Sunday, December 15, 2013

Friday's post produced skeptics.....

I have been posting ECGs like this for so long that I erroneously thought everyone would see the inferior STEMI here.  I was just posting it as another example.

One reader was exasperated and said, "this is too much," implying that it is simply beyond belief that this ECG, which I will post again here, unequivocally shows inferior MI.

Remember the patient had a previous anterior MI and an LAD stent.

Here is the ECG again:
The ECG clearly shows the Q-waves in V1 and V2 from the old MI.  But that was not the focus of the post.  The combination of proportionally huge T-waves in II, III, aVF and minimal ST elevation in these inferior leads, along with T-wave inversion and minimal reciprocal ST depression in aVL, as seen here, is diagnostic of inferior MI.  The T-wave inversion in V2 is highly supportive as it suggests concomitant posterior MI because the artery that supplies the inferior wall very frequently supplies the  posterior wall as well.


First, let me try to establish some trust here:
1) I have been intensely interested in the ECG for the last 25 years, and scrutinized many tens of thousands of them.  I am provided particularly difficult ECGs by hundreds of people, continuously, over the years, and so have seen not just thousands of ECGs, but thousands of ECGs which physicians find particularly difficult to analyze.  I have always verified my readings with angiographic outcomes.

2) I have read over a thousand papers on the ECG in acute MI; some of the older ones are described in the annotated bibiliography of each of thee 38 chapters of my book.

3) I have done a lot of formal research into ECG findings of MI, and written a lot of evidence-based chapters on the ECG in MI.  Here is a list of my cardiology publications, mostly on the ECG (see the first 5 pages on this google scholar search)  Much of what I've studied is in how to identify subtle occlusion and how to differentiate the ST elevation of STEMI from that of the look-alikes.

4) There are 3 studies in the literature that I am aware of which studied the incidence of an occluded infarct-related artery in patients who rule in for MI by troponin.  These are patients who rule in and therefore get an angiogram the next day.  In these studies, 30% have an occluded infarct-related artery that was not detected on the ECG.  They have higher mortality, worse LV function, and higher biomarkers.  They were ultimately diagnosed with "NonSTEMI" (the diagnosis came after the angiogram, even though the artery was occluded).  Could these have been identified on the ECG?  I believe from my experience that many of them could have been diagnosed.

5) There are a number of studies in which ECGs are given to "experts" who are told to say whether the patient has an occluded artery or not.  In one such study, accuracy by interventionalists was very low and the interrater reliability was even worse.

In another study here, the best accuracy was among interventionalists, but still they had a mean accuracy of only 79%.  This study published all the ECGs they used, so that you can take the test yourself.  I did so, and I had 100% specificity and 92% sensitivity, far higher than the mean for any group.  I state this not to brag but to illustrate that one can train and get very sensitive at finding the subtle occlusion without sacrificing specificity.

As for this ECG:
1) I was reading a stack of ECGs and I saw this one and I instantaneously knew without any doubt that it was an inferior MI.  I said to myself: "Wow, here is an inferior MI that most physicians would not recognize as such.  I wonder what happened."  So I investigated and found that the physicians had not, in fact, recognized the hyperacute T-waves.

2) We have studied the ECG in inferior ST Elevation (STEMI vs. early repol vs. pericarditis) and are about to submit a manuscript.  In our study, no patient with any ST elevation who did NOT have any reciprocal ST depression in aVL had an MI.  Zero.  And among (admittedly select, with ultimate diagnosis of STEMI) patients with some (even minimal) inferior ST elevation, any ST depression in aVL was 99% sensitive for inferior MI.  T-wave inversion had similar sensitivity and specificity.

3) I have seen countless such cases and recognize this ECG as if it is an old enemy.  For me, this morphology brings instantaneous recognition.  I am trying to find ways to codify this subjective morphology recognition of MI, for many kinds of morphologies, so that others can recognize it equally well.

4) Proportionality is the one of the most important aspects to reading ECGs.  ST elevation or T-wave size that is out of proportion to the size of the QRS is alarming.  In this ECG, it is particularly true of lead aVF.  It is not the size of the T-wave or the height of the ST elevation that indicates MI, it is their relative size, relative to the QRS.

5) The T-wave is just as important, or perhaps more important, than ST elevation.  QRS, and R-wave is very important to the diagnosis.

