Tuesday, March 19, 2019

A Pathognomonic ECG. What is it?

This patient presented with weakness, decreased urine output, and vomiting:
What is the ECG diagnosis?

There is a very long QT (computer says the QTc is 525 ms) due to a long ST segment.  This is pathognomonic for hypocalcemia.  The ionized Ca was 2.34 mg/dL (normal is 4.4-5.2)

The Cr was 12.1 indicating (new onset) of renal failure.

Calcium was given without much change.

The next AM the non-ionized Ca was 5.7 mg/dL (normal: 8.6-12.0).

Here was a repeat ECG:
QTc 523.  Long ST segment remains.

Although the QT is very long, long QT due to hypocalcemia is rarely associated with Torsades de Pointes.

Comment by KEN GRAUER, MD (3/19/2019):
There are a number of ECG patterns that should immediately suggest a clinical diagnosis. This is one of them! The value of recognizing this particular ECG pattern — is that it may expedite your clinical diagnosis even before laboratory results return.
  • To my reading — both of the ECGs in this case looked similar. I chose the 1st ECG — and for clarity, I’ve put it together with a user-friendly method I devised many years ago to rapidly estimate the QTc (Figure-1):
Figure-1: The initial ECG in this case — and a rapid method for estimating the QTc (See text).
COMMENT: I wanted to discuss a number of interesting aspects regarding the ECG in Figure-1. As per Dr. Smith — our attention is immediately captured by the very long QTc interval in ECG #1. Measurement of intervals is one of the tasks that the computerized ECG interpretation is usually very accurate with. The computer calculated a QTc 525 ms for the ECG #1.
  • I like the eyeball method” to tell at a glance if the QTc is likely to be prolonged. Assuming the heart rate is not too rapid (this method works less well with heart rates >90-100/minute) — one may suspect that the QTc will be long if the longest QT interval that you can clearly see on the tracing is more than half the R-R interval.
  • To quickly estimate a numerical value for the QTc — I developed a Correction Factor that has been surprisingly accurate for me in assessing too-numerous-to-count QTc values that I’ve estimated over the past 3+ decades.  As per the text under the ECG in Figure-1 — you only need to remember 3 values (ie, 1.1 for a rate ~75/min; 1.2 for ~85/min; and 1.3 for ~100/minute). With a little practice using this method — you can estimate the QTc within seconds.
  • Applying my method to the case at hand — the rhythm in ECG #1 is regular, with an R-R interval just under large boxes. Thus, the heart rate is just a bit over 75/minute (ie, 300÷4). I selected lead V3 as one of the leads where we can clearly define the onset and offset of the QT interval. I measure the QT in this lead to be ~2.4 large boxes = 480 msec. Using a correction factor of 1.1 (since the heart rate ~75/minute) — I estimate the QTc = 480 + [480 X .1 = 48) = 480 + 48 ~528 msec. For speed and ease of calculation — I usually round off values (it’s all an estimate anyway! ) — but I’ve enjoyed being able to get very close to computer-calculated QTc values by this simple correction factor method.
When the QTc is Prolonged:  Assuming there is no bundle branch block, ischemia or infarction — I suggest remembering the following short LIST whenever you recognize QTc Prolongation. Think ofiDrugs (many drugs prolong the QT interval — and combinations of drugs may result in marked prolongation)iiLytes” (ie, Think of low K+ — low Mg++  and/or — low Ca++)andiiiCNS Catastrophe (ie, stroke, bleed, coma, seizure, trauma, brain tumor).
  • Clinical correlation will typically suggest which one or more of these 3 causes of a prolonged QTc is operative for the case at hand. The patient in the case presented here had new-onset renal failure — so, assuming normal mentation and no potentially QT-altering drugs — electrolyte disturbance should be strongly suspected.
ECG Findings of HypoCalcemia:  Hypocalcemia generally prolongs the QT interval. It is therefore one of the entities on our short LIST to immediately think of whenever you recognize QT prolongation.
  • PEARL #1: In theory — pure hypocalcemia does not affect the ST segment! As a result — the characteristic ECG picture of hypocalcemia is that of a flat and prolonged ST segment, at the end of which occurs a surprisingly normal-looking T wave.
  • PEARL #2: Hypocalcemia and hyperkalemia may occur together in patients with renal failure. Clinically — this combined electrolyte disorder may occasionally be suspected by the ECG finding of peaked T waves with narrow base that occur at the end of a long and flat ST segment that produces a prolonged QT interval.
Final THOUGHT: We were not told what the serum K+ value was in this case. Given the very long QT interval in ECG #1 the remarkably flat ST segment in most leads the peaked and relatively narrow base for many T waves that look taller-than-they-should-be in leads II, III, aVF, and V2-V4 — I suspect combined Hypocalcemia and Hyperkalemia in this case.

