4 - Acid-Base Disorder | Manual of Nephrology. Diagnosis and Therapy 6e

Editors: Schrier, Robert W.

Title: Manual of Nephrology, 6th Edition

> Table of Contents > 4 - The Patient with an Acid-Base Disorder

The Patient with an Acid-Base Disorder

William D. Kaehny

Acid-base disorders occur commonly in hospitalized patients and are important markers of an underlying disordered physiology. Occasionally, acid-base disorders disrupt homeostasis sufficiently to move the arterial pH into a dangerous range (less than 7.10 or greater than 7.60). Depending on the overall status of the patient and the response of the cardiovascular system, the pH level may require direct attention. After the clinician is alerted to the presence of an acid-base disorder by clinical and laboratory clues, a logical progression of analysis leads to optimal management of the patient.
- Step 1. Measure pH. This identifies acidemia or alkalemia. The change in bicarbonate and partial pressure of CO₂ (PCO₂) indicates whether the primary process is metabolic or respiratory.
- Step 2. Check the compensatory or secondary response of the PCO₂ or HCO₃^- to see if the disorder is simple or mixed.
- Step 3. Calculate the serum anion gap (AG) to screen for an increase in organic anions such as lactate. Add any increase in AG ( AG) that is potential HCO₃^- to the serum total carbon dioxide (CO₂) content (tCO₂) to screen for a hidden metabolic alkalosis.
- Step 4. Determine the cause of the acid-base disorder from the clinical setting and laboratory tests.
- Step 5. Treat the underlying disorder, unless the pH is dangerous either acutely or chronically (such as acidosis affecting bone).
When to suspect acid-base disorders
- Clinical. The underlying cause of the acid-base disorder is most frequently responsible for a patient's signs and symptoms. Certain clinical settings and findings should alert the clinician to the likelihood of an acid-base disorder. Coma, seizures, congestive heart failure, shock, vomiting, diarrhea, and renal failure generate changes in the PCO₂ or HCO₃^- levels. Marked changes in the pH occasionally may cause direct clinical manifestations. Severe alkalemia causes an irritability of heart and skeletal muscle. Severe acidemia causes a depression of heart pump function and vascular tone. Although central nervous system dysfunction appears frequently with acid-base disorders, changes in pH do not appear responsible. Rather, altered plasma osmolality and PCO₂ appear to be the causative agents.
- Laboratory. A thoughtful measurement of serum electrolytes in patients with abnormal losses or gains of body fluids is good practice. An abnormal serum tCO₂ is definite evidence of an acid-base disorder; an abnormal serum AG is very suggestive; an abnormal serum potassium is suspicious.
  - Serum tCO₂. The HCO₃^- in blood can be estimated reasonably by measuring the tCO₂ in venous serum. Acid is added to the serum to release CO₂ from HCO₃^- plus dissolved CO₂, carbonate, and carbonic acid. The CO₂ diffuses into a buffer solution and causes the pH to change. This change is translated into tCO₂ in mmol per L. The serum tCO₂ is 1 to 3 mmol per L greater than the arterial HCO₃^- because it is from venous blood, which has more HCO₃^-, and it includes more than HCO₃^-. Normal sea level serum tCO₂ levels average 26 to 27 mmol per L. A value below 24 or above 30 likely marks a clinical acid-base disorder. An acid-base disorder of the mixed type may exist with a normal serum tCO₂.
  - The serum anion gap is calculated from the venous serum sodium, chloride, and tCO₂:
    
    The units are mEq per L, because this calculation estimates the charge difference between the so-called unmeasured anions (serum total anions represented by Cl^- and HCO₃^-) and unmeasured cations (total cations represented by Na⁺). The average normal value is 9 3 mEq per L, but may vary in a given laboratory. Albumin contributes most to the AG. A fall in serum albumin of 1 g per dL from a normal of 4.5 decreases the AG by 2.5 mEq per L.
    - Metabolic acidosis due to an organic acid such as lactic or acetoacetic acid is marked by an increased AG. An increase in the AG of 8 mEq per L to 17 or higher usually indicates the presence of organic acidosis, although at times the exact anion may not be identified. The anion of the organic acid replaces the HCO₃^- lost in the buffering of the hydrogen ion (H⁺) part of the acid and thus increases the unmeasured anions. Importantly, a normal or slightly elevated AG does not rule out the presence of organic metabolic acidosis, such as diabetic ketoacidosis, because a patient with good renal perfusion and ample urine flow may excrete the ketoanions at a rate sufficient to keep the serum AG from rising markedly.
    - Metabolic alkalosis. At times, a metabolic acidosis that increases the AG and lowers the HCO₃^- may coincide with a process that generates a metabolic alkalosis. For example, vomiting that generates a high HCO₃^- may be caused by diabetic ketoacidosis, which lowers the HCO₃^-. In this case, the serum tCO₂ (and arterial HCO₃^-) may be low or normal despite the elevating action of the metabolic alkalosis. The clue to the presence of such hidden metabolic alkalosis is derived in a holmesian fashion by adding the measured serum tCO₂ and the AG (measured AG 9). If this sum is greater than 30 mEq per L, a hidden metabolic alkalosis is likely present. The AG is a marker of lost or potential HCO₃^-, that which was titrated by the H⁺ of an organic acid. Pure metabolic alkalosis may directly increase the AG by up to 5 mEq per L due to effects on the albumin concentration and charge.
  - Serum potassium. Potassium metabolism is linked to acid-base metabolism at the levels of cell shifts, renal tubular functions, and gastrointestinal transport. Therefore, an abnormal serum potassium concentration alerts the clinician to the likelihood that an acid-base disorder is present also.

