Concise Definitive Review R. Phillip Dellinger, MD, FCCM, Section Editor

Management of severe hyperkalemia

Lawrence S. Weisberg, MD

Background and Objectives: Hyperkalemia is one of the few Results and Conclusions: A more complete understanding of potentially lethal electrolyte disturbances. Prompt recognition and potassium homeostasis in recent years has led to new ap- expeditious treatment of severe hyperkalemia are expected to proaches to the management of severe hyperkalemia. The phys- save lives. This review is intended to provide intensivists and iologically based sequential approach still applies. The efficacy, other interested clinicians with an understanding of the patho- pitfalls, and risks of the agents available for use at each step in physiology that underlies hyperkalemia, and a rational approach the sequence are critically reviewed. Rational use of the available to its management. tools will allow clinicians to successfully treat severe hyperkale- Methods: This article reviews and analyzes literature relevant mia. (Crit Care Med 2008; 36:3246–3251) to the pathophysiology and management of severe hyperkalemia. KEY WORDS: hyperkalemia; treatment; critical illness Methods include search of MEDLINE, and bibliographic search of current textbooks and journal articles.

yperkalemia is common in rium is modulated by insulin (3–5), cat- panies acute kidney injury, particularly in hospitalized patients, and echolamines (6, 7) and, to a lesser extent, the setting of mineralocorticoid defi- may be associated with ad- by acid-base balance (8–10), plasma to- ciency (13–15). Such mineralocorticoid verse clinical outcomes (1, nicity, and several other factors (3). The deficiency is often induced by drugs that 2).H Its prevalence and clinical impact in interfere with the renin-angiotensin- other system governs K homeostasis over critically ill patients are unknown. There the long-term by regulating external bal- aldosterone axis and commonly causes is no doubt, however, that severe hyper- ance: the parity between K intake and hyperkalemia in patients with chronic kalemia can be fatal. Proper treatment of elimination. In individuals with normal kidney disease, as well (16, 17). Sustained hyperkalemia depends on an understand- renal function, the kidneys are responsi- hyperkalemia is always attributable to in- ing of the underlying physiology. ble for elimination of about 95% of the adequate renal K elimination. A detailed The ratio of extracellular to intracel- daily K load with the remainder exiting discussion of the causes of hyperkalemia lular potassium (K) concentration largely through the gut. External K balance is in critically ill patients is beyond the determines the cell membrane resting maintained largely by modulating renal K scope of this article, but may be found in electrical potential that, in turn, regu- elimination. a recent review (18). lates the function of excitable tissues Almost all the K excreted by the kid- (cardiac and skeletal muscle, and nerve) ney comes from K secreted in the distal Clinical Manifestations of (1). Small absolute changes in the extra- nephron (connecting tubule and collect- Hyperkalemia cellular K concentration will have large ing duct) (11). Virtually all regulation of effects on that ratio, and consequently on Alterations in P have a variety of ad- K excretion takes place at this site in the K the function of excitable tissues. Thus, it verse clinical consequences, the expres- nephron, under the influence of two prin- is not surprising that the plasma K con- sion of which may be magnified in the ciple factors: the rate of flow and solute centration (P ) normally is maintained critically ill patient. The most serious of K (sodium and chloride) delivery through within very narrow limits. This tight reg- these manifestations are those involving that part of the nephron; and the effect of ulation is accomplished by two coopera- excitable tissues. tive systems. One system defends against aldosterone (11). K secretion is directly Cardiac Effects. Hyperkalemia depo- proportional to flow rate and sodium de- short-term changes in PK by regulating larizes the cell membrane, slows ventric- internal balance: the equilibrium of K livery through the lumen of the distal ular conduction, and decreases the dura- across the cell membrane. This equilib- nephron, and to circulating aldosterone tion of the action potential. These levels in the setting of an aldosterone- changes produce the classic electrocar- sensitive epithelium. This explains, in diographic (EKG) manifestations of hy- part, why the use of diuretic drugs that From the Division of Nephrology, Department of perkalemia including (in order of their Medicine, UMDNJ-Robert Wood Johnson Medical work proximal to the K secretory site usual appearance) peaked T waves, wid- School, Cooper University Hospital, Camden, NJ. (loop and thiazide diuretics) often is ac- ening of the QRS complex, loss of the P The author has not disclosed any potential con- companied by hypokalemia. K secretion wave, “sine wave” configuration, or ven- flicts of interest. is inversely proportional to the chloride For information regarding this article, E-mail: tricular fibrillation and asystole (19, 20). [email protected] concentration of the luminal fluid and is These EKG changes may be modified by a Copyright © 2008 by the Society of Critical Care stimulated, for example, by luminal deliv- multitude of factors such as extracellular Medicine and Lippincott Williams & Wilkins ery of sodium bicarbonate (12). Con- fluid pH, calcium concentration, sodium

