₃⁻ in extracellular fluid (hence the anion gap increases). Nonvolatile acids can be endogenously produced or ingested.

A. Failure to Excrete Endogenous Nonvolatile Acids

Endogenously produced organic acids are normally eliminated by the kidneys in urine (as described earlier). Glomerular filtration rates below 20 mL/min (kidney injury or failure) typically result in progressive metabolic acidosis from the accumulation of these acids.

B. Increased Endogenous Nonvolatile Acid Production

Severe tissue hypoxia following hypoxemia, hypoperfusion (ischemia), or an inability to utilize oxygen (cyanide poisoning) can result in lactic acidosis. Lactic acid is the end product of the anaerobic metabolism of glucose (glycolysis) and can rapidly accumulate under these conditions. Decreased utilization of lactate by the liver, and, to a lesser extent by the kidneys, is less commonly responsible for lactic acidosis; causes include hypoperfusion, alcoholism, and liver disease. Lactate levels can be readily measured and are normally 0.3 to 1.3 mEq/L. Acidosis resulting from D-lactic acid, which is not recognized by α-lactate dehydrogenase (and not measured by routine assays), may be encountered in patients with short bowel syndromes; D-lactic acid is formed by colonic bacteria from dietary glucose and starch and is absorbed systemically.

An absolute or relative lack of insulin can result in hyperglycemia and progressive ketoacidosis from the accumulation of β-hydroxybutyric and acetoacetic acids (diabetic ketoacidosis). Ketoacidosis may also be seen following starvation or alcoholic binges. The pathophysiology of the acidosis often associated with severe alcoholic intoxication and nonketotic hyperosmolar coma is complex and may represent a buildup of lactic, keto, or other unknown acids.

Some inborn errors of metabolism, such as maple syrup urine disease, methylmalonic aciduria, propionic acidemia, and isovaleric acidemia, produce a high anion gap metabolic acidosis as a result of accumulation of abnormal amino acids.

C. Ingestion of Exogenous Nonvolatile Acids

Ingestion of large amounts of salicylates may result in metabolic acidosis. Salicylic acid and other acid intermediates rapidly accumulate and produce a high anion gap acidosis. Because salicylates also produce direct respiratory stimulation, most adults develop mixed metabolic acidosis with superimposed respiratory alkalosis. Ingestion of methanol (methyl alcohol) frequently produces acidosis and retinitis. Symptoms are typically delayed until the slow oxidation of methanol by alcohol dehydrogenase produces formic acid, which is highly toxic to the retina. The high anion gap represents the accumulation of many organic acids, including acetic acid. The toxicity of ethylene glycol is also the result of the action of alcohol dehydrogenase to produce glycolic acid. Glycolic acid, the principal cause of the acidosis, is further metabolized to form oxalic acid, which may be deposited in the renal tubules and produce acute kidney injury.

Normal Anion Gap Metabolic Acidosis

Metabolic acidosis associated with a normal anion gap is typically characterized by hyperchloremia. Plasma [Cl⁻] increases to take the place of the HCO₃⁻ ions that are lost. Hyperchloremic metabolic acidosis most commonly results from abnormal gastrointestinal or renal losses of HCO₃⁻, or from excessive intravenous administration of 0.9% NaCl solution.

Calculation of the anion gap in urine can be helpful in diagnosing a normal anion gap acidosis.

The urine anion gap is normally positive or close to zero. The principal unmeasured urinary cation is normally NH₄⁺, which should increase (along with Cl⁻) during a metabolic acidosis; the latter results in a negative urinary anion gap. Impairment of H⁺ or NH₄⁺ secretion, as occurs in kidney failure or renal tubular acidosis (discussed below), results in a positive urine anion gap despite systemic acidosis.

