Chapter 5. Acid–Base Disorders

Essentials of Diagnosis

• Increase in plasma [HCO3−].
• Compensatory increase in arterial Paco2.
• Increase in arterial pH.

General Considerations

Metabolic alkalosis is an acid–base disorder in which a primary disease process leads to the net accumulation of base within or the net loss of acid from the extracellular fluid (ECF). When it occurs as a simple acid–base disorder, it is recognized as an increase in plasma [HCO3−] and a compensatory increase in arterial blood pH.

Unopposed by other primary acid–base disorders, the increase in arterial blood pH promptly and predictably depresses ventilation resulting in increased Paco2 and buffering of the magnitude of the alkalemia; an increase in Paco2 of 0.6–0.7 mm Hg for every 1.0 mEq/L increase in plasma [HCO3−] is predicted from empirical studies. Although a Paco2 greater than 55 mm Hg is uncommon, compensatory increases to 60 mm Hg or higher have been documented in severe metabolic alkalosis. The magnitude of the compensatory increase in Paco2 is directly related to the extent of the alkalosis and the degree of elevation of the plasma [HCO3−], whether or not there is an intracellular acidosis. This compensatory respiratory response is independent of hypoxia, hypokalemia, renal failure, and cause and is usually not detectable clinically because it is more dependent on a change in depth rather than the rate of ventilation.

The major clinically and pathophysiologically relevant classification is based on whether the metabolic alkalosis is dependent on Cl− depletion. The Cl−-depletion forms, also termed the Cl−-responsive alkaloses, are more common. The other major grouping is the Cl−-resistant alkaloses, most of which are due to K+ depletion with mineralocorticoid excess. Mixed K+ and Cl−-depletion metabolic alkaloses also occur. Several other relatively uncommon causes constitute the balance of etiologies of metabolic alkalosis (Table 5–1).

Clinical Findings

Symptoms and Signs

Metabolic alkalosis should be anticipated when conditions such as vomiting, diuretic use, or severe hypertension are present. Thus, a careful and complete history is vital to establishing the etiology. Although the generation and maintenance phases of metabolic alkalosis can be clearly delineated in animal models of metabolic alkalosis, in the clinical setting the genesis of the generation phase may be obscure, even after a careful history. This is especially true if the patient is concealing bulimia or diuretic or laxative abuse.

A physical examination is helpful primarily for assessing the status of body fluid volume. Both deficits and surfeits of volume can and often do accompany metabolic alkalosis depending upon etiology, but they are not causes.

Neurologic symptoms such as apathy, confusion, cardiac arrhythmias, and neuromuscular irritability are observed only when alkalosis is severe (arterial pH >7.55). Although neuromuscular irritability may be readily evident, Chvostek and Trousseau signs are uncommon.

Alkalosis per se has a mild positive inotropic effect on the heart and little or no effect on cardiac rhythm. Cardiac arrhythmias occur primarily because of hypokalemia (see Chapter 4).

Compensatory hypoventilation may contribute to hypoxia or pulmonary infection in very ill or immunocompromised patients.

Many of the clinical features of metabolic alkalosis are dependent largely on the effects of the electrolyte deficits and not on alkalosis per se. K+ depletion may be accompanied by cardiac arrhythmias, nephrogenic diabetes insipidus, and muscle weakness. Similarly, Cl− depletion may be associated with impaired urine concentrating ability, impaired response to loop diuretics, stimulation of renin via a macula densa mechanism, and a reduction in glomerular filtration rate (GFR).

Laboratory Findings

A serum electrolyte profile and an arterial blood gas are necessary to accurately diagnose metabolic alkalosis as outlined above. If the measured Paco2 is within ±3–5 mm Hg of the predicted value, a simple metabolic alkalosis is present. If not, a mixed disorder is present, respiratory acidosis if the Paco2 is higher and respiratory alkalosis if it is lower.

Disequilibrium occurs when generation of HCO3− and resultant elevation of plasma [HCO3−] exceed the capacity of the renal tubule to reabsorb HCO3−. Transient bicarbonaturia with concomitant Na+ or K+ loss ensues until a new steady state is achieved and urinary HCO3− excretion ceases. This ushers in so-called “paradoxic aciduria” sometimes described in stable chronic metabolic alkalosis because the effects of Cl− or K+ depletion prevent urinary excretion of HCO3−.

Additional laboratory tests including serum [Ca2+] and osmolality and urine creatinine, [Cl−], [K+], Na+], osmolality, and Ca2+ for the differential diagnosis are discussed below.

Differential Diagnosis

A urine [Cl−] of less than 10 mEq/L characterizes chloride depletion except when a chloruretic diuretic is present in the urine or accompanying profound K+ depletion produces severe tubule dysfunction that induces a Cl− leak. Cl− depletion can be generated by losses from the gut or the kidney; in chloridorrhea, the stool [Cl−] is greater than 90 mEq/L. Fully compensated respiratory acidosis is a Cl− depletion state that will persist after the successful treatment of chronic respiratory acidosis if Cl− repletion has not occurred. In cystic fibrosis (in which renal Cl− handling is normal), high sweat [Cl−] can contribute to Cl− losses with excessive sweating.

When the urine [Cl−] is greater than 20 mEq/L, K+-depletion alkalosis and other miscellaneous disorders are suggested. When K+ depletion is present, the urine K+ excretion is normally less than 20 mEq/day. Thus, in the differential diagnosis of K+-depletion alkalosis, urine K+ excretion greater than 30 mEq/day in the presence of hypokalemia establishes renal K+ wasting and indicates mineralocorticoid excess or an agent that promotes renal K+ wasting. Within this group, K+-depletion alkalosis can be further subdivided by the status of ECF volume and blood pressure. In those diseases characterized by persistent intravascular and ECF volume expansion and consequent hypertension, such as primary aldosteronism or syndromes of apparent mineralocorticoid excess, escape from the Na+-retaining effect of mineralocorticoids occurs but not from the K+-wasting effects thereby promoting alkalosis. The transtubular K+ gradient is a fast and useful test for determining the presence of renal K+ wasting; a value of more than 4 in the setting of hypokalemia is consistent with this abnormality.

In contrast, in Bartter syndrome, Na+ loss as well as both K+ and Cl− losses are associated with normotension. Gitelman's syndrome is similar but is less severe and is characterized by hypocalciuria not hypercalciuria seen in Bartter's syndrome. These uncommon syndromes could be confused with some of the commonly concealed causes of metabolic alkalosis such as diuretic or laxative abuse, bulimia, or surreptitious vomiting.

When the urine K+ excretion is less than 20 mEq/day, K+depletion due solely to dietary or gut losses is established. The alkalosis in these disorders is usually mild. More severe alkalosis should suggest additional causative factors, such as concomitant Cl− depletion or base ingestion.

Hypercalcemia, usually in the setting of suppressed parathyroid hormone such as malignancy or vitamin D excess, may uncommonly be associated with alkalosis. In milk-alkali syndrome, multiple factors likely contribute to the alkalosis including vomiting, hypercalcemia, and reduced GFR in addition to the excessive calcium intake.

Hypoalbuminemia may cause mild metabolic alkalosis because of the resultant shift in the buffering capacity of plasma. Patients can develop a metabolic alkalosis if they ingest alkali such as NaHCO3 (eg, baking soda) or calcium carbonate and cannot excrete the excess HCO3− because of renal insufficiency. Obligate excretion of anions with preferential retention of bicarbonate putatively explains the alkalosis that has rarely been associated with some antibiotics.

Transient states of alkalosis are common but usually inconsequential. Intravenous or oral base loading may occur during the transfusion of blood or blood products with a citrate anticoagulant or with the treatment of metabolic acidosis.

Treatment

General Principles

Specific treatment is indicated when the arterial pH is greater than 7.55 or the serum bicarbonate concentration is greater than 33 mEq/L. Existing deficits must be replaced and the continued generation of losses should be prevented or blunted to the extent possible.