DOI: 10.1097/CCM.0b013e31818f222b versely, hyperkalemia commonly accom- concentration, and the rate of rise of PK

3246 Crit Care Med 2008 Vol. 36, No. 12 Table 1. Emergency treatment of hyperkalemia

Agent Dose Onset Duration Complications

Membrane stabilization Calcium gluconate (10%) 10 mL IV over 10 min Immediate 30–60 min Hypercalcemia Hypertonic (3%) sodium chloride 50 mL IV push Immediate Unknown Volume overload hypertonicity Redistribution Insulin (short acting) 10 units IV push, with 25–40 g dextrose 20 min 4–6 hrs hypoglycemia (50% solution) Albuterol 20 mg in 4 mL normal saline solution, 30 min 2 hrs Tachycardia inconsistent nebulized over 10 min response Elimination Loop diuretics Furosemide 40–80 mg IV 15 min 2–3 hrs Volume depletion Bumetanide 2–4 mg IV Sodium bicarbonate 150 mmol/L IV at variable rate Hours Duration of infusion Metabolic alkalosis volume overload Sodium polystyrene sulfonate 15–30 g in 15–30 mL (70% sorbitol orally) Ͼ2 hrs 4–6 hrs Variable efficacy intestinal (Kayexalate, Kionex) necrosis Hemodialysis Immediate 3 hrs Arrhythmias (?)

IV, intravenously.

(19). Hospitalized patients with hyperka- and, thus, prevent correction of a meta- as arbitrary. Nonetheless, since the treat- lemia are reported to have a higher mor- bolic acidosis (29). ment for acute hyperkalemia is safe if tality rate than those without hyperkale- applied properly and hyperkalemia is po- mia (21, 22), but the high prevalence of Treatment of Severe tentially and unpredictably lethal, it is coexistent renal insufficiency in this pop- Hyperkalemia prudent to maintain a low threshold for ulation is a significant confounding vari- instituting emergency therapy. Because able that prevents attribution of the in- In general, the initial treatment of se- most patients manifest hyperkalemic creased mortality to the hyperkalemia vere hyperkalemia is independent of the EKG changes at PK greater than 6.7 itself. cause of the disturbance, whereas the ra- mmol/L (20), hyperkalemia should be Ͼ EKG changes may not accompany tional therapy of chronic hyperkalemia treated emergently for 1) PK 6.5 changes in PK. The sensitivity of the elec- depends on an understanding of its mmol/L or 2) EKG manifestations of hy- trocardiogram to reveal changes of hy- pathogenesis. perkalemia regardless of the PK (30). perkalemia is quite low (23). It does in- In considering when hyperkalemia Therapy of acute or severe hyperkale- crease in proportion to the severity of the constitutes an emergency, several points mia is directed at preventing or amelio- hyperkalemia (23), but normal electro- should be kept in mind. First, the elec- rating its untoward electrophysiologic ef- cardiograms have been seen even with trophysiologic effects of hyperkalemia are fects on the myocardium. The goals of extreme hyperkalemia (24) and the first directly proportional to both the absolute therapy, in chronologic order, are as fol- lows (Table 1): cardiac manifestation of hyperkalemia PK and its rate of rise (19). Second, con- may be ventricular fibrillation (25). (The current metabolic disturbances may ame- explanation for a normal electrocardio- liorate (e.g., hypernatremia, hypercalce- 1. Antagonize the effect of K on excitable gram in the setting of extreme hyperka- mia, alkalemia) or exacerbate (e.g., cell membranes. lemia is not entirely clear, but may relate hyponatremia, hypocalcemia, acidemia) 2. Redistribute extracellular K into cells. ͓ ͓ to a slow rate of rise in the PK 20, 24 ). the electrophysiologic consequences of 3. Enhance elimination of K from the Given this insensitivity of the electrocar- hyperkalemia (20, 24). Third, although body. diogram, EKG changes should not be the EKG manifestations of hyperkalemia considered necessary for the emergency are generally progressive and propor- Membrane Antagonism treatment of severe hyperkalemia. tional to the PK, ventricular fibrillation Neuromuscular Effects. Hyperkalemia may be the first EKG disturbance of hy- Calcium. Calcium directly antago- may result in paraesthesias and weakness perkalemia (25); conversely, a normal nizes the myocardial effects of hyperkale- progressing to a flaccid paralysis, which EKG may be seen even with extreme hy- mia without lowering PK (31, 32). It does typically spares the diaphragm. Deep ten- perkalemia (24). so by reducing the threshold potential of don reflexes are depressed or absent. Cra- With this in mind, it is apparent that cardiac myocytes, thereby restoring the nial nerves are rarely involved and sen- neither the EKG nor the PK alone is an normal gradient with the resting mem- sory changes are minimal (26, 27). adequate index of the urgency of hyper- brane potential, which is distorted by hy- Metabolic Effects. Hyperkalemia de- kalemia, and that the clinical context perkalemia (19, 20, 33). Calcium is ben- creases renal ammoniagenesis which by must be considered when assessing a hy- eficial even in patients who are itself may produce a mild hyperchloremic perkalemic patient. Thus, any pro- normocalcemic. Calcium for injection is metabolic acidosis (28), and will limit the nouncement on an absolute PK value available as the chloride or gluconate kidney’s ability to excrete an acid load constituting an emergency must be seen salt, both 10% by weight. The preferred

