Wednesday, December 18, 2019

Chest pain in high risk patient. Are these Hyperacute T-waves? What is going on here?

A 40-something woman with h/o obesity, hyperlipidemia, smoking, as well as cocaine and alcohol abuse presented with intermittent chest pain for 2-3 weeks that became constant in the last day.  There is also some generalized weakness.   There is some associated SOB, but no nausea and vomiting.

Here is the initial EKG:
What do you think?

There is sinus rhythm, a bit of ST depression (II, aVF, III, V2-V6) and apparently very wide and large T-waves.  However, if one looks closely, one sees a distinct and prominent U-wave in leads II, III, aVF.  Whether a T-wave or a U-wave, the interval (QT or QU) is very long.  I measure it at 700 ms.  The computer measured and QT interval of 450 ms and did not mention U-waves.

Look at V1 and V2:

Now notice that there are 2 tiny humps in these waves.  I point it out in the first complex with these arrows:

Here I draw a blue line down from the clear U-waves in II and III to lead II across the bottom (it is also obvious in lead aVF).    Then, I draw red lines from the same part of the QRSTU complex in lead II across the bottom up to the same part of the QRS complex in leads: (aVR, aVL, aVF), (V1-V3), (V4-V6).
This shows that those enormous T-waves are really TU-waves.
Is there a long QT?  Or only a long QU?
The QT interval appears to be measurable in lead III.

See lead III here:
Where does the T-wave end?  What is the QT interval?

The arrow points to what appears to be the end of the T-wave.
The QT interval is 480 ms, which is slightly long, but not dangerously so.
Thus what appears in other leads to be a long QT is a long QU!

As it turns out, the K was 2.6 mEq/L.

There was a previous ECG for comparison:

Ondansetron was given for nausea and vomiting AFTER this initial EKG.

Here are 2 more ECGs recorded before the K could be adequately repleted:
No significant difference

The computer measured this QT as 452 ms, QTc = 443 ms.
The computer made no mention of U-waves

Again, the "QT" is really a QU-wave.

U-waves are still prominent, and again, I measure the QU interval as very long at 640 ms
Computers are terrible at measuring the QT and QU!
--No ECG was recorded after K was replenished to 4.1 mEq/L. There was no EKG recorded confirming normalization of the QT/QU interval.
--The long QU was not recognized as a long QU.
--The supposed "long QT" was partly attributed to ondansetron, but this was given AFTER the first EKG.  

She ruled out for MI by serial troponins.

Found to be on methadone, and I was worried that this might be methadone effect, so we contacted the patient to have another ECG recorded with a normal K.  Here it is:
Thus, the long QU was probably all due to hypokalemia.

MY Comment by KEN GRAUER, MD (12/18/2019):
We have seen this picture that appears in the initial ECG in this case ( ECG #1 in Figure-1)  more than a few times before on Dr. Smith’s ECG Blog. The “picture” that I refer to is the following:
  • The patient is a set-up” for developing an electrolyte disorder (most notably low serum K+ and/or Mg++) — due to poor nutrition, neglect, recent nausea/vomiting; hypothermia, and all-too-often from alcohol abuse. Generalized “weakness” typically figures prominently in the recent history.
  • The ECG shows diffuse ST-T wave abnormalities — usually in the form of some ST-T wave flattening and/or depression, sometimes with some T wave inversion.
  • The QTc (or QU interval) is obviously prolonged.
  • Prominent U waves are present in a number of leads. These U waves may be readily visible as a distinct “hump” that follows the T wave — or U waves may be more subtle because they so gradually blend in with the preceding T wave. KEY Point — with practice looking for them — you will usually be able to find these U waves deforming, or at least producing some indentation in the preceding T wave. Learn to recognize U waves!

Figure-1: The first 2 ECGs shown in this case (See text).

I'll ADD the following thoughts to comments made above by Dr. Smith:
  • MTake” on the subject of HypoKalemia appears in My Comment at the bottom of the November 18, 2019 Post in Dr. Smith’s ECG Blog. Many of my observations regarding today’s case are THE SAME as I described on November 18 — so I will refer to that post here.
  • As was the case on November 18 — a distinct U wave was only seen in 1 or 2 leads. In most leads, the U wave fused with the terminal portion of the preceding T wave — and could therefore be easy to miss. Therefore — it's important to always assess all 12 leads when trying to determine if U waves might be hidden within the T waves that precede them.
  • My PEARL #1 from the Nov. 18 Post holds for today’s case. In my 30+ year experience of closely monitoring serial electrolyte levels in such patients — serum Mg++ is so very often also reduced when there is hypokalemia. I believe low serum Mg++ produces similar clinical effects (including weaknessand ECG changes as does hypokalemia. Correction of low serum K+ may be problematic until serum Mg++ is also corrected. We are not told what the serum Mg++ level was in this case ...
  • My PEARL #2 from the November 18 Post also holds true: Some patients with hypokalemia develop a “pseudo-P-pulmonale” pattern. Note that the P waves in each of the inferior leads in ECG #1 (especially in lead II) — are tall, peaked and pointed. This P wave pattern was not present in the previous tracing ( = ECG #2), which was obtained when K+ was normal.