A. Increased Gastrointestinal Loss of HCO₃⁻

Diarrhea is a common cause of hyperchloremic metabolic acidosis. Diarrheal fluid contains 20 to 50 mEq/L of HCO₃⁻. Small bowel, biliary, and pancreatic fluids are all rich in HCO₃⁻. Loss of large volumes of these fluids can lead to hyperchloremic metabolic acidosis. Patients with ureterosigmoidostomies and those with ileal loop neobladders that are too long or that become partially obstructed frequently develop hyperchloremic metabolic acidosis. The ingestion of chloride-containing anion-exchange resins (cholestyramine) or large amounts of calcium or magnesium chloride can result in increased absorption of chloride and loss of bicarbonate ions. The nonabsorbable resins bind bicarbonate ions, whereas calcium and magnesium combine with bicarbonate to form insoluble salts within the intestines.

B. Increased Renal Loss of HCO₃⁻

Renal wasting of HCO₃⁻ can occur as a result of failure to reabsorb filtered HCO₃⁻ or to secrete adequate amounts of H⁺ in the form of titratable acid or ammonium ion. These defects are encountered in patients taking carbonic anhydrase inhibitors, such as acetazolamide, and in those with renal tubular acidosis.

Renal tubular acidosis (RTA) is a disease of systemic acidosis resulting from inadequate renal compensation for systemic acid production. The kidneys are unable to adequately acidify the urine, and urinary pH is inappropriately high relative to the systemic acidemia. Kidney function is otherwise normal. RTA involves a defect in distal renal tubular H⁺ secretion (type 1 RTA), proximal renal tubular reabsorption of filtered HCO₃⁻ (type 2 RTA), or both (type 3 RTA). Type 4 RTA is the result of hypoaldosteronism or renal insensitivity to aldosterone.

C. Other Causes of Hyperchloremic Acidosis

Dilutional hyperchloremic acidosis may occur when extracellular volume is rapidly expanded with a bicarbonate-free, chloride-rich fluid such as normal saline. Plasma [HCO₃⁻] decreases in proportion to the amount of fluid infused as extracellular HCO₃⁻ is diluted, and this fall in [HCO₃⁻] is compensated by a rise in [Cl⁻]. This is a reason to prefer balanced salt solutions over 0.9% saline for fluid resuscitation. Amino acid infusions (parenteral hyperalimentation) contain organic cations in excess of organic anions and can produce hyperchloremic metabolic acidosis because chloride is commonly used as the anion for the cationic amino acids. Lastly, the administration of excessive quantities of chloride-containing acids, such as ammonium chloride or arginine hydrochloride (usually given to treat a metabolic alkalosis), can cause hyperchloremic metabolic acidosis.

Treatment of Metabolic Acidosis

Several general measures can be undertaken to control the severity of acidemia until the underlying processes are corrected. Any respiratory component of the acidemia should be corrected. Respiration should be controlled, if necessary; a PaCO₂ in the low 30s may be desirable to partially return pH to normal. If arterial blood pH remains below 7.20, alkali therapy, usually in the form of a 7.5% NaHCO₃ solution may be necessary. PaCO₂ may transiently rise as HCO₃⁻ is consumed by acids, emphasizing the need to control ventilation in severe acidemia. The amount of NaHCO₃ given is decided empirically as a fixed dose (1 mEq/kg) or is derived from the base excess and the calculated bicarbonate space (discussed next). In either case, serial blood gas measurements are mandatory to avoid complications (eg, overshoot alkalosis and sodium overload) and to guide further therapy. Raising arterial pH above 7.25 is usually sufficient to overcome the adverse physiological effects of the acidemia. Profound or refractory acidemia may require acute hemodialysis with a bicarbonate dialysate.

The routine use of large amounts of NaHCO₃ in treating cardiac arrest and low flow states is not recommended. Paradoxical intracellular acidosis may occur, particularly when CO₂ elimination is impaired, because the CO₂ formed readily enters cells, but the bicarbonate ion does not. Alternate buffers that do not produce CO₂, such as Carbicarb or tromethamine (THAM), may be theoretically preferable, but are unproven clinically.

