Must primary care clinicians be fluent in acid-base physiology? Yes, for 2 major reasons: First, many common disease states can be recognized through identification of acid-base disturbances. For example, respiratory alkalemia can be an early sign of septicemia. Second, acid-base abnormalities can point to rarer diseases. For example, a non-anion gap acidosis may suggest a rheumatic disease, such as Sjögren syndrome. A patient with multiple myeloma may present with a decreased or absent anion gap. Primary aldosterone excess-a rare cause of hypertension-may manifest itself through a combination of hypokalemia and metabolic alkalosis. The development of an elevated anion gap in a patient who has a mixed acid-base disorder may lead to the diagnosis of lactic acidosis, which portends a more severe illness.
Fortunately, the reputed complexity of interpreting acid-base disturbances can be greatly simplified. Here we present 5 easy rules for interpreting clinical acid-base physiology, along with straightforward algorithms that provide a step-by-step approach to various metabolic derangements. Together these rules and algorithms can help you solve complex cases of acid-base disturbances, as we demonstrate in "Cases in Point" on page 391.
THE 5 RULES
Homeostasis requires that hydrogen ions in body fluids be precisely regulated. Maintenance of the acid-base balance depends on lung and kidney adjustments-which regulate carbon dioxide and bicarbonate levels, respectively. Together, these organs maintain the chemical environment optimal for cellular function.
Acid-base problem solving begins with an arterial blood gas analysis and serum electrolyte measurements (sodium, potassium, chloride, and bicarbonate). You will use these values in the formulas presented in the 5 rules.
The arterial blood sample provides the arterial pH, the PO2, the PCO2, and a calculated bicarbonate level. These are used to assess both the primary acid-base status and the body's attempts to compensate for any disturbance in it. Serum electrolyte levels are measured at the time of the blood gas determination, and the anion gap is calculated with the resultant values.
Use the arterial pH to determine whether acidemia or alkalemia is present.
The normal pH of arterial blood is 7.40 to 7.44. A pH of less than 7.40 represents acidemia; any pH greater than 7.44 represents alka- lemia. Thus, acidemia and alkalemia cannot exist simultaneously. It is important not to use "acidemia" and "acidosis" or "alkalemia" and "alkalosis" interchangeably. Both "acidemia" and "acidosis" refer to processes in which acid is accumulated or alkaline reserves are depleted; however, "acidemia" is used here to refer only to a primary process, with the characteristic decrease in arterial pH. Similarly, "alkalemia" and "alkalosis" both refer to processes in which base is accumulated or acid is lost-but "alkalemia" is used here to refer only to a primary process, with the characteristic increase in arterial pH. This convention makes it possible to describe clearly situations such as that in which a patient has a significant underlying acidosis without acidemia (eg, an anion gap of 30 with an arterial pH of 7.48), or an underlying alkalosis without alkalemia.
Two or more metabolic processes can occur simultaneously. However, because the respiratory acid-base status is determined solely by the rate of elimination of carbon dioxide by the lungs, only one respiratory process can occur at a given time. Therefore, the most complex acid-base disturbance possible is a triple disturbance-and there can be only 2 types of triple disturbance:
Metabolic acidosis and metabolic alkalosis with a respiratory acidosis.
Metabolic acidosis and metabolic alkalosis with a respiratory alkalosis.
When acidemia or alkalemia is detected, determine whether the cause is respiratory, metabolic, or both.
2Alkalemia represents a respiratory process if the PCO2 is less than 40 mm Hg (normal range, 40 to 44 mm Hg). It represents a metabolic process if the bicarbonate level is greater than 28 mEq/L (normal range, 25 to 28 mEq/L). If the PCO2 is less than 40 mm Hg and the bicarbonate level is greater than 28 mEq/L, respiratory and metabolic alkalosis coexist.
Acidemia has a respiratory cause if the PCO2 is greater than 44 mm Hg. It has a metabolic cause if the bicarbonate level is less than 25 mEq/L. A PCO2 greater than 44 mm Hg concurrent with a bicarbonate level less than 25 mEq/L suggests that respiratory and metabolic acidosis coexist.
Always calculate the anion gap.
3The anion gap is the difference between the concentration of sodium cations in the serum and the sum of the serum concentrations of chloride anions and bicarbonate anions. Because the number of positive charges (cations) and negative charges (anions) in the serum must always be equal, the gap indicates the presence of a variety of unmeasured anions (albumin, organic ions, phosphate, and so forth).
