Sunday, March 31, 2019

Three EKGs shown to me - which if any need emergent reperfusion?

Written by Pendell Meyers

We commonly get feedback from readers who are skeptical that we can pick out the subtle NSTEMI occlusions from the endless onslaught of triage/EMS EKGs. Some believe that one cannot learn to pick out subtle true positive STE or hyperacute T-waves without sacrificing specificity.

We see countless EKGs constantly (as do all EM physicians) plus more because other providers constantly send us EKGs in addition. I see countless abnormal-looking EKGs to which I respond "I don't see any evidence of occlusion", and only a tiny percentage of diagnostic EKGs among those shown to me.

So take a look at these three EKGs I was shown during a single shift the other day. Imagine how you would handle these and see if you think it's impossible to differentiate the few OMIs from the countless abnormal but meaningless EKGs.

The only information I had in all three cases was a middle aged patient with chest pain. I could not see or evaluate the patient at the time of making the EMS/triage decisions (to activate the cath lab or not). 

Here they are:




What would you say in each of these cases brought to you in the middle of your shift?

My responses: 

#1 I suspect false positive ST elevation, do not activate the lab, and please get a 12-lead on a machine that does not cut off the QRS voltage so we can see the true ratios of QRS and ST deviations. Clear J-waves and typical morphology make BER extremely likely.

#2 STEMI equivalent because of hyperacute T-waves which are diagnostic in anterior leads. This patient was headed for a nearby hospital via ambulance, so I called them and told them my concerns because we are not able to activate their cath lab. If this patient were coming to my hospital, I would not activate the cath lab but instead we have a "heart alert" for situations that are deemed STEMI equivalents which do not meet formal criteria, in which we discuss immediately with the interventionalist, then decide the best course of action. I would have called a heart alert and argued strongly for emergent cath on arrival.

#3 False positive ST elevation, although technically this does meet "STEMI criteria" it is false positive because of the Q waves, abnormal and high voltage QRS, and ST morphology which is very unlike acute ischemic STE. The fact that there is indeed STD in aVL does not mean that it has to be acute ischemia - any process which causes focal STE must cause reciprocal STD. This case was brought to me after other providers had already activated the cath lab, so the patient was already proceeding emergently for cath.

So what happened?

Patient #1:


The patient (turned out to be 38 years old) was brought into the ED (not my zone) and had this EKG on arrival:

This is perhaps the lowest score I've ever heard of!

The providers were worried by the STE and activated the cath lab. Coronaries were completely normal (no CAD whatsoever), 3 troponins were undetectable, and the EKG remained unchanged (did not evolve).

Patient #2:


Although we lack V3, we can see Q-waves from V1-V4, with slight STE and very large symmetric T-waves which are hyperacute compared to their QRS complexes. There is STD in I and aVL. The forumula is contraindicated due to Q-waves because normal ST elevation never had Q-wave in V2-V4.

I called the receiving hospital and explained my concern about these EKG findings.

On arrival there, this was his EKG:

Findings may have improved slightly, however V3 still has diagnostic Q-wave, STE, and hyperacute T-wave. There is slight STD in V6, I, and aVL.

Days later I called to find out what happened to him.

Luckily, despite no STEMI criteria being met, the patient was taken fairly quickly to the cath lab where he was found to have a 99% thrombotic stenosis (TIMI-1 flow) of the mid-LAD, in the setting of severe three vessel disease. PCI was technically difficult, and the patient went for emergent CABG the next day. Peak trop and echo were not available unfortunately.

Patient #3: 

Triage EKG:

I found out later that this patient was a 57 yo M with history of CAD s/p CABG (LIMA-LAD), multiple PCI to LM, LAD, LCX, presenting with intermittent chest pain for the past 2 days. Just before arrival, he had an episode with associated diaphoresis. He was having chest pain during this ECG on arrival.

The providers activated the cath lab. He was of course found to have his previously known three vessel disease, and his LIMA-LAD was patent. He was found to have a 90% stenosis (TIMI 3 flow) of his ramus intermedius which was stented, however it does not seem to have been an acute culprit. 