Chloride Depletion

The Cl− deficit must be replaced with the selection of the accompanying cation based on (1) ECF volume status, (2) the degree of associated K+ depletion, and (3) the degree and reversibility of any depression of GFR. Because the magnitude of the loss of each of these cations is difficult to assess, an empirical clinical approach is recommended. When Cl−and severe K+ depletion coexist, both must be repleted to correct the alkalosis. In patients with overt signs of ECF volume contraction, administration of 3–5 L of 0.15 M NaCl at a minimum is usually necessary to correct volume deficits and metabolic alkalosis. When volume is repleted, further Cl− repletion should be accomplished with KCl unless contraindications are present.

In the clinical setting of ECF volume overload such as congestive cardiac failure in association with Cl− depletion alkalosis, NaCl administration is contraindicated. KCl is the preferred alternative in this setting but it too may be contraindicated or limited because of either concurrent hyperkalemia or renal insufficiency, which could precipitate hyperkalemia.

In addition to volume overload, other serious conditions such as hepatic encephalopathy, cardiac arrhythmias, digitalis cardiotoxicity, or altered mental status also may accompany severe alkalosis. No clinical studies specifically address the impact of the treatment of severe alkalosis on the outcome in these states. However, the high reported mortality of severe alkalosis suggests that to the extent that it may be a contributing factor, alkalosis should be corrected promptly. HCl (0.1 N) administration through a central venous catheter at rates up to 25 mEq/hour should be considered when either Na+ or K+ is not appropriate. The amount of HCl needed to correct alkalosis is calculated by the following formula: 0.5 × body weight (kg) × desired decrement in plasma bicarbonate (mEq/L) with additional amounts to replace any continuing losses of acid. Plasma bicarbonate concentration should be initially restored halfway to normal.

An alternative to HCl is ammonium chloride, which may be administered via a peripheral vein at a rate of not more than 300 mEq NH4+/day; it should be avoided in advanced renal or hepatic insufficiency. The HCl salts of lysine or arginine are available but have been associated with severe hyperkalemia; they are not recommended. Of these agents, HCl is preferred and should be used only as noted.

In the setting of volume overload, acetazolamide 250 mg orally two or three times daily or 5–10 mg/kg intravenously may be effective if GFR is adequate. Because the distal nephron can avidly reabsorb excess Na+ delivery promoted by acetazolamide, the carbonic anhydrase (CA) inhibitors are most effective when used in conjunction with diuretics that have more distal sites of action. Acetazolamide could also be used intermittently to avoid or decrease chloride-depletion matabolic alkalosis (CDA) in edema-forming states such as congestive heart failure treated with loop diuretics. Serum electrolyte composition should be followed serially during its administration since acetazolamide is usually associated with high urinary K+ losses.

CA in erythrocytes and along the pulmonary capillary endothelium participates in the dehydration of bicarbonate to CO2 and its subsequent excretion by the lung. Studies in critically ill patients have shown only minor CO2 retention. Particularly in patients with impaired respiratory function, clinicians should be alert to this potential for CO2 retention with CA inhibition. In such patients, the goal of treatment is the usual stable plasma bicarbonate concentration for that patient and not necessarily a normal concentration.

Other primary or adjunctive therapies should be considered. Villous adenomas require surgical removal. Congenital chloridorrhea does not respond to antidiarrheal agents and dietary repletion of fluid; Cl− and K+ losses are usually required. Omeprazole 20 mg twice daily may substantially decrease diarrhea and obviate the need for dietary electrolyte supplementation; the reduction in the intestinal Cl− load by inhibition of gastric Cl− secretion is presumably the mechanism. When continuing gastric acid losses cannot be prevented, eg, Zollinger–Ellison syndrome, omeprazole or a histamine 2-receptor blocker, eg, cimetidine or ranitidine, will reduce output. These blocking agents have also been used to augment Cl− replacement in patients with unremitting losses due to gastrocystoplasty; on occasion, only surgical revision can correct the alkalosis in these patients.

If renal insufficiency limits the effectiveness of the above therapies or in patients on maintenance dialysis, hemofiltration with replacement infusions of NaCl, exchange of bicarbonate for high bath [Cl−] during hemodialysis, or peritoneal dialysis is an effective means for correcting metabolic alkalosis.

Potassium Depletion

The magnitude of K+ depletion can be estimated from the serum [K+]. The decrease in serum [K+] evoked by alkalosis per se (approximately a 0.5 mEq decrement for each 0.1 increment in arterial pH) accounts for only a modest overestimation of the K+ deficit.

Oral replacement will often suffice unless ileus is present. Oral KCl given in the liquid form diluted with fruit juice or in the slow-release form can be given in doses of up to 40–60 mEq four or five times per day. K+ salts such as citrate, gluconate, or bicarbonate are not appropriate.

If, however, a serious cardiac arrhythmia or generalized paralysis is present, intravenous KCl in concentrations no greater than 60 mEq/L at rates as high as 40 mEq/hour should be preferred. The downregulation of Na+-K+-ATPase in skeletal muscle with K+ depletion states may slow the clearing of the K+ load. Thus, monitoring with electrocardiograms and frequent determinations of serum [K+] are mandatory. Glucose should be omitted from infusions initially because stimulated insulin secretion may cause serum [K+] to decrease even further. However, once repletion is begun, infused glucose will facilitate cellular K+ repletion. If chloruretic diuretics or laxatives are contributing, they should be stopped.

Correction of the K+ deficit reverses the alkalinizing effects of K+ depletion but blockade or removal of the source of mineralocorticoid excess is essential for definitive correction. If the source of aldosterone excess cannot be removed, K+-sparing diuretics will blunt its effects. Amiloride 5–10 mg daily, triamterene 100 mg twice daily, or spironolactone 25–50 mg in single or divided doses daily all are useful. Restriction of Na+ and addition of K+ to the diet also ameliorate the alkalosis and associated hypertension.

In Bartter's syndrome, the focus of therapy is to prevent urinary K+ loss. Converting enzyme inhibitors have been shown to be effective and are a reasonable first approach. Because of concern for hypotension, low doses should be used initially. Other interventions may also partially correct the alkalosis. The K+-sparing diuretics mentioned above are effective but dietary K+ supplementation is usually also needed. Spironolactone may produce unacceptable gynecomastia in men. Because renal production of prostaglandin E2 is increased and may contribute to Na+, Cl−, and K+ wasting, prostaglandin synthetase inhibitors, such as indomethacin or ibuprofen, may blunt but not completely correct the hypokalemic alkalosis. Since magnesium depletion may contribute to the increase in urinary K+ wasting, hypomagnesemia should be corrected and magnesium stores repleted, if clinically feasible. Oral magnesium oxide in doses of 250–500 mg four times daily (12.5–25 mEq Mg2+) is recommended. However, the degree to which the correction of magnesium depletion blunts the alkalosis is uncertain and magnesium salts often produce an unacceptable degree of diarrhea that can worsen electrolyte imbalance.

Several of the primary disorders of mineralocorticoid excess are treated definitively by tumor ablation or by medication when that cannot be accomplished.

Miscellaneous

In acute milk-alkali syndrome, cessation of alkali ingestion and the calcium sources (milk, antacids, etc.) and repletion of Cl− and volume usually will lead to the prompt resolution of these abnormalities. The treatment of hypercalcemia is discussed in Chapter 6. For those alkaloses due to alkali loading, cessation of alkali administration and continuation of normal electrolyte intake will usually suffice.

Prognosis

Data on the prevalence and outcome of metabolic alkalosis are sparse. Metabolic alkalosis is common and, when severe, is associated with high morbidity and mortality; in one study, it comprised half of all acid–base disorders. A mortality rate of 45% in patients with a pH of 7.55 and of 80% when the pH was greater than 7.65 has been recorded and confirmed in a separate study (48.5% for alkalemia greater than 7.60). While this relationship between alkalemia and mortality is not necessarily causal, severe alkalosis should be viewed with concern and should be promptly treated.

Essentials of Diagnosis

• Reduction in plasma bicarbonate concentration.
• Reduction in blood pH.
• Decreased concentration of carbon dioxide (the compensatory respiratory response).