A normal gap ranges between approximately 3 and 10 mEq/L. For many years, a gap between 8 and 16 mEq/L was considered normal. However, because the method used to measure serum electrolytes has changed-from flame photometry to more accurate ion-selective electrodes-the normal range has changed as well.1,2
The serum anion gap is a critical branch point in determining the presence and cause of metabolic acidosis.
An anion gap greater than 10 mEq/L suggests metabolic acidosis.
An anion gap greater than 25 mEq/L indicates severe organic acid metabolic acidosis.3,4
Adjust the anion gap for the patient's serum albumin level. Albumin represents a significant portion of the unmeasured anions in serum. A 1 g/dL drop in the serum albumin level leads to a decrease of 2.5 mEq/L in the anion gap. Therefore, an apparently normal anion gap of 10 mEq/L-when corrected for a 2 g/dL decrease in the serum albumin level-becomes an anion gap of 15 mEq/L (10 + [2 × 2.5]), which is consistent with an anion gap metabolic acidosis.
Always check for the degree of compensation.
4To compensate for the increase in hydrogen ion concentration during the development of acidemia, the body attempts to return the pH to normal through the addition of a base or the removal of acid. The converse is true during the development of alkalemia.
If the primary process is metabolic, the body's compensation will be respiratory. If the primary process is respiratory, the body's compensation will be metabolic. However, compensation never fully corrects the primary disturbance. Appropriate compensation for each type of primary disturbance is summarized below.
Compensation for metabolic acidemia.In metabolic acidemia, the compensatory respiratory change-hyperventilation-removes carbon dioxide (and thus acid) and results in a decrease in the PCO2. The compensatory mechanism is the same whether or not the metabolic acidemia is associated with an increased anion gap.
The decrease in PCO2 should be approximately 1.3 times the decrease in bicarbonate resulting from the metabolic acidemia. Thus, if the bicarbonate decreases to 15 mEq/L from a normal value of 25 mEq/L (a decrease of 10), the PCO2 should decrease by 10 × 1.3, or 13 mm Hg. A decrease of 13 mm Hg from the normal level of 40 would result in a PCO2 of 27 mm Hg (40 − [1.3 × 10] = 27). A measured PCO2 within 2 mm Hg of this figure would indicate appropriate respiratory compensation for the degree of metabolic acidemia present. However, a PCO2 of 30 mm Hg or higher would indicate the simultaneous presence of a respiratory acidosis; a PCO2 of 24 mm Hg or lower would indicate the simultaneous presence of a respiratory alkalosis.
Compensation for metabolic alkalemia.The compensatory mechanism for a metabolic increase in bicarbonate is hypoventilation; because carbon dioxide (and thus acid) is retained, the PCO2 increases. An obvious drawback to this means of compensation is the potential for hypoxia.
In most clinical settings, the increase in PCO2 should be 0.6 times the increase in bicarbonate resulting from the metabolic alkalemia. Thus, if the bicarbonate increases to 38 mEq/L from a high-normal value of 28 mEq/L (an increase of 10), the PCO2 should increase 10 × 0.6, or 6 mm Hg. An increase of 6 mm Hg from the normal level of 40 would result in a PCO2 of 46 mm Hg (40 + [0.6 × 10] = 46). A measured PCO2 within 2 mm Hg of this figure would indicate appropriate respiratory compensation for the degree of metabolic alkalemia present. However, a PCO2 of 49 mm Hg or higher would indicate the simultaneous presence of a respiratory acidosis; a PCO2 of 43 mm Hg or lower would indicate the simultaneous presence of a respiratory alkalosis.
Compensation for primary respiratory acid-base disturbances.Respiratory acidemia occurs when hypoventilation leads to the retention of carbon dioxide. Conversely, respiratory alkalemia results when hyperventilation leads to an acute or chronic decrease in carbon dioxide. The degree of compensation for respiratory disturbances varies according to whether the situation is acute (of 48 to 72 hours' duration or less) or chronic (of more than 72 hours' duration).
In the acute stage, respiratory acidemia or alkalemia is countered by an increase or decrease in the bicarbonate level, which represents the work of the bicarbonate buffer system. If the respiratory acid-base disturbance is not corrected within 48 to 72 hours, the kidneys begin to compensate further for the lungs' addition or removal of acid. In chronic respiratory acidemia, the kidneys synthesize and reabsorb bicarbonate. In chronic respiratory alkalemia, the kidneys decrease production of bicarbonate and increase its excretion in the urine.
