Saturday, May 9, 2020

Adding to the many faces of Hypokalemia....

These 2 cases came in on one shift:

Patient 1

This patient was weak with a K of 2.4 mEq/L

See the U-waves that are most prominent in V2 (which is the usual lead) and in V3-V6. 

I magnify the precordial leads here:
Now the U-waves are much easier to see

Patient 2

This patient was weak with a K of 2.0 mEq/L
This ECG appears to have an incredibly long QT in V2-V4, but that apparent T-wave really is stretched out by a 2nd hump which is the U-wave.

Hyperkalemia is called the "syphilis of ECG findings" because it comes in so many forms, but hypokalemia ECGs are multiform also.

However, they usually some variation of U-waves.

MY Comment by KEN GRAUER, MD (5/8/2020):
I thought this to be an excellent case for reinforcing a number of important ECG concepts that we have recently reviewed. These include:
  • How BEST to measure the QTc (Being sure to look at all 12 leads — and selecting THAT lead in which you can clearly see the end of the T wave — and in which the QT interval is longest).
  • Appreciating that if YOU are having difficulty determining the end of the QT interval (either because of baseline artifact and/or indistinct boundaries) — then it will also be difficult for the computer! In such cases — Be AWARE that you can not trust the computer evaluation of the QTc, and YOU need to assess the QTc yourself!
  • Being aware of the short LIST of Common Causes of QT Prolongation (ie, Drugs/Lytes/CNS  Please see My Comment in the May 3, 2020 post).
  • Appreciation of the History (Neither patient in this case had chest pain. Instead — both patients presented with weakness).

It turns out that both patients in this case had marked hypokalemia — which accounted for their presenting symptom of weakness. For the purpose of review — Figure-1 illustrates sequential ECG changes that may be anticipated with increasing severity of hypokalemia. To the contents of this Figure — I’d add the following points:
  • Not all patients read the textbook. I’ve seen many patients with all of the ECG findings shown in Figure-1 who had normal serum K+. I’ve also seen patients with mild-to-moderate hypokalemia who did not manifest any of the signs in Figure-1. Thus, in my experience — the ECG is less than optimally sensitive or specific as a tool for detecting mild-to-moderate hypokalemia. That said — with more severe degrees of hypokalemia, the ECG tends to do much better in its prediction, and the ECG is very unlikely to be “normal”.
  • PEARL #1 — The finding of U waves is not specific for hypokalemia! U waves may also be found in patients with LVH and/or bradycardiaand occasionally as a normal variant. That said — when U waves are present, and the clinical history is consistent with (or suggestive of) hypokalemia (as in a patient with vomiting, or the weakness that both patients in this case had) — prompt assessment of serum electrolytes becomes essential!
  • PEARL #2 — IF U waves are so large that they become bigger than (or almost as big asT waves — then the ECG becomes much more reliable for predicting significant hypokalemia (Panels E and F in Figure-1).
  • PEARL #3 — In my experience — Hypomagnesemia produces virtually identical ECG changes as hypokalemia. This is important — because serum Mg++ levels are not necessarily included in chem profiles unless specifically ordered.

Figure-1: Sequential development of ST-T wave changes of hypokalemia. Note increasing U wave amplitude (See text).

Let’s return to the ECGs from the 2 patients in this case. For clarity — I’ve labeled the tracings, and I’ve put them together in Figure-2.

  • HOW would you interpret these 2 ECGs?
  • For each tracing — WHICH lead(s) would you use to measure the QT interval?
  • WHICH leads show U waves?

Figure-2: The 2 ECGs in this case (See text).

My Thoughts on the ECG from Patient #1: The rhythm in the TOP tracing (from Patient #1) is sinus at ~90/minute. The PR interval is normal (ie, not more than 1 large box = 0.20 second in duration) — and, the QRS is narrow (ie, not more than half a large box = 0.10 second in duration).
  • Confession: I found it very difficult to measure the QT interval. This is because the end point of the T wave is not distinct ...
  • Given the fairly rapid heart rate — one might think from lead aVL that the QT is prolonged. But (as per Dr. Smith) — U waves are present!
  • The problem I had was distinguishing T wave from U wave. I believe we clearly see the beginning of the P wave in the long lead II (RED arrow). I therefore drew a vertical BLUE line at precisely the onset of the P wave. Note in simultaneously-obtained leads V1V2 and V3 — that there is a small, upsloping deflection seen before the onset of the P wave (BLUE arrows that clearly occur before the vertical BLUE line). This upsloping deflection must be part of the U wave (since it clearly occurs after the T wave — but before the P wave).
  • But there is also an extra “hump” in lead V2 (GREEN arrow in this lead). Does the T wave have a double hump? — or, is this 2nd ‘hump” the initial part of the U wave? Or, is the U wave biphasic (neg-pos)? My Guess — is that the answer is some combination of these possibilities ...
  • BOTTOM Line: U waves are present in this ECG — and either the QT and/or QU interval is markedly prolonged — which for practical purposes means that we need to think of the short LIST of Potential Causes of QT Prolongation = i) Drugs; ii) Lytes (ie, low K+/low Mg++ and/or low Ca++)and/oriii) some CNS Catastrophe (For more on this LIST, see My Comment in the March 19, 2019 post).