Monday, March 18, 2019

IVCD, Saddleback STE in III, with reciprocal STD in aVL: Is it pseudoOMI or OMI? Echo with Speckle Tracking gives the answer.

A 77 y.o. woman with a history of hypertension and congestive heart failure presented for acute onset chest pain and shortness of breath. She stated that she woke up in the morning with a central chest pressure with associated shortness of breath. She had been feeling well the day before.

She had no h/o CAD but had a history of "unspecified cardiomyopathy"

No old EKGs or angiogram were available.

This ECG was texted to me with no information:
What do you think?

There is ST elevation in inferior leads, with reciprocal ST depression in aVL, so one must strongly suspect acute inferior MI.

However, 3 features made me think that it could be a false positive (pseudoOMI):
1) There is LVH
2) There is an intraventricular conduction defect, with QRS duration of 125 ms
3) There is an RSR' (saddleback) in lead III

My response was that the diagnosis of inferior OMI is probable but not certain.

I was looking at it on my phone when I gave that opinion; looking at the full size version I am more convinced of OMI than I was at the time.

Case continued:

The emergency physicians were also not certain.

So they did a bedside point of care cardiac ultrasound,

Here is the parasternal short axis:
The area on the lower left of the image has poor contractility

Apical 4-chamber

This shows, to my interpretation, an apical wall motion abnormality, and of the distal septum

This one was done with Speckle Tracking Strain Echocardiography.

The lavender line is associated with the lavender sector where it says "inferior".
That sector is clearly not moving as well as the upper right.
But you don't have to depend on your Gestalt.
You can see the graph.
The deeper the graph (deepest is yellow, which is anterior), the better the contraction.
The lavender sector contracts less than 10 on the scale, so there is indeed an inferior wall motion abnormality.

You might wonder why the inferior wall looks so medial.
The location depends on the operator who tells the machine what wall is what by placing dot markers on the endocardium.  So if the inferior wall is not accurately marked, it will identify the wrong wall.

Here are more cases with Speckle Tracking

Here is an article we wrote on the topic:

Rowland-Fischer A. Smith SW.  Laudenbach A.  Reardon R.  Diagnosis of acute coronary occlusion in patients with Non–STEMI by point-of-care echocardiography with speckle tracking. American Journal of Emergency Medicine 34(9):1914e3-1914e6; Sept 2016.

Here is the case we describe in the report: Ultrasound Before ECG for Chest pain? Whoever gets there first. Use of Speckle Tracking.

Case Continued:

The initial troponin I was 0.012 ng/mL (URL = 0.030 ng/mL)

The Cath lab was activated.

A 2nd troponin returned at 0.047 ng/mL (elevated)

The resident wrote that the cath showed only an OM occlusion.  

She asked 2 very astute questions: 
"Could an OM occlusion cause an inferior MI?"
and related:
"Could the EKG abnormalities be baseline and he really had an occlusion that did not manifest on the EKG?"

My answer:

In a left dominant system, an OM could supply the inferior wall, but that would be unlikely.  In this case, due to the EKG abnormalities, it is likely.  But the proof would be in the subsequent ECGs: did they evolve?  If the EKG abnormalities are a result of the ischemia, the ECG will always evolve.  The ST segments will resolve, or there will be T-wave inversion, or both.

Did the EKG evolve?

First, the full angiogram:

LM normal
LAD normal
LCx large codominant, normal appearing until point of occlusion in mid-vessel
    --So this OM does indeed supply the inferior wall.
RCA supplies PDA only, normal.

PTCA of LCx OM2 branch performed. Distal vessel difficult to visualize.  Flow eventually restored after multiple passes with thrombectomy aspiration catheter.

Here is the post PCI ECG:
The ST elevation is resolved (and the reciprocal ST depression)
Therefore the ECG abnormalities were definitely a result of the ischemia.