Crit Care Med 2008 Vol. 36, No. 12 3247 agent is the gluconate salt, since it is less dextrose has been shown to be inadequate hypokalemic effect of albuterol (42, 48). likely than calcium chloride to cause tis- to prevent hypoglycemia at 60 mins (42). The mechanism for this resistance is un- sue necrosis if it extravasates (34). The It is interesting to note that when insulin known, and there is currently no basis for recommended dose is 10 mL intravenous was given by continuous intravenous in- predicting which patients will respond. Ͻ over 10 mins. The onset of action is 3 fusion for 4 hrs to normal volunteers, PK For that reason, albuterol should never mins. The EKG should be monitored fell over the first 90 mins and rose there- be used as a single agent for the treat- continuously. The dose may be repeated after (45). Based on that observation, ment of urgent hyperkalemia in patients in 5 mins if there is no improvement in there seems to be no advantage of a con- with renal failure. the EKG, or if the EKG deteriorates tinuous infusion over a bolus injection. Bicarbonate. The putative benefits of a after an initial improvement. The dura- Insulin should be used without dex- bolus injection of sodium bicarbonate in tion of action is 30–60 mins, during trose in hyperglycemic patients; indeed, the emergency treatment of hyperkale- which time further measures may be the cause of the hyperkalemia in those mia pervaded the literature until the past undertaken to lower PK (30). patients may be the hyperglycemia itself decade. Ironically, this dogma was based There are several case reports of sud- (46). The administration of hypertonic on studies using a prolonged (4–6 hrs) den death in patients given intravenous dextrose alone for hyperkalemia is not infusion of bicarbonate (53). It has now calcium while also receiving digitalis gly- recommended for two reasons: first, en- been clearly demonstrated that short- cosides (35, 36). Although these anec- dogenous insulin levels are unlikely to term bicarbonate infusion does not re- dotes do not provide clear guidance, it is rise to the level necessary for a therapeu- duce PK in patients with dialysis-depen- wise to administer intravenous calcium tic effect; and second, there is a risk of dent kidney failure, implying that it does under very close supervision to patients exacerbating the hyperkalemia by induc- not cause K shift into cells. Infusion of a known or strongly suspected to have ing hypertonicity (46). hypertonic or an isotonic bicarbonate so- toxic levels of digitalis glycosides. ␤-adrenoceptor Agonists. An appreci- lution for 60 mins has been shown to

