Potassium (K) is the most abundant intracellular cation. Only about 2% of total body K is extracellular. Since most intracellular K is contained within muscle cells, total body K is roughly proportional to lean body mass. An average 70-kg adult has about 3500 mEq of
K is a major determinant of intracellular osmolality. The relationship between intra- and extracellular fluid K concentrations strongly influences cell membrane polarization, which in turn influences important cell processes, such as the conduction of nerve impulses and muscle (including myocardial) cell contraction. Thus, relatively small alterations in plasma K concentration can have major clinical manifestations.
In the absence of serious metabolic disturbances, the plasma K level provides a reasonable clinical estimate of total body K content. Assuming a constant plasma pH, a decrease in plasma K concentration from 4 to 3 mEq/L indicates a total K deficit of 100 to 200 mEq. A fall in plasma K to < 3 mEq/L indicates a total K deficit of about 200 to 400 mEq. In many disease states, the plasma K concentration becomes an unreliable guide to total body K content because of processes resulting in shifts of K into or out of cells.
Internal Potassium Balance
Numerous factors affect the movement of K between the intracellular and extracellular fluid compartments. Among the most important is circulating insulin level. In the presence of insulin, K moves into cells, thus lowering plasma K concentration. When circulating insulin is lacking as in diabetic ketoacidosis, K moves out of cells, thus raising plasma K even in the presence of total body K deficiency. Stimulation of the sympathetic nervous system also affects transcellular K movement. -agonists, especially selective 2-agonists, promote cellular uptake of K, whereas -blockade or stimulation by -agonists appears to promote movement of K out of cells. Plasma K can also be affected significantly by plasma pH. Acute metabolic acidosis promotes the movement of K out of cells into the ECF. Acute metabolic alkalosis promotes the transfer of K in the opposite direction. However, changes in plasma HCO3 concentration may be more important than changes in pH in this regard. Thus, acidosis caused by accumulation of mineral acids (non-anion gap, hyperchloremic acidosis) is more likely to show an elevation of plasma K due to transcellular shifts. In contrast, metabolic acidosis due to accumulation of organic acids (increased anion gap acidosis) does not cause hyperkalemia. Thus, the hyperkalemia that frequently accompanies diabetic ketoacidosis results from insulin deficiency and ECF hypertonicity rather than from acidosis per se. Acute respiratory acidosis and alkalosis appear to have less of an effect on plasma K concentration than do metabolic disturbances. Nonetheless, the plasma K concentration should always be interpreted in the context of the plasma pH (and HCO3 concentration).
External Potassium Balance
Dietary intake of K normally varies between 40 and 150 mEq/day. In the steady state, fecal losses are relatively constant and small (roughly 10% of intake). Urinary excretion is regulated to approximate K intake so that balance is maintained. However, when a K load is ingested acutely, only about 50% appears in the urine over the next several hours. A rise in plasma K is minimized by transfer of most of the remaining K load into the intracellular compartment. If elevated intake continues, renal K excretion rises, probably due to K-stimulated aldosterone secretion. In addition, the K absorption from the stool appears to be under some degree of regulation and may fall by 50% in chronic K excess.
When the dietary K intake falls, intracellular K again serves to buffer against wide swings in plasma K concentration. Renal K conservation develops relatively slowly in response to decreases in dietary K and is far less efficient than the kidneys’ ability to conserve Na. Urinary K excretion of 10 mEq/24 h represents near maximal renal K conservation and, therefore, implies significant K depletion.
Plasma K is freely filtered at the glomerulus. Most of the filtered K is reabsorbed in the proximal tubule and loop of Henle. Normally, K is secreted into the filtrate in the distal tubule and collecting duct. Net renal K excretion is regulated primarily by changes in K secretion in the distal nephron. Distal K secretion is regulated by aldosterone, acid-base status, the rate of urine flow in the distal nephron, and membrane polarity. High-circulating aldosterone levels lead to increased K secretion and kaliuresis. Deficiency or suppression of aldosterone decreases distal nephron K secretion and results in renal K conservation. Acute acidosis impairs K excretion, whereas chronic acidosis and acute alkalosis can lead to kaliuresis (see Disturbances of Acid-Base Metabolism, below). Increased delivery of Na to the distal nephron and high distal nephron urine flow rates favor K secretion. The reabsorption of Na in the distal nephron increases luminal electrical negativity, a factor that further favors secretion of K. Thus, increased delivery of Na to the distal nephron, as occurs with high Na intake or loop diuretic therapy, is associated with increased K excretion.