General Considerations

Metabolic acidosis is one of the four primary acid–base disorders and is caused by a reduction in plasma bicarbonate concentration. It is associated with a reduction in blood pH (termed acidemia) and a decrease in the concentration of carbon dioxide (Pco2), termed the compensatory respiratory response.

Metabolic acidosis can result from several different mechanisms including (1) excessive systemic H+ loads as with ketoacidosis or lactic acidosis, (2) impairment in renal HCO3− generation as with renal failure, (3) extrarenal HCO3− loss as with gastrointestinal HCO3− loss with small bowel diarrhea, and (4) renal HCO3− loss as with proximal renal tubular acidosis (RTA).

Pathophysiology

Normal Control of Acid–Base Balance

The plasma bicarbonate concentration is normally maintained at a constant level of 24–25 mEq/L in males and 22–23 mEq/L in nonpregnant females. Plasma bicarbonate concentration is maintained at these levels, despite ongoing H+ production resulting from metabolism of dietary constituents, because of compensatory equimolar generation of bicarbonate by the kidney (˜70 mmol/day). In addition, since the kidney filters a large quantity of bicarbonate each day, ˜4500 mEq, it must reclaim most of this bicarbonate to maintain a normal plasma bicarbonate concentration.

Approximately 85% of filtered bicarbonate is reclaimed in the proximal tubule. As depicted in Figure 5–1, this bicarbonate is reabsorbed indirectly via the apical sodium–hydrogen exchanger NHE3, and exits the cell via the sodium–bicarbonate cotransporter kNBC1. Membrane-bound carbonic anhydrase IV in the apical and basolateral membranes and cytoplasmic carbonic anhydrase II are necessary for efficient absorption of filtered bicarbonate from the tubular fluid to the systemic circulation.

In addition to absorbing the filtered bicarbonate load, the kidney needs to generate new bicarbonate to match the loss of HCO3 resulting from neutralization of acid generated in the liver from the daily metabolism of dietary protein. The proximal tubule in the kidney generates the new HCO3 primarily from the metabolism of glutamine. In addition to HCO3, NH4+ is also produced. The HCO3 is transported across the basolateral membrane of the proximal tubule cell to the systemic circulation. Were all the NH4+ produced in the proximal tubule transferred along with HCO3 to the systemic circulation, the HCO3 produced in the proximal tubule would be converted to urea in the liver and therefore would be unavailable to buffer the dietary H+ load. As depicted in Figure 5–1, this futile cycle is prevented by intrarenal transport mechanisms that ensure that a portion of the NH4+ is trapped in the lumen of the collecting duct as the result of proton transport by an apically located vacuolar H+-ATPase and possibly H+-K+-ATPase. Luminal proton transport by the H+-ATPase is coupled to basolateral bicarbonate exit via the anion exchanger AE1, with the NH4+ subsequently excreted in the urine.

Collecting duct proton secretion by the H+-ATPase (the primary proton transporter) is modulated by the action of aldosterone, in part by enhancing sodium reabsorption via the epithelial sodium channel (ENaC) to produce a favorable electrical gradient. The kidney generates the remaining new HCO3 by a process called titratable acid (TA) formation/excretion. In this process, secreted H+ (via the proximal tubule NHE3, collecting duct H+-ATPase, and possibly H+-K+-ATPase transporters) bind to HPO42− in the tubule lumen and generates intracellular HCO3 that is transported to the systemic circulation. Clinically, the total effective new bicarbonate generated by the kidney can be quantified by measuring a parameter called net acid excretion, which is equal to the urinary excretion of NH4+ + TA − HCO3−. Under normal acid–base conditions, 40 mEq NH4 and 30 mEq TA are excreted daily while HCO3− excretion is negligible.

Physiologic Response to an Increment in Acid Load or Extrarenal Bicarbonate Loss

The normal response of the body to an H+ load or HCO3− loss involves four processes: (1) Extracellular buffering, (2) intracellular buffering, (3) respiratory compensation, and (4) enhanced renal HCO3− generation. Immediately upon an increase in an acid load, H+ is buffered by plasma HCO3− followed by H+ influx into cells (intracellular buffering). The latter process occurs more slowly, and is completed in 2–4 hours. Approximately 60% of an H+ load is buffered intracellularly, but this can increase dramatically with more severe degrees of metabolic acidosis as bicarbonate buffers are depleted. The degree of intracellular buffering can be indirectly determined by the quantity of bicarbonate required to raise plasma bicarbonate concentration to a certain level (bicarbonate deficit × bicarbonate space). The bicarbonate space is the apparent volume of distribution of administered bicarbonate and is calculated according to the following equation: Bicarbonate space = [0.4 + (2.6/plasma bicarbonate) × lean body weight]. The bicarbonate space can increase from 50% body weight with mild to moderate metabolic acidosis (12–23 mEq/L) to more than 100% body weight with severe metabolic acidosis (plasma bicarbonate concentration <5 mEq/L).

The kidney plays the dominant role in regulating acid–base balance by increasing the quantity of new HCO3− generated. New HCO3− generation increases immediately and achieves a maximal level in approximately 4 days. The quantity of HCO3− generated can increase several fold and is due primarily to enhanced glutamine metabolism.

Respiratory Compensation

A fall in plasma bicarbonate concentration and blood pH stimulates receptors in the periphery and central respiratory center increasing alveolar ventilation. The increase in alveolar ventilation creates an inequality between mitochondrial CO2 production and pulmonary CO2 excretion. Pco2 decreases until a new steady state is reached, a process that requires approximately 12–24 hours. The magnitude of the decrease in Pco2 for any given level of sustained metabolic acidosis has been empirically determined and is given by the following relationship: For a given decrease in HCO3 of 10 mEq/L, the Pco2 decreases by approximately 12 mm Hg. When the Pco2 decreases appropriately, the metabolic acidosis is called “compensated” and a simple metabolic acidosis is said to be present. If the Pco2 is above the predicted value, the patient has a coexisting defect in ventilation and a mixed acid–base disturbance, ie, respiratory acidosis and a metabolic acidosis. Conversely, if Paco2 falls below the expected value, the patient has a mixed metabolic acidosis and respiratory alkalosis.

Serum Anion Gap

Metabolic acidosis is subdivided into those disorders in which the serum anion gap is normal and those in which it is elevated. The serum anion gap represents the concentration of unmeasured anions minus unmeasured cations: Na+ + K+ + unmeasured cations = Cl− + HCO3− + unmeasured anions. Since the concentration of serum K+ is low, it is not considered in the calculation of the serum anion gap, which is then calculated as Na+ − (Cl− + HCO3). The normal serum anion gap ranges between 8 and 18 mEq/L, with an average of 10–12 mEq/L. However, introduction of a new autoanalyzer method for measuring serum chloride in some clinical laboratories has resulted in a higher value of serum chloride and therefore a lower value for the mean serum anion gap (average of 6–8 mEq/L).

An increase in the serum anion gap is usually the result of retention of unmeasured anions in the blood, although rarely a decrease in the concentration of unmeasured cations can also be the cause. A decrease in the anion gap may be due to a decrease in the concentration of unmeasured anions (primarily negative charges on albumin) or an increase in unmeasured cations. An important cause of the latter is overproduction of cationic proteins seen in some cases of myeloma.

The differential diagnosis of metabolic acidosis is facilitated by examination of the serum anion gap. In certain forms of metabolic acidosis (eg, lactic acidosis, ketoacidosis) H+ is infused into the circulation with a non-Cl− anion such as lactate or β-hydroxybutyrate (see Table 5–2). The increase in the anion gap reflects the increase in the concentration of lactate or β-hydroxybutyrate. By contrast, in patients in whom HCO3− is lost from the body (eg, small bowel diarrhea, proximal RTA), the serum anion gap remains stable because the fall in plasma HCO3 concentration is matched by an equivalent rise in serum Cl−. C8 Metabolic acidosis associated with a high anion gap is called an elevated or high anion gap metabolic acidosis. Metabolic acidosis associated with a normal anion gap is also called hyperchloremic metabolic acidosis.