The post PCI troponin was 22.4 ng/mL

Formal echo later:

Moderate to severe decrease in left ventricular systolic function with an estimated EF of 30%.
Regional wall motion abnormality--apical septal, mid anteroseptal, mid to apical anterior, and apical lateral hypokinesis.
Regional wall motion abnormality--basal to mid inferior and inferolateral hypokinesis.

Comment by KEN GRAUER, MD (3/18/2019):
There are a number of reasons why I really like this case: iAs per the title chosen by Dr. Smith — it’s a great example of how combined use of ECG + Echo with Speckle Tracking makes the diagnosis; iiIt illustrates some important points about ECG assessment when the initial ECG shows an unusual form of intraventricular conduction defect (IVCD);  andiiiIt shows how the follow-up (ie, 2nd) ECG confirms the diagnosis.
  • For clarity — I’ve put both tracings in this case together, and have labeled some key findings (Figure-1):
Figure-1: The 2 ECGs shown in this case (See text).
COMMENT: As per Dr. Smith, it’s important to emphasize that in this older patient with underlying heart disease and new-onset chest pain — acute OMI must be assumed until proven otherwise. That said — I was initially less certain of this diagnosis from inspection of ECG #1 alone, without the benefit of a prior tracing for comparison ( = My Opinion).
  • In favor of acute OMI until proven otherwise from ECG #1 — all 3 inferior leads show subtle-but-real ST elevation — and, there is mirror-image opposite reciprocal ST depression in lead aVL. I’ve added horizontal RED and BLUE lines to these leads to facilitate recognition.
  • My reservation about whether or not these changes were acute stemmed from how atypical the IVCD in ECG #1 was. This patient clearly has underlying structural heart disease — and that can affect ST-T wave appearance appearance in a way that is hard to determine without seeing a baseline ECG. The QRS complex is wide. Although the QRS does not look overly wide in a number of leads — one takes the widest QRS that you are clearly able to see for measurement of QRS duration, and I measure 0.12 second for QRS width in leads V2 and V3. Therefore, there is a conduction defect. However, the triphasic ( = rsR’) complex in lead I (quadriphasic in lead aVL) is distinctly atypical for LBBB — as is the narrow R wave in lead V6. Typical LBBB also does not produce a 6mm R wave as early as lead V2, as we see here in ECG #1. This defines the conduction defect in ECG #1 as a nonspecific IVCD — because it does not conform to either typical RBBB or LBBB. Assessment of ST-T wave changes, as well as chamber enlargement is often more difficult in the setting of IVCD. (NOTE: I did not say “impossible” — but I did say that it’s often more difficult to assess acute ST-T wave changes in the setting of an unusual IVCD, such as the one we see here).
  • As per Dr. Smith — there is also LVH in ECG #1, and the presence of LVH can be a confounder for recognizing acute MI on ECG. I will emphasize that ECG criteria for diagnosis of LVH are different, and often much more difficult to ascertain in the presence of any conduction defect (RBBB, LBBB or IVCD). This is because the sequence of both ventricular depolarization and repolarization is altered when there is a conduction defect. As a result — recognition of ST-T wave changes of LV “strain” is less reliable — and, the usual numerical criteria established for ECG diagnosis of LVH are different and not well established. That said — the finding of very deep anterior S waves (ie, ≥25-30mm in V1, V2 or V3) has been correlated with high statistical likelihood of LVH. This criterion is met by the 32mm S wave in lead V2 of ECG #1.
  • Finding a prior ECG on this patient would have been invaluable for proving that the limb lead changes in ECG #1 were acute and not longstanding. Unfortunately, a prior ECG on this patient was not available. This highlights the value of Echo in the ED — with demonstration of a localized wall motion abnormality confirming the need for immediate cath (Our THANKS to Dr. Smith for these beautiful Speckle Tracking Echo videos! ).
COMMENT on ECG #2: There are a number of important findings to highlight on ECG #2, performed after successful PCI:
  • As per Dr. Smith — the inferior lead ST elevation and reciprocal ST depression in lead aVL has resolved in ECG #2.
  • NOTE: QRS morphology in leads I and aVL in ECG #2 is even more unusual than it was in ECG #1 — in that both leads I and aVL now show quadriphasic (rsR’s’) complexes, with a near isoelectric deflection. Extra-notched QRS complexes (ie, fragmenting) — is also seen in leads III, aVF; and in the S waves of leads V5 and V6. Such fragmentation often indicates “scar” — either from underlying cardiomyopathy, prior infarction or both.
  • It should be appreciated that R wave progression is markedly different in ECG #2, compared to ECG #1. That is, transition (where the R wave becomes taller than the S wave is deepoccurred between leads V3-to-V5 in ECG #1 — but it never occurs in ECG #2 (ie, the S wave remains predominant through to lead V6 in ECG #2). I would generally not expect this marked change in R wave progression to be the result of reperfusion — which raises the question of whether chest lead electrodes were placed in the same position when recording both tracings? Awareness of this difference in R wave progression is essential when attempting to compare serial ST-T wave changes!
  • That said, despite this marked difference in R wave progression between these 2 tracings — there should be little doubt that there has been evolution of the ST-T wave. Note the much more modest T wave amplitude (within the dotted BLUE rectangles — with tiny, almost flat T waves in the other leads) in ECG #1. In contrast — the T waves in all chest leads of ECG #2 are clearly more peaked and manifest a narrower base (within the RED rectangles). I interpret this difference as showing evolutionary reperfusion ST-T wave changes — that add further support to the acuity of the findings we saw in ECG #1.
Our THANKS to Dr. Smith for providing this superb teaching case!