Identifying the major acid-base disorders. When the clinician suspects that an acid-base disorder might be present and that patient management might be adjusted, a set of acid-base variables should be obtained: pH, PCO_2, and HCO₃^-.

Chemistry and physiology of acid-base. Cellular, tissue, and organ systems apparently function best at an extracellular fluid (ECF) pH of around 7.40. Intracellular fluid (ICF) pH is heterogeneous within the cell, depending on organelles and metabolic activity, but averages around 7.00. ECF pH is a function of the state of available buffers, those molecules that respond to changes in pH by binding or releasing H⁺ to keep pH close to 7.40. Thus, buffers prevent extreme shifts in pH in the face of the gain or loss of acids or bases.
- Blood pH is the mathematical expression of the intensity of acidity or H⁺ activity. It can be translated into H⁺ concentration in mol per L:
  
  P.49
  
  H⁺concentration usually is expressed in nmol per L. H⁺ concentration is 100 nmol per L at pH 7.00 and 40 nmol per L at pH 7.40. Within the pH range of 7.26 to 7.45, H⁺ concentration is estimated accurately as (80 the decimal of pH). For example, at pH 7.32, H⁺ equals 80 minus 32 or 48 nmol per L. The pH is measured at body temperature with a glass, flow-through electrode.
- The partial pressure of carbon dioxide in blood, PCO₂, represents the respiratory component in blood. The respiratory system determines the level at which the PCO₂ is set. PCO₂ substitutes for the buffer carbonic acid, H₂CO₃, in the acid-base equation. H₂CO₃ is formed from water and CO₂. PCO₂ is measured in whole blood with a pH electrode that detects the change caused by the diffusion of CO₂ from the sample into a buffer solution.
- HCO₃^- is the metabolic component of the acid-base equation, serving as the base of the H⁺-binding partner in the buffer pair. HCO₃^- concentration is controlled by the buffering state, metabolic processes, and the kidneys. HCO₃^- concentration is calculated from the pH and PCO₂ using the Henderson-Hasselbalch equation. The fact that it is calculated makes it no less reliable a value than the serum tCO₂, which is also calculated.
- The acid-base equation allows the determination of the state of ECF acid-base balance, the presence of an acid-base disorder, the nature of the disorder, and the presence of a simple or mixed disorder:
  
  Thus, the pH level depends on the ratio or mathematical relationship between the HCO₃^- and the PCO₂. An acid-base disorder is generated by an alteration from normal of either of these two factors. The resultant change in pH results in chemical shifts in the buffers, which mitigate the change in pH somewhat. A physiologic response occurs in the respiratory system for a metabolic disorder and in the kidneys for a respiratory disorder. A new steady state ensues, with the new pH set by the new values of the HCO₃^- concentration and PCO₂.
Measurement of acid-base variables. The determination of the acid-base state usually is based on an analysis of arterial blood, although arterialized venous blood analysis is equally valid. After warming the extremity, blood is drawn without air mixing from an artery or from a forearm vein without tourniquet. Although experimental studies show that ICF pH and mixed venous acid-base measurements correlate with organ function, arterial blood measurements are easily available and provide a readily interpretable view of the metabolic state of organs and their function. Keep in mind that tissue hypoperfusion, as in cardiopulmonary arrest or profound shock, makes tissue acidosis worse than that reflected by arterial blood acid-base values.
- Calculation of HCO₃^- from pH and PCO₂. Arterial HCO₃^- is 1 to 3 mmol per L less than venous tCO₂. To calculate the HCO₃^- or verify the reported value, convert the pH to H⁺ and use a simplified form of the Henderson equation:

Identification of a major acid-base disorder. The basis of this approach is to determine the direction (up or down) in which the measured values differ from the arbitrary normal values for pH (7.40), PCO₂ (40 mm Hg), and HCO₃^- (24 mmol per L). First, determine if acidemia (pH down) or alkalemia (pH up) is present. Then determine if the primary generating change was in the HCO₃^- or in the PCO₂ (Table 4-1). The compensating

P.50

factor should change in the same direction as the generating factor to yield a simple acid-base disorder.

Table 4-1. Simple Acid-Base Disorders


	Metabolic acidosis	Metabolic acidosis	Respiratory acidosis	Respiratory acidosis

Primary change	HCO₃^-	HCO₃^-	Pco₂	Pco₂
Compensation	Pco₂	Pco₂	HCO₃^-	HCO₃^-
Effect on pH	pH	pH	pH	pH

decreased;
increased.