Therefore, during acute respiratory acidemia, every 10 mm Hg increase in the PCO2 produces an increase of 1 mEq/L in the bicarbonate level; however, when the process is chronic, the same 10 mm Hg increase in the PCO2 results in an increase of 4 mEq/L in the bicarbonate level. In respiratory alkalemia, a 10 mm Hg decrease in the PCO2 leads-in the acute phase-to a decrease of 2 mEq/L in the bicarbonate level; once the process has become chronic, the same 10 mm Hg decrease in the PCO2 results in a decrease in the bicarbonate level of 4 to 5 mEq/L.
Determine whether quantitative changes in the different groups of anions in the blood are in a 1:1 relationship.
5This concept is based on the premise that there can never be an excess of positive or negative charges in the blood. To maintain electroneutrality when an acid is added to the blood, the bicarbonate level will decrease-by an amount proportionate either to the increase in the anion gap (in elevated anion gap metabolic acidosis) or to the increase in chloride concentration (in non- anion gap metabolic acidosis). Calculation of the decline in the bicarbonate level can help you identify an occult metabolic alkalosis not detected by the application of the previous 4 rules. If a metabolic alkalosis has already been diagnosed from the first 4 rules, it is not necessary to confirm the 1:1 relationship.
If acidosis is associated with an increased anion gap, then for every 1 mEq/L increase in the anion gap, there should be a 1 mEq/L decline in the bicarbonate level. If the bicarbonate level declines less than predicted from the rise in the anion gap, there is an underlying metabolic alkalosis. For example, if the anion gap increases from 10 to 20 mEq/L, the bicarbonate level should decrease by 10 (ie, from 25 to 15 mEq/L). If the bicarbonate level is higher than 15 mEq/L, then it was elevated before acid was added to the serum. This establishes the presence of an occult metabolic alkalosis.
If acidosis is associated with a normal anion gap, for every 1 mEq/L increase in chloride concentration, there should be a 1 mEq/L decrease in the bicarbonate level. If the bicarbonate level declines less than predicted from the increase in chloride concentration, there is an underlying metabolic alkalosis. Note that determinations made by using the "one-to-one" rule are more reliable in the setting of increased anion gap metabolic acidosis than in non-anion gap metabolic acidosis.
DETERMINING THE CAUSE OF METABOLIC ACIDOSIS
To establish the cause of a metabolic acidosis, first calculate the anion gap (remember to correct for the serum albumin level). There are 2 basic types of metabolic acidosis, which are defined by the size of the anion gap. These are:
Elevated anion gap metabolic acidosis.
Non-anion gap metabolic acidosis.
Elevated anion gap metabolic acidosis. Once you have determined that the anion gap is elevated, calculate the osmol gap, which is the difference between the actual and calculated serum osmolarity (Algorithm I). Use the following formula to determine the calculated serum osmolarity: 2(serum sodium level) + (serum glucose level)/18 + (blood urea nitrogen)/2.4 + (blood ethanol level)/4.6.
If the osmol gap is greater than 10, consider ethylene glycol or methanol poisoning. Given the morbidity and mortality associated with such toxicities, early diagnosis and treatment are essential.
If the osmol gap is less than 10, the differential diagnosis includes ketoacidosis, lactic acidosis, renal failure, salicylate ingestion, and D-lactic acidosis. The clinical presentation guides evaluation and treatment.
Non-anion gap metabolic acidosis. In elevated anion gap metabolic acidosis, the decline in bicarbonate level is a direct result of the buffering process that occurs when an organic acid, such as lactic acid, is added to the blood (the same process that leads to the accumulation of unmeasured anions, such as lactate, that constitute the increase in the anion gap). In non-anion gap metabolic acidosis, a physical process leads to the loss of bicarbonate. Keep in mind that there are only 2 organs capable of wasting bicarbonate-the gut (through diarrhea) and the kidney (in renal tubular acidosis).
An approach to the patient with non-anion gap acidosis is shown in Algorithm II. The first step is to measure electrolytes in a random urine sample and calculate the difference between positive and negative charges.
An excess of negative charges indicates that bicarbonate has been lost through the gut. (The renal tubular response to acidosis is intact; the kidneys are acidifying the urine through generation of ammonium [NH4+], a titratable acid that combines with chloride in the urine.)
The absence of excess negative charges indicates that the acidosis is of renal origin. The differential diagnosis is narrowed to renal tubular acidosis, use of carbonic anhydrase inhibitors, early renal failure, and ureteral diversions.