Friday, March 15, 2019

A 50-something with chest pain presents to a Non-PCI capable facility

An avid reader sent me this case.  He learned to read subtle ECGs here.


A 50-something presented with chest pain.  A triage ECG was recorded and shown to the ED physician:
What do you think?
The QTc-B was 427 ms

This was texted to me with no information, and here is my reply:

"I think it looks like an LAD occlusion."  

Why did I say this?  
--There is minimal STE in V2, but in the presence of a tiny QRS.
--The ST segment in V2 is convex upward, or at best straight
--The QRS starts with a Q-wave 
--There is a fragmented QRS (down-up-down-up-down-up)
--There is ST elevation in V3 with a rather straight ST segment

If you ignore the convexity, and use the 4-variable formula, we get:

STE60V3 = 3.0
QTc = 427
RAV4 = 12.5
QRSV2 = 2

Value = 21.73.  This is a very high value and strongly supports LAD occlusion.  The best cutoff was 18.2 with approximately 85-90% sensitivity and specificity.

The reader then told me the details:

"The ECG was shown to my partner, who thought it was normal.  He signed it as "No STEMI" (which is technically true)."

"I then signed up for the patient on the board without having seen the ECG."

"Then I saw the ECG and thought to myself "Oh, my God!""

"I too thought it looked like an LAD occlusion."

"I then saw the patient clutching his chest."

"I gave aspirin and heparin bolus, and called for a helicopter to transfer to the PCI facility."

"I was afraid they would not even accept the patient."

"The helicopter arrived and we recorded another ECG 25 minutes after the first, just as he was being loaded onto the stretcher:
Now it is an obvious STEMI (see leads V3 and V4)

Then I gave 50 mg (full-dose) TNK-tPA (tenecteplase) 


The cardiologist at the receiving institution reported that the patient was pain free on arrival.  [No EKG data was provided. This is important because if the ECG does not show evidence of reperfusion, immediate angio +/- PCI is indicated regardless of pain status]

"He said the patient was nearly pain free on arrival to their ER because of the TNK-tPA and thus, since he got TNK, protocol is to wait 3-24 hours because of the risk of bleeding.  If he were still having significant chest pain, they would have taken him immediately."

Because he was pain free, they waited 14 hours (until the next AM) to take him to CATH:  


100% occlusion of the old stent in the LAD (with a 95% stenosis proximal to the stent), 90% stenosis of RCA, and 70% stenotic circumflex, so he stented all three.

Whether there was TIMI 3 flow was not stated.

His troponin I peaked at 76 ng/mL.

Great work by the reader!!!