6) Finally, we are only beginning to understand these relationships.  The old research is very poor and not angiographically based.  ST elevation is a crude measure of STEMI, paradoxically.  For instance, in our study comparing early repolarization to subtle anterior STEMI, the R-wave amplitude was the best differentiator, NOT ST elevation.  There are 6 rules of ST elevation cutpoints: we studied all 6 and the best accuracy was only 60%!

Finally, one must simply look at a lot of these, over and over, until they become like a recognizable face.  Perhaps one day facial recognition software will do a good job of making the diagnosis.  Until then, we need humans who have the experience, or at least the rules, to do so.

12 comments:

  1. First - I would like to second all that Dr. Steve Smith says. I'm not sure if there is anyone in this country who has looked at as many difficult-to-interpret acute tracings in patients presenting with chest pain - for which they meticulously follow-up with serial tracings and catheterization confirmation.

    In my teaching of ECG interpretation - I stress the concept of "pattern recognition". There is a similar concept in radiology, which I believe is described as, "Aunt Minnie". You can verbally describe what aunt Minnie looks like for pages - but until you see her - you may not recognize her. So it is with pattern recognition in ECG interpretation. It is a "picture" - and the T waves in the inferior leads of this tracing convey that picture that I like to call, "HYPERACUTE T waves". These T waves are taller than they "should be" (ie, with respect to the relatively small QRS complexes in these leads). Note how these T waves are broader - almost as if these T waves are "trying" to elevate the ST segment. This is NOT normal. A look at lead aVL, which shows deeper-than-expected and more prominent negative T at the end of this sagging ST segment confirms that what is seen in the inferior leads is likely to be real and probably acute (esp. if the patient is having new-onset chest pain). BOTTOM LINE: Hyperacute T waves may be subtle - they are often a short-lived (transient) ECG sign - but one that is extremely important to recognize, because in the patient with new chest pain (esp. in association with reciprocal ST-T wave changes - as are seen here in aVL) - it means your patient is acutely infarcting from acute occlusion of a major coronary artery. THANK YOU for presenting this case Dr. Smith!

    ReplyDelete
  2. So I'm a paramedic with a few years experience now. I've looked at a couple of 12 leads but am really trying to get better and work at a higher level. I see a lot of "chatter" about hyper-acute T-waves. Can you describe the difference between a hyper-acute T-wave as an early change in MI and the peaked T-waves indicative of hyperK? Let's limit the conversation, if possible, to EKG changes and not return to the obvious importance of pt history and lab values.
    Thanks
    Jason

    ReplyDelete
    Replies
    1. Hyperkalemia manifests primarily in precordial leads, so hyperacute T-waves and hyperkalemia are only a major diagnostic differential if seen in V2-V5. For medics, it is more difficult because you can't measure ST elevation. However, they really do look different but if is hard to explain. Hyperkalemia really has sharply peaked T-waves, as if pulled up by a string. They have a flatter ST segment.

      See these two posts:

      http://hqmeded-ecg.blogspot.com/2009/02/hyperacute-t-waves.html

      http://hqmeded-ecg.blogspot.com/2010/01/peaked-t-waves-hyperacute-stemi-vs.html

      Delete
  3. Dear Dr. Smith,

    I am very glad to have you as an excellent teacher of ECG. I read your every new post, and have started to read the older posts as well. I recommended your website to many people with praise. I also own your book from which I benefitted tremendously, and which I’d like to have on the market.

    Thank you very much.

    Mehmet K. Celenk, MD

    ReplyDelete
  4. Steve,

    What I can say in support of the skeptics is this:
    I look at this ECG and it is an obvious inferior wall MI, and that is entirely a result of reading this blog over the years. Before I discovered your site, I never would have thought anything but non-specific findings as the intrerpretation.

    ReplyDelete
  5. Hi Dr. Smith , I thank you once again for the wonderful job that allows me and many of my colleagues to improve their knowledge. I visit your site every day and every day I discover something new.
    Thanks for everything.
    Vittorio

    ReplyDelete
  6. hi doctor
    please keep posting ECGs, i've learned so much, i recognized immediately the Inferior MI at the first glance when my sight " fell" on the disproportion of the QRS and T wave in DII, my though was enhaced by looking at the other leads, all that is due to reading your posts on this wonderful blog.

    greatings from algeria.

    ReplyDelete

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.