Hypertonic Saline. Intravenous hy- ation for the effect of catecholamines on have no effect on PK in dialysis patients, pertonic sodium chloride has been shown internal potassium balance recently has despite a substantial increase in serum to reverse the EKG changes of hyperka- been applied to the clinic. Patients with bicarbonate concentration (40, 54–56). ␤ lemia in patients with concurrent hypo- renal failure given the selective 2- Only after a 4-hr infusion was a small (0.6 natremia (37). This effect seems to be adrenoceptor agonist, albuterol, by intra- mmol/L) but significant decrease in PK is mediated by a change in the electrical venous infusion (0.5 mg over 15 mins) detectable (57). Whether bicarbonate in- properties of cardiomyocytes rather than show a significant decline in PK (about 1 fusion might enhance insulin-mediated by a reduction in PK (38). Whether hy- mmol/L) that is maximal between 30 and cellular K uptake remains unresolved by pertonic saline is effective in the treat- 60 mins (47). Because injectable albu- two contradictory studies (54, 56). The ment of eunatremic patients has not been terol is unavailable in the United States, absence of a demonstrable effect of bicar- established. Until such benefit has been it is encouraging to note that nebulized bonate to shift K into cells over the short demonstrated, the use of hypertonic (3%) albuterol in a high dose, administered to term is not to imply that bicarbonate saline should be restricted to hyponatre- patients with end-stage renal disease, has might not be useful in the emergency mic patients with hyperkalemia, with an a similar effect: PK declines by 0.6 treatment of hyperkalemia; rather, that awareness of the volume overload that mmol/L after inhalation of 10 mg of al- its onset and mechanism of action are may ensue. buterol, and by about 1.0 mmol/L after 20 quite different from what conventional mg (41, 42, 48, 49). Note that the effec- wisdom has held (see below). Further- Redistribution of Potassium into tive dose is at least four times higher than more, the foregoing is not meant to im- Cells that typically used for bronchodilation ply that sodium bicarbonate should be (50), although a smaller decline in PK withheld from the hyperkalemic patient Insulin. Insulin reliably lowers PK in (about 0.4 mmol/L after 60 mins) is seen with metabolic acidosis; rather, that no patients with end-stage renal disease even with a metered-dose inhaler (51). short-term effect on the PK should be (39–43), confirming its effect to shift K The effect of high-dose therapy is appar- anticipated. into cells. The effect of insulin on potas- ent at 30 mins and persists for at least 2 sium is dose dependent from the physio- hrs (48). The effect of insulin is additive Elimination of Potassium from logic through the pharmacologic range with that of albuterol, with the combina- the Body (5). It is mediated by activation of Na, tion reported to result in a decline in PK K-ATPase, apparently by recruitment of by about 1.2 mmol/L at 60 mins (42). Enhanced Renal Elimination. Hyper- intracellular pump components into the More recently, subcutaneous terbutaline kalemia occurs most often in patients plasma membrane (4, 44). An intrave- (7 ␮g/kg body weight) has been shown to with renal insufficiency. However, renal nous dose of ten units of regular insulin reduce PK in selected dialysis patients by K excretion may be enhanced even in given as a bolus along with an intrave- an average of 1.3 mmol/L within 60 mins patients with significant renal impair- nous bolus of dextrose (25 g as a 50% (52). Mild tachycardia is the most com- ment by increasing the delivery of sol- solution) to anephric adult patients low- mon reported side effect of high-dose ute to the distal nephron, the site of K ers the PK by about 0.6 mmol/L (42). The nebulized albuterol or terbulatine. Pa- secretion. onset of action is Ͻ15 mins and the effect tients taking nonselective ␤-adrenocep- Studies using acetazolamide show is maximal between 30 and 60 mins after tor blockers will be unlikely to manifest that bicarbonate delivery to this site in a single bolus (41, 42). After the initial the hypokalemic effect of albuterol. Even the nephron has a particular kaliuretic bolus, a dextrose infusion should be among patients not taking ␤-blockers, as effect (58), even in patients with renal started, since a single bolus of 25 g of many as 40% seem to be resistant to the insufficiency (59). It would be unwise to