DETERMINING THE CAUSE OF METABOLIC ALKALOSIS
The evaluation ofmetabolic alkalosis starts with measurement of the urinary chloride level (Algorithm III).
If the urinary chloride level is low (less than 20 mEq/L), the metabolic alkalosis is most likely the result of a volume-responsive condition. Volume resuscitation with saline typically leads to resolution.
If the urinary chloride level is high (greater than 20 mEq/L), measure the patient's blood pressure.
If the patient is hypertensive, the differential diagnosis includes:
Aldosterone excess (an aldosterone-to-renin ratio greater than 20:1 indicates primary hyperaldosteronism).
Excessive licorice ingestion (the active component of licorice, glycyrrhetinic acid, inhibits 11β-hydroxysteroid dehydrogenase, the enzyme that converts cortisol to cortisone5; the resultant excess cortisol acts much like aldosterone, suppressing both renin and aldosterone and raising blood pressure).
Cushing syndrome, which involves excess cortisol.
Liddle syndrome, which is characterized by a renal tubular sodium channel defect that results in excess sodium reabsorption and potassium wasting6; the resultant volume expansion also suppresses renin and aldosterone.
Hydroxylase deficiencies (rare); signs and symptoms may include hirsutism and clitoromegaly, decreased renin and aldosterone levels, and (usually) hypertension; elevated levels of 17-ketosteroids and dihydroepiandrosterone sulfate are diagnostic.
Diuretic use, which produces elevated renin and aldosterone levels.
Renal artery stenosis, which results in elevated renin and aldosterone levels.
Renin-secreting tumors, which result in an increase in the aldosterone as well as the renin level.
Measurement of renin and aldosterone levels can help narrow the differential.
If the patient is normotensive and has metabolic alkalosis and a high urinary chloride level, the differential diagnosis includes:
Bartter syndrome (not a single disorder), which is characterized by hypokalemia, metabolic alkalosis, and elevated renin and aldosterone levels. (Patients who have classic Bartter syndrome, which occurs during early childhood, also have increased urinary prostaglandin excretion and normal urinary calcium excretion; rarely, this condition is associated with hypomagnesemia.7)
Gitelman syndrome (a variant of Bartter syndrome), which is often associated with low serum magnesium levels, normal prostaglandin excretion, and decreased urinary calcium excretion.8,9
Villous adenoma (unlikely if diarrhea is not present).
1. WinterSD, Pearson JR, Gabow PA, et al. The fall of the serum anion gap. Arch Intern Med. 1990;140: 311-313.
2. Paulson WD. How to interpret the anion gap. J Crit Illness. 1997;12:96-99.
3. Rutecki GW, Whittier FC, Lappin J, Jarjoura D. An elevated anion gap as a predictor of metabolic acidosis [abstract]. Presented at: the International Congress of Nephrology; May 25-29, 1997; Sydney, Australia.
4. Emmet ME, Narins RG. Clinical use of the anion gap. Medicine. 1977;56:38-54.
5. Farese RV Jr, Biglieri EG, Shackleton CH, et al. Licorice-induced hypermineralocorticoidism. N Engl J Med. 1991;325:1223-1227.
6. Palmer BF, Alpern RJ. Liddle's syndrome. Am J Med. 1998;104:301-309.
7. Guay-Woodford LM. Bartter syndrome: unraveling the pathophysiologic enigma. Am J Med. 1998; 105:151-161.
8. Luthy C, Bettinelli A, Iselin S, et al. Normal prostaglandinuria E2 in Gitelman's syndrome, the hypocalciuric variant of Bartter's syndrome. Am J Kidney Dis. 1995;25:824-828.
9. Simon DB, Nelson-Williams C, Bia MJ, et al. Gitelman's variant of Bartter's syndrome, inherited hypokalaemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nat Genet. 1996;12:24-28.
10. Oh MS, Linbarn J, Carroll HJ. Electrolyte case vignette: a case of unusual organic acidosis. Am J Kidney Dis. 1988;11:80-82.
11. Bartholow C, Whittier FC, Rutecki GW. Hypokalemia and metabolic alkalosis: algorithms for combined clinical problem solving. Comp Ther. 2000;26:114-120.
12. Goldstein MB, Bear R, Richardson KMA, et al. The urine anion gap, a clinically useful index of ammonium secretion. Am J Med Sci. 1986;292:198-202.