Continuing with my Descriptive Analysis of the ECG from Patient #1:
  • There is slight left axis deviation (about -10 degrees, given greater negativity than positivity in lead aVF).
  • Regarding chamber enlargement — there is LAA (deep negative component to the P wave in lead V1). If the patient is not a young adult — then there is probable LVH, given the very deep anterior S waves (IF interested in My Take on the ECG diagnosis of LVH  CLICK HERE).
  • Regarding Q-R-S-T Changes — there are small sepal q waves — normal transition (that occurs between leads V3-to-V4) — and nonspecific ST-T wave flattening in multiple leads + a long QT and/or a long QU, as previously discussed.
  • CLINICAL IMPRESSION: There are no acute changes on this tracing. As emphasized above — in light of this patient’s presenting complaint of “weakness” — this tracing is perfectly consistent with moderate-to-severe hypokalemia.
  • PEARL #4 — I have found it insightful (when possible) to follow patients like this who have severe electrolyte disorders, with serial ECGs. Doing so allows you to follow the evolution of ST-T wave and U wave changes as serum electrolytes are corrected. As is the case for hyperkalemia — You have NO IDEA as to what this patient’s true “baseline” ECG looks like until you repeat the ECG after you have corrected serum electrolytes (ie, It could be that the reason for ST flattening rather than ST depression in this case — is that baseline ST elevation has been attenuated by the ST segment lowering effect of superimposed hypokalemia!).

My Thoughts on the ECG from Patient #2: The underlying rhythm is irregularly irregular. There is significant baseline artifact in a number of leads — but no clear sign of atrial activity. Therefore, the rhythm is AFib with a controlled ventricular response. Beats #1 and 3 are PVCs.
  • Continuing with Descriptive Analysis — there is no PR interval (since the rhythm is AFib) — the QRS is narrow — and the QT interval appears prolonged.
  • There is low voltage in the limb leads (QRS amplitude in each of the 6 limb leads does not exceed 5 mm).
  • The frontal plane axis is difficult to determine — since there are bi- (or tri-) phasic QRS complexes of very small amplitude in each of the 6 limb leads. That said — the axis looks normal.
  • There is no chamber enlargement.
  • Transition is slightly delayed (ie, the R wave does not become taller than the S wave is deep until between leads V5-to-V6).
  • Once again — there is nonspecific ST-T wave flattening in multiple leads, but there are no acute changes.
  • PEARL #5 — Once again I am hard pressed to identify the end of the T wave. However, clinically — the important point is that this does not matter! I completely agree with Dr. Smith — that a substantial U wave does appear to be present, but I cannot reliably identify its boundaries. That said — using the rule that one measures the QT (or QU) interval by selecting THAT lead in which you are able to identify the end of the T (or U) wave, and in which the interval looks longest — the PURPLE arrows that are seen in lead V4 tell us that the QT (which is really the QU) interval is markedly prolonged — which leads us to the same conclusion I described in the BOTTOM Line that I wrote for interpretation of the ECG from Patient #1 (See above).

FINAL Learning Point: As per Dr. Smith — The ECG picture of hypokalemia comes in many forms. Most of the time — these different forms end up being some variation of the U wave.


  1. Great post and explanation Dr Sminth and Ken Grauer.Thanks a lot my teachers!!! GOOD statement by Dr Smith "Hyperkalemia is called the "syphilis of ECG findings" because it comes in so many forms, but hypokalemia ECGs are multiform also" I never will not forget it. But Regarding ECG 1 there is a notched R in the lower descending part of the R wave in I, aVL, V5 and V6. Also there are J waves in lead II . What this findings means in this case?

    1. THANKS for the positive feedback! I do NOT think the notches you are referring to in ECG #1 are at the J-point. The “J”-point “joins” the end of the QRS with the beginning of the ST segment — yet I think the notches you point out are part of the QRS complex. I don’t know what they mean. They may reflect “fragmentation” from scarring (ie, due to either cardiomyopathy or prior injury). As I pointed out in My Comment — this patient DOES have LVH — so probably there is some significant underlying cardiac disease … But to better be able to address your excellent question — I’d need to know more about the History AND I’d like to see a repeat ECG after correction of all electrolyte disturbances to see if that nothcing persisted … — :)


DEAR READER: We welcome your Comments! Unfortunately — due to a recent marked increase in SPAM — we have had to restrict commenting to Users with a GOOGLE Account. If you do not yet have a Google account — it should not take long to register. Comments give US feedback on how well Dr. Smith’s ECG Blog is addressing your needs — and they help to clarify concepts of interest to all readers. THANK YOU for your continued support!

Recommended Resources