Example of a simple disorder. Arterial blood analysis revealed the following values: pH 7.55, HCO₃^- 18 mmol per L, PCO₂ 21 mm Hg.
- Step 1. The pH is up. Thus, alkalemia is present and must be due to an increased HCO₃^- (as in metabolic alkalosis) or to a decreased PCO₂ (as in respiratory alkalosis).
- Step 2. The HCO₃^- is low and cannot be responsible for an increased pH.
- Step 3. Because the PCO₂ is low, it can account for the increased pH; this is respiratory alkalosis.
- Step 4. The HCO₃^- change is in same direction as that of the PCO₂; this is consistent with a simple respiratory alkalosis.
Example of a mixed acid-base disorder. Sampling of arterial blood yielded the following: pH 7.55, HCO₃^- 30 mmol per L, and PCO₂ 35 mm Hg.
- Step 1. The pH is up. Thus, alkalemia is present.
- Step 2. The HCO₃^- is increased and may be responsible for the increased pH.
- Step 3. The PCO₂ is low and it, too, can account for an increased pH.
- Step 4. The two acid-base determinants are changed from normal in opposite directions. Thus, this is mixed metabolic and respiratory alkalosis. The metabolic alkalosis is dominant because the percent change in HCO₃^- is 6/24 or 25% whereas the percent change in PCO₂ is 5/40 or 12.5%.

Judging whether an acid-base disorder is simple or mixed. When an underlying process generates an acid-base disorder by perturbing one member of the HCO₃^- PCO₂ buffer pair (remember that PCO₂ represents H₂CO₃), the other partner is adjusted by the physiologic response of the body and changes in the same direction as the primary partner in order to reduce the magnitude of the change in pH. The time-honored term for this physiologic response is compensation. However, the physiologic response mechanisms may be activated by stimuli other than pH and actually may contribute to the maintenance of the abnormal pH. Thus, some have termed these responses maladaptive, because they are not always truly compensatory. For example, a low PCO₂ in response to metabolic acidosis actually causes the kidneys to reduce HCO₃^- reabsorption. Importantly, compensation does not restore the pH to normal, because that would shut off the stimulus for the compensatory mechanism.

Steps in judging whether an acid-base disorder is simple. After the major disorder is identified, determine whether the compensation for the primary event is appropriate.

Check directions of changes from normal of HCO₃^- and PCO₂. The acid-base buffer pair change from normal in the same direction in all simple acid-base disorders. If they change in opposite directions, the disorder must be mixed. This step was discussed in section III.C.2.d.

Compare the magnitude of the compensation of the PCO₂ or HCO₃^- with the primary change in the HCO₃^- or PCO₂. In metabolic disorders,

P.51

the primary change occurs in the HCO₃^-, with the compensation occurring in the PCO₂. The opposite is true in the respiratory disorders. Table 4-2 contains guidelines or rules that can be used to judge whether compensation is appropriate. The respiratory disorders have two stages of compensation: acute, when only tissue buffering slightly changes the HCO₃^-, and chronic (after 24 hours), when the kidneys cause major changes in the HCO₃^- concentration. If the measured change in the compensating factor does not approximate the change predicted, a mixed disorder is likely. Two methods of predicting compensation appear in Table 4-2. One describes the expected changes in the buffer partner for a given change in the generating partner. For example, a fall in HCO₃^- of 10 mmol per L in metabolic acidosis is expected to result in hyperventilation that drops the PCO₂ by 10 to 15 mm Hg to 25 to 30 mm Hg. The other method relates the primary, generating factor change to a change in pH. For example, a fall in HCO₃^- of 10 mmol per L in metabolic acidosis is expected to decrease the pH by 0.1 to 7.30.

Table 4-2. Appropriate Compensations in the Acid-Base Disorders


	Change in PCO₂ per change in HCO₃^-	Change in pH per change in HCO₃^-
Metabolic acidosis	1.0 1.5 per 1	0.010 per 1
Metabolic alkalosis	0.25 1.00 per 1	0.015 per 1

	Change in HCO₃^- per change in PCO₂	Change in pH per change in PCO₂
Respiratory acidosis
Acute	1 per 10	0.08 per 10
Chronic	4 per 10	0.03 per 10
Respiratory alkalosis
Acute	1 per 10	0.08 per 10
Chronic	4 per 10	0.03 per 10

Check the anion gap for evidence of a hidden metabolic disorder. An increase in the AG of more than 8 mEq per L to greater than 17 suggests the presence of metabolic acidosis due to an organic acid. Also if the AG is added to the measured serum tCO₂, the theoretical maximum serum tCO₂ can be estimated. A value greater than 30 mmol per L suggests metabolic alkalosis.

Application of the rules
- The primary event in metabolic acidosis is a fall in HCO₃^-; the compensation is a fall in the PCO₂, due to the stimulation of central nervous system receptors by the low pH. Hyperventilation increases the excretion of CO₂, and PCO₂ falls. For example, if the HCO₃^- falls from 24 mmol per L by 10 to 14 mmol per L, the PCO₂ should fall by 1.0 to 1.5 times as much, or 10 to 15 mm Hg, to a level of 25 to 30 mm Hg (40 10 = 30; 40 15 = 25).
- The primary event in metabolic alkalosis is a rise in HCO₃^-. The respiratory system responds to the rise in pH with hypoventilation, which reduces carbon dioxide excretion and results in a rise in PCO₂. For example, if HCO₃^- rises by 16 mmol per L from 24 to 40 mmol per L, the PCO₂ should rise by 0.25 to 1.00 multiplied by the rise in HCO₃^- of 16, or by 4 to 16 mm Hg, to a level of 44 to 56 mm Hg (40 + 4 = 44; 40 +16 = 56). This response is tempered by the body's response to the concomitant hypoxemia resulting from hypoventilation.
- The primary event in respiratory acidosis is a rise in PCO₂. During the acute phase (up to 24 hours), only buffering contributes measurably
  P.52
  