High Anion Gap

The disorders producing a high anion gap metabolic acidosis are shown in Table 5–2. They include renal failure, ketoacidosis, either diabetic or alcoholic, lactic acidosis, salicylate intoxication, and methanol and ethylene glycol intoxication.

A high anion gap metabolic acidosis can be observed with acute or chronic renal failure. With chronic renal failure, the acidosis is generally mild to moderate in degree, with plasma bicarbonate concentrations ranging from 12 to 22 mEq/L. Values below 12 mEq/L should raise the suspicion of superimposition of other acid–base disorders. Epidemiologic studies have indicated that the acidosis can first be detected when glomerular filtration rate (GFR) falls below 20–30 mL/minute of normal. Therefore, the presence of metabolic acidosis at higher levels of GFR should alert the physician to potential additional renal tubular disorders such as hyporeninemic hypoaldosteronism (Type IV RTA) that can cause a non-anion gap metabolic acidosis. The metabolic acidosis of chronic kidney disease remains the most common cause of chronic metabolic acidosis, ie, metabolic acidosis lasting for more than several days or weeks. Although chronic renal disease is often considered a paradigm of high anion gap metabolic acidosis, various studies have indicated that these patients can manifest a wide range of anion gap levels.

Metabolic acidosis is also frequent in oliguric acute renal failure. The severity of the metabolic acidosis in acute renal failure is dependent upon the level of residual renal function, duration of renal failure, and catabolic state of the patient. Thus, the acidosis will be more severe in catabolic patients with minimal residual renal function that has been present for several days.

Ketoacidosis either due to diabetes or alcohol intoxication is one of the most common causes of acute metabolic acidosis, ie, metabolic acidosis lasting a few hours to days. Diabetic ketoacidosis is detected by noting increased urinary and/or blood concentrations of ketoacids. The nitroprusside reaction detects acetoacetate and will be positive in the majority of cases of diabetic ketoacidosis. However, if diabetic ketoacidosis is complicated by lactic acidosis, or if there is alcoholic ketoacidosis, the reaction may be trace positive or even negative reflecting the predominance of β-hydroxybutyrate over acetoacetate in the body fluids. Therefore, a negative nitroprusside reaction does not exclude ketoacidosis. Most cases of diabetic ketoacidosis are associated with marked hyperglycemia. In contrast, alcoholic ketoacidosis is often characterized by a normal or low blood sugar, reflecting impaired glucose release from the liver. Another clue to the presence of alcoholic ketoacidosis is an elevated serum osmolal gap (see below). As noted with chronic renal failure, it has been recognized that patients with ketoacidosis can have a wide range of anion gap levels. However, patients with higher serum anion gap values are usually volume depleted, reflecting a decreased ability of the kidney to excrete the ketone bodies.

Lactic acidosis, another common cause of acute metabolic acidosis and possibly the most frequent cause of severe metabolic acidosis, is indicated by a blood pH <7.1 and plasma bicarbonate concentration <8–10 mEq/L. Indeed, ketoacidosis and lactic acidosis together account for more than 90% of the cases of metabolic acidosis in which the anion gap is >30 mEq/L. Type A lactic acidosis associated with tissue hypoxia is the most frequent type and can be readily suspected by the presence of hypotension and reduced tissue perfusion. The diagnosis of lactic acidosis is confirmed by a serum lactate concentration greater than 5 mEq/L.

Salicylate intoxication can cause both metabolic acidosis and respiratory alkalosis. Therefore, a high anion gap metabolic acidosis in association with respiratory alkalosis could indicate the presence of this disorder. Ketoacids accumulate and can be detected with the nitroprusside reaction. Although this acid–base disturbance has frequently been encountered in young adults trying to commit suicide, recent studies have found it in adults taking salicylates for treatment of rheumatic conditions. Another clue to its presence includes a prolonged prothrombin time. Treatment includes forced diuresis to increase urinary excretion of the salicylates and hemodialysis. The latter procedure is indicated when the serum levels of salicylates are extremely high or the patient has marked central nervous system (CNS) abnormalities.

Methanol (wood alcohol) and ethylene glycol (antifreeze) are rare, but serious causes of metabolic acidosis. Since both disorders can be rapidly lethal, it is critical to recognize their presence. Both substances are low-molecular-weight moieties and can increase serum osmolality. Therefore, measurement of serum osmolality and comparison of this value to the estimate of serum osmolality derived from consideration of the usual, osmotically active substances in blood can be of great value. Quantitatively the most important moieties contributing to serum osmolality are sodium (and its counterbalancing anions chloride and bicarbonate), glucose, and urea. Serum osmolality can rapidly be estimated using the following formula: 2 × [Na+] + [glucose]/18 + [BUN]/2.8, where [glucose] and [BUN] (blood urea nitrogen) are expressed in mg/dL. An osmolal gap (defined as the difference between the measured and estimated serum osmolality) greater than 10 mOsm/kg H2O indicates the presence in serum of additional osmotically active particles.

Alchoholic ketoacidosis is the most common cause of a high anion gap metabolic acidosis associated with an increased osmolal gap, but methanol and ethylene glycol intoxication are other important causes. A slight increase in the osmolal gap ˜10 mOsm/kg H2O has been reported with lactic acidosis or chronic renal failure and might confound the diagnosis in some instances. If methanol and ethylene glycol are metabolized completely to their toxic byproducts, formic acid and glycolic acid, respectively, little or no increment in the osmolal gap might be detected. Other clues to the diagnosis of these disorders include optic papillitis in methanol intoxication and renal failure with oxalate crystals in the urine in patients with ethylene glycol intoxication. Treatments of methanol and ethylene glycol intoxication include infusion of alcohol to retard their metabolism and/or hemodialysis, which is very effective in removing these substances from the body.

In summary, rapid diagnosis of the cause of high anion gap metabolic acidosis can be facilitated by measuring serum creatinine, urine and blood ketones, serum osmolality, serum sodium, glucose, and urea nitrogen necessary for calculation of the serum osmolal gap, and serum salicylate levels. If an elevated osmolal gap is found, then measurement of methanol, ethylene glycol, and alcohol levels is warranted to detect one of these intoxications.

Normal Anion Gap (Hyperchloremic)

As shown in Table 5–3, disorders causing a normal anion metabolic acidosis are often subdivided into those in which serum potassium concentration is low and those in which serum potassium concentration is normal or elevated to facilitate diagnosis of the underlying cause. Alternatively, the disorders causing a normal anion gap acidosis can be categorized based on the predominant organ involved in their pathogenesis as discussed below.

Gastrointestinal Causes

The most common cause of a normal anion gap metabolic acidosis associated with hypokalemia is diarrhea. Some studies have indicated this is more common with diarrhea emanating from the distal bowel. Other causes of gastrointestinal bicarbonate wasting include intestinal or pancreatic fistulas in which the bicarbonate-rich fluids are lost from the body. Construction of a conduit from the ureter to the sigmoid or ileal segments of the bowel is often done after removal of the bladder in patients with bladder cancer. The former procedure is regularly accompanied by the development of hypokalemic metabolic acidosis, whereas with the latter procedure this disorder usually indicates blockage of the ileal conduit.

Administration of solutions containing amino acids that are metabolized in the liver to produce hydrochloric acid (arginine, lysine) or sulfuric acid (cysteine, methionine) can also produce a normal anion gap acidosis with hyperkalemia. In this regard, the administration of total parenteral nutrition solutions containing cationic and sulfur-containing amino acids causes a metabolic acidosis. Addition of sufficient organic anions, such as acetate, that are metabolized into HCO3− has eliminated or reduced the severity of this problem.

