Friday, August 12, 2011

ECG Interpretation Review #27 (ST-T Wave Changes - QT-U Wave - Hypokalemia-Ischemia)

Interpret the ECG below, obtained from a patient with a history of alcohol abuse and atypical chest pain.  Is there ischemia?  — an electrolyte disturbance?

Figure 1 – 12-lead ECG obtained from a patient with atypical chest pain and a history of alcohol abuse. What might the ST-T wave changes be due to?

INTERPRETATION:  There is sinus arrhythmia. The PR and QRS intervals are normal. However — the QT interval is long (clearly more than half the R‑R interval).  The axis is normal (about +65°).  There may be LAA (left atrial abnormality) given the fairly deep negative component to the P wave in lead V1 — but otherwise no sign of chamber enlargement.   
  • QRST Changes:  There is a Q wave in aVL, and a QS in V1,V2.  Transition is slightly delayed.  The most remarkable finding is diffuse ST-T wave flattening/depression with in addition symmetric T inversion in leads V4,V5,V6. The QT interval is markedly prolonged, and there are U waves in multiple leads (best seen in V3, as shown by the RED arrow in Figure 2 below).
Figure 2 – Blowup of leads V2,V3 from Figure 1. U waves are best seen in lead V3 (red arrow).  It is impossible to tell if there is QT or Q-“U” prolongation.

CLINICAL Impression:  The diffuse ST-T wave changes seen in Figure 1 may be due to any of the common causes of ST depression.  To facilitate recall — these common causes are listed in Table 1 and include ischemia; “strain” from LVH; electrolyte disturbance (hypokalemia; hypomagnesemia); digoxin effect; and/or tachycardia (See also ECG Blog #26). Given the history of chest pain — one has to consider ischemia that may be acute (difficult to know IF the T wave inversion in Figure 1 is a new finding without availability of a prior ECG for comparison) 

Table 1 – List of the most common causes of ST segment depression.

The markedly long QT (or “Q‑U”) interval in Figure 1 should suggest one or more of the common causes of QT prolongation.  To facilitate recall — these common causes are listed in Table 2 and include “Drugs – Lytes – and CNS catastrophes” (See also ECG Blog #4).  The ECG signs and history of alcohol abuse in this case should place hypokalemia (and/or hypomagnesemia) high on your list.  Electrolyte disturbance is further supported as a contributing factor to the ST-T wave changes in this case by the finding of U waves in multiple leads.

Table 2 – List of the common causes of QT prolongation.

Clinical Correlation / USE of the “LISTS”:  Clinical correlation is needed to determine the likely cause(s) of ST-T wave abnormalities in Figure 1.  At the least — We suspect ischemia and hypokalemia (and/or hypomagnesemia).  Serum electrolytes, serial troponins and follow-up ECGs/comparison with prior tracings should be revealing, although at times it may not be possible to precisely determine each contributing factor …
  • The ECG in this Blog post provides an excellent example of how we use our “Lists” to assist with ECG interpretation.  We intentionally limit both the number and length of each of our 6 “Lists” to facilitate recall.  On recognizing a particular ECG finding (such as QT prolongation or ST depression) — recall of the entities on the relevant list help us to expediently hone in on the differential diagnosis (See Tables 1 and 2 above plus Table 1 in Blog #23 and Table 1 in Blog #26).

ECG Changes of HypoKalemia:  We conclude this ECG post by brief review of the ECG changes of Hypokalemia.  In contrast to hyperkalemia — the ECG (in our experience) is not an overly reliable tool for assessing for assessing mild-to-moderate hypokalemia, as both sensitivity and specificity of ECG findings for less-than-severe hypokalemia are relatively low.  That said — the changes that one looks for are sequentially illustrated in Figure 3.
  • Ais a normal ST-T wave.    
  • Bshows flattening of the T wave, which typically is the earliest change.
  • C and DIn association with ST-T wave flattening (and sometimes with slight ST depression) — a U wave develops.  A "pseudo-P-pulmonale" pattern (with P wave peaking) may be seen.     
  • E and FST depression is more noticeable and the U wave increases in amplitude (arrow), until ultimately the U wave overtakes the T wave. At this point, distinguishing between T wave and U wave may be almost impossible (ie, there may be Q-U" rather than “Q-T prolongation as in F).
Figure 3 – Sequential development of ST-T wave changes of hypokalemia. Note increasing U wave amplitude.