  to the response. The HCO₃^- should increase, but not to as high as 30 mmol per L. In contrast, the kidneys respond to chronic elevations of PCO₂ by generating sufficient HCO₃^- to prevent the pH from falling to less than 7.20, even in the most severe cases of chronic respiratory acidosis.
- The primary event in respiratory alkalosis is a fall in PCO₂. Initially, buffering occurs as a result of the release of H⁺ from cells; later (hours) the kidneys dump HCO₃^- into the urine, with a resultant fall in blood HCO₃^-, as defined in Table 4-2.
Effects of respiratory responses to metabolic disorders. The kidneys respond to changes in PCO₂ regardless of the pH. A fall in PCO₂ causes renal HCO₃^- loss; a rise in PCO₂ causes renal HCO₃^- generation. Thus, in chronic metabolic acidosis (lasting days), some of the reduction in bicarbonate actually is due to the compensatory fall in PCO₂ and not directly to the process causing the metabolic acidosis. Similarly, the increase in PCO₂ in chronic metabolic alkalosis contributes to the hyperbicarbonatemia.

Examples of mixed acid-base disorders. Four combinations of the double mixed acid-base disorders are possible. Two are important because they cause drastic changes in pH: metabolic and respiratory acidosis and metabolic and respiratory alkalosis. The other two disorders tend to be associated with pH values close to normal and are not dangerous per se; however, they are important markers of underlying disease. Two other mixed disorders, so-called triple disorders, also have been described. The anion gap points to both metabolic acidosis and alkalosis developing simultaneously or sequentially in these. The imposition of a respiratory disorder yields the infamous triple acid-base disorder.

Metabolic acidosis and respiratory acidosis. A patient with emphysema and carbon dioxide retention (chronic respiratory acidosis) develops diarrhea (metabolic acidosis). Note how the reduction in HCO₃^- to normal results in severe acidemia (Table 4-3).

Table 4-3. Example of a Mixed Acid-Base Disorder

Health Emphysema Emphysema and diarrhea

pH 7.40 7.32 7.10

Pco₂ 40 80 80

HCO₃^- 24 40 24
Metabolic alkalosis and respiratory acidosis. The same emphysematous patient is given a diuretic for cor pulmonale. The bicarbonate level rises from 40 to 48 mmol per L, which, with the PCO₂ at 80 mm Hg, sets the pH at 7.40. Although this is a normal pH, some believe it is better for carbon dioxide-retaining patients to be mildly acidemic to keep ventilation stimulated.

Triple acid-base disorder. A more common mixture of disorders involves metabolic acidosis developing in a patient with metabolic alkalosis and superimposed respiratory alkalosis. For example, a patient with metabolic alkalosis (HCO₃^- 32) from nasogastric suction becomes septic, which generates both lactic acidosis and pronounced hyperventilation, thus causing independent respiratory alkalosis due to endotoxin (Table 4-4). Note that both the metabolic and respiratory alkaloses should cause only small increases in the anion gap. The lactic acidosis of septic shock results in a fall of from 32 to 24 mmol per L in the HCO₃^- with a reciprocal increase in anion gap. The anion gap of 33 is diagnostic of organic acidosis. The AG of 26 (35 9) added to the serum tCO₂ of 9 yields 35 mmol per L, an estimate of the value before acidosis, indicative of metabolic alkalosis. The evidence for the presence of respiratory alkalosis is the high pH and low PCO₂ due to the hyperventilation caused by endotoxemia.

Table 4-4. Example of a Triple Mixed Acid-Base Disorder


	Health	Nasogastric suction	Septic shock	Endotoxemia

pH	7.40	7.49	7.14	7.44
Pco₂	40	44	24	12
HCO₃^-	24	32	8	8
Anion gap	9	11	33	35
Venous total carbon dioxide	26	35	9	9
Disorder		Metabolic alkalosis	Metabolic alkalosis	Metabolic alkalosis
			Metabolic acidosis	Metabolic acidosis
				Respiratory alkalosis

P.53

Identify the underlying cause of an acid-base disorder. Usually, the cause of an acid-base disorder is obvious from the history, examination, and clinical course. However, on occasion, careful review of a thoughtful differential diagnosis is necessary to identify a remote causative disorder.

Causes of metabolic acidosis. The anion gap is used to divide the causes of metabolic acidosis into those with influx of organic acid into plasma (increased anion gap) and those with external losses of bicarbonate (normal anion gap; hyperchloremic). Some disorders belong to both groups at different stages (diabetic ketoacidosis) or are generated by mechanisms other than those described (renal failure). A list of causes is given in Table 4-5.

