Tuesday, September 15, 2020

A 58 year old collapses in the hot sun

A 58 yo male was out working in the hot sun for 2-3 hours. He stated he almost passed out, and bystanders called 911. They give him water with salt, as he thought he was dehydrated.
When medics arrived, he was alert, sweating, and felt weak.  He walked to the ambulance for evaluation.  He denied headache, chest pain, nausea / vomiting and dyspnea. 

He had no cardiac history, meds, or risk factors. Vitals were obtained, and placed on cardiac monitor, including this 12 lead prehospital ECG: 
QTc =  320 ms; (QTc = 374 ms)
The computer measures the ST Elevation at the J-point for you.
Here it is 4.08 mm in V2, and 2.84 in V3, as well as 2.34 mm in V4.

This looks worrisome for anterior MI, and with ST elevation in aVL and reciprocal ST depression in inferior leads, it looks like a proximal LAD occlusion.

Or is it normal variant ST Elevation (often known as early repolarization?)

This has been thoroughly studied by me and Dr. Emre Aslanger (an interventionalist) in a series of articles:
Calculate the 4-variable formula at MDcalc, or get the iPhone app ("SubtleSTEMI"), or the Android app ("ECG Smith")

In these studies differentiating Subtle LAD occlusion from normal variant ST Elevation, we excluded patients with ECGs that had "obvious" LAD occlusion.

This included:
Absence of upward concavity in V2-V6, even though it is found in 40-50% of LAD occlusion
STE at the J point of 5 mm or more
Inferior or precordial ST depression
Terminal QRS distortion (absence of S-wave and J-wave) in V2 or V3.
Q-waves in any of V2-V4

In this ECG:
There is no STE at the J-point of 5 mm
There is upward concavity in V2-V6  
There is no Terminal QRS distortion.
There are no Q-waves.
There is no ST depression in precordial leads.

Since there is ST depression in inferior leads
it is hazardous to use the formulas to differentiate normal variant from LAD occlusion. 
This patient would have been excluded from the study because it is an "non-subtle" (or obvious) LAD occlusion.

Nevertheless, let's do the calculation just to see what the value is: 

4-variable formula variables:

R-wave amplitude in V4 (RAV4) = 13 mm
Total QRS amplitude in V2 = 15 mm
QTc = 374 
ST elevation at 60 ms after the J-point (STE60V3) = 6 mm

Formula value = 20.07, which is substantially above the most accurate cutpoint of 18.2, and strongly suggests LAD occlusion. 

Note on the cutpoint: Above a cutpoint of 18.2, the formula was 89% sensitive and 95% specific in the derivation (and 83% sensitive and 88% specific in the external validation study). In the derivation, the specificity only gets above 97% at a value of 19.0).  

As with all dichotomous rules, the closer the value is to the cutpoint, the less accurate.  If there is an overall specificity of 88% for a value above 18.2, that specificity will be less for values close to 18.2 and higher for values far above 18.2.

Here is the simplified formula (QT = 320, or 8 mm): 
(R-wave amplitude in lead V4 +QRS amplitude in V2) minus
(QT interval in millimeters (not milliseconds!)  +  STE60 in V3).  

QT = 320 ms = 8 mm

Calculate: (13 + 15) - (8 + 6) = 28 - 14 = 14.

The most accurate cutoff is 12, with values above 12 suggesting LAD occlusion, and values below 12 suggesting benign ST Elevation.

Case Continued

At the Emergency Department, a cardiologist reviewed and activated the Cath lab for an exploratory cath.

cath was negative. Troponins were and remained negative. Patient left AMA after cath.

Dominance: Right
LM: A 5 mm vessel which bifurcates into the LAD and LCx coronary artery. The LM coronary artery is free of disease 
LAD: A type 3 LAD, which gives rise to usual septal perforators and diagonal branches. The LAD and its major branches are free of disease
LCX: A non-dominant vessel that is moderate caliber in size, which gives rise to several OM branches and continues to complete its course in the AV grove as a small vessel. The LCx and its major branches are free of disease
RCA: A dominant vessel witch gives rise to the PDA and PLA. The RCA and its major branches are free of disease

Learning Points

1. Pay attention to the inclusion criteria for the formula.  

    A. Inclusion: STE of at least 1 mm at the J-point in at least 1 of V2-V4 (measured relative to the PQ junction (QRS onset)

2. Pay attention to the exclusion criteria for the formula. These are features which are very rare in benign ST elevation

    B. Exclusion: Wide QRS (such as LBBB or RBBB), STE of at least 5 mm in just one lead, convex ST segment in any of leads V2-V6, Q-waves in any of V2-V4, inferior or precordial ST depression, Terminal QRS distortion in either V2 or V3.