Specific therapy for diabetic ketoacidosis includes replacement of the existing fluid deficit resulting from a hyperglycemic osmotic diuresis first, as well as insulin, potassium, phosphate, and magnesium. The treatment of lactic acidosis should be directed first at restoring adequate oxygenation and tissue perfusion. Alkalinization of the urine with NaHCO₃ to a pH greater than 7.0 increases elimination of salicylate following salicylate poisoning. Treatment options for methanol or ethylene glycol intoxication include ethanol infusion or fomepizole administration, which competitively inhibit alcohol dehydrogenase and hemodialysis or hemofiltration.

Bicarbonate Space

The bicarbonate space is defined as the volume to which HCO₃⁻ will distribute when administered intravenously. Although this theoretically should equal the extracellular fluid space (approximately 25% of body weight), in reality, it ranges anywhere between 25% and 60% of body weight, depending on the severity and duration of the acidosis. This variation is at least partly related to the amount of intracellular and bone buffering that has taken place.

Example: Calculate the amount of NaHCO₃ necessary to correct a base deficit (BD) of –10 mEq/L for a 70-kg man with an estimated HCO₃⁻ space of 30%:

In practice, only 50% of the calculated dose (eg., 105 mEq) is usually given, after which another blood gas is measured.

ANESTHETIC CONSIDERATIONS IN PATIENTS WITH ACIDOSIS

Acidemia can potentiate the depressant effects of most sedatives and anesthetic agents on the central nervous and circulatory systems. Because most opioids are weak bases, acidosis can increase the fraction of the drug in the nonionized form and facilitate opioid penetration into the brain, potentiating its sedative effect. The circulatory depressant effects of both volatile and intravenous anesthetics can also be exaggerated. Moreover, any agent that rapidly decreases sympathetic tone can potentially allow unopposed circulatory depression in the setting of acidosis. Halothane is more arrhythmogenic in the presence of acidosis. Succinylcholine should generally be avoided in acidotic patients with hyperkalemia to prevent further increases in plasma [K⁺].

PHYSIOLOGICAL EFFECTS OF ALKALOSIS

Alkalosis increases the affinity of hemoglobin for oxygen and shifts the oxygen dissociation curve to the left, making it more difficult for hemoglobin to give up oxygen to tissues. Movement of H⁺ out of cells in exchange for the movement of extracellular K⁺ into cells can produce hypokalemia. Alkalosis increases the number of anionic binding sites for Ca²⁺ on plasma proteins and can therefore decrease ionized plasma [Ca²⁺], leading to circulatory depression and neuromuscular irritability. Respiratory alkalosis reduces cerebral blood flow. In the lungs, respiratory alkalosis increases bronchial smooth muscle tone (bronchoconstriction), but decreases pulmonary vascular resistance.

RESPIRATORY ALKALOSIS

Respiratory alkalosis is defined as a primary decrease in PaCO₂. The mechanism is usually an inappropriate increase in alveolar ventilation relative to CO₂ production. Table 50–5 lists the most common causes of respiratory alkalosis. The distinction between acute and chronic respiratory alkalosis is not always made, because the compensatory response to chronic respiratory alkalosis is quite variable: Plasma [HCO₃⁻] usually decreases 2 to 5 mEq/L for each 10 mm Hg decrease in PaCO₂ below 40 mm Hg.

Treatment of Respiratory Alkalosis

Correction of the underlying process is the only treatment for respiratory alkalosis. For severe alkalemia (arterial pH >7.60), intravenous hydrochloric acid, arginine chloride, or ammonium chloride may be indicated (see later discussion).

METABOLIC ALKALOSIS

Metabolic alkalosis is defined as a primary increase in plasma [HCO₃⁻]. Most cases of metabolic alkalosis can be divided into (1) those associated with NaCl deficiency and extracellular fluid depletion, often described as chloride-sensitive, and (2) those associated with enhanced mineralocorticoid activity, commonly referred to as chloride-resistant (Table 50–6).

