Saturday, January 24, 2015

Anatomy of a Missed LAD Occlusion (classified as a NonSTEMI)

A male in his 50's called 911 for constant 8/10 midsternal chest pressure.  Here was his prehospital ECG:
Computerized QTc is 423 ms.  The ST elevation at the J-point was measured by the computer (see right side), and is less than the "criteria" for anterior STEMI (2 mm in males over age 40).  There is also almost a saddleback morphology in V2.  I say "almost" because the R' wave is not tall enough to be a typical saddleback.

By the 4-variable subtle LAD occlusion vs. Early Repol formula, this is clearly LAD occlusion: STE60V3 = 2, RAV4 = about 7, QRS = 9 mm; value = 20.88 (greater than 18.2 indicates probable LAD occlusion)

Here is a typical saddleback morphology, which is rarely due to STEMI:
This saddleback ST elevation prompted a false positive cath lab activation.  It was due to LVH. 

Notice that there is a significant preceding S-wave
Notice the tall preceding R' wave. 

Contrast these S- and R'-waves with the above ECG
Such saddleback ST elevation is rarely due to STEMI (I have never seen one that was!). 

This nearly meets ECG criteria for type 2 Brugada morphology (I will post on that difficult topic soon).  Full text link:  Current electrocardiographic criteria for diagnosis of Brugada pattern: a consensus report

Case continued:

He was given 2 sublingual NTG, with improvement of pain to 4/10.  Here was the second prehospital ECG:
Computerized QTc is 421 ms.  There is no significant change.

He arrived in the ED and had this ECG recorded:
QTc is 435 ms.  There is still less than 2 mm STE at the J-point, so it does not meet STEMI criteria.  The formula, using 1.5, 435, and 11 gives a value of 23.8, which, greater than 23.4 is unequivocally positive.  LAD occlusion must be assumed until proven otherwise.

The ECG was read as normal by both the computer and the physician.

Comment: Readers of this blog may be critical of this, but that is because you have been sensitized to this diagnosis.  This is the normal assessment throughout the world!  To do better than this is the exception, not the norm.

That is why in study after study, 20-30% of angiograms done for "rule-in MI" by troponins find an 100% occluded artery at next day angiogram.

Don't be critical of this assessment; rather, pass the word and help your fellow emergency physicians and cardiologists to see these findings.

The initial contemporary, sensitive troponin (but not high sensitivity troponin -- these are not yet available in the U.S.) was less than 0.10 ng/mL (undetectable).

The first troponin is negative in 50% of acute STEMI.

A repeat ECG was done 140 minutes later:
QTc is 444 ms.  There is no more STE than before.  T-waves are slightly different, but not larger.
Q-waves are beginning to form in V2 and V3.  This proves it is an acute anterior MI.

A troponin drawn almost 4 hours after the first one was 0.18 ng/mL.  It doesn't sound very high, does it?  The Q-waves were not noticed.

Case continued

At 4.5 hours, another ECG was recorded:

At this point, the Q-waves were noticed and the cath lab was activated.

A 100% LAD occlusion was stented.

No further followup is available.

Learning point

I repeat this theme over and over: Acute coronary occlusion may be very subtle.  It is frequently missed.  Readers of this blog probably would not miss this.  The person who sent it to me does read the blog, and he was doing QA when he noticed this and immediately recognized that it was a missed subtle occlusion.  It would be classified as a NonSTEMI.  He states that it is difficult to convey to his colleagues how to recognize these.


  1. first prehospital ECG with 7 mm (RAV4)
    first ED ECG 11 mm (RAV4)
    explication feasible ?

    merci, "toujours le meilleur"


  2. Hi Dr. Smith ~ This was my case! Great analysis that proved to be very accurate. I certainly will apply your insights in the future. I would only add that this patient at bedside had 1/10 pain that he reported went away with rubbing along the sternal border. In the moment, the mildness of his pain clearly biased me towards thinking early repol. What is more, on bedside echo the basal portion of the septal wall, that we typically image at the level of the papillary muscle, was normal and global l EF was normal. Again biasing against a complete LAD occlusion. On review of this POC echo there was clearly an evolving wall motion abnormality of the mid to distal LAD territory which is less easily appreciated on standard ED POC echo views that emphasize the basal portion of the LV. Thanks for assistance to learn from this case!