3.  When a patient does not have a high pretest probability, a "positive" ECG has a much lower positive predictive value (this patient had no CP or SOB, no risk factors, and had a good explanation for his symptoms).  This is why our prehospital protocol is: 1) computer diagnoses ***STEMI*** AND patient has chest pain.

4. The specificity of the formula is not perfect.  The specificity of the simplified formula is 92%.  The specificity of the 4-variable formula was 95%, with specificity above 97% if the value was greater than 19 (here it was 20.07).  The specificity of the 4-variable formula in the validation at a cutoff of 18.2 was 87%, and would be higher for a value of 20.07.

5.  I do not use a negative formula value to dismiss an ECG that I am worried about.  I use it to identify an LAD occlusion that I otherwise would have missed.

6. When the value is high and you do not believe that it is LAD occlusion, I recommend very intensive evaluation using serial ECGs, troponins, and (this is best) high quality contrast echocardiogram.

7. Even 58 year old patients may have marked normal variant ST elevation.


MY Comment by KEN GRAUER, MD (9/15/2020):


I like this case — because it highlights that as helpful as Dr. Smith’s formulas may be — they are not perfect. As Dr. Smith emphasizes in his 3rd Learning Point — “I do not use a negative formula value to dismiss an ECG that I am worried about."

  • I limit my comments to WHY the “correct answer” in today’s case was to activate the cath lab — even though the cath turned out to be normal. To facilitate discussion — I’ve reproduced the initial ECG obtained in the field in Figure-1.

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

The patient in today’s case was a previously healthy 58yo man with overexertion after working outside in the hot sun for several hours. He was dehydrated and “almost passed out” — but did not complain of chest pain.

MY Thoughts on the HISTORY: A history of “chest pain” is not reported in all patients who are found to have myocardial infarction. That said, as per Dr. Smith — the pre-test probability of acute MI in today’s case (ie, the likelihood of acute MI based just on history before you even look at the ECG!) — is relatively low. This is because this patient was previously healthy — he did not have chest pain — and there was a good explanation for the symptoms he had.

  • That said — this patient is of a certain age (ie, 58yo) — he did almost pass out, and he was both weak and diaphoretic after the incident. Therefore — although pre-test probability of acute MI is clearly much less than if the history had been new-onset chest pain — there clearly is a possibility of MI associated with non-chest-pain equivalent symptoms.
  • PEARL #1: As per Dr. Smith — the fact that the “prevalence” of disease (ie, the incidence of acute MI among patients with this type of presentation) is lower — ECG abnormalities will be less specific for acute OMI. This is known as Bayes’ Theorem. Our interpretation of the findings in ECG #1 remains the same — BUT, the relatively low prevalence of disease simply means that abnormal ECG findings may be less accurate for predicting acute OMI, than if identical ECG findings were seen in a patient with a much more worrisome history of new-onset, cardiac-sounding chest pain.

PEARL #2: Not all patients with acute MI report chest pain. The Framingham studies from many years ago taught us that the incidence of Silent MI” is as high as ~30% of all MIs.

  • The interesting part of this data was that in about half of this 30% (ie, ~15% of all patients with MI) — patients found on yearly follow-up ECGs to manifest clear evidence of infarction had NO symptoms at all — therefore truly “silent” MIs.
  • In the other half of this 30% (ie, in ~15% of all patients with MI) — patients found on follow-up ECG to have had infarction did not have chest pain — but they did have “something else” thought to be associated with their MI.
  • The most common “something else” symptom was shortness of breath. Other non-chest-pain equivalent symptoms included — abdominal pain — “flu-like” symptoms (ie, myalgias; not “feeling” good) — excessive fatigue — mental status changes (ie, as might be found in an elderly patient wandering from home).
  • BOTTOM Line: Be aware of the entity of “Silent MI” — which can either be completely “silent” — or, associated with a non-chest-pain equivalent symptom. The incidence of both types of silent MI is more common than is sometimes appreciated.

MY Thoughts on ECG #1: The rhythm in ECG #1 is sinus at 85-90/minute. All intervals and the axis are normal. Criteria for chamber enlargement are not seen (although the S wave is cut off in lead V3). Regarding Q-R-S-T Changes:

  • There are no Q waves.
  • R wave progression is normal — with transition (where S wave depth exceeds R wave height) occurring normally between leads V3-to-V4.
  • As noted by Dr. Smith — there is significant ST elevation (especially in leads V2 and V3, attaining 4 mm in V3). In all — ST elevation is noted in no less than 8 leads (ie, leads I,aVL; and V1-thru-V6).
  • There is ST-T wave depression in lead III — with a suggestion of this also in lead aVF.

