Sunday, November 15, 2020

A Covid patient with cough and Fever. Why does the ST-T wave look so abnormal?

I was reading through a stack of ECGs and saw this one.  

What is going on here?

At first glance, it looks like a low atrial rhythm, with a negative P-wave in inferior leads.  

There appear to be inferior QS-waves and and intraventricular conduction delay (computer measure QRS at 120 ms).

There appear to be very strange down up T-waves.

On closer inspection, the P-wave in V1 appears biphasic, which should not happen in a low atrial rhythm.

Then we see that there is another P-wave inbetween, superimposed on the T-wave, in V1.

Then we see that in lead II, the negative P-waves have another negative P-wave between the QRS and the T-wave, which accounts for the abnormal appearing ST-T.

So this is an extremely slow atrial flutter with 2:1 conduction.  Atrial rate 146, ventricular rate 73.

It turns out that this patient has cardiac amyloidosis.  I suspect that the amyloid slows the conduction of the atrial flutter.

I later found out that the patient complained of cough and fever, and was diagnosed with Covid.

It turned out that he had a history of slow atrial flutter.  His cough and fever was from mild Covid pneumonia.  All troponins were negative.

Diagnosis: Extremely slow Atrial flutter


MY Comment by KEN GRAUER, MD (11/15/2020):


Some of the most interesting ECG cases I have seen have been picked up “reading through a stack of ECGs” — in the same manner that Dr. Smith picked up today’s case.

  • In my previous role of overseeing the ECG interpretation of other providers — the Questions I was always faced with when confronted with a large “stack” of tracings, but without clinical information were: i) What was going on with the patient (ie, WHY was this ECG obtained?) — andii) DID the provider recognize the abnormal findings?

The ECG in today’s case is one in which Dr. Smith immediately knew he needed to find out what was going on with this patient!

  • It turned out the patient had cardiac amyloidosis. The presenting complaint was cough and fever from mild Covid pneumonia. There was no chest pain — and all troponins were negative. The chart revealed that the arrhythmia was not new.

So, even though we know “the Answers” (Dr. Smith interpreted this ECG above) — I think it worthwhile to take another LOOK at the ECG in today’s case (Figure-1).


  • What is the most time-efficient way to diagnose the rhythm and the QRST abnormalities that we see in this case?

Figure-1: The initial ECG in this case (See text).

ANSWERS: Use of a systematic approach to ECG interpretation is the most time-efficient way to interpret a complex rhythm and 12-lead tracing, such as that seen in Figure-1.

  • Regardless of whatever “system” you use in your interpretation — Look FIRST at the cardiac rhythm! (For those interested  CLICK HERE for review of the Systematic Approach I favor).
  • PEARL #1: You can tell in an instant that the rhythm in ECG #1 is not sinus — because the P wave is not upright in lead II (RED arrow in lead II of Figure-2). Unless there is dextrocardia or lead misplacement — the P wave in lead II should always be upright IF there is sinus rhythm!

PEARL #2: The PR interval of this negative P wave in lead II looks long (ie, it is ~0.24 second in duration). While possible for a low atrial rhythm to have a long PR interval — this is unusual.

  • The PEARL is that when you see an unusual-looking (ie, not sinus) P wave in lead II with a longer-than-expected PR interval — Consider that there may be extra P waves!
  • Use Calipers! It literally takes me no more than 3-5 seconds to verify with calipers that there are 2 equally-spaced negative deflections in the long lead II rhythm strip in Figure-2 (slanted BLUE lines in the long-lead rhythm strip).
  • It’s easy to confirm the 2:1 AV relationship in ECG #1 — because 2:1 atrial activity is also seen in multiple other leads (thin, slanted PURPLE lines that I’ve drawn in several of these other leads showing 2:1 atrial activity).
  • NOTE: In cases in which it is not as obvious as it is in Figure-2 that there are 2 atrial deflections within each R-R interval — I simply set my calipers at precisely HALF the R-R interval — and then look for extra atrial deflections in any of the 12-leads that allow me to “walk out” 2:1 atrial activity.

Figure-2: I’ve labeled atrial activity in ECG #1 (See text).

PEARL #3: My systematic approach to rhythm interpretation entails assessment of 5 Parameters that are EASY to recall by use of this memory aid = “Watch your Ps & Qs, and the 3 Rs”. This memory aid brings to mind:

  • Looking for P waves (which we’ve already done in Figure-2).
  • Is the QRS wide or narrow? (ie, Is the rhythm supraventricular?).
  • Rate (looking for both the atrial and ventricular rate, if these are different).
  • Is the rhythm Regular (looking at both the atrial and ventricular rhythms).
  • And, if P waves are present — Are P waves Related to neighboring QRS complexes?