3248 Crit Care Med 2008 Vol. 36, No. 12 administer acetazolamide alone to most be 1.8% among postoperative patients re- kidney function, although it use for this patients with hyperkalemia, since they ceiving SPS (75). Thus, the slow onset of purpose has not been systematically eval- tend to present with a concomitant met- action and serious, albeit infrequent, toxic- uated. SPS resin has a slow onset of ac- abolic acidosis that would be exacerbated ity make SPS a poor choice for the treat- tion and debatable efficacy. Furthermore, by the drug. But a sodium bicarbonate ment of urgent hyperkalemia. it carries the small risk of intestinal ne- infusion administered during 4–6 hrs at Dialysis. Hemodialysis is the method crosis. Hemodialysis remains the most a rate designed to alkalinize the urine of choice for removal of potassium from reliable tool for removing K from the may enhance urinary K excretion (53), the body. PK falls by over 1 mmol/L in the body in patients with kidney failure. and would be desirable especially in pa- first 60 mins of hemodialysis and a total tients with metabolic acidosis. The risk of of 2 mmol/L by 180 mins, after which it REFERENCES volume expansion with the bicarbonate reaches a plateau (3, 76). Rebound always infusion can be mitigated by the use of occurs after dialysis, with 35% of the re- 1. Gennari FJ: Disorders of potassium ho- loop-acting diuretics, which would be duction abolished after an hour and meostasis. Hypokalemia and hyperkalemia. likely to further enhance the kaliuretic nearly 70% after 6 hr; the magnitude of Crit Care Clin 2002; 18:273–288, vi 2. Stevens MS, Dunlay RW: Hyperkalemia in effect. Loop-acting diuretics alone or in the postrebound PK is proportional to the hospitalized patients. Int Urol Nephrol 2000; combination with a thiazide diuretic will predialysis PK (77). There is controversy induce a kaliuresis and will be beneficial as to whether dialysis for severe hyperka- 32:177–180 3. Ahmed J, Weisberg LS: Hyperkalemia in di- in the volume expanded patient. Diuretic- lemia precipitates serious ventricular ar- alysis patients. Semin Dial 2001; 14:348–356 induced volume contraction must be rhythmias (78–84). Because of that pos- 4. Clausen T, Everts ME: Regulation of the Na, avoided since this will lead to decreased sibility, patients dialyzed for severe K-pump in skeletal muscle. Kidney Int 1989; distal nephron flow and reduced K excre- hyperkalemia should have continuous 35:1–13 tion (30). EKG monitoring (3). 5. DeFronzo RA, Felig P, Ferrannini E, et al: Exchange Resin. Sodium polystyrene The rate of potassium removal with Effect of graded doses of insulin on sulfonate (SPS, Kayexalate, Kionex) is a peritoneal dialysis is much slower than splanchnic and peripheral potassium me- cation-exchange resin that is prepared in with hemodialysis. Indeed, much of the tabolism in man. Am J Physiol 1980; 238: E421–E427 the sodium phase but has a higher affin- decrement in PK with peritoneal dialysis ity for potassium than sodium (60). In seems to be due to translocation of po- 6. DeFronzo RA, Bia M, Birkhead G: Epineph- rine and potassium homeostasis. Kidney Int the lumen of the intestine, it exchanges tassium into cells as a result of the glu- 1981; 20:83–91 sodium for secreted potassium. Most of cose load rather than extracorporeal dis- 7. Rosa RM, Silva P, Young JB, et al: Adrenergic this exchange takes place in the colon, posal. This modality may be used for modulation of extrarenal potassium disposal. the site of most potassium secretion patients on maintenance peritoneal dial- N Engl J Med 1980; 302:431–434 in the gut (61, 62). Each gram of resin ysis who have modest hyperkalemia (3). 8. Adrogue HJ, Madias NE: Changes in plasma binds approximately 0.65 mmol of potas- potassium concentration during acute acid- sium in vivo, although the effect is highly CONCLUSION base disturbances. Am J Med 1981; 71: variable and unpredictable (63). The resin 456–467 causes constipation and, hence, almost Hyperkalemia is one of the few poten- 9. Magner PO, Robinson L, Halperin RM, et al: always is given with a cathartic. It may be tially lethal electrolyte disorders. Its ra- The plasma potassium concentration in met- abolic acidosis: A re-evaluation. Am J Kidney given orally or by retention enema, al- tional treatment has evolved as a result of Dis 1988; 11:220–224 though the oral route is considered to be our more complete understanding of the 10. Perez GO, Oster JR, Vaamonde CA: Serum more effective because of the longer tran- physiology of potassium homeostasis. potassium concentration in acidemic states. sit time through the gut lumen (3). The sequential approach to the treatment Nephron 1981; 27:233–243 There are two concerns with the use of of urgent hyperkalemia still pertains. Cal- 11. Mount DB, Yu ASL: Transport of inorganic SPS for the treatment of urgent hyperka- cium gluconate is the preferred agent to solutes: Sodium, chloride, potassium, mag- lemia. The first is its slow effect. When immediately reverse the adverse electro- nesium, calcium and phosphorus. In: Bren- given orally, the onset of action is at least physiologic effects of hyperkalemia, al- ner and Rector’s The Kidney. Brenner B (Ed). 2 hrs and the maximum effect may not be though hypertonic saline may be used in Eighth Edition. Philadelphia, Saunders, seen for 6 hrs or more (60). The effect of selected circumstances. Insulin is the 2007, pp 181–183 12. Velazquez H, Ellison DH, Wright FS: Lumi- SPS as a retention enema is more rapid most reliable agent for translocating K ␤ nal influences on potassium secretion: Chlo- but of lesser magnitude. One recent study into cells, but -adrenoceptor agonists ride, sodium, and thiazide diuretics. Am J in normokalemic hemodialysis patients provide some additional benefit in about Physiol 1992; 262:F1076–F1082 failed to show any effect on PK during 12 60% of patients. Terbutaline may have 13. Dunn MJ: Nonsteroidal antiinflammatory hrs after an oral dose of 30 g of SPS with some utility in this regard, but its use has drugs and renal function. Annu Rev Med cathartic (64). Indeed, early studies with never been studied in patients with heart 1984; 35:411–428 this agent showed very little effect over and disease. ␤-adrenoceptor agonists should 14. Mathews A, Bailie GR: Acute renal failure and above that of sorbitol alone (60). The sec- never be used without insulin for this hyperkalemia associated with triamterene ond concern with SPS is its possible toxic- purpose, since about 40% of patients will and indomethacin. Vet Hum Toxicol 1986; ity. There are numerous case reports of have no response. Sodium bicarbonate 28:224–225 15. Paladini G, Tonazzi C: Indomethacin- patients who have developed intestinal ne- seems to have no effect to shift K into induced hyperkalemia and renal failure in crosis after exposure to SPS in sorbitol as cells, even after several hours. It is likely multiple myeloma. Acta Haematol 1982; 68: an enema (63, 65–69), and as an oral agent to be effective, especially in combination 256–260 (66, 70–74). A retrospective study esti- with a diuretic drug, in enhancing uri- 16. Juurlink DN, Mamdani MM, Lee DS, et al: mated the prevalence of colonic necrosis to nary K elimination in patients with some Rates of hyperkalemia after publication of