Chloride-Sensitive Metabolic Alkalosis

Depletion of extracellular fluid causes the renal tubules to avidly reabsorb Na⁺. Because not enough Cl⁻ is available to accompany all of the Na⁺ ions reabsorbed, increased H⁺ secretion must take place to maintain electroneutrality. In effect, HCO₃⁻ ions that might otherwise have been excreted are reabsorbed, resulting in metabolic alkalosis. Physiologically, maintenance of extracellular fluid volume is therefore given priority over acid–base balance. Because secretion of K⁺ ion can also maintain electroneutrality, potassium secretion is also enhanced. Moreover, hypokalemia augments H⁺ secretion (and HCO₃⁻ reabsorption) and will also propagate metabolic alkalosis. Indeed, severe hypokalemia alone can cause alkalosis. Urinary chloride concentrations during a chloride-sensitive metabolic alkalosis are characteristically low (<10 mEq/L).

Diuretic therapy is the most common cause of chloride-sensitive metabolic alkalosis. Diuretics, such as furosemide, ethacrynic acid, and thiazides, increase Na⁺, Cl⁻, and K⁺ excretion, resulting in NaCl depletion, hypokalemia, and usually mild metabolic alkalosis. Loss of gastric fluid is also a common cause of chloride-sensitive metabolic alkalosis. Vomiting or continuous loss of gastric fluid by gastric drainage (nasogastric suctioning) can result in marked metabolic alkalosis, extracellular volume depletion, and hypokalemia. Gastric secretions contain 25 to 100 mEq/L of H⁺, 40 to 160 mEq/L of Na⁺, about 15 mEq/L of K⁺, and about 200 mEq/L of Cl⁻. Rapid normalization of PaCO₂ after plasma [HCO₃⁻] has risen in chronic respiratory acidosis results in metabolic alkalosis (posthypercapnic alkalosis; see previous section). Infants being fed formulas containing Na⁺ without chloride readily develop metabolic alkalosis because of the increased H⁺ (or K⁺) secretion that must accompany sodium absorption.

Chloride-Resistant Metabolic Alkalosis

Increased mineralocorticoid activity commonly results in metabolic alkalosis, even when it is not associated with extracellular volume depletion. Inappropriate increases in mineralocorticoid activity cause sodium retention and expansion of extracellular fluid volume. Increased H⁺ and K⁺ secretion takes place to balance enhanced mineralocorticoid-mediated sodium reabsorption, resulting in metabolic alkalosis and hypokalemia. Urinary chloride concentrations are typically greater than 20 mEq/L in such cases.

Other Causes of Metabolic Alkalosis

Metabolic alkalosis is rarely encountered in patients given even large doses of NaHCO₃ unless renal excretion of HCO₃⁻ is impaired. The administration of large amounts of blood products and some plasma protein–containing colloid solutions frequently results in metabolic alkalosis because citrate, lactate, and acetate contained in these fluids are converted by the liver into HCO₃⁻. Patients receiving high doses of sodium penicillin (particularly carbenicillin) can develop metabolic alkalosis. Because penicillins act as nonabsorbable anions in the renal tubules, increased H⁺ (or K⁺) secretion must accompany sodium absorption. For reasons that are not clear, hypercalcemia that results from nonparathyroid causes (milk-alkali syndrome and bone metastases) is also often associated with metabolic alkalosis.

Treatment of Metabolic Alkalosis

As with other acid–base disorders, correction of metabolic alkalosis is never complete until the underlying disorder is corrected. When ventilation is controlled, any respiratory component contributing to alkalemia should be corrected by decreasing minute ventilation to normalize PaCO₂. The treatment of choice for chloride-sensitive metabolic alkalosis is administration of intravenous saline (NaCl) and potassium (KCl). H₂-blocker therapy is useful when excessive loss of gastric fluid is a factor. Acetazolamide may also be useful in edematous patients. Alkalosis associated with primary increases in mineralocorticoid activity readily responds to aldosterone antagonists (spironolactone). When arterial blood pH is greater than 7.60, treatment with intravenous hydrochloric acid (0.1 mol/L), ammonium chloride (0.1 mol/L), arginine hydrochloride, or hemodialysis should be considered.