Four serial troponins were undetectable, and this was the only repeat ECG on file (5 hours after cath):

Despite the difference in the QRS complex, there is still STE in the inferior leads (this morphology is actually closer to ischemic changes than the initial EKG). Thus, the ECG has not evolved down the progression of ischemia. Troponins remained undetectable during this persistent STE and during pain. Echo showed no wall motion abnormalities, no LVH, and EF 65%. This is all strong evidence that there was no acute culprit and no ACS. 

Learning Points:

Some EMS 12-leads have software which cut off the QRS complexes to avoid overlap on the display. This is important to know because all findings on EKG are proportional, and so you must be able to see the full QRS complex in order to interpret the ST segments and T-waves correctly.

Advanced ECG interpretation involves improvement in sensitivity and specificity over the current STEMI vs. NSTEMI paradigm. With practice you can diagnose hyperacute T-waves and all the other subtle OMI findings without overcalling the majority of cases.

Comment by KEN GRAUER, MD (4/2/2019):
Instructive post by Dr. Pendell Meyers — regarding 3 ECGs shown to him during a routine shift in the ED. The only clinical information known about each case — was that the patient was “middle-aged”, and that they had “chest pain”.
  • For clarity — I’ve reproduced these 3 ECGs in Figure-1.
  • My “bottom line” impressions about each tracing (regarding whether or not to activate the cath lab) — were the same as those of Dr. Meyers.
  • I did have a few additional thoughts ...
Figure-1: The 3 ECGs reviewed by Dr. Meyers in this case (See text).
ECG #1: Great point by Dr. Meyers (!), regarding the importance of using a 12-lead ECG machine that does not cut off QRS voltage. Without this — we have no idea of either S wave depth or R wave height in 5 (if not all 6) of the chest leads (RED arrows).
  • The ST-T wave appearance in leads V4-6 clearly suggests a repolarization variant (concave-up = “smiley”-configuration shape to the ST segment, with very prominent J-point notching in V4, V5). Determining whether or not T wave amplitude looks taller-than-it-should-be in leads V2 and V3 would be greatly facilitated by knowing how deep the S waves are in these leads — which we unfortunately are unable to do from this tracing alone.
  • Even without being able to see true QRS amplitudes — I suspected this was most probably a repolarization variant because of: igeneralization of the ST elevation and its shape in virtually all chest leads (a stemi would be more likely to localize); andiithe unremarkable ST-T wave appearance in the limb leads.
  • That said, although I would not activate the cath lab on the basis of ECG #1 — I would want more information — and, I’d want to see a follow-up ECG, done with a machine allowing full visualization of QRS amplitude before making my final decision. (As discussed by Dr. Meyers in follow-up of this case — the 2nd ECG done on this patient did confirm that actual QRS amplitude in chest leads was dramatically increased).
  • I have no doubt that the ST-T wave appearance in leads V4 and V5 of this tracing represents a repolarization variant. But, if anterior S waves were no deeper in ECG #1 than what we see here — my concern regarding the ST-T wave appearance in leads V2 and V3, would be to ensure that we are not looking at new changes superimposed on a repolarization variant.
ECG #2: It’s hard to imagine not promptly performing cardiac cath on this patient, given a history of “chest pain” and the appearance of ECG #2. As per Dr. Meyers — there are large Q waves in leads V1-thru-V4 and dramatic, hyperacute T waves in leads V2-thru-V6. Although J-point ST depression is lacking — this T wave pattern otherwise evokes the image of DeWinter T waves, in how disproportionately tall most chest lead T wave are in relation to R wave amplitude in their respective lead.
  • Without more information (and ideally, a prior ECG for comparison) — it would be difficult to “date” the occurrence of LAD occlusion — but regardless of whether the “millimeter amount” of ST elevation in the “stemi definition” is met — the cardiologist-on-call needs to immediately evaluate this patient.
There are 2 additional findings that concerned me about ECG #2:
  • There is incomplete RBBB (Qr’ pattern in both leads V1 and V2 + terminal S waves in both leads I and V6LAHB (rS pattern, with predominant negativity in all 3 inferior leads). Is this bifascicular conduction disturbance a new finding? — related to a large acutely evolving anterior STEMI?
  • The ST-T wave appearance in lead II of ECG #2 is remarkable! The image of the tiny QRS complex in this lead — with strikingly flat ST segment — and obviously fat-at-its-peak and disproportionately tall T wave — is one that should be embedded in memory as saying, “I am acute until you prove otherwise”.
ECG #3: I found ECG #3 the most interesting — and the most challenging of these 3 tracings! Rather than a “yes” or “no” determination, I wondered — What was going on? My thoughts, as I systematically assessed ECG #3 without the benefit of any clinical information other than that the patient was “middle-aged with chest pain”:
  • There is sinus tachycardia — which is a finding of concern.
  • The intervals (PR, QRS, QT) and mean QRS axis both seem to be normal.
  • Chamber enlargement: This is a bit more difficult to assess than at first might appear. There is no sign of LVH. There is a predominant R wave in lead V1 — which should always make one consider the possibility of RVH (See LINK provided below). That said — I did not think there was RVH here. But P wave morphology is clearly abnormal! It is rare in my experience to see notched inferior P waves with an initial negative (not positive) component. Rather than atrial enlargement — I interpreted this as an intra-atrial conduction defect — which is relevant, in suggesting underlying coronary disease.
  • Beyond-the-Core: There appear to be PTa deviations (elevation or depression of the T wave component of the preceding P wave, during which time atrial repolarization takes place = “PTa” wave). Specifically, the PTa segment in leads III, aVF and aVR appears to be elevated. It appears to be depressed in other leads. Given our concern in assessing this patient with chest pain and sinus tachycardia for a possible acute event — I contemplated whether these PTa deviations might be acute (Please see Figure-2 below — for thoughts on recognizing the usually ignored entity of acute atrial infarction).
Assessment of QRST Changes — is complex:
  • Q waves (big and smallare seen in multiple leads (ie, leads II, III, aVF; V1 — and small q waves in V3, V4, V5, V6). The surprisingly wide (relatively speaking) Q wave in lead V1, as well as the q waves in V3 and V4 raise the question of prior anterior infarction (the Q in V1 is not normal; and normal septal q waves shouldn’t go as far anterior as lead V3).
  • R Wave Progression — As noted earlier, the R wave is predominant (albeit with a qRs complex) in lead V1. This is followed by a very large R wave in lead V2. This raises the possibility of prior posterior infarction. 
  • Consistent with the idea of prior posterior infarction — is the virtual certainty of prior inferior infarction. The question I had was whether these large and wide inferior Q waves with associated ST elevation were “old” — vs recent — vs “new-on-top-of” old? The huge Q in III with modest ST elevation looked “old” to me, as did the minimal reciprocal changes in lead aVL. But I was not at all sure about the ST elevation in leads II and aVF — which to me looked like they certainly might reflect new injury.
  • BOTTOM Line about ECG #3: I would not immediately activate the cath lab for this patient based on the appearance of ECG #3. That said — the multiple ECG abnormalities noted tell us this patient has significant underlying coronary disease. I thought that although the changes on ECG #3 most likely were not the result of acute coronary occlusion — that with a history of new chest pain, the cardiologist-on-call should be seeing this patient sooner-rather-than-later — that more information was needed — and, that another cath might be in this patient’s “near future”.

Figure-2: Recognition of Atrial Infarction (Excerpted from Grauer K: ECG-2014-ePub).

Our THANKS to Dr. Meyers for this highly insightful case!
  • Regarding ST segment Shape (ie, “smiley” vs “frowny” configuration— CLICK HERE (Please see My Comment at the bottom of the page).
  • For more on the Tall R Wave in Lead V1 — CLICK HERE.