Laboratory determination of plasma K concentration is usually accurate. Older methods using flame photometry have largely been replaced by measurement with ion-specific electrodes. Newer colorimetric tests are now available for rapid bedside determination of plasma K. These are reasonably accurate and, although not a replacement for clinical laboratory measurements, are useful, particularly in an ICU because of the rapid availability of the results.
Several conditions lead to spurious values of K concentration. Falsely low serum K (pseudohypokalemia) occasionally occurs in patients with myeloid leukemia with an extremely elevated WBC count (> 105/µL) if the specimen is allowed to sit at room temperature before being processed due to uptake of plasma K by abnormal leukocytes in the sample. Pseudohypokalemia can be prevented by prompt separation of plasma or serum in blood samples obtained for electrolyte determination. Falsely elevated serum K (pseudohyperkalemia) can also occur, most commonly from hemolysis and release of intracellular K from RBCs in the sample. For this reason, phlebotomy personnel should take care not to rapidly aspirate blood through a narrow-gauge needle or excessively agitate samples of blood. Pseudohyperkalemia can result from thromobocytosis (platelet count > 106/µL) due to release of K from platelets during clotting. In cases of pseudohyperkalemia, the plasma K (unclotted blood), as opposed to serum K, will be normal.
DISORDERS OF POTASSIUM METABOLISM
A decrease in the serum potassium concentration below 3.5 mEq/L caused by a deficit in total body potassium stores or abnormal movement of potassium into cells.
Etiology and Pathogenesis
Hypokalemia can be caused by decreased intake of K but is usually caused by excessive losses of K in the urine or from the GI tract. Abnormal gastrointestinal K losses occur in chronic diarrhea and include that due to chronic laxative abuse or bowel diversion. Other causes of gastrointestinal K losses include clay pica, vomiting, and gastric suction. Rarely, villous adenoma of the colon can cause massive K loss from the GI tract. Gastrointestinal K losses may be compounded by concomitant renal K losses due to metabolic alkalosis and stimulation of aldosterone due to volume depletion.
The transcellular shift of K into cells may also cause hypokalemia. This can occur in glycogenesis during TPN or enteral hyperalimentation or after administration of insulin. Stimulation of the sympathetic nervous system, particularly with 2-agonists, such as albuterol or terbutaline, may produce hypokalemia due to cellular K uptake. Similarly, severe hypokalemia occasionally occurs in thyrotoxic patients from excessive -sympathetic stimulation (hypokalemic thyrotoxic periodic paralysis). Familial periodic paralysis is a rare autosomal dominant disease characterized by transient episodes of profound hypokalemia thought to be due to sudden abnormal shifts of K into cells (see Hyperkalemia, below). Episodes are frequently associated with varying degrees of paralysis. They are typically precipitated by a large carbohydrate meal or strenuous exercise, but variants have been described without these features.
Various disorders can cause increased renal K wasting. Kaliuresis can occur in adrenal steroid excess due to direct mineralocorticoid effects on K secretion by the distal nephron. Cushing’s syndrome, primary hyperaldosteronism, rare renin-secreting tumors, glucocorticoid-remediable aldosteronism (a rare inherited disorder), and congenital adrenal hyperplasia can all cause hypokalemia from excess mineralocorticoid formation. Inhibition of the enzyme 11-hydroxysteroid dehydrogenase (11-HSDH) prevents the conversion of cortisol, which has some mineralocorticoid activity, to cortisone, which does not. Substances such as glycyrrhetinic acid (found in licorice and chewing tobacco) inhibit 11-HSDH, resulting in high circulating levels of cortisol and renal K wasting.