OTHER Points that I’d Make:
  • I’m surprised that the computerized interpretation in this case missed the U waves in ECG #1. I suspect the computer in this case used lead III, or possibly lead aVF for its calculation of the QTc. Dr. Smith and I both (independently) used lead III — but selection of lead aVF by the computer (considering the bradycardia) would yield a corrected QTc value close to the 450 msec reported. In my experience — the computer is generally quite good in estimating the QTc — because this is a calculated value, and the computer excels at calculations. That said — errors are possible if the wrong data point for the end of the T wave is selected by the computer. BOTTOM LINE  It is essential to always oversee what the computer says. It should be obvious in ECG #1 that either the QT and/or the QU interval is markedly prolonged. This finding of a long QT (QU) interval should not be overlooked. STEP BACK a few inches, and look at ECG #1 from a short distance away. Then fix in your brain” the picture we see in ECG #1 of a disproportionately long QT (or QU) interval.

PEARL #3  The easiest way to ensure that you never forget to assess the QT interval — is to use a Systematic Approach in your interpretation of each-and-every ECG you encounter. Failure to comply with this recommendation — is a guarantee that you will periodically forget critical ECG findings, such as a long QTc(I illustrate use of a Systematic Approach to ECG Interpretation in My Comment at the bottom of the June 28, 2019 SSmith Post).
  • Once a long QT (or QUinterval is recognized — Remember the simple LIST that I suggested in the November 18 Post for the causes of a long QTc. In a patient with a history like the one in this case — the presumed diagnosis in ECG #1 (until you prove otherwise) should be low K+/Mg++.

FINAL Points:
  • Although the possibility of ischemia should enter into your differential diagnosis for the cause of the ST-T wave flattening/depression that we see in ECG #1 — given this patient’s history, the very long QT (or QU) interval, and our presumed diagnosis of low K+/Mg++ until proven otherwise — I would not have activated the cath lab at this point. My rationale for saying this, is that even IF this patient did have an acute coronary syndrome — IF serum K+/Mg++ are dangerously low — then performing an invasive procedure such as emergency cardiac cath is not advised until you’ve improved serum electrolyte status. Slight delay for this purpose allows: i) A little time to see if chest discomfort lessens, and ECG abnormalities begin to correct as you replace serum electrolytes; andii) A chance to follow the patient closely to better assess IF new-onset cardiac chest pain is truly the principal problem (Sounds like it may not have been in this case ...).
  • Did you note that voltage criteria for LVH are met in both ECGs #1 and #2? Small and narrow q waves are also seen in multiple leads on both tracings. These narrow q waves are unlikely to be clinically significant (they are probably “septal” q waves, which can also be seen in inferior leads when the frontal plane axis is relatively vertical, as it is in ECG #2).
  • Of interest — Comparison of ECG #1 with the previous tracing on this patient obtained when serum electrolytes were normal ( = ECG #2) showed that: i) the QTc was previously normal; ii) Other than perhaps a tiny U wave in lead V2 — U waves were absent in the previous tracing; andiii) ST-T wave flattening and depression was not seen in the previous tracing. These differences between ECG #1 and the prior (baseline) tracing confirm that ECG changes of electrolyte imbalance in ECG #1 are new!

Our THANKS to Dr. Smith for presenting this case!


  1. pointed p wave (II) which can go well with Hypokalemia

    merci - Al

    1. @ Anonymous — Thanks for your comment. As per PEARL #2 above — a “pseudo-P-pulmonale” pattern will not always be seen in lead II with hypokalemia — but sometimes (as in this case) it will, and recognition of this can be helpful — :)

  2. Dr. Amal Attu discusses quite eloquently how the T wave 'gets no respect.' Lol. This example by Dr. Smith with the QT/QU impresses the importance of the T wave, QTi and QU wave. As a pre-hospital provider, our educators have 'shoved' dogmatic STEMI/NSTEMI, ST segment changes, etc, down our throats. There are so many other ECG indicators that relate to serious pathology that have NOTHING to do with the ST segment. EMS/EM education needs to shift it's focus from it's dogmatic approach to ECG interpretation that endorses true interpretation and understanding. Great case study!

    1. @ Ivarre911 — Thanks for your comment! The EASY thing that EMS/EM education could do (so as to follow your suggestion about shifting its focus for ECG interpretation) — is simply to endorse use of a SYSTEMATIC APPROACH. Doing so would result in routine assessment of “Intervals” ( = PR/QRS/QTc) at an early point in the process — which (as per my PEARL #3 above) should lead to immediate recognition of the markedly long QTc (or QT) interval in ECG #1 — and recognition of a long QTc (or QU) should immediately bring to mind the LIST of the 3 Common Causes = i) Certain DRUGS; ii) “LYTES” (ie, low K+/Mg++/Ca++) and/or iii) CNS Catastrophies (as in My Comment in the Nov. 18 Dr. Smith post). Failure to use a Systematic Approach is the reason seemingly obvious findings (like the long QT/QU interval here) get missed. THANKS again for your comment! — :)

  3. When can we use JTC interval vs QTC interval ?? Any particular situations or cases?

    1. @ Jamshid Khan — The JTc interval has been proposed as a way to compensate for QTc prolongation caused by a prolonged QRS. The formula for calculating this interval is: JTc = QTc - QRS duration. Use of the JTc would seem to have benefit in patients with marked QRS prolongation (ie, LBBB). That said — a potential problem with using the JTc, is that there is far less outcome data. I’ve not used this interval clinically. Input from others who have is welcome!

  4. Having grown up in a coastal city, when I see hypokalemia on an ECG tracing, the QRS complexes in the smoothly undulating baseline caused by the Q-U intervals always remind me of buoys bobbing up and down in the waves. In my advanced ECG classes I now refer to this as the "buoys on the bay" sign.

    Thanks, Ken and Steve, for a great discussion.

    1. Thanks so much for your comment Jerry. "Buoys on the bay sign" — I LIKE that! — :)


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