Table 4-5. Causes of Metabolic Acidosis

High anion gap type

Ketoacidosis

Diabetic

Alcoholic

Starvation

Lactic acidosis

Type A

Type B

D-Lactic acidosis

Poisonings

Ethylene glycol

Methanol

Salicylate

Advanced renal failure

Normal anion gap type

Gastrointestinal HCO₃^- loss

Diarrhea

External fistulae

Renal HCO₃^- loss

Acetazolamide

Proximal renal tubular acidosis (RTA)

Distal RTA

Hyperkalemic RTA

Miscellaneous

NH₄Cl ingestion

Sulfur ingestion

Toluene inhalation

Pronounced dilution

Metabolic acidosis with increased anion gap. Severe metabolic acidosis is caused by only three broad groups of disorders: ketoacidosis, lactic acidosis, and poisonings. In addition, renal failure may cause mild to moderate acidosis. This yields the mnemonic KlaPR (pronounced clapper) for ketoacidosis, lactic acidosis, poisonings, renal failure.
- Ketoacidosis arises when glucose is not available to cells because of insulin lack, cell dysfunction, or glucose depletion, and fatty acids are oxidized to yield energy, acetone (not an acid), and the two ketoacids (acetoacetic and beta-hydroxybutyric). The H⁺ produced are consumed (buffered) by HCO₃^-, producing carbonic acid, which dehydrates into water and carbon dioxide. The ketoanions accumulate in the serum in place of the HCO₃^-, further increasing the anion gap. The diagnostic test for ketoacidosis consists of testing the serum with a nitroprusside reagent, which only reacts with the acetoacetate. In diabetic ketoacidosis, the beta-hydroxybutyrate-acetoacetate ratio averages 5 to 2, whereas in alcoholic ketoacidosis, it may reach as high as 20 to 1. In these cases, only urine, a concentrate of serum, may have a concentration of acetoacetate high enough to give a positive purple reaction with the reagent. Diabetic ketoacidosis occurs because of insulin deficiency. Hyperglycemia may be corrected by volume re-expansion, but insulin is needed to stop ketogenesis. Volume expansion enhances the renal excretion of ketoanions, thus correcting the increased anion gap. However, the kidneys take time to generate new HCO₃^- to replace that lost earlier in buffering H⁺. Thus, early in diabetic ketoacidosis, the anion gap usually is increased; during correction, the anion
  P.54
  
  gap may return to normal despite a low HCO₃^-. Chloride replaces the ketoanions, and this stage is therefore called hyperchloremic or normal anion gap metabolic acidosis. Alcoholic ketacidosis occurs because of volume depletion, which causes the -adrenergic suppression of insulin release. The patient relates a story of severe vomiting following recently increased alcohol intake. Urine ketones are usually positive. Blood glucose ranges between 50 and 250 mg per dL. Starvation ketoacidosis occurs because of the use of fatty acids for energy maintenance. The degree of acidosis is mild, with arterial HCO₃^- not less than 18 mmol per L.
- Lactic acidosis arises when oxygen delivery to cells is inadequate for the demand (type A) or cell processes cannot use oxygen (type B). In this situation, glucose is metabolized via anaerobic glycolysis to pyruvate and then to the dead-end metabolite lactate. The H⁺ produced from nicotinamide adenine dinucleotide (NADH) (one per lactate) are buffered by HCO₃^-, which is replaced in the blood by lactate. Thus, the anion gap is increased. Type A lactic acidosis is caused by the primary inadequate delivery of oxygen to tissues. Shock is the most common mechanism. Hypovolemia, heart failure, and sepsis cause shock. Because carbon monoxide binds more avidly to hemoglobin than does oxygen, carbon monoxide poisoning can cause varying degrees of lactic acidosis. Type B lactic acidosis occurs when tissue oxygenation is normal but tissues cannot use the oxygen normally or need excessive amounts of oxygen. Causes of type B lactic acidosis include hepatic failure, malignancy drugs, and seizures. Metformin is a biguanide
  P.55
  
  hypoglycemic agent that rarely (8 per 100,000) causes lactic acidosis in contrast to its relative, phenformin. Renal, liver, and heart failure are risk factors. Reverse transcriptase inhibitors for AIDS also are rare causes of lactic acidosis due to injury to cell mitochondria. Lactic acidosis has been seen in patients receiving large intravenous doses of lorazepam and diazepam due to the propylene glycol solvent. D-lactic acidosis is generated when colon bacteria metabolize malabsorbed sugars into both L- and D-lactate, which accumulates in the blood. The clinical manifestation is metabolic encephalopathy. At least two dozen inborn errors of metabolism result in pediatric lactic acidosis. The diagnosis is usually established by exclusion of ketoacidosis, intoxications, and advanced renal failure as causes for a high anion gap metabolic acidosis. L-lactate can be measured by an automated assay, but measurement is often unnecessary.
- Three modern poisonings cause high anion gap metabolic acidosis: ethylene glycol ingestion, methanol ingestion, and salicylate intoxication. Methanol and ethylene glycol are low-molecular-weight alcohols that readily enter cells. Metabolism generates H⁺ that cause acidosis and formate (with methanol) or glycolate (with ethylene glycol) that causes a high anion gap. A clue to the presence of early stages of acidosis with elevated alcohol levels is an increased osmolal gap. This gap is the difference between the measured serum osmolality and the calculated osmolality (calc Sosm).
  
