Hypokalemia is the serum manifestation of a whole-body potassium deficit. That potassium is largely intracellular, and so the serum represents a tiny portion of the total. If one were for whatever reason potassium-depleted, one would make an attempt to conserve daily potassium loss (in which the renal loss plays the greatest role).  There is a certain obligatory daily loss of potassium; people whose potassium intake was restricted still produced a certain minimum output. The minimal urinary potassium concentration seems to be about 2-5mmol/L, which means that in the absence of dietary potassium your intracellular reserves will gradually dwindle away (Squires and Huth, 1959).

Surprisingly, hypokalemia has never been directly asked about in the CICM fellowship exam. It has however come up indirectly, as various hypokalemic blood gases have been shown to the candidates. As a question topic, it lends itself very well to those "list six of something" questions, which are very popular with the examiners these days (because they are easy to mark).

Oh's Manual mentions hypokalemia briefly, and makes it the subject of a couple of Blue Boxes. It is not an ideal resource for hypokalemia reading, and may not be adequate to answer any SAQ except for the most superficial "list six of something" sort of questions. For detailed knowledge of the topic, the best starting point would probably be the 1998 NEJM article by FJ Gennari.  It probably contains more detail than you need to know in the text, but the tables are particularly useful, as they list all the possible and impossible causes of hypokalemia. This article was used as the backbone for my summary.

Diagnosis of the causes of hypokalemia

Oh's Box 93.5 (page 955) sorts the causes of hypokalemia according to their underlying physiological disturbance. You are either losing potassium, or you are failing to gain enough of it, or it has shifted into the cells somehow.

Causes of Hypokalemia, Sorted According to Physiology

Gastrointestinal potassium loss

Gastrointestinal loss

  • Vomiting and nasogastric aspiration
  • Tumour-related diarrhoea
    • Villous adenoma of the colon
    • VIPoma
    • Zollinger-Ellison syndrome
  • Infectious diarrhoea
    • Cholera
    • Salmonella
    • Strongyloides
    • AIDS-associated diarrhoea
  • Chemical diarrhoea
    • Laxative abuse
  • Medical and surgical complications
    • Enteric fistula
    • High-output fistula
    • Radiation enteropathy

Inadequate potassium intake

  • Strange potassium-poor diet
  • Cation exchange media
    • Intentional, eg. Resonium and other cation-exchange resins
    • Unintentional, eg. geophagia with ingestion of clay
  • Malabsorption:
    • Short gut syndromes
    • Iron transport defects
  • Inadequate parenteral potassium supplementation (eg. poorly compounded TPN)

Renal potassium loss

  • Malignant hypertension
  • Renal artery stenosis
  • Mineralocorticoid-associated effects
    • Conn’s syndrome
    • Cushing’s syndrome
    • Bartter’s syndrome
  • Ectopic ACTH syndrome
    • Small-cell carcinoma of the lung
    • Pancreatic carcinoma
    • Carcinoma of the thymus
  • Renal effect of drugs....See a more detailed list below
    • Diuretics
    • Corticosteroids
    • Carbapenems
    • Aamphotericin B
    • Gentamicin
    • Cisplatin   

Compartment shift

  • Alkalosis
  • Insulin
  • Na + /K + ATPase stimulation
    • Sympathomimetic agents with β2 effect
    • Methylxanthines
  • Barium poisoning
  • Hypothermia
  • Toluene intoxication
  • Hypokalaemic periodic paralysis
  • Delayed following blood transfusion

An alternative classification can be offered, where the causes of hypokalemia are organised by urinary potassium.
Generally speaking, the "low" threshold for urinary potassium is under 2mmol/L, whereas anything over 5mmol/L is considered "high" when you are hypokalemic.