Liddle’s syndrome is a rare autosomal dominant disorder characterized by severe hypertension and hypokalemia. Liddle’s syndrome is caused by unrestrained Na reabsorption in the distal nephron due to one of a variety of mutations found in genes encoding for epithelial Na channel subunits. Inappropriately high reabsorption of Na results in both hypertension and renal K wasting.
Bartter’s syndrome is an uncommon disorder of uncertain cause characterized by renal K and Na wasting, excessive production of renin and aldosterone, and normotension.
Lastly, renal K wasting can be caused by numerous congenital and acquired renal tubular diseases such as the renal tubular acidoses and Fanconi’s syndrome, an unusual syndrome resulting in renal wasting of K, glucose, phosphate, uric acid, and amino acids.
Diuretics are by far the most commonly used drugs that cause hypokalemia. K-wasting diuretics block Na reabsorption proximal to the distal nephron and include thiazides, loop diuretics, and osmotic diuretics. Spironolactone, amiloride, and triamterene block Na reabsorption in the distal tubule and collecting duct and thus are not associated with renal K wasting. By inducing diarrhea, laxatives, especially when abused, can cause hypokalemia. Surreptitious diuretic and/or laxative abuse is a frequent cause of persistent hypokalemia, particularly among patients preoccupied with weight loss and among health care workers with access to prescription medications.
Other drugs that can cause hypokalemia include amphotericin B, antipseudomonal penicillins (such as carbenicillin) and high-dose penicillin. Finally, hypokalemia is observed in both acute and chronic theophylline intoxication.
Symptoms, Signs, and Diagnosis
Severe hypokalemia (plasma K < 3 mEq/L) may produce muscle weakness and lead to paralysis and respiratory failure. Other muscular dysfunction includes muscle cramping, fasciculations, paralytic ileus, hypoventilation, hypotension, tetany, and rhabdomyolysis. Persistent hypokalemia can impair renal concentrating ability, producing polyuria with secondary polydipsia. Metabolic alkalosis is often present, although hypokalemia can also occur with metabolic acidosis as in diarrhea or renal tubular acidosis. Generally, GFR, water, and Na balance are unaffected by hypokalemia. However, a state resembling nephrogenic diabetes insipidus can occur with severe K depletion.
Cardiac effects of hypokalemia are usually minimal until plasma K levels are < 3 mEq/L. Hypokalemia may produce premature ventricular and atrial contractions, ventricular and atrial tachyarrhythmias, and second or third degree atrioventricular block. Patients with significant preexisting heart disease and/or those receiving digitalis are at risk for cardiac conduction abnormalities even from fairly mild hypokalemia. The characteristic ECG changes of ST segment depression, increased U-wave amplitude, and T-wave amplitude less than U-wave amplitude (in the same lead) are shown in Fig. 12-1.
The diagnosis of hypokalemia is made on the basis of a plasma or serum K level < 3.5 mEq/L (see Potassium Metabolism, above).
Prophylaxis and Treatment
Routine K replacement is not necessary in most patients receiving diuretics. However, avoidance of hypokalemia is particularly important in patients receiving digitalis, in asthmatic patients receiving 2-agonists, and in non-insulin-dependent diabetics. Such patients should receive the lowest effective dose of a diuretic of moderate duration of action; their dietary Na intake should be restricted (< 2 g/day); and their plasma K should be monitored closely following the initiation of therapy. Once stable K concentration has been documented, less frequent monitoring is needed unless the dose is increased or symptoms of hypokalemia or other problems occur. If hypokalemia develops, K supplementation is indicated and the diuretic should be discontinued if possible. Addition of triamterene 100 mg/day or spironolactone 25 mg qid may be useful in occasional patients who become hypokalemic with diuretic therapy but should be avoided in patients with renal failure, diabetes, or other interstitial renal disease associated with hyperkalemia due to hyporeninemic hypoaldosteronism (type 4 renal tubular acidosis). K deficiency should be corrected very carefully in patients with renal insufficiency.