Friday, March 29, 2019

Epigastric pain radiating to the chest for 18 hours. ECG makes the Dx. Troponin makes the Dx. CT makes the Dx!

I was shown this ECG with no other information:
What do you think?
Hint: try to see through the artifact!

I answered immediately: "High lateral MI with posterior MI. OMI." (Occlusion Myocardial Infarction)

I asked, "Did the patient present with chest pain?"

Here is the history: "A middle-aged male complained of about 18 hours of epigastric pain that radiated to the chest.  He also had an apparently new facial droop of equal duration.  A stroke code was called, NIH stroke scale was only 1, and attention was turned to the chest pain."  BP was 148/83.

How did I make this ECG diagnosis?

There is subtle STE in aVL with reciprocal STD in II, III, aVF, and STD in V3 and V4.

While there are ECGs that have STE in aVL with reciprocal STD in II, III, aVF in the absence of OMI, they do not also have STD in right precordial leads.  Furthermore, the morphology here is perfect for OMI.

(See the bottom of the post for 3 ECGs from this week with STE aVL and STD III, for which my answer was "no ischemia.")

They performed a point of care cardiac ultrasound (parasternal short axis):

What do you think?

Notice the upper right (septum) contracts much more vigorously than the lower right and right (lateral wall)

The physician did not appreciate this because he was thinking about the patient's pain combined with the facial droop, and so he appropriately obtained a chest CT aortogram to look for aortic dissection.  It did not show dissection, but did show the following images of his heart:
LV on the right. Lower right wall has dark area where there is no contrast in the myocardium (no perfusion)

These show hypoperfusion of the lateral and posterior walls.
Dark areas have no contrast because they are not perfused.

Our radiology chief, Gopal Punjabi, is a CT genius.  He loves to use Spectral CT, and has a spectral CT blog:

He applied spectral CT:
Anterior is to the left, and posterior is to the right
See the dark parts of the posterior wall.  This means there is no contrast there.  It is transmural infarction.

He diagnosed acute MI from his CT reading room. 

For more cases of CT diagnosis of myocardial ischemia via perfusion defect (all by Dr. Punjabi), see here:

--A middle-aged man with severe syncope, diffuse weakness

Case continued
At the very same time that the radiologist was calling to alert of acute MI, the first troponin I returned at 39.4 ng/mL (very high).  In almost any scenario, a troponin this high (even without the ECG and CT) should be considered diagnostic of OMI until proven otherwise by angiography.  The presence of acute refractory chest pain is an indication for emergent angiography regardless of the ECG or any imaging evidence of ischemia.

Could it be something else?  Possible, but unlikely.  Myocarditis, takotsubo, type 2 MI and others almost never have troponin this high.

IV nitroglycerine was started.  Aspirin and heparin were given.

A 2nd ECG was obtained later while the patient had ongoing pain:
OMI looks more obvious

Cardiology was consulted.

A second troponin returned at 55.6 ng/mL.  Pain continued.  A third troponin returned at 57.1 ng/mL.  Pain continued.

The ED physicians decided to give nitroglycerin.

The BP dropped to 96/54 and the pain did not go away (even though hydromorphone was also given -- this can only do harm by masking the pain).  

A 4th troponin returned at 63.8 ng/mL and the patient was taken to angiogram.

At angiogram, he had a 100% occluded circumflex.

It was opened and stented.

The troponin peaked 2.5 hours after the angiogram, and 12 hours after arrival, at 135.6 ng/mL.  This is very high.

Echo next day:
Normal left ventricular size, mild concentric LV hypertrophy, and normal systolic function.
The estimated left ventricular ejection fraction is 60%.
Regional wall motion abnormality-inferolateral, hypokinetic.

(Fortunately, the patient's EF remained quite good.)

Here is the post PCI ECG:
STE and STD remain

Learning points:
1. Learn to recognize STE elevation in aVL, with reciprocal STD in inferior leads.
2. Learn to differentiate it from look-alikes (see below)
3. Pay attention to perfusion of the heart with contrast on CT scan.
4. Very high troponin is almost always a type 1 MI due to OMI
5. Use echo to help the diagnosis.