Crit Care Med 2008 Vol. 36, No. 12 3249 the Randomized Aldactone Evaluation Study. port of two deaths. JAMA 1936; 106: 52. Sowinski KM, Cronin D, Mueller BA, et al: N Engl J Med 2004; 351:543–551 1151–1153 Subcutaneous terbutaline use in CKD to re- 17. Palmer BF: Managing hyperkalemia caused 36. Shrager MW: Digitalis intoxication; a review duce potassium concentrations. Am J Kidney by inhibitors of the renin-angiotensin- and report of forty cases, with emphasis on Dis 2005; 45:1040–1045 aldosterone system. N Engl J Med 2004; 351: etiology. AMA Arch Intern Med 1957; 100: 53. Fraley DS, Adler S: Correction of hyperkale- 585–592 881–893 mia by bicarbonate despite constant blood 18. Faridi AB, Weisberg LS: Acid-base, electro- 37. Garcia-Palmieri MR: Reversal of hyperkale- pH. Kidney Int 1977; 12:354–360 lyte and metabolic abnormalities. In: Critical mic cardiotoxicity with hypertonic saline. 54. Allon M, Shanklin N: Effect of bicarbonate Care Medicine: Principles of Diagnosis and Am Heart J 1962; 64:483–488 administration on plasma potassium in dial- Management in Adults. Parrillo JE, Dellinger 38. Ballantyne F III, Davis LD, Reynolds EW Jr, ysis patients: Interactions with insulin and RP (Eds). Third Edition. Philadelphia, et al: Cellular basis for reversal of hyperkale- albuterol. Am J Kidney Dis 1996; 28:508–514 Elsevier, 2008, pp 1203–1243 mic electrocardiographic changes by so- 55. Gutierrez R, Schlessinger F, Oster JR, et al: 19. Fisch C: Relation of electrolyte disturbances dium. Am J Physiol 1975; 229:935–940 Effect of Hypertonic versus isotonic sodium- to cardiac arrhythmias. Circulation 1973; 47: 39. Sterns RH, Feig PU, Pring M, et al: Disposi- bicarbonate on plasma potassium concentra- 408–419 tion of intravenous potassium in anuric man: tion in patients with end-stage renal-disease. 20. Surawicz B: Electrolytes and the electrocar- A kinetic analysis. Kidney Int 1979; 15: Miner Electrolyte Metab 1991; 17:297–302 diograim. Postgrad Med 1974; 55:123–129 651–660 56. Kim HJ: Combined effect of bicarbonate and 21. Moore ML, Bailey RR: Hyperkalaemia in pa- 40. Blumberg A, Weidmann P, Shaw S, et al: insulin with glucose in acute therapy of hy- tients in hospital. N Z Med J 1989; 102: Effect of various therapeutic approaches on perkalemia in end-stage renal disease pa- 557–558 plasma potassium and major regulating fac- tients. Nephron 1996; 72:476–482 22. Paice B, Gray JM, McBride D, et al: Hyperkal- tors in terminal renal failure. Am J Med 57. Blumberg A, Weidmann P, Ferrari P: Effect aemia in patients in hospital. Br Med J (Clin 1988; 85:507–512 of prolonged bicarbonate administration on Res Ed) 1983; 286:1189–1192 41. Lens XM, Montoliu J, Cases A, et al: Treat- plasma potassium in terminal renal failure. 23. Montague BT, Ouellette JR, Buller GK: Ret- ment of hyperkalaemia in renal failure: Sal- Kidney Int 1992; 41:369–374 rospective review of the frequency of ECG butamol v. insulin. Nephrol Dial Transplant 58. Carlisle EJ, Donnelly SM, Ethier JH, et al: changes in hyperkalemia. Clin J Am Soc 1989; 4:228–232 Modulation of the secretion of potassium by Nephrol 2008; 3:324–330 42. Allon M, Copkney C: Albuterol and insulin accompanying anions in humans. Kidney Int 24. Szerlip HM, Weiss J, Singer I: Profound hy- for treatment of hyperkalemia in hemodialy- 1991; 39:1206–1212 sis patients. Kidney Int 1990; 38:869–872 59. Kamel KS, Ethier JH, Quaggin S, et al: Stud- perkalemia without electrocardiographic 43. Allon M, Takeshian A, Shanklin N: Effect of ies to determine the basis for hyperkalemia manifestations. Am J Kidney Dis 1986; insulin-plus-glucose infusion with or with- in recipients of a renal transplant who are 7:461–465 out epinephrine on fasting hyperkalemia. treated with cyclosporine. J Am Soc Nephrol 25. Dodge HT, Grant RP, Seavey PW: The effect Kidney Int 1993; 43:212–217 1992; 2:1279–1284 of induced hyperkalemia on the normal and 44. Hundal HS, Marette A, Mitsumoto Y, et al: 