Comment by KEN GRAUER, MD (3/15/2019):
This is an important example of shape recognition. The case was submitted by an avid follower of Dr. Smith’s ECG Blog. The patient was a man in his 50s, who presented to the ED with chest pain. His initial triage tracing is shown in ECG #1. Forty-five minutes later — a 2nd ECG (ECG #2) was obtained. For clarity — I have put both tracings together in Figure-1.
  • Dr. Smith has expertly detailed the reasons why the QRST appearance in leads V2 and V3 of ECG #1 is diagnostic (until proven otherwise) of acute LAD occlusion.
QUESTION: Aside from leads V2 and V3 — How many other leads in ECG #1 are abnormal?
  • ALSO  Do you think there may be a problem with lead placement in ECG #1?
Figure-1: The 2 ECGs shown in this case (See text).
COMMENT: As per Dr. Smith — considering that this ECG was obtained in an ED, leads V2 and V3 are all but diagnostic of acute LAD occlusion. That said, this tracing was interpreted as “normal” when first seen. As a result — it is worth dissecting the findings in ECG #1.
  • In addition to the findings noted by Dr. Smith — the SHAPE of the QRST complex in lead V2 should “jump out” at you! The frowny” shape (ie, coved) ST segment that is topped off by a fatter-than-it-should-be T wave peak is simply disproportionate to the tiny QRS complex in this lead. This is a “picture” that in a patient with new chest pain — should be saying to you, “I’m an acute OMI until you prove that I am not."
  • To support the presumption that lead V2 is acute — we need to find at least one neighboring lead with ST elevation. As noted by Dr. Smith — the 2mm of J-point ST elevation with “rather straight” takeoff in lead V3 is clearly abnormal. NO more than these 2 neighboring leads in this patient with new chest pain should be needed to justify prompt transfer to the closest PCI facility.
BUT — it is Important to Recognize the OTHER Findings:
  • With proximal (as opposed to mid- or distal) LAD occlusion, in addition to anterior chest lead ST elevation — there will often be associated limb lead changes in the form of ST elevation (which may be subtle) in lead aVL inferior lead ST-T wave depression. In ECG #1 — there is suggestion of slight ST elevation in lead aVL — and, there are subtle-but-real ST segment flattening changes in each of the 3 inferior leads. Of these, lead aVF is the most marked — in that it clearly shows a straightened (if not slightly depressed) ST segment, with abrupt angulation that leads into a taller-than-it-should-be T wave given the tiny amplitude of the QRS complex in this lead aVF. NOTE: These limb lead changes are subtle! Butin the context of clearly abnormal ST elevation in V2 and V3 in this patient with new chest pain — I believe they strongly support the likelihood of acute proximal LAD occlusion.
  • Next in the context of clearly abnormal ST elevation in leads V2 and V3 — I interpreted the slight-but-real ST elevation in neighboring leads V1 and V4 as abnormal. I felt this offered further support that acute proximal LAD occlusion was ongoing. Retrospectively — I felt a look at ECG #2 (in Figure-1confirmed that the subtle ST elevation that I just alluded to for leads V1 and V4 of ECG #1 was real — because abnormalities in both leads V1 and V4 have clearly become more evident in this 2nd ECG obtained 45 minutes later.
Finally — I suspect there is malposition of at least lead V2 in ECG #1. This is because R wave progression as we move from lead V1-to-V2-to-V3 in ECG #1 just doesn’t make physiologic sense (ie, practical disappearance of the S wave in lead V2, with this multiphasic, almost null QRS complex — that then shows return of a substantial S wave in lead V3). Realizing that much has changed in ECG #2 with evolution of the infarct and development of deep anterior Q waves (QS complexes) — Doesn’t progression of QRS complexes and ST-T waves across chest leads in ECG #2 appear to be more logical?
  • Given that ECG abnormalities were initially not recognized in ECG #1 — it is quite possible that suspecting malposition of one or more chest leads, and immediately repeating the ECG might have resulted in a repeat tracing with more leads showing abnormal findings that might have been picked up earlier.
BOTTOM LINE: The fact that ECG #1 was initially interpreted as “normal” in this patient with new chest pain — means there is still “work to be done” in the area of ECG interpretation education.Rather than strict adherence to numerical stemi guidelines:
  • Remember that the prevalence of acute cardiac disease is greatly increased in an adult of a certain age who presents to an ED with new-onset chest pain.
  • When the amount of ST segment deviation is modest — ST segment SHAPE takes on much greater importance.
  • Although changes may be subtle — it’s essential to scrutinize all 12 leads, to see if a consistent “story” is being told. Doing so reveals that no less than 8 of the 12 leads in ECG #1 show at least subtle abnormal findings — which, when interpreted together in association with the clearly abnormal appearance of leads V2 and V3, make a convincing case for acute LAD occlusion until proven otherwise.
Our THANKS to the avid reader who shared with us this insightful tracing!

Recommended Resources