High Lateral Mimics

Here is a previous post on STE in aVL, and when reflects high lateral OMI and when it does not
True Positive ST elevation in aVL vs. False Positive ST elevation in aVL

These are some other ECGs I was shown this week in patients with chest pain:

There is STE in aVL, with reciprocal STD in III
What do you think?

There is STE in aVL, with reciprocal STD in III. What do you think?

There is STE in aVL, with reciprocal STD in III
What do you think?

My immediate response to all was "normal." And they were normal.  They just do not have the "look" of ischemia.  I'm sure Ken Grauer can tell us exactly what makes these non-ischemic when, in contrast, the one at the top of the post is indeed ischemic and we can recognize it as such.  All I can say is that they are a face I can recognize.

Comment by KEN GRAUER, MD (3/29/2019):
Great post by Dr. Smith — which should be entitled, A Face I Can Recognize. Dr. Smith’s astute step-by-step analysis of serial ECGs in this case + supplemental material (Echo, CT angiogram, spectral CT) needs no further explanation. But I did want to comment on those 3 other ECGs he was asked to assess this week — for which the diagnosis was not OMI.
  • As per Dr. Smith, despite ST elevation in lead aVL with reciprocal ST depression in lead III for each case — all 3 of these tracings were “faces” Dr. Smith instantly recognized as not being due to acute ischemia/OMI.
  • I have included 4 ECGs in my Figure-1 below. For clarity — I put as the TOP tracing ECG #3 = the 3rd ECG done in this case, which was obtained post-PCI for treatment of acute LCx occlusion. The 3 ECGs below this top tracing — are the 3 “Mimics” that manifest ST elevation in aVL + reciprocal ST depression in lead III, but which are not the result of acute OMI.
Figure-1: The post-PCI tracing ( = ECG #3) + the 3 Mimics not due to OMI (See text).
COMMENT: Dr. Smith’s ability to instantly recognize the “face” of acute OMI — and to distinguish this from mimics of acute ischemia, is truly exceptionalIt is the result of his years of assessing case-after-case, accompanied by his unsurpassable follow-up of each case. He is almost never wrong.
  • I am admittedly not as fast as Dr. Smith in arriving at a definitive diagnosis. In the hope that my approach to each tracing may prove insightful — I offer the following reflections on the 4 tracings shown in Figure-1.
ECG #3: For illustrative purposes — I chose the best quality tracing of the 3 ECGs performed on the patient in this case. This was the ECG done following PCI of acute LCx occlusion ( = ECG #3). Although the amount of ST segment deviation in any given lead in ECG #3 is modest — this tracing should be immediately identified as potentially new or very recent acute ischemia until proven otherwise:
  • Lead aVL manifests an “acute-looking” shape of ST elevation. Despite tiny amplitude of the QRS complex — this shape alone is highly suspicious of acute disease. There is also a definite Q wave in this lead — that is disproportionally deep compared to the tiny R wave.
  • Not only does lead III manifest a mirror-image picture of reciprocal ST depression (to the ST elevation seen in aVL) — but both of the other inferior leads (II, aVFshow subtle-but-definitely-abnormal ST flattening and depression!
  • While not elevated — the T wave in lead V2 looks taller-and-fatter-than-it-should-be given the tiny amplitude of the QRS complex in this lead.
  • There is subtle-but-real ST segment flattening with ST depression in leads V3, V4, V5 and V6.
BOTTOM Line regarding ECG #3: There is abnormally-shaped ST elevation in lead aVL — with subtle-but-real ST-T wave abnormalities in no less than 8 of the remaining 11 leads. In this patient with new-onset chest pain — a tracing like this has to be assumed acute or recent until proven otherwise.
  • I wanted to contrast these findings in ECG #3 — with what we don’t see in the other 3 ECGs ...
Tracing A: Despite ST elevation in lead aVL with reciprocal ST depression in lead III — this tracing looks far less likely to be acute. Reasons for having a low index of suspicion for acute OMI in Tracing A include:
  • There is marked LVH! The finding of very deep anterior S waves (~23mm in V2; ~20mm in V3) + a surprisingly tall R wave in V6 (despite the late transition) clearly satisfies voltage for LVH. Although ST-T wave changes of LV “strain” are most often seen in lateral leads — sometimes LV “strain” may manifest itself as ST elevation in anterior leads, especially in association with unusually deep anterior S waves. The magnified insert to the right of Tracing A shows that the mirror-image of the ST-T wave in lead V2 conveys an ST-T wave shape identical to that expected with marked LV “strain”.
  • The isolated finding of T wave inversion in lead III (even when deep and symmetric, as seen in Tracing A) — is not necessarily abnormal. This is especially true when the QRS complex in this lead is predominantly negative, as seen here.
  • The shape of the upright T wave and slightly elevated ST segment in lead aVL looks similar to the ST-T wave appearance in 5 of the 6 chest leads — and this shape in those chest leads with deep S waves seems most likely to be due to LVH.
  • BOTTOM Line: The predominant finding in Tracing A is marked LVH + LV “strain” in the anterior leads. LVH is a common mimic of acute ischemia. While I would not be 100% certain from Tracing A alone that nothing acute is going on (ie, the T wave in lead V5 seems a bit-taller-than-expected given the relatively small r wave in this lead) — almost all of the findings described above are explainable by the marked LVH.
Tracing B: The QTc is prolonged in Tracing B (I estimate it to be ~460-470msec). Although the ST segment is coved and slightly elevated in leads I, aVL and V2 — the arc of the ST segment looks very gradual in these leads (in contrast to the ST-T wave in leads aVL and V2 of ECG #3). This does not look acute. In support of low likelihood that this ST coving is acute — the mirror-image downward arcing of the ST segment in leads III and aVF is equally gradual — and ST-T waves are simply “flat” (not depressed) in 5 other leads (leads II, V3-V6).
  • This tracing just doesn’t look acute ...
Tracing C: The finding of T wave inversion in either lead III and/or in lead aVF is not necessarily abnormal. For this reason — I place great weight in decision-making on the ST-T wave appearance in the 3rd inferior lead = lead II. Although in Tracing C, the ST-T wave in lead II is somewhat flattened — there is no hint of T wave inversion. Given that the QRS complex is predominantly negative in these 2 leads with T wave inversion and, the lack of J-point ST depression in these 2 leads — I don’t think any conclusion can be reached as to whether or not T wave inversion in leads III and aVF of Tracing C is acute.
  • The slight ST elevation in lead aVL is concave-up (ie, “smiley” configuration) — and totally consistent with a repolarization abnormality. The q wave in lead aVL is consistent with a normal “septal” q wave. And, the shape of the ST segment in lead III is the mirror-image opposite of the elevated ST segment in aVL that looks like a repolarization variant.
  • The overall picture in the limb leads does not look acute.
  • In support of low likelihood that these limb lead findings are acute — the chest leads are inconclusive. While one might wonder about the size of the T wave in lead V2 relative to the small r wave in this lead — neither neighboring lead (V1 or V3) gives even a hint of abnormality — and, the ST-T waves in V4-6 are unimpressively flat. So, while I might not be 100% certain from Tracing C alone that nothing acute is going on (depending on how suspicious the history was) — my index of suspicion for acute coronary disease would clearly be low.
Additional Learning Points:
  • Pay attention to ST-T wave shape. Shape is often more important than the amount of ST-T wave deviation.
  • The more leads that show subtle-but-real ST-T wave abnormalities consistent with the overall picture — the more likely these changes are to be acute.
  • You don’t have to make a definitive decision on the basis of a single tracing. Clinical correlation with history, serial tracings and additional diagnostic modalities can usually provide a more confident answer in a time-efficient fashion.
Our THANKS to Dr. Smith for this highly insightful case!

Recommended Resources