60. Scherr L, Ogden DA, Mead AW, et al: Man- abnormal electrocardiogram. Am Heart J Insulin induces translocation of the alpha 2 agement of hyperkalemia with a cation- 1953; 45:725–740 and beta 1 subunits of the Naϩ/K(ϩ)-ATPase exchange resin. N Engl J Med 1961; 264: 26. Weiner ID, Wingo CS: Hyperkalemia: A po- from intracellular compartments to the 115–119 tential silent killer. J Am Soc Nephrol 1998; plasma membrane in mammalian skeletal 61. Frohnert PP, Johnson WJ, Mueller GJ, et al: 9:1535–1543 muscle. J Biol Chem 1992; 267:5040–5043 Resin treatment of hyperkalemia. II. Clinical 27. Weiner M, Epstein FH: Signs and symptoms 45. Minaker KL, Rowe JW: Potassium homeosta- experience with a cation exchange resin (cal- of electrolyte disorders. Yale J Biol Med 1970; sis during hyperinsulinemia: Effect of insulin cium cycle). J Lab Clin Med 1968; 71: 43:76–109 level, beta-blockade, and age. Am J Physiol 840–846 28. Tannen RL: Relationship of renal ammonia 1982; 242:E373–E377 62. Frohnert PP, Johnson WJ, Mueller GJ, et al: production and potassium homeostasis. Kid- 46. Goldfarb S, Cox M, Singer I, et al: Acute Resin treatment of hyperkalemia. I. Ex- ney Int 1977; 11:453–465 hyperkalemia induced by hyperglycemia: change properties of a cation exchange resin 29. Szylman P, Better OS, Chaimowitz C, et al: Hormonal mechanisms. Ann Intern Med (calcium cycle). J Lab Clin Med 1968; 71: Role of hyperkalemia in the metabolic acido- 1976; 84:426–432 834–839 sis of isolated hypoaldosteronism. N Engl 47. Montoliu J, Lens XM, Revert L: Potassium- 63. Lillemoe KD, Romolo JL, Hamilton SR, et al: J Med 1976; 294:361–365 lowering effect of albuterol for hyperkalemia Intestinal necrosis due to sodium polysty- 30. Weisberg LS: Potassium homeostasis. In: in renal failure. Arch Intern Med 1987; 147: rene (Kayexalate) in sorbitol enemas: Clini- Principles and Practice of Medical Intensive 713–717 cal and experimental support for the hypoth- Care. Carlson RW, Geheb MA (Eds). Philadel- 48. Allon M, Dunlay R, Copkney C: Nebulized esis. Surgery 1987; 101:267–272 phia, Saunders, 1993 albuterol for acute hyperkalemia in patients 64. Gruy-Kapral C, Emmet M, Santa Nan CA, 31. Bisogno JL, Langley A, Von Dreele MM: Ef- on hemodialysis. Ann Intern Med 1989; 110: et al: Effect of single dose resin-cathartic fect of calcium to reverse the electrocardio- 426–429 therapy on serum potassium concentration graphic effects of hyperkalemia in the iso- 49. Montoliu J, Almirall J, Ponz E, et al: Treat- in patients with end-stage renal disease. JAm lated rat heart: A prospective, dose-response ment of hyperkalaemia in renal failure with Soc Nephrol 1998; 9:1924–1930 study. Crit Care Med 1994; 22:697–704 salbutamol inhalation. J Intern Med 1990; 65. Rogers FB, Li SC: Acute colonic necrosis 32. Chamberlain MJ: Emergency treatment of 228:35–37 associated with sodium polystyrene sulfonate hyperkalaemia. Lancet 1964; 18:464–467 50. Nair S, Thomas E, Pearson SB, et al: A ran- (Kayexalate) enemas in a critically ill patient: 33. Ettinger PO, Regan TJ, Oldewurtel HA: Hy- domized controlled trial to assess the opti- Case report and review of the literature. perkalemia, cardiac conduction, and the mal dose and effect of nebulized albuterol in J Trauma 2001; 51:395–397 electrocardiogram: A review. Am Heart J acute exacerbations of COPD. Chest 2005; 66. Rashid A, Hamilton SR: Necrosis of the gas- 1974; 88:360–371 128:48–54 trointestinal tract in uremic patients as a 34. Semple P, Booth C: Calcium chloride; a re- 51. Mandelberg A, Krupnik Z, Houri S, et al: result of sodium polystyrene sulfonate minder. Anaesthesia 1996; 51:93 Salbutamol metered-dose inhaler with (kayexalate) in sorbitol: An underrecognized 35. Bower J, Mengle H: The additive effect of spacer for hyperkalemia: How fast? how safe? condition. Am J Surg Pathol 1997; 21:60–69 calcium and digitalis. A warning with a re- Chest 1999; 115:617–622 67. Scott TR, Graham SM, Schweitzer EJ, et al:

3250 Crit Care Med 2008 Vol. 36, No. 12 Colonic necrosis following sodium polysty- kayexalate-sorbitol. South Med J 2000; 93: and efficacy of low-potassium dialysate. Am J rene sulfonate (kayexalate)-sorbitol enema in 511–513 Kidney Dis 1989; 13:137–143 a renal transplant patient. Report of a case 73. Gardiner GW: Kayexalate (sodium polysty- 79. Karnik JA, Young BS, Lew NL, et al: Cardiac and review of the literature. Dis Colon Rec- rene sulphonate) in sorbitol associated with arrest and sudden death in dialysis units. tum 1993; 36:607–609 intestinal necrosis in uremic patients. Can J Kidney Int 2001; 60:350–357 68. Wootton FT, Rhodes DF, Lee WM, et al: Co- Gastroenterol 1997; 11:573–577 80. Kovesdy CP, Regidor DL, Mehrotra R, et al: lonic necrosis with kayexalate-sorbitol ene- 74. Roy-Chaudhury P, Meisels IS, Freedman S, Serum and dialysate potassium concentra- mas after renal transplantation. Ann Intern et al: Combined gastric and ileocecal toxicity tions and survival in hemodialysis patients. Med 1989; 111:947–949 (serpiginous ulcers) after oral kayexalate in Clin J Am Soc Nephrol 2007; 2:999–1007 69. Chatelain D, Brevet M, Manaouil D, et al: sorbital therapy. Am J Kidney Dis 1997; 30: 81. Lafrance JP, Nolin L, Senecal L, et al: Pre- Rectal stenosis caused by foreign body reac- 120–122 dictors and outcome of cardiopulmonary re- 75. Gerstman BB, Kirkman R, Platt R: Intestinal suscitation (CPR) calls in a large haemodi- tion to sodium polystyrene sulfonate crystals necrosis associated with postoperative orally alysis unit over a seven-year period. Nephrol (kayexalate). Ann Diagn Pathol 2007; 11: administered sodium polystyrene sulfonate Dial Transplant 2006; 21:1006–1012 217–219 in sorbitol. Am J Kidney Dis 1992; 20: 82. Morrison G, Michelson EL, Brown S, et al: 70. Abraham SC, Bhagavan BS, Lee LA, et al: 159–161 Mechanism and prevention of cardiac ar- Upper gastrointestinal tract injury in pa- 76. Blumberg A, Roser HW, Zehnder C, et al: rhythmias in chronic hemodialysis patients. tients receiving kayexalate (sodium polysty- Plasma potassium in patients with terminal Kidney Int 1980; 17:811–819 rene sulfonate) in sorbitol: Clinical, endo- renal failure during and after haemodialysis; 83. Redaelli B, Locatelli F, Limido D, et al: scopic, and histopathologic findings. Am J relationship with dialytic potassium removal Effect of a new model of hemodialysis po- Surg Pathol 2001; 25:637–644 and total body potassium. Nephrol Dial tassium removal on the control of ventric- 71. Cheng ES, Stringer KM, Pegg SP: Colonic Transplant 1997; 12:1629–1634 ular arrhythmias. Kidney Int 1996; 50: necrosis and perforation following oral so- 77. Zehnder C, Gutzwiller JP, Huber A, et al: 609–617 dium polystyrene sulfonate (resonium Low-potassium and glucose-free dialysis 84. Sforzini S, Latini R, Mingardi G, et al: Ven- A/kayexalate in a burn patient. Burns 2002; maintains urea but enhances potassium re- tricular arrhythmias and four-year mortality 28:189–190 moval. Nephrol Dial Transplant 2001; 16: in haemodialysis patients. Gruppo Emodi- 72. Dardik A, Moesinger RC, Efron G, et al: Acute 78–84 alisi e Patologie Cardiovascolari. Lancet abdomen with colonic necrosis induced by 78. Hou S, McElroy PA, Nootens J, et al: Safety 1992; 339:212–213

Crit Care Med 2008 Vol. 36, No. 12 3251