Sunday, May 27, 2018

Weakness, head trauma, and an abnormal ECG

Written by Pendell Meyers, with edits by Steve Smith

A man in his 50s with history of CAD s/p CABG, CHF, and COPD presented after several falls attributed to acute generalized weakness.  Several had reportedly resulted in head trauma.  There was a normal neurologic exam.

Here is his ECG:

What do you think?

Sinus rhythm at around 60 bpm. There is STD with "down-up" T-waves in V2-V6, or more accurately T-wave inversion followed by large U-waves. The morphology is classic for hypokalemia. The computerized QT interval and QTc were 676ms and 663ms, which is really a measure of the Q-U interval instead of the QT interval. In other words, when the QT interval looks impossibly long, then you should check to see if what you thought was a T-wave is really a U-wave.

This ECG is not consistent with posterior ischemia given the overall morphology and U-waves, however you should remember to include posterior ischemia in your differential of STD that is maximal in V3-V4. More importantly, the patient has generalized weakness and no chest pain, SOB, or clear anginal equivalent.

Intracranial hemorrhage and/or Takotsubo ("stress") cardiomyopathy may present with STD and prolonged QT interval, but this is very different because of the U-waves.

We just presented an abstract discussing our ECG findings in hundreds of ICH cases (traumatic and non traumatic), and prolonged QTc and STD were the findings with the highest correlation to low GCS on arrival and death in this setting. I will be discussing these results in a separate post later.

The potassium returned at 2.6 mEq/L. Magnesium was 1.6 mEq/L. Head CT was negative. Two serial troponins were undetectable. Potassium and magnesium were supplemented, and his dose of diuretic was decreased. He did well.

Here is the patient's post-treatment ECG, with potassium = 3.8 mEq/L several hours prior to the ECG:

Computerized QT and QTc = 464 and 474 ms. There is still a tiny bit of dowsloping STD in V5-V6 with flat/inverted T-wave. This could be baseline, but it would not be possible to tell without a baseline ECG.

Here is the patients prior baseline ECG, two months prior to the presentation ECG:
Almost identical to the post-treatment ECG above.

See more examples to lock in the pattern recognition:

A woman in her 20s with syncope

Are These Wellens' Waves??

Look at These "T"-Waves

Why is this patient weak?

Diabetic Ketoacidosis: is there hypokalemia?


  1. Excellent example of hypokalemia presented by Drs. Meyers & Smith. I think the 3 of us each recognized within no more than 2-3 seconds the high likelihood of hypokalemia as probable cause for the ECG abnormalities on this tracing. That said, I’ll play a bit of “Devil’s Advocate”, while adding the following points. i) The principal ECG finding in this ECG with sinus bradycardia (rate just under 60/minute) is a markedly long QT (or QU) interval. The OTHER findings are ST segment flattening with slight depression in multiple leads (with ST elevation in leads aVR and V1) — AND — incomplete RBBB, with a Q in V1 and that slight ST elevation in this lead. RBBB morphology is suggested by the QR in V1 in association with narrow terminal s waves in leads I and V6 (with the QRS not being wide enough for this to be complete rbbb). ii) The causes of a LONG QT (or QU) INTERVAL — makes up one of my “LISTS” — whereby the interpreter’s life is simplified by remembering to consider the most common causes of a Long QT/QU = 1-Drug effect; 2-Electrolyte abnormality (ie, low K+/low Mg++/low Ca++); and 3-CNS catastrophe (ie, stroke, seizure, coma, tumor, trauma, bleed, etc). Ischemia/Infarction/Conduction defects are other common reasons for seeing a long QT/QU — but these entities will usually be suggested by other associated findings on the ECG. When the ONLY thing wrong with an ECG is a long QT/QU — Think Drugs/Lytes/CNS, or some combination of these as the cause. iii) As per Dr. Meyers — this ECG just “looks” like moderate-to-severe hypokalemia — because of the very long QT/QU with ST flattening and slight depression in multiple leads + large U waves. There is no obvious history for coronary disease, and lab tests will quickly confirm serum K+ & serum Mg++ status. But one should ALSO entertain in the differential diagnosis the possibility of contribution to these ECG abnormalities from ischemia (there is after all, diffuse ST depression with ST elevation in aVR) and/or CNS disturbance (this patient HAS been “falling”, which DID result in head trauma). This does NOT necessarily mean that troponin and CT scanning are immediately (or at all) needed — but rather that the BEST way not to miss a subtle diagnosis (ie, more than just electrolyte imbalance might be wrong with this patient … ) — is to start with a broad differential diagnosis, and to individualize considerations based on patient presentation/full exam/serial follow-up. iv) I favor obtaining at least a few serial ECGs on this patient as he is being treated. Chances are electrolyte imbalance is the primary (and perhaps only) cause of these ECG abnormalities. With K+ (and Mg++ if needed) replacement — I’d expect ST-T wave flattening and ST depression to get significantly less — and one will probably begin to see clear separation of the fused T-U wave into distinct T wave and U wave components — and you’ll have more opportunity to better assess if weakness from low K+ was the cause of the falls, or if there might also be some underlying neurologic problem … THANKS again to Drs. Smith and Meyers for presenting this insightful case!

  2. thank you to the "three ecg-kateers".
    another cool case.
    am awaiting Pendell's blog on the head trauma-ekg cases.

  3. would it be safe to diagnose incomplete rbbb also here

    1. Probably "incomplete" RBBB. Although the ECG does not have great resolution, I believe the QRS duration is < 120 ms


DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.

Recommended Resources