As emphasized above — U waves per se are not specific for hypokalemia.  U waves may also be found in patients with LVH and/or bradycardia, or occasionally as a normal variant.  However, the setting and ECG findings in this case (given the history of alcohol use with diffuse ST-T wave flattening and relatively large amplitude U waves in multiple leads) strongly suggests the possibility of electrolyte disturbance.
  • PEARL #1: When the clinical setting is "right" (ie, potentially predisposing to hypokalemia) — by the time U waves become so large that they are bigger than (or almost as big as)  T waves — specificity of the ECG for significant hypokalemia is greatly increased.
  • PEARL #2:  Hypomagnesemia produces virtually identical ECG changes as hypokalemia.  Low body magnesium is often encountered in association with other electrolyte abnormalities (ie, low sodium, potassium, calcium or phosphorus); acute MI; cardiac arrest; digoxin/diuretic use; alcohol use and abuse; renal impairment.
- For more information  GO TO:



  1. what is the correct manifestation of hypokalemia is it ST segment depression or prolonged PR interval?

  2. Hello Darwin. For those patients who "read the textbook" when they drop their serum K+ values - the serial ECG changes shown in Figure 3 are what is expected. The ST segment becomes flat, may be depressed - and a U wave develops. To my knowledge - the PR interval does not prolong. I hope that answers your question!

  3. There is a "Mitral P" in lead II?

  4. Thank you Stillcho for your comment. In this particular example — P wave notching is seen in multiple leads — so assuming correction of electrolyte problems doesn't resolve this diffuse P wave abnormality — it would see more likely that an intra-atrial conduction defect (rather than left atrial enlargement) is responsible). I discuss ECG diagnosis of atrial abnormality more in Blog #75 — GO TO — — Hope that answers your question!

  5. How do we measure the QT interval when there are prominent U waves that runs into the U wave ?

    1. Excellent question Mostafa — that I would answer by saying that precise determination of the "QTc" when you have prominent U waves may just not be possible (since the U waves may hide the point where the T wave ends ... ). That said — precise measurement should really not be needed. For example, in the case here in which the QTc is clearly lengthened (most probably due to severe electrolyte disturbance with low K+/low Mg++ in this patient with a history of alcohol abuse) — Management really will not change depending on any slight difference that there might be due to not being able to clearly see the end of the T wave ... Treatment of the underlying condition (electrolyte repletion; nutrition; hydration; etc.) takes priority — and once the patient's condition improves: i) You WILL be able to accurately measure the QTc, because U waves will decrease in amplitude (and may even resolve); and ii) The QTc will approach more normal values, so concern about a long QTc will probably cease to be an issue ...

  6. When should one consider the QS complex as exactly QS, in other words which criteria are for the QS complex, because I see in V1-V2 tiny r waves.Thanks for the answer, Sir.

    1. GOOD question! — And I fully admit that sometimes it is difficult for me to decide IF there is or is not a tiny R wave. For example — in lead aVR of Figure-1, I’m sure we BOTH agree that there is a definite wide Q wave (of course, not a “QS” — because there is an R wave after the wide Q). But the question is whether that “tiny notch” at the very onset of the QRS in leads V1 and V2 of Figure-2 is truly an “r” wave — or artifact — or in the case of lead V1, perhaps some T wave of the P wave depression, with rise to the baseline before the QRS begins … And then note that in leads like V4,5,6 (especially V5) — there is some artifactual variation in the baseline. So I think it is difficult to know for certain if there truly is or is not an initial r wave in leads V1 and V2 of Figure-1 … THANKS again for your question! — :)

  7. How does digoxin cause ST depression? thank you sir

    1. The ECG features of digoxin effect are seen with therapeutic doses of digoxin and are due to: Shortening of the atrial and ventricular refractory periods — producing a short QT interval with secondary repolarisation abnormalities affecting the ST segments, T waves and U waves — :)