Renal Causes

Proximal Renal Tubular Acidosis (Type 2)

Proximal RTA (Table 5–4) is the result of impaired reabsorption of filtered HCO3 leading to urinary bicarbonate wasting. This can be due to selective dysfunction of proteins present in the proximal tubule essential for bicarbonate absorption such as the basolateral Na–HCO3 cotransporter (kNBC1), defective cytoplasmic carbonic anhydrase (CAII) activity, or metabolic abnormalities that alter cellular adenosine triphosphate (ATP) production. The causes of proximal RTA are either genetic or acquired as shown in Table 5–4. When generalized proximal tubule malabsorption is present, bicarbonate wasting may be accompanied by glycosuria, phosphaturia, aminoaciduria, and hyperuricosuria (Fanconi's syndrome).

During the generation phase of proximal RTA, an excessive amount of HCO3 is delivered to the distal nephron (which is incapable of reclaiming it), resulting in a net renal excretion of bicarbonate. The excretion of bicarbonate causes the plasma HCO3 to decrease, resulting in a metabolic acidosis. The decrease in plasma HCO3 lowers the filtered load of HCO3 (GFR × plasma bicarbonate). Once the plasma HCO3 concentration falls below the reabsorptive threshold for the patient, the proximal tubule will again be able to reclaim the majority of the luminal HCO3 and the excessive excretion of bicarbonate will no longer occur. A new steady state is achieved, albeit at a plasma HCO3 concentration lower than normal.

The reabsorptive threshold in proximal RTA can vary, resulting in a steady-state plasma HCO3 concentration ranging from 15 mEq/L to 22 mEq/L. Urinary bicarbonate wastage will also occur as bicarbonate is administered to patients to raise the plasma HCO3 concentration. Urine pH is elevated during both the generation phase and reparative phases of proximal RTA, but is appropriately acidic during the steady state.

Proximal RTA is diagnosed by measuring the fractional excretion of HCO3 (FEHCO3) when the plasma HCO3 concentration has been normalized using the following formula:

A value greater than 20% is diagnostic of proximal RTA. The defect in proximal tubule HCO3 bicarbonate absorption may be mild. In this case, the FEHCO3 will be less than 20%. Since the nephron segments distal to the proximal tubule normally absorb ˜20% of the filtered HCO3 load, an FEHCO3− less than 20% does not necessarily implicate the proximal tubule as the site of defective HCO3 absorption. Hypokalemia is prominent in proximal RTA, although it is often less severe than in distal RTA. Urinary K+ losses are in large part associated with urinary HCO3 and Na+ losses, and therefore are most severe during generation and treatment of proximal RTA.

Distal Renal Tubular Acidosis

Distal RTA has traditionally been divided into two types: Type I (hypokalemic) and Type IV (hyperkalemic) (Table 5–4). In both forms, there is decreased collecting duct net H+ secretion and consequent reduced titratable acid (H2PO42−) excretion and NH4+ excretion leading to decreased new renal bicarbonate generation. Defective net proton transport can be due to (1) impaired hydrogen secretion from a dysfunction of one of the subunits of the H+-ATPase and possibly the H+-K+-ATPase (Figure 5-1), (2) increased luminal H+ efflux into the cells because of increased apical cellular permeability (occurring with exposure to amphotericin B), or (3) impaired hydrogen secretion because of a less favorable electrical gradient due to the abnormal function of aldosterone, decreased sodium absorption via ENaC (sodium channel), or augmented chloride entry (chloride shunt).

Type I RTA can result from genetic diseases affecting the transporters that play a role in collecting duct proton secretion or from diseases that damage the collecting duct. Genetic causes have been ascribed to mutations in the apical H+-ATPase and basolateral Cl–HCO3 exchanger, AE1, and carbonic anhydrase II, which is expressed in collecting duct intercalated cells. Although many features are shared among these disorders, the clinical phenotype differs somewhat depending upon the transporter that is affected. Thus, patients with mutations in AE1 sometimes have red cell abnormalities. Type I RTA due to AE1 mutations is either autosomal dominant or autosomal negative. Some patients with mutations in specific H+-ATPase subunits have concomitant sensorineural deafness, since the proton pump is expressed in the inner ear. Patients with carbonic anhydrase II mutations can have a combined proximal and distal RTA and bony abnormalities.

Many patients with distal RTA in addition to metabolic acidosis have hypercalciuria, nephrocalcinosis, and hypokalemia. In some patients osteomalacic bone disease can also be detected.

Although the hypercalciuria is often attributed to the metabolic acidosis and the buffering of H+ by bone, this explanation is likely incomplete, given that many patients with Type IV RTA or patients with extrarenal causes of metabolic acidosis do not have hypercalciuria. The hypokalemia has been attributed to impaired collecting duct H+-K+-ATPase rather than impaired vacuolar H+-ATPase function. This explanation may not be correct, since patients with distal RTA due to genetic defects in the vacuolar H+-ATPase also have hypokalemia. An alternate explanation is that Na+ wasting results in volume depletion and enhanced aldosterone secretion with subsequent increased collecting duct K+ secretion and renal K+ excretion. Importantly, since these disorders often appear in childhood and metabolic acidosis affects bone metabolism, stunted growth may be present.

Acquired disorders such as systemic lupus erythematosus (SLE), interstitial nephritis, and Sjögren's syndrome can injure the collecting duct and produce Type I RTA. Immunohistochemical staining of some patients with SLE have documented a decrease in H+-ATPase pumps in the collecting duct.

Patients with distal RTA are often recognized by documenting an elevated urine pH (>5.5) in the presence of a normal anion gap metabolic acidosis in the absence of diarrhea (see below). More sophisticated studies of collecting duct proton secretion such as a urine minus blood Pco2 difference after bicarbonate infusion or urinary acidification with sodium sulfate administration can be helpful in determining the precise mechanism of the disease, but are often not required in clinical practice.

Hyperkalemic Distal Renal Tubular Acidosis

Hyperkalemia associated with distal RTA was first recognized in patients with hyporeninemic hypoaldosteronism, and was designated Type IV RTA to distinguish the syndrome from Types I, II, and III RTA associated with hypokalemia. Subsequently, other causes of distal RTA associated with hyperkalemia were recognized in which a low serum aldosterone concentration was not the pathogenic mechanism. The term Type IV RTA is used when referring to patients with hyperkalemic distal RTA due to an abnormality in the renin–aldosterone axis resulting in aldosterone deficiency or aldosterone resistance (pseudohypoaldosteronism). Importantly, these patients are able to acidify their urine appropriately. Patients with distal RTA and hyperkalemia where aldosterone does not play a role (Table 5–4) are diagnosed as having hyperkalemic distal RTA (HDRTA). These patients are unable to acidify their urine appropriately.

The causes of Type IV RTA are listed in Table 5–5. It is important to realize that not all patients with Type IV RTA have hyporeninemic hypoaldosteronism. Depending on the cause, patients with Type IV RTA can have elevated aldosterone and renin levels as, for example, in genetic or acquired syndromes resulting in aldosterone resistance (Table 5–5). Furthermore, some patients with Type IV RTA have high renin and low aldosterone levels [angiotensin-converting enzyme (ACE) inhibition, adrenal abnormalities, heparin].

As shown in Table 5–6, based on the pathogenesis of the disorder patients with HDRTA are further subcategorized into those with genetic or acquired diseases that decrease ENaC activity and those with a genetic disease called Gordon's syndrome, also referred to as a chloride shunt defect. Acquired causes of decreased ENaC activity include treatment with amiloride, pentamidine, trimethoprim, or triamterene, which all bind to and block sodium absorption via ENaC. Patients with Gordon's syndrome have hypercalciuria and hypertension, distinguishing them clinically from patients with HDRTA due to decreased ENaC activity. Gordon's syndrome has recently been shown to be due to mutations in WNK1 and WNK4 kinases.