Causes of Hypokalemia, Sorted According to Urinary Potassium

Low urinary potassium

High urinary potassium with acidosis

  • Distal tubular acidosis (Type 1)
  • Proximal tubular acidosis (Type 2)
  • Diabetic ketoacidosis

High urinary potassium with alkalosis

  • Diuretics
  • Gastrointestinal secretion losses (eg. vomiting, excessive NG suction, etc)
  • Bartter syndrome
  • Gitelman syndrome
  • Derangement of renin-aldosterone axis
    • Low renin, high aldosterone
      • Primary hyperaldosteronism
    • High renin, high aldosterone
      • Cardiac failure
      • Renal artery stenosis
      • Renin-secreting tumours
    • Low renin, low aldosterone
      • Adrenal hyperplasia
      • Cushing syndrome
      • Exogenous corticosteroid excess
      • Liddle syndrome
    • Cushings disease
    • Exogenous corticosteroid excess
    • Adrenal hyperplasia

This classification has the advantage of being directly related to the tests which you will do to discriminate between causes of hypokalemia. The disadvantages is a failure to account for some of the causes of compartment shift.

Furthermore, the abovedescribed classification systems fail to acknowledge that drugs are probably responsible for 99% of hypokalemia seen in hospital clinical practice, and particularly in the ICU. The following adopted table from the 1998 NEJM article by FJ Gennari attributes appropriate credit to drugs as causes of hypokalemia. Some adjustment was required to Australinise the 1998 American drug list (all epinephrines and albuterols have been adrenalined and salbutamolled).

Drugs Which Cause Hypokalemia

Drugs which cause a compartment shift

  • Insulin
  • β-2 agonists
    • Salbutamol, etc
    • Adrenaline
    • Isoprenaline
    • Ephedrine
    • Pseudoephedrine
  • Methylxanthines
    • Theophylline
    • Aminophylline
    • Caffeine
  • Verapimil (in toxicity)
  • Chloroquine (in toxicity)

Drugs which cause increased potassium loss

Drugs which cause increased intestinal loss

  • Resonium and similar cation exchange resins
  • Phenolphthalein, which is used as a laxative

Drugs which cause increased renal loss

  • Diuretics
    • Thiazides
    • Acetazolamide
    • Loop diuretics
  • Steroids
    • Fludrocortisone
    • Corticosteroids
  • Substances with mineralocorticoid effect
    • Licorice
  • Antibiotics at high dose
    • β-lactams (pretty much all of them)
  • Drugs which deplete magnesium
    • Aminoglycosides
    • Cisplatin
    • Doscarnet
    • Amphtericin

An unusual finding from reading around this topic is the potential for high dose antibiotics to cause hypokalemia. Turns out, it is a real thing. The mechanism appears to be due to the anionic property of such antibiotics. Acting as an unresorbable anion, the whatever-cillin tends to "drag" the cationic potassium out of the tubular cells, preventing its reabsoprtion by a brutally stupid electroneutrality mechanism.  Lipner et al (1975) described this with carbenicillin, but it appears to be relevant to all renally excreted β-lactam antibiotics.

Clinical Features of Hypokalemia

Physical examination findings

  • Weakness
  • Decreased muscle tone
  •  

Complications of prolonged hypokalemia

  • Rhabdomyolysis
  • Polyuria due to nephrogenic DI

Features of History

  • Ileus
  • Failure to wean from ventilation
  • Coma

ECG findings

  • Ventricular tachycardia: classically, torsades de pointes
  • Atrial tachycardias
  • PR interval prolongation (>200 msec)
  • P wave amplitude increased (>2.5 mm in limb leads, >1.5 mm in chest leads)
  • P wave width increased (>120 msec)
  • u-waves
  • T-wave inversion
  • Ectopics (ventricular and atrial)

As an illustration from a real-life scenario, here is an ECG of an alcoholic who spent three days vomiting relentlessly as a consequence of pancreatic inflammation. Relevant ABG electrolyte data is offered alongside the ECG image. U-waves are prominent in V2-V4.

severe hypokalemia- ECG changes with u-waves

Safe potassium replacement

What should we accept as a safe level of serum potassium?