  3. All of the voltages are less. They are difficult to compare ECG to ECG, especially prehospital to hospital, as technique may be very different. One must compare all the findings "internally" - proportions of R-wave to QRS to T-wave etc.

    Merci a vous!! Toujours un plaisir!


    1. Andrew,
      Thanks for the additional information, which is very interesting. Ultrasound, unless done by a super expert with the use of contrast, has its limitations. So does the ECG, of course. Were you using Definity contrast? Do you have Speckle tracking, which also helps? As for chest pain, there was a study this year (which I can't find now) showing that severity of pain does not correlate with MI vs. benign diagnosis.
      Thanks for contacting me!
      Steve Smith

  4. One high sens assay is available in the us

    1. Scott,
      There are no high sensitivity assays that are FDA cleared. The contemporary troponins are very sensitive and are sometimes mistaken for hs assays. But they do not reach hs criteria, which is that > 50% of normals have a detectable troponin by an hs assay.

  5. Thanks for posting Steve. You have truly collected an amazing number of these types of tracings - all emphasizing the same points. Your collection is GOLDEN - truly insightful!

    The findings of concern on the initial (prehospital) ECG are once again: i) ST coving and elevation in V1 (subtle-but real); ii) Anterior T waves that not only are peaked - but also are taller than expected (given height of the R wave in V2,V3,V4) as well as "fatter"-than-expected for simple repolarization abnormality; and iii) abnormal ST-T waves in the inferior leads (subtle but real ... ).

    THANKS again for posting and your tireless work in this area! - :)

  6. Is there any correlation with the T waves in AVR and V1? Seems that a lot of LAD occlusions have inverted T waves in AVR with positive T waves in V1.

    1. T-wave inversion in aVR is the norm. An upright T-wave in V1 is associated with LAD occlusion. In our study of 143 subtle LAD occlusion vs. 171 early repol, 35% of LAD occlusion had an upright T-wave in V1 and only 14% of early repol. However, it did not add anything to the multivariate model we used to create the logistic regression formula. So just use the formula! The upright T-wave is certainly something you could use to heighten your suspicion.
      Steve Smith

    2. Amal Mattu had a lecture whereby he was showing NTTV1 or TV1>TV6 is a indicator for LAD occlusion.

    3. He bases that on no data, but opinion only. See my data above.

  7. Steve...

    Great post and I appreciate your comment about containing the criticism. Most EM physicians (actually, most physicians in general) really haven't had very extensive ECG training. I do have a question regarding your formula - specifically, the R height in V4. When you look at the pre-hospital ECGs, it's evident that there is some variability in the height of the R wave in V4. Which R wave height would you recommend measuring for the formula? It seems to me that this choice may skew the answer one way or the other.

    Thanks for the blog!

    Jerry W. Jones, MD FACEP FAAEM

    1. Jerry,
      The findings on any one ECG can change by lead location, patient position, etc., but the important thing is the internal relationships among all the variables. It is the R-wave amplitude relative to all the other findings on THAT ECG which are most important.
      So use the 3 variables on the ECG you have.
      Do look for changes from one to the next, but if R-wave amplitude is less, is ST elevation also less? Is T-wave size also less?
      If R-wave amplitude is lower but ST elevation is higher and T-wave is larger, then you have an evolving MI.

    2. Steve...

      Thanks for such a prompt reply, but I'm not sure if you understood my question or if I may have misunderstood your response. In the first pre-hospital ECG, within Lead V4, there is variation of R wave height. Within that lead are R waves with different heights. Which would you recommend using in the calculation - the taller R wave or the shorter R wave?

      Also, is your formula available as a phone app?

      Thanks again,


    3. Jerry,
      Sorry, I did misunderstand.
      Best to just take the average R-wave height.
      free iPhone app: subtleSTEMI

  8. Dear Dr. Smith,
    Just a query. Isn't there a J wave in the lead V2 which could be due to early repolarisation?

    1. Yes, there is a J-wave. But anterior STEMI may also have J-waves. In our study comparing 355 consecutive LAD occlusion (143 subtle) with 240 early repol, the presence or absence of a J-wave did not improve the prediction accuracy of the above formula. Without looking back at the data, as I recall the indicence of J-waves was similar in both groups.


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.

Recommended Resources