  If this difference is greater than 25 mOsm per kg of serum, the presence of a toxic alcohol is likely. Needlelike or envelope-shaped calcium oxalate crystals in the urine suggest ethylene glycol ingestion. The combination of high anion gap metabolic acidosis and a high osmolal gap is an indication for specific analysis for methanol and ethylene glycol. The decision to measure levels of these alcohols, of course, must be tempered by the clinical setting. Salicylate intoxication is an important frequent unintentional chronic or intentional acute overdose that causes metabolic acidosis, respiratory alkalosis, or a mixed disorder. It should be suspected at the extremes of age.
- Renal failure. Failure to excrete the daily acid load of 1 mmol per kg of body weight that is generated by metabolism results in metabolic acidosis. Bone buffers take up some hydrogen ions during chronic renal failure, and, thus, the degree of acidosis is moderated until the end stages of disease. Arterial bicarbonate usually remains above 15 mmol per L. In acute renal failure, venous tCO₂ or arterial HCO₃^- falls by about 0.5 mmol per L per day unless hypercatabolism increases daily acid production. The anion gap increases less than the HCO₃^- falls, resulting in hyperchloremic metabolic acidosis in early and middle stages of chronic renal failure. In advanced chronic renal failure, the serum anion gap rises about 0.5 mEq per L for each 1.0 mg per dL rise in serum creatinine. Retention of sulfate, phosphate, and organic anions causes the increase in anion gap.
Metabolic acidosis with normal anion gap (hyperchloremic) can be caused by three groups of disorders: gastrointestinal HCO₃^- loss, renal HCO₃^- loss or acid retention, and inorganic acid intake.
- Gastrointestinal bicarbonate loss. The gastrointestinal tract distal to the stomach has the capacity to absorb chloride and secrete bicarbonate. Thus, diarrhea and external drainage of pancreatic, biliary, or small-bowel juices can cause external losses of bicarbonate-rich
  P.56
  
  fluid. Relatively chloride-rich fluid remains behind. This generates hyperchloremic metabolic acidosis (normal anion gap). An interesting variety of this disorder occurs when normal urine, rich in sodium chloride (NaCl) from dietary sources, is drained into the gut via ureterosigmoidostomy or ileal loop conduit (both bladder replacement constructions). If contact time with mucosa is excessive, the gut reabsorbs the chloride in exchange for bicarbonate, resulting in hyperchloremic metabolic acidosis.
- Renal bicarbonate loss. The proximal renal tubule reabsorbs the bulk (85%) of filtered HCO₃^-. The carbonic anhydrase inhibitor acetazolamide blocks much of this reabsorption, resulting in urinary bicarbonate losses until arterial HCO₃^- falls to 16 to 18 mmol per L. The filtered load of HCO₃^- at this concentration can be completely reabsorbed by the distal nephron. Thus, the urine becomes bicarbonate-free, with an acidic pH, in this new steady state. Proximal renal tubular acidosis (old type II RTA), a defect in proximal tubular HCO₃^- reabsorption, has identical features. Proximal RTA is unusual but may occur with Wilson's disease, multiple myeloma, transplant rejection, and in other disease states. Distal RTA (old type I) differs in that it is a defect of the collecting duct, in which the daily metabolic acid load is not excreted totally and a small HCO₃^- leak occurs every day. This leads to a mild-to-moderate normal anion gap, hyperchloremic metabolic acidosis, and hypercalciuria with calcium stones or nephrocalcinosis. Two varieties occur: hypokalemic distal RTA and hyperkalemic distal RTA. Hypokalemic distal RTA occurs when collecting duct potassium secretion is intact and in fact enhanced by the small amount of bicarbonaturia. Hyperkalemic distal RTA occurs due to two distinct mechanisms when collecting duct hydrogen ion and potassium secretion are impaired: hypoaldosteronism (old type IV RTA) or tubular defect. Hypokalemic distal RTA occurs with amphotericin B toxicity, cirrhosis of the liver, medullary sponge kidney, and many other diseases. Chronic obstruction of the kidney, lupus erythematosus, and sickle cell disease can cause the tubular defect type of hyperkalemic distal RTA. Diabetes mellitus, mild chronic renal failure, and old age are associated with the hyporeninemic hypoaldosteronism type of hyperkalemic distal RTA. Diagnosis of RTA is made by eliminating non-renal causes for a normal anion gap metabolic acidosis (e.g., diarrhea). Proximal RTA is characterized by urinary wasting of more than 5% to 15% of the filtered load of HCO₃^- when serum levels are maintained close to normal. Hypokalemic distal RTA is characterized by the inability to decrease urine pH to less than 5.4 with an oral acid load of ammonium chloride. Hyperkalemic normal ion gap metabolic acidosis almost always is due to distal RTA. The tubular defect type is marked by an inability to acidify maximally (urine pH usually above 6.0), in contrast to the hypoaldosteronism type, in which the intensity of acidification is intact (urine pH can fall lower than 5.5). In both types, renal ammonium excretion is reduced, and the urine anion gap often is positive (see section V.A.2.d).
- Inorganic acid intake. The ingestion of ammonium chloride to reduce appetite or acidify urine produces hyperchloremic metabolic acidosis. Inorganic sulfur, such as flowers of sulfur cathartic, is oxidized to form H₂SO₄. The hydrogen ions titrate the HCO₃^- down, and the sulfate is excreted rapidly with sodium. This leaves a low HCO₃^- with a normal anion gap. A similar process happens with toluene inhalation from paint or glue sniffing. Toluene is metabolized to hippuric acid, and the hippurate is excreted rapidly.
- On occasion, the urinary anion gap (UAG) may help to distinguish gastrointestinal from renal loss of HCO₃^- as the cause of hyperchloremic
  P.57
  
  metabolic acidosis:
  