KEY: It does not matter in sequence you assess the above 5 parameters — and I often change the sequence I use, depending on which of these features are easiest to identify in the tracing I am looking at. But the “beauty” of the Ps, Qs & 3Rs” Memory Aid — is that IF you regularly use it, you’ll never forget to assess each of these 5 essential parameters. As a result, even when you may not be certain of the rhythm diagnosis — you will systematically narrow down your diagnostic possibilities in the most time-efficient manner. Applying the Ps, Qs & 3Rs Approach to the rhythm in Figure-2:

  • P waves are negative in lead II — so the rhythm is not sinus.
  • There is 2:1 atrial activity — with an atrial Rate just under 150/minute (and a ventricular Rate just under 75/minute).
  • Both the atrial and ventricular rhythms are Regular.
  • The QRS complex is wide (ie, about 0.12 second) — but despite QRS widening, the rhythm in ECG #1 is supraventricular because there is a constant (albeit prolonged) PR interval before each QRS complex. This tells us that every-other P wave is conducting! (ie, every-other P wave in Figure-2 is Related to the QRS by a fixed PR interval).
  • Putting It All Together: There is 2:1 AV conduction in ECG #1, with an atrial rate ~150/minute. The differential diagnosis is between a slow AFlutter vs ATach with 2:1 block. We are not told if this patient is on any long-term antiarrhythmic agents that might be slowing the rate of AFlutter. Or perhaps, as postulated by Dr. Smith — amyloid infiltration of atrial tissue might be slowing conduction of flutter activityBOTTOM Line: There is a regular atrial rhythm at ~150/minute. Emergency management will not be altered not knowing if this rhythm represents 2:1 AFlutter vs 2:1 ATach.

The REST of the 12-Lead:

Now that we have assessed the rhythm — we can interpret the rest of the 12-lead.

  • As already noted — the QRS complex of this supraventricular rhythm is wide (ie, ~0.12 second) — so there is some type of conduction defect.
  • The QTc looks long (ie, I estimate ~480 msec).
  • There is marked left axis deviation in the frontal leads — which is the result of deep and wide inferior lead QS complexes.
  • Assessment for LVH is always challenging in the presence of conduction defects that alter the sequence of ventricular depolarization. That said — the very deep S waves in leads V3 and V4 (>25 mm in lead V3) suggest there probably is LVH.
  • In addition to those very deep and wide QS complexes in the inferior leads — there is extremely poor R wave progression in the chest leads. More than this, there is loss of R wave amplitude from lead V3 onward — with no more than a tiny r wave remaining in lateral leads V5 and V6. This is extremely unusual!
  • Assessment of ST-T wave changes is often problematic when there is 2:1 AV conduction. That said, after “mentally subtracting” for 2:1 atrial deflections (that superimpose on various parts of the cardiac cycle) — I believe we are left with somewhat “voluminous” and peaked T waves in multiple leads that I suspect do not reflect acute ST changes.

Putting It All Together for ECG #1:

  • There is 2:1 AV conduction with an atrial rate ~150/minute (could be either slow AFlutter vs ATach with 2:1 block).
  • QRS morphology is bizarre. The QRS complex is wide — but does not resemble any known form of bundle branch block. While possible that this tracing could represent extensive prior infarction in multiple lead areas — this nonspecific IVCD (IntraVentricular Conduction Defect) pattern that we see in ECG #1 should raise concern for some unusual type of cardiomyopathy or other form of significant underlying structural heart disease.
  • I suspect LVH (those very deep S waves in V3 and V4).
  • My hunch is that there are no acute ST-T wave changes.
  • Clinical correlation is essential for understanding what the above combination of ECG findings might mean.

ECG Findings with Cardiac Amyloidosis:

As noted above — Dr. Smith found out that the patient in today’s case had cardiac amyloidosis. It’s helpful to remember that "amyloidosis" does not refer to a single diagnostic entity — but rather to a collection of disorders in which there is a protein-based infiltration in cardiac tissues.

  • There are a variety of subtypes of cardiac amyloidosis — depending on the specific nature of the protein being deposited. Extending beyond the scope of this emergency medicine ECG Blog — is that clinical and ECG manifestations of the various subtypes of cardiac amyloidosis may vary depending on the specific type of this deposited protein.

PEARL #4: Without getting into specific amyloid subtypes — I thought it might be useful to mention some of the ECG manifestations described in Cardiac Amyloidosis, realizing that not all patients with “amyloidosis” will manifest each of these ECG findings. Among the ECG findings you may see in a patient with cardiac amyloidosis are the following (Cheng et al and O’Donnell et al):

  • Low voltage in limb leads.
  • Atrial arrhythmias (especially AFib or AFlutter).
  • AV block.
  • Prolonged QTc interval.
  • Bundle branch block (or other form of conduction defect).
  • Pseudo-infarction patterns.