IMPRESSION (Putting It All Together): This 58yo man does have new symptoms — albeit the lack of chest pain and explainable circumstances surrounding his presenting symptoms suggest a lower pre-test likelihood of disease.

  • There is no prior tracing for comparison.
  • The amount of ST elevation seen in leads V2 and V3 is more than is usually seen with repolarization variants. This is accompanied by definite ST elevation in lead aVL — which is often seen with acute proximal LAD occlusion.
  • The shape of the depressed ST-T wave in lead III is the mirror-image opposite of the shape of the elevated ST-T wave in lead aVL. Concern that this is a “real” finding is heightened by subtle-but-real ST segment straightening, with a hint of ST depression in lead aVF.

BOTTOM Line: IF we had to judge this case solely on the history and ECG #1 — then cardiac cath is clearly indicated. There simply is no way to rule out acute ongoing OMI from proximal LAD occlusion from this single ECG alone.

  • Sometimes the cath will be normal, as it was in today’s case. That’s why caths are done! If all cath lab activations that we order are positive — then we are not ordering enough cath lab activations. Of course, our goal is to limit the number of negative catheterizations that we order as much as possible — but it’s important to appreciate that the negative cardiac cath in today’s case did provide important diagnostic information.
  • As we’ve discussed many times on Dr. Smith’s ECG Blog — among additional tools available in the ED for sorting out which patients with tracings such as the one we see in ECG #1 need prompt cath include: — serial high-sensitivity troponin values — serial ECGs — searching for a prior ECG for comparison purposes — bedside Echo in the ED during chest pain. To know which of these tools are needed for which patients before deciding on the need for prompt cath — “Ya gotta be there...”.

Some Final THOUGHTS on Today’s Case: There are some important lessons to be learned from today’s case. I’ll add the following to the Learning Points put forth by Dr. Smith above.

  • In answer to Dr. Smith’s rhetorical question = “Is there an upper age limit to where we shouldn’t suspect a repolarization variant?” — My ambulatory experience of interpreting all ECGs from 35 medical providers over 30 years taught me that repolarization variants can occasionally be seen in patients in the age range of the 58-year old man in today’s case. And, repolarization variants can be seen in both males and females of any race.
  • I indicated that there was no evidence for chamber enlargement in ECG #1. But there are a number of voltage criteria for LVH that utilize S wave depth in lead V3 — and the S wave in lead V3 is cut off! (See My Comment at the bottom of the June 20, 2020 post in Dr. Smith’s Blog). As a result — we have NO idea if voltage for LVH might not be present. And IF the S wave in lead V3 was significantly deeper than what we see in ECG #1 — then the amount of ST elevation that we see in leads V2 and V3 might not be disproportionately increased.
  • Several ECG features in ECG #1 are consistent with a repolarization variant. These include: i) Upward-sloping (ie, “smiley”-configuration) of the ST segments that are elevated; ii) Presence of this same shape of ST elevation in so many (8 of 12) leads, with J-point notching characteristic of repolarization variants in 4 leads (ie, leads I, aVL, V5 and V6); andiii) Lack of reciprocal ST-T wave depression in the 1 inferior lead with a predominantly positive QRS complex ( = lead II).
  • That said — Even if we attribute all of the ST elevation we see in ECG #1 to a “repolarization variant” — one usually does not see mirror-image opposite ST-T depression in lead III (compared to lead aVL) as we see here. And while true that both leads III and aVF may at times normally manifest T wave inversion when the QRS in these leads is predominantly negative — we still usually do not see ST depression (as in lead III here) or ST straightening (as in aVF). THEREFORE — I would have really liked to see a follow-up ECG on this patient after resting up, electrolyte repletion and rehydration, to see if this resulted in resolution of these ST-T wave changes in leads III and aVF.
  • At least the providers knew that the cath was negative before this patient signed out AMA. Had the patient snuck out before his cath — it would have been incumbent on ED staff to ensure the patient was fully aware of potential consequences of leaving.
  • Finally — Had the patient not left AMA, and had repeat ECG after rest and rehydration shown identical findings as seen in ECG #1 — I would have made a miniaturized copy of his ECG for him to carry in his wallet to show providers in the event that he ever again presented to an ED with new symptoms.


  1. What about lead misplacement in V1-V2 based on abnormal P wave morphology? (quite negative in V1 and biphasic in V2). Could this be responsible for the STEMI pattern?