Correction of the underlying cause may suffice when hypokalemia is mild. When deficits and hypokalemia are more severe (plasma K < 3 mEq/L) or when continued therapy with K-depleting agents is necessary, KCl can be given po (10% potassium chloride). Usually, 20 to 80 mEq/day in excess of ongoing K losses given in divided doses over several days corrects K deficits. However, the need for K supplementation may continue for several weeks during refeeding after prolonged starvation.
A variety of oral K supplements are available. Liquid potassium chloride given orally is poorly tolerated in doses above 25 to 50 mEq due to bitter taste. Enteric-coated K preparations have been found to lead to small-bowel ulceration. Wax-impregnated potassium chloride preparations appear to be safe and well tolerated. GI bleeding may be even less common with microencapsulated potassium chloride preparations. Several preparations containing 8 or 10 mEq/capsule are available.
When hypokalemia is severe, symptomatic, or unresponsive to oral therapy, K must be replaced parenterally. The rate of correction of hypokalemia is limited because of the lag in K disposal into cells. In cases of K deficit with high plasma K concentration, as in diabetic ketoacidosis, one should wait until the plasma K starts to fall before administering IV K. Even when K deficits are severe, it is rarely necessary to give > 80 to 100 mEq of potassium in excess of continuing losses in a 24-h period. Modern, accurate IV infusion pumps have decreased the risk of administering highly concentrated KCl solutions. However, in most situations, the potassium concentration of IV solutions need not exceed 60 mEq/L, and infusion rates should not exceed 10 mEq/h. Occasionally, it may be necessary to give IV KCl solutions more rapidly to prevent progressive, severe hypokalemia. Infusion of > 40 mEq potassium chloride/h should be undertaken only with continuous cardiac monitoring and hourly plasma K determinations to avoid severe hyperkalemia and/or cardiac arrest. Glucose solutions are not ideal choices for administering KCl, because subsequent elevation in the patient’s plasma insulin level could result in transient worsening of hypokalemia, leading to worsened symptoms particularly in digitalized patients. Lastly, when hypokalemia is associated with hypomagnesemia, it is usually necessary to correct magnesium deficiency to stop renal K wasting and to facilitate K repletion (see Hypomagnesemia, below).
An increase in the serum potassium concentration above 5.5 mEq/L (plasma potassium above 5.0) caused by an excess in total body potassium stores or abnormal movement of potassium out of cells.
Etiology and Pathogenesis
Since the kidneys normally excrete K loads eventually, sustained hyperkalemia usually implies diminished renal K excretion. Hyperkalemia also may be caused by transcellular movement of K out of cells in metabolic acidosis; hyperglycemia in the presence of insulin deficiency; moderately heavy exercise, particularly in the presence of -blockade; digitalis intoxication; acute tumor lysis; acute intravascular hemolysis; or rhabdomyolysis. Hyperkalemic familial periodic paralysis is a rare inherited disorder characterized by episodic hyperkalemia due to sudden movement of K out of cells, usually precipitated by exercise.
Hyperkalemia from total body K excess is particularly common in oliguric states (especially acute renal failure) and is associated with rhabdomyolysis, burns, bleeding into soft tissue or the GI tract, and adrenal insufficiency, which is increasingly recognized in patients with AIDS (see Hyponatremia, above). In chronic renal failure, hyperkalemia is uncommon until the GFR falls below 10 to 15 mL/min unless dietary K intake is excessive or another source of excess K load is present, such as oral or parenteral K therapy, GI bleeding, tissue injury, or hemolysis. Other potential causes of hyperkalemia in chronic renal failure are hyporeninemic hypoaldosteronism (type 4 renal tubular acidosis), ACE inhibitors, K-sparing diuretics, fasting (suppression of insulin secretion), -blockers, and NSAIDs. If sufficient KCl is ingested orally or given parenterally, severe hyperkalemia may result even with normal renal function. Nonetheless, iatrogenic hyperkalemia is most commonly seen in patients with some degree of renal impairment. Other drugs that may limit renal K output, thereby producing hyperkalemia, include cyclosporine, lithium, heparin, and trimethoprim.