Of the above ECG findings — all but the 1st bullet are seen in ECG #1!
  • PEARL #5: Among all of the cardiomyopathies that we commonly see (ie, dilated = congestive, ischemic, Takotsubo, hypertrophic) — it's not all that often that we see one of the more unusual forms from entities such as from muscular dystrophy, infiltrative (ie, sarcoid, amyloid), other congenital disorders, etc. Seeing bizarre QRS morphology not explained by the common conduction disorders of LBBB, RBBB, IVCD — especially when there is a pseudo-infarction pattern — has been one of the most helpful clues for me to suspect one of these less common forms of cardiomyopathy.
  • For example — I reviewed clinical and ECG features of Duchenne and other Muscular Dystrophies HERE — with another example of Duchenne Muscular Dystrophy published by Drs. Walsh & Smith in our January 4, 2017 post — in which there are Q waves suggesting a pseudo-infarction pattern.


  1. Thank you for the excellent case and comment.
    Is the atrial rate too low (<200/min) to diagnose an atrial flutter?
    Best regards

    Aldo Cannata, MD
    Milan, Italy

  2. Ciao Aldo! Bella domanda! (Good question). Normally, the atrial rate of untreated flutter in adults is ~300/minute (usual range ~250-350/minute). A number of factors can potentially result in atrial rates that are slow than this (ie, use of rate-slowing antiarrhythmic medications). The patient in today’s case has an infiltrative cardiomyopathy due to amyloidosis. As a result, Dr. Smith suggested that a potential reason for a slower rate of atrial flutter might be amyloid infiltration in the atria, which might result in slowed atrial conduction (and the slower rate of flutter). That said — we are not told how the diagnosis of AFlutter was made. If this was not made by EP study — then I would think an equally reasonable diagnosis might be ATach with 2:1 AV block, because the atrial rate is slower-than-expected for untreated flutter and the baseline between P waves seems flat (rather than sawtooth). We are also not told if this patient was on some type of antiarrhythmic treatment. That said — there is OVERLAP between ECG diagnosis of AFlutter vs ATach — and sometimes the only way to distinguish between the two is EP study. The final point I tried to emphasize in My Comment above, is that from a practical (ie, management) standpoint — it simply doesn’t matter in the ED whether the 2:1 AV conduction with atrial rate ~150/minute that we see here is AFlutter or ATach.

  3. you guys never cease to amaze me. extraordinary case, and analyses. thanks once again.

    i have a (perhaps silly) question: is it possible that the ecg represents second degree heart block, either Mobitz 1 or 2?

    again, thanks guys.


    1. Excellent question Tom! — and it is a question that is often asked! The atrial rate in today’s case is ~150/minute — and although I’ve never seen this in writing — I generally don’t talk about “2nd-Degree AV Block” (of either the Mobitz I or Mobitz II type) when there is an underlying atrial rate this fast. As per My Comment above (as well as in my answer to Aldo Cannata’s question just above yours) — I would not be sure from just this ECG alone if we were dealing with ATach with 2:1 block - OR - AFlutter with 2:1 AV conduction.

      So we get into some SEMANTICS here (and terms may vary depending on whose opinion/approach you favor). I generally do NOT use the term “block” when discussing AFlutter — because the reason for 2:1 AV conduction (especially when the atrial rate is >250/minute) is PHYSIOLOGIC (ie, we are happy that we don’t conduct 1:1 — which I think of as the result of the beneficial “built-in” AV node refractory period that serves the useful purpose not limiting conduction to the ventricles when the atrial rate is very fast). That said — we will on occasion see AFlutter at faster rates (ie, close to the usual untreated flutter rate ~300/minute) in which there is Wenckebach conduction out of the AV node. The “primary” rhythm disturbance here is NOT “AV block” — but rather AFlutter, in which Wenckebach-type conduction occurs either as a response to medication or as a result of the effect of the very fast atrial rate on refractory periods.

      In contrast — with ATach — when you see 2:1 or 3:2 AV conduction, this is more likely the result of “block” (ie, especially if the patient is taking Digoxin) — although it is possible with ATach to see Wenckebach conduction simply due to the fast rate.

      In general — I favor reserving use of the term “2nd-degree AV block” (with 2:1 or 3:2 or whatever conduction) for when the AV block is the PRIMARY process — in which case the atrial rate should not be “that fast” (ie, usually ≤110/minute or so …).

      Hopefully the above makes sense. THANKS again for your question — :)


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.