    1. Yes it is true that the finding of a negative P wave in leads V1 or V2 may be the result of faulty lead placement (ie, placing leads V1,V2 by 1 or 2 interspaces too high on the chest). I review this and 2 other clues to V1,V2 misplacement in My Comment at the bottom of the November 4, 2018 post (https://hqmeded-ecg.blogspot.com/2018/11/chest-pain-and-q-waves-in-v1-and-v2-is.html ). But I also emphasize in this Nov. 4 post in my section on “Potential Caveats regarding the 3 Clues” — that these clues are not infallible indicators of lead misplacement. These clues work best when ALL 3 of the findings I suggest are present — especially when the P wave, QRS and T wave in leads aVR, V1 and V2 all look very similar to each other (as they do in my Figure-1 in the Nov. 4, 2018 example). As per the Caveats I list — one may sometimes normally see a prominent negative component to the P wave in lead V1 — especially when there is left atrial enlargement. And when there is an intra-atrial conduction defect — you can see anything regarding morphology in many of the P waves on the tracing.

      So, in the Sept. 15, 2020 case — although there IS a negative component to the P waves in all 3 of the leads I list (aVR, V1, V2) — P wave morphology looks quite different (especially the biphasic P in lead V2) — the only one of these 3 leads with an rSr’ complex is aVR (and aVR so commonly shows this as a normal finding) — and QRS morphology looks completely different in each of these 3 leads. As to P wave morphology — at least 4 of the 12 leads in the Sept. 15 case show notched P waves, and that suggests there may be an intra-atrial conduction defect. BOTTOM Line: I didn’t think there was lead misplacement of V1 and V2 in this case.

  2. Excellent presentation and I totally agree.

    A question for Steve, but if you (Ken) know the answer then feel free to respond. In the study in which Steve ascertained that 40% (or more) of anterior STEMIs have upwardly concave ST segments, did any of those segments have the J-point notch commonly associated with normal variant (early repol)?

    Another question: I really don't see any ST depression in the inferior leads except for Lead III. But Lead III has a special relationship with Leads I and II and also with Leads aVL and aVF that - I feel - is unassociated with the issue of "reciprocal change." Based on Einthoven's Equation, Lead III = Lead II - Lead I and Lead III also equals Lead aVF - Lead aVL. The presence of early repol in Leads I and aVL would be expected to result in the morphology seen in Lead III - without any issue of reciprocity or acute epicardial ischemia. Early repolarization is said to not exhibit any reciprocal changes and, in this ECG, I think that still holds true. However, I certainly agree that there was a lot going on that favored caution over assumption.

    Excellent case, by the way, and (thankfully) much better management than the previous case you posted.

    Jerry W. Jones, MD FACEP FAAEM

    1. THANKS so much for your comments Jerry — with which I totally agree! The SHAPE of the ST-T wave in lead III (especially given the notch in lead III !!!) could simply be a “mirror image” of a repolarization variant change for the reasons you cite related to Einthoven Equation relationships. But even though the ST segment is not depressed in lead aVF — I thought the SHAPE of the ST segment in aVF was abnormal (ie, to me, clearly straighter than it should be). Of course, because the amplitude in lead aVF is tiny — everything is “miniaturized” … — but I don’t think the shape of the ST-T wave in aVF is “normal”. And the problem with inferior lead reciprocal changes is that while ideally we would see them in all 3 of the inferior leads, sometimes they are only seen in leads III and aVF … — therefore (as you so appropriately state) — Best to favor “caution over assumption”.

    2. Jerry,
      2 things:
      1) In the study in JEM, I did not look for notching. In the study where we derived the 3 variable formula (it is very interesting and I recommend reading the whole study), we found J-point notching in 14% of subtle LAD occlusion.
      2) Early repol usually has STE in inferior leads and in lateral leads V4-V6, but much less commonly in I and aVL. This is because the ST vector in early repol is most commonly towards lead II (which is similar to V5 and V6), resulting in STE in II > aVF > III, but present in all three leads. There is NOT usually STE in I and aVL. There are cases of normal variant STE in I and aVL, but they are not common and therefore are much more likely to mimic OMI. So any ST depression in lead III is likely to be OMI: early repol would usually manifest ST ELEVATION in lead III.


DEAR READER: We welcome your Comments! Unfortunately — due to a recent marked increase in SPAM — we have had to restrict commenting to Users with a GOOGLE Account. If you do not yet have a Google account — it should not take long to register. Comments give US feedback on how well Dr. Smith’s ECG Blog is addressing your needs — and they help to clarify concepts of interest to all readers. THANK YOU for your continued support!

Recommended Resources