Diagnosis

Diagnosis of the cause of normal anion gap metabolic acidosis can often be made with the use of clinical information and routine laboratory studies. However, since defects in renal acidification are often prominent causes, measurement or estimates of urine NH4+ excretion can be helpful. In general, in patients with extrarenal metabolic acidosis, the urine NH4+ excretion is increased several fold. Failure to detect an appropriate increase in urine NH4+ excretion will implicate the kidney as the cause of the metabolic acidosis. The urine NH4+ concentration can be estimated from calculation of either the urine anion or osmolal gap. The former is calculated from [Na+] − (Cl− + HCO3−) in a urine with a pH <6.5, which is free of nonreabsorbable anions such as ketones. In an acidemic patient, the urine anion gap should be negative (˜ −30 mEq/L). A positive urine anion gap indicates a low NH4+ concentration. In patients in whom there is increased excretion of unmeasured anions, such as ketones or hippurate, the urine anion gap can be positive despite ample quantities of NH4+ in the urine. In these rare cases, the urine osmolal gap can be calculated from the measured osmolality − (2[Na+ + K+] + [urea nitrogen]/2.8 + [glucose]/18), where [urea nitrogen] and [glucose] are measured in mg/dL. An appropriate urine osmolal gap in an acidemic patient is greater than 150–200 mOsm/kg H2O (the NH4+ concentration is half this value), whereas in patients with a decreased urine NH4+ concentration as in renal failure or distal RTA, the osmolal gap is usually less than 50–100 mOsm/kg H2O.

Measurement of urine pH in patients with normal anion gap metabolic acidosis can complement the estimation of urinary ammonium excretion for further characterization of metabolic acidosis resulting from defects in renal acidification. For optimal measurement of urine pH and ammonium, a sample of urine should be obtained under oil (to prevent CO2 loss) when the patient is acidemic. If the urine pH is below 5.5, bicarbonate is administered until the urine becomes alkaline or the serum bicarbonate returns to normal. The development of an alkaline urine (pH >6.5) prior to normalization of serum bicarbonate indicates that proximal RTA (defective proximal tubule bicarbonate absorption) is present. If urine pH is high (>6.0) and remains relatively constant despite bicarbonate administration, distal RTA is potentially present. It is necessary to be cautious in making a diagnosis of distal RTA without initially ruling out diarrhea as a cause of a non-gap acidosis and an elevated urine pH. Importantly, hypokalemia due to diarrhea can result in an inappropriately elevated urine pH that is due to a decrease in collecting duct H+ secretion (distal RTA), but results from the fact that hypokalemia is a potent stimulus of renal NH3 production. The excess NH3 in the urine binds to secreted H+, thereby elevating the urine pH despite normal tubular H+ secretion. Importantly, when the hypokalemia is treated, NH3 production by the kidney decreases, and the urine pH decreases appropriately. In contrast, in Type I distal RTA, correction of hypokalemia has no effect on the urine pH.

The classification of metabolic acidosis into high anion gap and normal anion gap forms is extremely useful in determining the cause of the acidosis. The reciprocal fall in serum bicarbonate, termed the (delta) ΔHCO3, and the rise in the anion gap, termed the Δ anion gap, have also been useful in the identification of mixed forms of metabolic acidosis. It has been suggested that there is a strict 1:1 relationship between the rise in the serum anion gap and fall in serum bicarbonate concentration in patients with high anion gap metabolic acidosis. However, the ratio between the Δ anion gap and ΔHCO3 can range between 1 and 2 and may differ with different acid–base disorders. Thus, with lactic acidosis the rise in the serum anion gap exceeds the fall in bicarbonate concentration, reflecting in part differences in the volume of distribution of H+ and lactate ions (H+ is buffered in intracellular and extracellular compartments, whereas lactate anion is confined to the extracellular space). By contrast, the Δ anion gap to ΔHCO3 ratio in diabetic ketoacidosis is ˜1:1. However, irrespective of the precise nature of the relationship between the ΔHCO3 and the Δ anion gap, the sum of the value for the Δ anion gap and the prevailing blood HCO3 concentration allows an approximation of the basal value of the blood HCO3 concentration existing prior to the development of the high anion gap metabolic acidosis. This concept is important both for distinguishing between a high anion gap metabolic acidosis and a mixed high and normal anion gap metabolic acidosis, and for detecting the presence of a mixed high anion gap metabolic acidosis and metabolic alkalosis.

Clinical Findings

The clinical findings associated with metabolic acidosis are relatively limited and primarily related to the underlying cause. Hyperventilation reflecting the respiratory response to the metabolic acidosis might be observed. With severe degrees of metabolic acidosis organ dysfunction such as impaired cardiac output and hypotension might be apparent. With chronic metabolic acidosis, bone disease and muscle wasting may be present, which produce clinical abnormalities.

The nature of the adverse effects of metabolic acidosis depends on both the duration of the metabolic acidosis and its severity. It is valuable to consider the effects of acute and chronic metabolic acidosis separately.

Adverse Effects of Acute Metabolic Acidosis

The adverse effects of acute metabolic acidosis include decreased cardiac output and hypotension, impaired glucose control, decreased tissue perfusion, reduced oxygen delivery, and induction of an inflammatory state. There is a direct relationship between the severity of the acidemia and the appearance of these changes, with most appearing when the blood pH is less than 7.1–7.2. Although not definitively proven, there is a correlation between the severity of acidemia and mortality, increasing at lower values of blood pH. These findings have an important bearing on the approach to treatment taken by most clinicians (see below).

Adverse Effects of Chronic Metabolic Acidosis

The adverse effects of chronic metabolic acidosis are distinctly different from those of acute metabolic acidosis. Chronic metabolic acidosis has been shown to impair bone metabolism contributing to the genesis of osteomalacia and/or osteitis fibrosa and might also play a role in the genesis of osteoporosis. Muscle wasting has been demonstrated in both experimental and clinical studies of chronic metabolic acidosis, which improves with correction of the acidosis. Similarly, chronic metabolic acidosis can contribute to the genesis of hypoalbuminemia. Insulin resistance due to impaired ligand binding may contribute to abnormal glucose tolerance. In experimental studies in animals, metabolic acidosis can exacerbate renal disease, but this remains to be shown in humans. Cardiac disease so prominent in acute metabolic acidosis has not been shown to be present by chronic metabolic acidosis, although mortality in dialysis patients was increased in the presence of acidosis. Of interest, many of these abnormalities can be seen with even mild metabolic acidosis, a finding that has important implications for treatment.

Treatment

Treatment of acute metabolic acidosis remains one of the more controversial issues in clinical medicine. Although severe acidemia (blood pH <7.1 to 7.2) has been shown to have important adverse effects on organ function, base administration to raise blood pH has not been demonstrated to improve the outcome of lactic acidosis or ketoacidosis, the two disorders in which this issue has been examined. Moreover, in some studies, when bicarbonate is given, a decrease in cardiac output has been found in patients with lactic acidosis and exacerbation of cerebral edema has been noted in children with diabetic ketoacidosis. Benefits and possible adverse effects of treatment of normal anion gap metabolic acidosis have not been examined in a prospective way. These adverse effects of bicarbonate administration have been attributed in part to a bicarbonate-induced reduction in the intracellular pH rise in cellular sodium and fall in ionized calcium. We presently recommend administration of bicarbonate when blood pH is less than 7.1 as a constant infusion rather than a bolus. A calcium infusion might be indicated if ionized calcium falls. Moreover, we target a blood pH ˜7.2 but not higher initially. We also recommend considering alternative bases such as tris (hydroxymethyl) aminomethane (THAM) or the use of continuous renal replacement therapy.

Treatment of chronic metabolic acidosis is less controversial because the side effects of treatment are less severe. Since even mild metabolic acidosis can contribute to abnormalities of bone and muscle metabolism, we recommend complete normalization of acid–base balance. This can be achieved with administration of either oral bicarbonate or other bases that are metabolized to produce bicarbonate, such as Shohl's solution (sodium citrate). The latter is preferred because oral bicarbonate can produce gas that patients do not tolerate well. Possible consequences of base administration include potentiation of vascular calcifications and volume excess and exacerbation of hypertension (from accompanying sodium).

The respiratory acid–base disorders, respiratory alkalosis and respiratory acidosis, are commonly seen in intensive care units and emergency rooms, as well as in general practice. Both are caused by changes in alveolar ventilation that lead to a rise or fall in the partial pressure of CO2 in arterial blood (Pco2). The clinical importance of these disorders, however, is very different. While respiratory alkalosis rarely requires specific treatment, respiratory acidosis accompanies some of the most dramatic presentations of illness a physician will see.