  The UAG is an estimate of urinary ammonium that is elevated in gastrointestinal HCO₃^- loss but low in distal RTA. UAG is a negative value if urine ammonium is high (as in diarrhea; average, -20 mEq per L), whereas it is positive if urine ammonium is low (as in distal RTA; average, +23 mEq per L).

Causes of metabolic alkalosis. Metabolic alkalosis is characterized by a low urine Cl^-, indicating the avidity of the kidney for solute, largely NaCl. During the generation phases of metabolic alkalosis, bicarbonaturia may occur and necessitate Na⁺ and K⁺ excretion. Thus, urine Cl^-is a better marker than urine Na⁺ of the stimulation of renal salt reabsorption by volume depletion. During the maintenance stage of metabolic alkalosis, bicarbonaturia is minimal, urine pH is acid, and urine Na⁺ is low. Volume-depleted metabolic alkalosis is due to external losses of hydrogen ion or chloride. Volume-replete metabolic alkalosis is characterized by spot urine Cl^-of usually well over 20 mmol per L. The kidney is not avid for salt because of volume expansion (mild) and thus excretes the daily Na⁺ and Cl^-load without difficulty. This group of disorders is due to hypermineralocorticoidism or profound potassium depletion.

Metabolic alkaloses of the volume-depleted variety have in common the external loss of fluids rich in H⁺ or Cl^-. The stomach, kidney, or skin may be the culprit (Table 4-6).

Table 4-6. Causes of Metabolic Alkalosis

Volume-depleted type

Gastric acid loss

Vomiting

Gastric suction

Renal chloride loss

Diuretics

Post-hypercapnia

Cystic fibrosis

Volume-replete type

Mineralocorticoid excess

Hyperaldosteronism

Gitelman's syndrome

Bartter's syndrome

Cushing's syndrome

Licorice excess

Profound potassium depletion

Metabolic alkaloses of the volume-replete variety are characterized by enhanced renal H⁺ secretion despite normal or increased ECF volume. The stimulus for this sustained H⁺secretion is aldosterone (or a relative) or major cellular potassium depletion (see Table 4-6). Gitelman's syndrome is an autosomal recessive disorder usually appearing in adults as hypokalemic, hypomagnesemic metabolic alkalosis. The distal convoluted tubule Na⁺Cl^-cotransporter is defective. In contrast, Bartter's syndrome appears in childhood as hypokalemic metabolic alkalosis. In this syndrome, defects in loop of Henle Na⁺Cl^-reabsorption lead to normotensive secondary hyperaldosteronism. Diuretic abuse resembles these disorders.

P.58

Causes of respiratory acidosis. Two ventilatory abnormalities allow CO₂ retention and increased PCO₂: alveolar hypoventilation and severe ventilation perfusion mismatching. Hypoxemia occurs in both settings. Causes according to level of respiratory system affected are listed in Table 4-7. Renal compensation for chronic respiratory acidosis produces very high HCO₃^- levels. If the PCO₂ is reduced by artificial ventilation, the high HCO₃^- may persist if not enough chloride is provided to replace it. This results in post-hypercapnic metabolic alkalosis. Acetazolamide may be useful in this setting.

Table 4-7. Causes of chronic respiratory acidosis

Abnormal respiratory drive

Obesity hypoventilation syndrome

Sleep apnea

Brainstem infarcts

Myxedema

Abnormal nerve conduction

Poliomyelitis

Multiple sclerosis

Amyotrophic lateral sclerosis

Phrenic nerve injury

Guillain-Barr syndrome

Thoracic cage problems

Kyphoscoliosis

Ankylosing spondylitis

Muscular dystrophy

Polymyositis

Pleural disease

Fibrothorax

Hydrothorax

Pulmonary disease

Advanced interstitial fibrosis

Chronic obstructive lung disease

Causes of respiratory alkalosis. Disorders that drive ventilation independent of PCO₂ can cause hyperventilation and hypocapnia. Inflammatory and mass lesions of the brain, psychiatric disorders, and certain central-acting drugs and chemicals increase central respiratory drive and produce hypocapnia. Importantly, salicylates, endotoxin, and progesterone are among this group of drugs and chemicals. Disorders that cause hypoxemia are common causes of the hyperventilation that causes hypocapnia. Disorders that reduce lung or chest compliance, such as mild pneumonia or pulmonary edema; vascular disorders, such as emboli; and mixed disorders, such as hepatic cirrhosis or failure, can cause hypocapnia. Volume depletion is a primary stimulus to hyperventilation and hypocapnia.