Symptoms, Signs, and Diagnosis
Although flaccid paralysis occasionally occurs, hyperkalemia is usually asymptomatic until cardiac toxicity supervenes. The first ECG changes seen with progressive hyperkalemia (plasma K > 5.5 mEq/L) are shortening of the QT interval and tall, symmetric, peaked T waves. Progressive hyperkalemia (plasma K > 6.5 mEq/L) produces nodal and ventricular arrhythmias, widening of the QRS complex, PR interval prolongation, and disappearance of the P wave. Finally, the QRS complex degenerates into a sine wave pattern and ventricular asystole or fibrillation ensues.
In hyperkalemic familial periodic paralysis, weakness frequently develops during attacks and can progress to frank paralysis.
The diagnosis of hyperkalemia is made by a plasma or serum K level > 5.5 mEq/L (plasma > 5.0 mEq/L) (see Potassium Metabolism, above).
Mild hyperkalemia (plasma K < 6 mEq/L) may respond to diminished K intake or discontinuance of drugs such as K-sparing diuretics, -blockers, NSAIDs, or ACE inhibitors. Addition of a loop diuretic can also enhance renal K excretion. A plasma K > 6 mEq/L requires more aggressive therapy. However, in acute or chronic renal failure, especially in the presence of hypercatabolism or tissue injury, treatment should be initiated when the plasma K level exceeds 5 mEq/L.
If there are no ECG abnormalities and the plasma K is not greatly elevated (< 6 mEq/L), sodium polystyrene sulfonate in sorbitol can be given (15 to 30 g in 30 to 70 mL of 70% sorbitol po q 4 to 6 h). Na polystyrene sulfonate acts as a cation exchange resin and removes K through the GI mucosa. Sorbitol is administered with the resin to ensure passage through the GI tract. Patients unable to take medications orally because of ileus or other reasons may be given similar doses by rectal retention enema. About 1 mEq of K is removed per gram of resin given. Resin therapy is slow and often fails to lower plasma K significantly in hypercatabolic states. Since Na is exchanged for K when Na polystyrene sulfonate is used, Na overload may occur, particularly in oliguric patients with preexisting volume overload.
In emergencies such as cardiac toxicity or plasma K level > 6 mEq/L, the following three measures should be performed immediately in rapid sequence without waiting to repeat plasma K values after each:
- IV administration of 10 to 20 mL 10% calcium gluconate (or 5 to 10 mL 22% calcium gluceptate) over 5 to 10 min. Caution should be used when giving calcium to patients taking digitalis because of the risk of precipitating hypokalemia-related arrhythmias. If the ECG has deteriorated to a sine wave or asystole, the calcium gluconate may be given rapidly IV (5 to 10 mL over 2 min).
- IV administration of 5 to 10 U regular insulin by IV push followed immediately by rapid infusion of 50 mL 50% glucose. This should be followed by 10% D/W at 50 mL/h to prevent hypoglycemia. An effect on plasma K occurs within 15 min.
- Inhalation of a high-dose -agonist such as albuterol (10 to 20 mg) over 10 min (5 mg/mL concentration). This has been shown to be efficacious and safe in treating hyperkalemia. Onset of action is within 30 min. Duration of effect is 2 to 4 h.
Note: NaHCO3 has purposely been omitted from this algorithm. The effectiveness of empirical administration of NaHCO3 for the treatment of life-threatening acute hyperkalemia has recently been questioned.
In addition to the above strategies for lowering K by shifting it into cells, maneuvers to remove K from the body should also be performed early in the treatment of severe or symptomatic hyperkalemia. Removal of K can be accomplished via the GI tract by administration of Na polystyrene sulfonate or by hemodialysis. Hemodialysis should be instituted promptly after emergency measures in patients with renal failure or if emergency treatment is ineffective. Peritoneal dialysis is relatively inefficient at removing K but may benefit the acidotic patient, especially if volume overload is likely to occur with NaHCO3.