CO2 is produced by metabolism and eliminated by ventilation. Alveolar CO2 concentration is conveniently measured as Pco2. Pco2 is inversely proportional to alveolar ventilation (VA), so that anything that increases VA (an increase in respiratory rate or tidal volume, or improved V̇/Q̇ matching) causes a decrease in Pco2, and anything that decreases VA (a decrease in respiratory rate or tidal volume, or impaired V̇/Q̇ matching leading to increased dead space) causes a decrease in Pco2.

Understanding the diagnosis and treatment of the respiratory acid–base disorders requires knowledge of the buffer systems that serve to protect against alterations in hydrogen ion concentration and pH. As CO2 enters the blood, it combines with H2O to form carbonic acid (H2CO3), which dissociates into bicarbonate (HCO3−) and hydrogen ions:

Most of the H+ ions produced by this addition of CO2 to the blood combine with intracellular buffers including hemoglobin, and this tissue buffering minimizes the elevation in hydrogen ion concentration and the corresponding fall in pH.

In respiratory acid–base disturbances, a primary rise or fall in Pco2 (respiratory acidosis or alkalosis, respectively) will result in a change in pH unless a proportional change in HCO3− occurs to compensate. The initial response to a primary change in Pco2 is tissue buffering of the change with movement of intracellular HCO3− to or from the extracellular fluid. Subsequently, renal HCO3− excretion is adjusted, and serum HCO3− changes further to defend against pH changes.

Respiratory Alkalosis

Essentials of Diagnosis

• Low Pco2 and high pH.
• Acute respiratory alkalosis: HCO3− falls 2 mEq/L for each 10 mm Hg fall in Pco2 (in minutes).
• Chronic respiratory alkalosis: HCO3− falls 5 mEq/L for each 10 mm Hg fall in Pco2 (over days). pH may return to normal!
• If calculated compensation is too little or too much, another acid–base abnormality (a mixed disorder) must be present.

General Considerations

Respiratory alkalosis is the result of hyperventilation produced by a variety of influences. It rarely requires specific treatment, other than for the underlying condition. In fact, overly aggressive efforts to treat respiratory alkalosis itself are often fruitless or dangerous, and occasionally cause respiratory acidosis.

Pathogenesis

With hyperventilation, CO2 is eliminated out of proportion to its production, the equilibrium described above shifts to the left, hydrogen ions are utilized, and pH rises. This decrease in Pco2 (hypocapnia) and rise in pH constitute respiratory alkalosis.

Acute reduction in Pco2 releases H+ from tissue buffers, titrating HCO3− and decreasing its concentration. Eventually, decreased Pco2 also inhibits renal tubular reabsorption and generation of HCO3−, the serum level falls further, and pH returns toward normal.

Prevention

The only common cause of respiratory alkalosis that can be prevented occurs in a mechanically ventilated patient when the chosen ventilator settings produce too high a minute ventilation. This may be difficult to distinguish from the situation in which a ventilated patient develops a respiratory alkalosis because of dyspnea, pain, anxiety, or the underlying disease (see Treatment below).

Clinical Findings

Symptoms and Signs

Chronic respiratory alkalosis is generally asymptomatic, as the blood pH is near normal (Table 5–7). In acute respiratory alkalosis, patients may experience dyspnea, dizziness, anxiety, and acral or circumoral paresthesias. Symptoms are related to both decreased ionized calcium and reduced cerebral blood flow (see below).

Laboratory Findings

Respiratory alkalosis is diagnosed by arterial blood gas analysis showing a high pH, decreased Pco2, and variably decreased serum HCO3−. It must be distinguished from metabolic acidosis in which the Pco2 and HCO3− are also decreased, but pH is low. Accurate diagnosis requires knowledge of the magnitude of the expected compensation (fall in HCO3−) and the duration of the abnormality. An initial decrease in HCO3− occurs in minutes in response to respiratory alkalosis, but full renal compensation for chronic respiratory alkalosis takes days to develop. The data describing appropriate compensation for acute and chronic respiratory alkalosis come from studies of hyperventilation in normal volunteers.

1. Acute respiratory alkalosis: HCO3− falls 2 mEq/L for each 10 mm Hg fall in Pco2 (in minutes).

2. Chronic respiratory alkalosis: HCO3− falls 5 mEq/L for each 10 mm Hg fall in Pco2 (over days). pH may return to normal!

3. If calculated compensation is too little or too much, another acid–base abnormality (a mixed disorder) must be present.

Imaging Studies and Special Tests

Imaging studies and specialized testing are generally not helpful in the diagnosis and management of respiratory alkalosis, although such testing may be appropriate in the management of the underlying disorder.

Differential Diagnosis

Most conditions causing hyperventilation and respiratory alkalosis (with the exception of mechanical ventilation) do so via central respiratory stimulation, increasing the minute ventilation and alveolar ventilation and therefore lowering the Pco2. Table 5–8 lists the common causes.

Anxiety and pain are common causes of hyperventilation and respiratory alkalosis; the act of obtaining a blood gas sample is likely to produce sufficient hyperventilation to demonstrate an acute respiratory alkalosis. Hypoxia in the patient with respiratory disease is also very common and often overlooked as a cause of respiratory alkalosis.

The cause of respiratory alkalosis in both liver failure and pregnancy is believed to be elevated levels of both progesterone and estradiol. Progesterone increases ventilation by acting on central nervous system progesterone receptors, while estradiol is thought to increase the number of these receptors. Increased levels of progesterone and estradiol are part of the normal physiologic milieu of pregnancy, while in liver failure they are caused by the inability of the diseased liver to metabolize free hormones.

Aspirin, although its effect in causing hyperventilation is well known, commonly causes a mixed acid–base pattern: Respiratory alkalosis and metabolic acidosis. The same pattern is seen in gram-negative sepsis. The hyperventilation seen with gram-negative sepsis, or even with fever alone, is caused by the effects of inflammatory mediators, most prominently tumor necrosis factor and the interleukins.

Complications

Alkalosis directly enhances neuromuscular excitability and modestly decreases serum calcium. As mentioned above, these effects cause paresthesias, numbness, twitching, and with severe alkalosis, tetany. Severe alkalosis and hypocapnia can cause dizziness, confusion, and loss of consciousness due to cerebral vasospasm with decreased cerebral blood flow. In fact, intentional production of respiratory alkalosis (using mechanical ventilation) and subsequent cerebral vasospasm is a short-term treatment for increased intracranial pressure.

Treatment

Respiratory alkalosis itself is rarely a clinically important problem. Sedation, analgesia, and antipyretics (for anxiety, pain, and fever) are often sufficient therapy.

In other cases, treating the underlying problem (oxygen for hypoxia, hemodialysis for aspirin overdose, treating infections) is necessary. In patients with respiratory alkalosis complicating mechanical ventilation, simply adjusting ventilator settings or changing the mode of ventilation [eg, to synchronized intermittent mechanical ventilation (SIMV)] should not be considered an adequate “treatment.” These maneuvers can produce an apparent improvement in arterial blood gases, but at the cost of a fatigued or dyspneic patient whose reason for hyperventilation remains undiagnosed and untreated. While settings should be checked for appropriateness, a more effective approach to treating alkalosis in this setting is to search for underlying causes of hyperventilation and treat them. Typically, pain, anxiety, or dyspnea due to a concurrent or new condition (pneumonia, pulmonary edema, bronchospasm, retained secretions) is the cause of a new respiratory alkalosis in a previously stable mechanically ventilated patient.

Prognosis

Prognosis in respiratory alkalosis is entirely dependent on the underlying cause; patients may recover from an episode of anxiety-related hyperventilation in minutes or remain chronically subject to increased ventilatory drive from severe lung diseases such as interstitial fibrosis.