The greater proportion of the human population walks around with a serum potassium around 3.5mmol/L, which calls into question the usual ICU practice of obsessively correcting it to a value closer to 4.0-4.5 mmol/L. How did this practice develop, and why do we ignore the usual "normal" laboratory value?

What we do in ICU is an example of "protocol creep", these recommendations having crept in from the cath lab.  The 2005 AHA recommendations suggest a K+ level above 4.0 as the ideal level for patients recovering from acute coronary syndromes (to protect them from ventricular arrhythmia). Of the acute coronary syndrome patients to which the AHA was referring, many would have been in the ICU. It is easy to translate the replacement protocol for ACS patients to the rest of the ICU population, considering that many non-cardiac ICU patients end up having ACS.  The relatively harmless nature of potassium correction plays a role (why not do it for everybody, one might ask). More recently, Goyal et al (2012) have demonstrated that there is in fact a U-shaped relationship between potassium levels and mortality in cardiology patients, with the lowest mortality seen among patients whose potassium is between 3.5 and 4.5 mmol/L. In short, there does not appear to be any strong evidence for maintaining potassium levels in a supra-physiological range for the majority of ICU patients. However, it is universally acknowledged that ICU patients are at a greater risk of both developing hypokalemia and developong complications from hypokalemia. Because of such thinking, routine replacement to a level of 4.5-5.0mmol/L continues locally, and at many other units.

Methods of potassium replacement

We have all heard the trite bullshit about "feed'em more bananas", and for most experienced ICU staff hearing this joke is a tooth-gratingly painful experience. Fortunately, FJ Gennari had published a NEJM article in 1998 which contains within it. In short, seaweed is king. 

Foods with High Potassium Content

Extreme (25mmol/100g) Very high (12.5mmol/100g) High (6.2mmol/100g)
  • Dried figs
  • Molasses
  • Seaweed
  • Dried fruits, eg. prunes
  • Nuts
  • Avocados
  • Bran cereals
  • Wheat germ
  • Lima beans
  • Spinach
  • Tomatoes
  • Broccoli
  • Squash
  • Beetroot
  • Carrots
  • Cauliflower
  • Potatoes
  • Bananas
  • Kiwi fruit
  • Oranges
  • Mangos
  • All sorts of meat

In the ICU, you are unlikely to encounter a situation where increasing a patient's seaweed intake will help you manage their dangerous hypokalemia. Of greater relevance probably will be this table of potassium content among (locally relevant, frequently seen) enteric nutritional supplements:

Potassium Content of Enteral Supplements

Supplement brand name Potassium content in mmol/L

Jevity (1.0 kcal/ml)

40.2

Jevity Plus (1.2 kcal/ml)

47.7

Jevity HiCal (1.5 kcal/ml)

42.2

Nepro (2.0 kcal/ml)

27.1

Pulmocare (1.5 kcal/ml)

50.1

Promote (1.0 kcal/ml)

53.7

Of even greater relevance would be a listing of the potassium content of commonly used drugs and oral potassium supplements. Locally, the major enteral supplement for the outpatient population is Slow-K, which offers 8mmol of KCl per waxy slow-release tablet. Of greater interest to the nasogastrically intubated group are the effervescent Chlorvescent tablets. These contain 14mmol of potassium of which 8mmol is KCl and the rest is potassium bicarbonate which fizzes with citric acid when dunked into water. The result of this is "an acceptable drink", according to the manufacturer. In the author's opinion, it only barely meets this description.

What is a safe rate of rapid  IV potassium replacement?

Routinely, the local practice is to replace potassium via a central line at a rate of 20mmol/L. Oh's Manual recommends that the maximum rate of potassium replacement should not exceed 40 mmol/L. As the reference for this statement, Finfer and Delaney offer this 1977 article by J.R Stockigt which is not available in free full text. The latter article makes reference to several studies which have explored irresponsibly rapid replacement rates, with (largely) surviving patients. For instance, Hamill et al (1991) replaced at 40mmol/hr via central lines, into ICU patients some of whom had renal failure, and found that such rates were "safe to administer and effectively increased serum potassium levels in a dose-dependent and predictable fashion". In patients with diabetic ketoacidosis, the potassium ends up sucked up into cells at a remarkably rapid rate and a 1999 article by P. Glover ("Hypokalemia", which is hosted on the CICM site) reports that up to 100mmol of potassium had been safely given to such patients over 1 hour.