Cerebral ischemia can occur from marked reduction in cerebral blood flow during respiratory alkalosis, particularly during hypotension. The combination of alkalemia and hypokalemia can precipitate severe atrial and ventricular arrhythmias. Reports of effects of alkalemia on neuromuscular blockers are inconsistent.

Interpretation of acid–base status from analysis of blood gases requires a systematic approach. A recommended approach follows (see Figure 50–3):

1. Examine arterial pH: Is acidemia or alkalemia present?

2. Examine PaCO₂: Is the change in PaCO₂ consistent with a respiratory component?

3. If the change in PaCO₂ does not explain the change in arterial pH, does the change in [HCO₃⁻] indicate a metabolic component?

4. Make a tentative diagnosis (see Table 50–1).

5. Compare the change in [HCO₃⁻] with the change in PaCO₂. Does a compensatory response exist (Table 50–7)? Because arterial pH is related to the ratio of PaCO₂ to [HCO₃⁻], both respiratory and renal compensatory mechanisms are always such that PaCO₂ and [HCO₃⁻] change in the same direction. A change in opposite directions implies a mixed acid–base disorder.

6. If the compensatory response is more or less than expected, by definition, a mixed acid–base disorder exists.

7. Calculate the plasma anion gap in the case of metabolic acidosis.

8. Measure urinary chloride concentration in the case of metabolic alkalosis.

An alternative approach that is rapid, though less precise, is to correlate changes in pH with changes in CO₂ or HCO₃. For a respiratory disturbance, every 10 mm Hg change in CO₂ should change arterial pH by approximately 0.08 U in the opposite direction. During metabolic disturbances, every 6 mEq change in HCO₃ also changes arterial pH by 0.1 in the same direction. If the change in pH is greater or less than predicated, a mixed acid–base disorder is likely to be present.

Values obtained by routine blood gas measurement include oxygen and carbon dioxide tensions (PO₂ and PCO₂), pH, [HCO₃⁻], base excess, hemoglobin, and the percentage oxygen saturation of hemoglobin. As a rule, only PO₂, PCO₂, and pH are measured. Hemoglobin and percentage oxygen saturation are measured with a cooximeter. [HCO₃⁻] is derived using the Henderson–Hasselbalch equation and base excess from the Siggaard–Andersen nomogram.

Sample Source & Collection

Arterial blood samples are most commonly utilized clinically, though capillary or venous blood can be used if the limitations of such samples are recognized. Oxygen tension in venous blood (normally 40 mm Hg) reflects tissue extraction, not pulmonary function. Venous PCO₂ is usually 4 to 6 mm Hg higher than PaCO₂. Consequently, venous blood pH is usually 0.05 U lower than arterial blood pH. Despite these limitations, venous blood is often useful in determining acid–base status. Capillary blood represents a mixture of arterial and venous blood, and the values obtained reflect this fact. Samples are usually collected in heparin-coated syringes and should be analyzed as soon as possible. Air bubbles should be eliminated, and the sample should be capped and placed on ice to prevent significant uptake of gas from blood cells or loss of gases to the atmosphere. Although heparin is highly acidic, excessive heparin in the sample syringe usually lowers pH only minimally, but decreases PCO₂ in direct proportion to percentage dilution and has a variable effect on PO₂.

Temperature Correction

Changes in temperature affect PCO₂, PO₂, and pH. Decreases in temperature lower the partial pressure of a gas in solution—even though the total gas content does not change—because gas solubility is inversely proportionate to temperature. Both PCO₂ and PO₂ therefore decrease during hypothermia, but pH increases because temperature does not appreciably alter [HCO₃⁻] and the dissociation of water decreases (decreasing H⁺ and increasing pH). Because blood gas tensions and pH are always measured at 37°C, controversy exists over whether to correct the measured values to the patient’s actual temperature. “Normal” values at temperatures other than 37°C are not known. Many clinicians use the measurements at 37°C directly (α-stat), regardless of the patient’s actual temperature (see Chapter 22).