Respiratory Acidosis

Essentials of Diagnosis

• Low Pco2 and high pH.
• Acute respiratory acidosis: HCO3− rises 1 mEq/L for each 10 mm Hg rise in Pco2 (in minutes).
• Chronic respiratory acidosis: HCO3− rises 3.5 mEq/L for each 10 mm Hg rise in Pco2 (over days).
• If calculated compensation is too little or too much, another acid–base abnormality (a mixed disorder) must be present.

General Considerations

Unlike respiratory alkalosis, respiratory acidosis, particularly acute respiratory acidosis caused by severe pulmonary disease or sedative overdose, may produce dangerous hypercapnia and acidosis and often requires urgent treatment.

Pathophysiology

With hypoventilation, elimination of CO2 is unable to keep pace with its metabolic production, CO2 is retained, and the equilibrium above shifts to the right. Excess hydrogen ions are produced, and pH falls. This rise in Pco2 (hypercapnia) and fall in pH constitute respiratory acidosis.

Decreased ventilation quickly results in increased Pco2 because metabolic production of CO2 is so rapid. Acutely, tissue buffering slightly raises HCO3, limiting the pH drop. Eventually, renal acid excretion increases, HCO3− reabsorption is stimulated, serum HCO3− rises, and pH returns toward normal.

Two pathophysiologic mechanisms produce hypercapnia and respiratory acidosis: Severe ventilation–perfusion (V̇/Q̇) mismatch of the dead space (high V̇/Q̇) type and alveolar hypoventilation. Again, since Pco2 is inversely proportional to VA, any process that decreases VA (alveolar hypoventilation or V̇/Q̇ mismatch) causes a rise in Pco2.

In patients with chronic respiratory disease and hypercapnia, excess supplemental oxygen can cause hypoventilation and V̇/Q̇ mismatch (via the release of hypoxic vasoconstriction), causing worse hypercapnia and respiratory acidosis (see Prevention and Differential Diagnosis, below).

Prevention

Few of the causes of respiratory acidosis can be prevented. One exception (see Pathogenesis, above and Differential Diagnosis, below) is the iatrogenic acute respiratory acidosis that can result from overzealous oxygen administration in patients with severe, chronic respiratory disease who are adapted to hypercapnia and dependent on hypoxia to stimulate ventilation. In these patients, oxygen should be administered by low flow (nasal cannulas) or controlled dose (Venturi mask) methods to deliver the minimal FiO2 necessary to correct hypoxemia.

In addition, extreme care should be used in administering sedative drugs to patients with underlying hypercapnic lung disease.

Clinical Findings

Symptoms and Signs

The symptoms and signs of respiratory acidosis (Table 5–9) depend on how quickly the acidosis develops (because a rapid rise in brain Pco2 is not quickly compensated for by a rise in brain HCO3−) and are related to its effect on brain pH. Hypercapnia lowers brain pH and produces cerebral vasodilation, increased cerebral blood flow, and increased intracranial pressure with symptoms and signs similar to the effects of narcotic agents. Early symptoms may include blurred vision, headache, restlessness, tremors, and delirium. These may progress to drowsiness, lethargy, and coma as Pco2 rises and pH falls.

Laboratory Findings

Respiratory acidosis is diagnosed by arterial blood gas analysis showing a low pH, elevated Pco2, and variably elevated serum HCO3−. Respiratory alkalosis must be distinguished from metabolic alkalosis in which the Pco2 and HCO3− are also elevated, but pH is high. When a blood gas is not immediately available, or when questioning the duration of respiratory acidosis, the measured serum HCO3− may also be helpful. In an appropriate clinical setting, a significantly elevated serum HCO3− from a previous arterial blood gas (at least a few days prior) is indirect evidence of a chronic respiratory acidosis. As mentioned, an elevated serum HCO3− may also represent metabolic alkalosis.

As for respiratory alkalosis, accurate diagnosis requires knowledge of the magnitude of the expected compensation (rise in HCO3−) and the duration of the abnormality. An initial increase in HCO3− from tissue buffering occurs in minutes in response to respiratory acidosis, but maximal renal compensation for chronic respiratory acidosis takes days to develop. The data describing appropriate compensation for acute and chronic respiratory acidosis come from several studies of dogs, normal humans, and patients with severe underlying pulmonary disease.

1. Acute respiratory acidosis: HCO3− rises 1 mEq/L for each 10 mm Hg rise in Pco2 (in minutes).

2. Chronic respiratory acidosis: HCO3− rises 3.5 mEq/L for each 10 mm Hg rise in Pco2 (over days).

3. If calculated compensation is too much or too little, another process (a mixed disorder) must be present.

Imaging Studies and Special Tests

Imaging studies and specialized testing are generally not helpful in the diagnosis and management of respiratory acidosis, although such testing may be appropriate in the management of the underlying disorder.

Differential Diagnosis

Common causes of respiratory acidosis are listed in Table 5–10.

Acute Respiratory Acidosis

Acute respiratory acidosis most frequently results from iatrogenic or intentional overdose of a sedative drug (opiates, benzodiazepines). These drugs suppress central respiratory drive causing alveolar hypoventilation, hypercapnia, and respiratory acidosis. Severe, acute exacerbations of any respiratory disease (eg, asthma) can also cause acute respiratory acidosis. These conditions produce severe V̇/Q̇ mismatch and a high work-of-breathing with respiratory muscle fatigue, both leading to hypercapnia and respiratory acidosis.

One important, preventable cause of acute respiratory acidosis is suppression of the hypoxic drive to breathe by administered oxygen in patients with chronic respiratory disease and hypercapnia (see Prevention and Pathogenesis, above).

Chronic Respiratory Acidosis

The most important and most common cause of chronic respiratory acidosis is severe emphysema [chronic obstructive pulmonary disease (COPD)]. In this condition, severe V̇/Q̇ mismatch and respiratory muscle weakness/dysfunction produce hypoventilation, hypercapnia, and respiratory acidosis.

Other severe respiratory (eg, pulmonary fibrosis), neuromuscular (Guillain–Barré syndrome), and chest wall (kyphoscoliosis) diseases are less common causes of chronic respiratory acidosis. These conditions are characterized by severe V̇/Q̇ mismatch, respiratory muscle dysfunction, and/or hypoventilation, leading to hypercapnia.

“Acute‐on-Chronic” Respiratory Acidosis

Patients with severe but compensated respiratory diseases and chronic respiratory acidosis may develop an acute respiratory acidosis when an acute insult (pneumonia, pulmonary embolus, flare of the underlying disease) occurs. The pathophysiologic mechanisms at work here are usually severe V̇/Q̇ mismatch due to the underlying disease and hypoventilation due to respiratory muscle fatigue and an increased work of breathing.

Complications

Unlike respiratory alkalosis, respiratory acidosis is often clinically important. Severe respiratory acidosis and hypercapnia simulate the effects of opiates. They can cause hypotension, confusion and obtundation, and ultimately coma (see Table 5–11 and Symptoms and Signs above). Hypercapnia causes cerebral vasodilation, which may result in increased intracranial pressure and papilledema.

Since hyperventilation is the primary defense against metabolic acidosis, patients with chronic respiratory disease who develop metabolic acidosis are at greater risk and may need early mechanical ventilation.

Treatment

Treatment is focused on improving ventilation. When chronic respiratory acidosis results from respiratory disease, optimally treating the underlying condition (eg, COPD) will simultaneously treat the acidosis.

When acute (or acute-on-chronic) respiratory acidosis is caused by acute exacerbations of chronic respiratory disease or drug overdose, mechanical ventilation may be necessary as well as specific antidotes for the drug ingested.

Although CO2 production is one of the determinants of alveolar and arterial Pco2, it is rare for increased CO2 production to play a significant role in the development of respiratory acidosis. Still, it is occasionally possible to improve hypercapnia in cases of end-stage pulmonary disease by decreasing CO2 production with a low carbohydrate diet. This measure is rarely necessary or useful in other clinical settings.

Prognosis

Prognosis in respiratory acidosis depends on the underlying etiology of the disorder. Patients may recover fully from acute acidosis associated with drug overdose or exacerbation of asthma. Patients with severe, hypercapnic obstructive lung disease generally follow an inexorably downhill course.