The importance of replacing magnesium together with potassium

It is a widely acknowledged truth that replacement of magnesium must take place at the same time as replacement of potassium. Indeed, hypokalemia will remain refractory to replacement until you replace the magnesium alongside it, a phenomenon which will lead to bewilderment if you haven't measured the magnesium level (Whang et al, 1992). Magnesium deficiency aggravates potassium deficiency by several mechanisms, explored in great detail by Huang et al (2007).

In brief, magnesium deficiency impairs Na-K-ATPase, which would decrease cellular uptake of K+ (because magnesium is an important co-factor for the use of ATP by this exchange pump). A decrease in cellular uptake of K+ leads to serum hypokalemia if it is associated with increased renal or gastrointestinal losses (because cellular reserves are not replenished, and serum reserves are rapidly depleted by the wasting process). Magnesium also decreases renal potassium excretion. It does this by influencing the action of the ROMK inward-rectifying K+ channel, which is responsible for potassium flow into the lumen (to maintain luminal electroneutrality when the ENaC channel reabsorbs sodium under the influence of aldosterone).

References

Assadi, Farahnak. "Diagnosis of hypokalemia: a problem-solving approach to clinical cases." Iranian journal of kidney diseases 2.3 (2008): 115-122.

Weiner, I. David, and Charles S. Wingo. "Hypokalemia--consequences, causes, and correction." Journal of the American Society of Nephrology 8.7 (1997): 1179-1188.

Links, Thera P., et al. "Familial hypokalemic periodic paralysis: clinical, diagnostic and therapeutic aspects." Journal of the neurological sciences 122.1 (1994): 33-43.

Gennari, F. John. "Hypokalemia." New England Journal of Medicine 339.7 (1998): 451-458.

Huang, Chou-Long, and Elizabeth Kuo. "Mechanism of hypokalemia in magnesium deficiency." Journal of the American Society of Nephrology 18.10 (2007): 2649-2652.

Stockigt, J. R. "Potassium metabolism." Anaesthesia and intensive care 5.4 (1977): 317-325.

AHA 2005 American Heart Association. "American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care: Part 8: Stabilization of the Patient With Acute Coronary Syndromes." Circulation 112 (2005): 89-110.

Goyal, Abhinav, et al. "Serum potassium levels and mortality in acute myocardial infarction." Jama 307.2 (2012): 157-164.

Gennari, F. John. "Hypokalemia." New England Journal of Medicine 339.7 (1998): 451-458.

Kasama, Richard, and Alfred Sorbello. "Renal and electrolyte complications associated with antibiotic therapy." American family physician 53 (1996): 227-236.

Lipner, Henry I., et al. "The behavior of carbenicillin as a nonreabsorbable anion." The Journal of laboratory and clinical medicine 86.2 (1975): 183-194.

Hamill-Ruth, Robin J., and Robb McGory. "Magnesium repletion and its effect on potassium homeostasis in critically ill adults: results of a double-blind, randomized, controlled trial." Critical care medicine 24.1 (1996): 38-45.

Whang, Robert, David D. Whang, and Michael P. Ryan. "Refractory potassium repletion: a consequence of magnesium deficiency." Archives of internal medicine 152.1 (1992): 40-45.

Huang, Chou-Long, and Elizabeth Kuo. "Mechanism of hypokalemia in magnesium deficiency." Journal of the American Society of Nephrology 18.10 (2007): 2649-2652.

Glover, P. "Hypokalaemia." Critical Care and Resuscitation 1999; 1: 239-251.

HAMILL, ROBIN J., et al. "Efficacy and safety of potassium infusion therapy in hypokalemic critically ill patients." Critical care medicine 19.5 (1991): 694-699.