Treating acid-base disorders. As discussed, acid-base disorders are markers of underlying diseases, and these diseases should be the targets of treatment.
- Step 1. Correct volume and electrolyte deficiencies.
- Step 2. Direct specific treatment at the underlying cause.
- Step 3. Manipulate the bicarbonate or PCO₂ only if the pH is adversely affecting organ function or if pH is less than 7.10 or greater than 7.60.
- Treatment of metabolic acidosis with alkali has not been shown to be efficacious in acute situations, including cardiopulmonary resuscitation, possibly because the HCO₃^- reaction with H⁺ generates CO₂ at the tissue level and lowers cell pH. However, a mixture of carbonate and bicarbonate that
  P.59
  
  produces less CO₂ has not been shown to be of clinical benefit. At present, we cannot recommend alkali therapy for any acute metabolic acidosis. In chronic distal renal tubular acidosis, alkali therapy reduces bone loss, hypercalciuria, and nephrocalcinosis. Oral alkali can be given as sodium bicarbonate tablets, 500 mg or 6 mmol, or as sodium or potassium citrate solution, 1 mmol per mL. Usually 3 mmol per kilogram of body weight is the starting dose. Oral HCO₃^- is accompanied by a sodium or potassium load that must be watched for adverse effects.
  - Insulin is the specific treatment for diabetic ketoacidosis. HCO₃^- and phosphate administration are unnecessary, but potassium repletion is important. Volume and electrolyte repletion, plus glucose and thiamine, suffice to correct alcoholic ketoacidosis. Starvation requires only calories. Note that metabolism of ketoanions, as with all organic anions, produces HCO₃^-; metabolism corrects about half of the HCO₃^- deficit. If alkali is given unwisely, an overshoot metabolic alkalosis may result.
  - In the treatment of lactic acidosis, the restoration of tissue perfusion and oxygenation is desirable, but often difficult to attain. Attention to potassium and calcium levels is important.
  - Treatment of poisonings. Ethylene glycol and methanol poisoning require immediate fomepizole infusion to retard metabolism of the alcohol to toxic products. Hemodialysis is started if renal failure is present. An alternative approach is to infuse ethanol to maintain a blood level of 100 mg per dL to compete for alcohol dehydrogenase activity. The loading dose of ethanol is 0.6 to 1.0 g per kg of body weight followed by a maintenance infusion of 10 to 20 g per hour. Blood alcohol levels must be monitored frequently. Salicylate intoxication should be treated with urinary alkalinization by infusing a 5% glucose solution containing NaHCO₃, 150 mmol added per liter, at 375 mL per hour, for 4 or more hours. Hemodialysis should be used to remove salicylate in patients with prominent renal failure, worsening mental status, recalcitrant acidosis, or general deterioration.
- Correction of metabolic alkalosis very rarely requires the administration of acid. If renal failure prohibits renal excretion of HCO₃^-, the patient usually requires dialysis for other reasons. A low-bicarbonate dialysate can be used. If heart failure precludes the use of NaCl, then acetazolamide, 500 mg by mouth or intravenously, consistently reduces the serum tCO₂ by about 6 mmol per L. Rarely, dilute hydrochloric acid, 150 mmol per L, can be infused at 250 mL per hour into a central venous catheter or, preferably, in parenteral nutrition solution. Acid infusion is fraught with the potential complications of hemolysis and vascular necrosis and is best avoided. Ammonium chloride can be substituted for hydrochloric acid, but it causes gastric distress even when given intravenously and may cause ammonia intoxication.
  - Volume-depletion metabolic alkalosis is corrected by providing ample chloride with sodium or potassium. Prevention, however, is preferable. Proton-pump inhibitors minimize gastric acid loss in patients with nasogastric suction. Use of the potassium-sparing diuretics spironolactone, triamterene, and amiloride reduces the frequency and severity of diuretic-induced alkalosis.
  - Treatment of volume-replete metabolic alkalosis. If possible, the cause of increased mineralocorticoid production should be removed. For example, a functioning adrenal adenoma should be surgically excised. In the interim, the use of spironolactone in doses up to 400 mg per day with large amounts of potassium chloride may be effective. Indomethacin may be beneficial in Bartter's syndrome.
- Respiratory acidosis per se does not require direct treatment. Even at chronic PCO₂ levels above 100 mm Hg, the kidneys generate and maintain HCO₃^- levels sufficient to keep the pH above 7.20. However, adequate oxygenation is the critical issue in both acute and chronic respiratory acidosis.
- Definitive treatment of respiratory alkalosis again requires the correction of the underlying condition causing hyperventilation. The provision of oxygen is essential for the hypoxemic patient.
- Treatment of mixed acid-base disorders
  - Metabolic acidosis and respiratory acidosis. The most rapid treatment is to provide assisted or controlled ventilation. The administration of base is not warranted. The correction of the cause of metabolic acidosis is a priority.
  - In metabolic alkalosis and respiratory acidosis, the pH is often alkalemic. Acetazolamide given daily or every other day may be used to keep the pH near 7.35 to 7.40, which is a good level to avoid respiratory suppression.
  - Metabolic alkalosis and respiratory alkalosis in combination may produce severe alkalemia with dangerous arrhythmias. The most expedient treatment consists of intravenous morphine and a benzodiazepine, with immediate access to airway intubation and mechanical ventilation.