Hypokalemia
The T flattens and a small U appears
Hypokalemia
criticalFalling potassium slows repolarisation: the T flattens, the ST sags, a U wave grows into a long QU, and premature beats landing on it can start torsades de pointes.
On the trace
The tracing above is K⁺ 3.0, the earliest form; the Variant control beside the title steps down through the rest. Turn on the Normal switch under the tracing to draw the same heart at a normal potassium in grey behind it.
- Look at the T in II and V3.
Show
Lower than normal: 1.5 mm in II against 2.3, and 3.5 mm in V3 against 4.8. A falling T is the earliest change. - Look just after the T.
Show
A small second wave, the U: 0.5 mm in II, 0.7 mm in V3, about a quarter of the T. Tap a lead and choose T in the box that opens: it gives the U's height, and its height against the T's. - Look at the ST.
Show
Slightly below the baseline: 0.4 mm in II and 0.6 mm in V3 at the J point. - Measure the QT.
Show
382 ms, QTc 420 ms: longer than normal, not yet long. - Check the rest.
Show
Sinus, 80 per minute, PR 182 ms, QRS 95 ms: unchanged. Hypokalaemia slows repolarisation, not conduction.
How to recognise it
The changes come in order as the level falls. Each row is the form of that name beside the title.
| Serum K⁺ | What appears | On this tracing |
|---|---|---|
| 3.0 to 3.4 | Often nothing. Then a lower T and a small U | K⁺ 3.0: T 1.5 mm in II, U 0.5 mm |
| 2.5 to 2.9 | ST depression; the U as tall as half the T or more, best in V2 and V3 | K⁺ 2.5: ST 0.5 mm down in II, U 1.1 mm in V3 |
| Below 2.5 | The T flat or inverted, the U taller than it; the two fuse into a long QU | K⁺ 2.0: U 0.9 mm against a T of 0.9 in II |
| Severe | Premature beats, then torsades de pointes, VT or VF | Torsades |
The QRS does not widen and the PR barely moves: if they do, look for another cause.
Mechanism
Two things go wrong in the cell. The potassium channels that carry phase 3 conduct less when there is less potassium outside, even though the push for potassium to leave is greater. And the sodium-potassium pump, which needs potassium outside to run, slows by about half at 2.7 mmol/L.
Go deeperShow less
Why the action potential lengthens unevenly
The repolarising currents are not spread evenly. The Purkinje fibres have the least of the inward rectifier current, and the mid-wall cells the longest action potentials and the least of the slow delayed rectifier. Their reserve runs out first, so they lengthen most, and the gap between the first cells to finish and the last grows.
The U wave
Where the U comes from is still debated: the late repolarisation of the Purkinje fibres, or of the mid-wall cells. Both put it in the tissue that recovers last, which is the tissue hypokalaemia lengthens most. As it separates from the rest, its repolarisation becomes a wave of its own, and then outgrows the flattening T.
Why the ST sags
The ST depression of hypokalaemia is the least explained part of the picture. It sits under the flattened T and is not ischaemic: it corrects with the potassium.
The pump, calcium and the second upstroke
A slower pump leaves sodium in the cell; the sodium-calcium exchanger then removes less calcium, and calcium builds up. That switches on calcium-calmodulin kinase II, which increases the late sodium current and the L-type calcium window current. Both push against repolarisation, so the reserve falls further. Where it runs out, the cell depolarises again before it has finished: an early afterdepolarisation. In isolated hearts, blocking that kinase or the late sodium current prevented the arrhythmias of hypokalaemia; a drug that only lengthened the QT did not cause them at normal potassium.
Clinical impact
Hypokalaemia is common and usually mild, and an ECG with only a flat T needs no more than the cause found and the potassium replaced. A long QU, frequent premature beats, digoxin, a drug that lengthens the QT, or heart disease change that: the risk is then a polymorphic VT. Low magnesium travels with low potassium and makes both the arrhythmias and the hypokalaemia itself harder to correct.
Management
Replace potassium, and find why it is low
By mouth when the ECG is normal and the level mild; intravenously, with ECG monitoring and at the rate local protocol sets, when the level is severe, the ECG abnormal or the patient cannot swallow.
Check and replace magnesium
Low magnesium makes the kidney lose potassium, so hypokalaemia will not correct until magnesium does.
Remove what lengthens the QT
Stop drugs that prolong the QT where possible, and be cautious with digoxin until potassium is normal.
Treat torsades with magnesium
Intravenous magnesium sulfate 2 g, whatever the serum magnesium; raise potassium to 4.5 to 5 mmol/L; pace or speed the rate to prevent pauses; defibrillate if it is sustained or pulseless.
Differential
ST depression in a coronary territory, often with chest pain, and changing with it. No U wave. Hypokalaemia's ST depression is widespread and comes with a U.
A scooped, sagging ST with a short QT. Hypokalaemia lengthens the QU instead. The two often coexist, and low potassium makes digoxin toxic.
Also a long QT, but from a long, flat ST with a normal T and no U. In hypokalaemia the ST is short and the length is in the T-U.
A long QT with a normal U, the T itself long or notched. The same risk of torsades, with or without hypokalaemia; often both.
References
- The electrophysiology of hypo- and hyperkalemia — Circulation: Arrhythmia and Electrophysiology, 2017
- Prevention of torsade de pointes in hospital settings: a scientific statement from the AHA and ACCF — Circulation, 2010
- Molecular basis of hypokalemia-induced ventricular fibrillation — Circulation, 2015