Monday, September 21, 2020

Interventionalist at the Receiving Hospital: "No STEMI, no cath. I do not accept the transfer."

Are Some Cardiologists Really Limited by Strict Adherence to STEMI millimeter criteria? Yes. We don't know how many though.

I was texted these ECGs by a recent residency graduate after they had all been recorded, along with the following clinical information:

A 50-something with no cardiac history, but with h/o Diabetes, was doing physical work when he collapsed. He was found in ventricular fibrillation and defibrillated, then brought to a local ED which does not have a cath lab.

Here is the initial ED ECG:

This is pretty obviously and inferior posterior OMI, right?
There is slight inferior ST Elevation, with reciprocal ST depression in I & aVL, and an inverted T-wave.
There is deep ST depression in V2-V4, also with an inverted T-wave.
There is slight ST elevation in V5 and V6.
This is diagnostic of an infero-postero-lateral OMI

Some claim that posterior MI always has an upright T-wave (along with ST depression) in V2-V4;  THIS IS FALSE.  It may be inverted or upright or biphasic during the acute phase.

This right sided ECG was recorded about 30 minutes after the first:

V1R here = V2
V2R = V1
V3R is further right, etc.
There is no right sided STE
V1 and V2 have substantially less ST depression.  There may be some spontaneous reperfusion.
There is no STE in right sided leads -- no RV OMI

This ECG, in which V4-V6 are on the posterior thorax (V7, V8, V9) was recorded 2 hours later:

First, notice that the ST depression in V2 and V3 is almost all gone now.
So you would not expect there to be ST Elevation in V7-V9!!
For a posterior ECG to enlighten the standard ECG, they must be recorded at the same time.
1) inferior leads no longer show OMI; there has been reperfusion
2) V2 and V3 have only minimal ST depression, further supporting reperfusion
3) The T-waves in V2, V3 are now upright and large.   This is what is to be expected when there is reperfusion of the posterior wall.  See discussion below of "posterior reperfusion T-waves"
4) V7-V9 have T-wave inversion, which further supports reperfusion of the posterior wall.

Case Continued

The emergency physician was worried about OMI. He called the interventionalist at the receiving institution.  This is the response he got:

Interventionist: "No STEMI, no cath.  I do not accept the transfer."

After stabilizing the patient and recording more ECGs, he tried again:

Interventionalist: "It isn't a STEMI."

Emergency physician: "I know. It is a STEMI equivalent.  He was occluded and now he has reperfused."

Interventionalist: "We don''t cath these patients. And we wouldn't do it tomorrow either."

The troponin I rose to 33 ng/mL, which is a typical level for STEMI/OMI.  He was treated medically for ACS and did not get an angiogram within 72 hours.  Then he did not awaken after 72 hours, so was put on comfort cares.  But until that point, no one could have known his neurologic outcome, and the stated reason for not accepting was not worry for neurologic outcome but that the ECG did not meet STEMI criteria.


Pendell and I just did a study of the ECG in OMI, and submitted the manuscript to JACC.  I read over 4000 EKGs on 808 patients, many with OMI, many with Non-OMI, and many with MI ruled out (MIRO).

I read serial ECGs and, with each read, would decide "OMI" or "Not OMI" (both NOMI and MIRO were "Not OMI").  Many had serial ECGs; for each patient, after reading and interpreting ECG-1, and moving on to ECG-2, and -3, and -4, etc., I was unable to go back and change my read of each preceding ECG. I would also measure the ST segments and determine if it was a STEMI by millimeter criteria (as outlined in the 4th Universal Definition of MI), or OMI without STEMI.  The outcome measure was whether the patient ultimately had an OMI or not.  The details of the methods and results are of course complex, and you will see them when the paper is published, but suffice it to say that STEMI criteria were about 40% sensitive for OMI and my interpretation was about 90% sensitive (and equally specific), and I interpreted OMI a mean of 3 hours before the ECG which the treating clinicians determined was positive.  Pendell read about a third of the ECGs, and his results were the same.    

The cardiology reviewers said that "No one strictly follows STEMI criteria. The 4th universal definition mentions ST depression, posterior MI, and T-wave changes."  And the article was rejected.

There is no literature, to my knowledge, of how cardiologists and interventional cardiologists interpret the acute MI guidelines in the 4th Universal Definition of MI.  But anectotally, many follow the millimeter STEMI criteria strictly.   This blog is full of cases in which OMI that did not meet STEMI millimeter criteria were dismissed.

Posterior Reperfusion T-waves

Why do T-waves in V2 and V3 become large with reperfusion? Answer: Vectors.  1) The anterior wall has positive T-wave vectors (pointing anterior).  2) The posterior wall now has a negative T-wave vector (pointing posterior), which is the same as having a positive vector towards anterior leads.  So both vectors add up and result in abnormally large upright T-waves in V2 and V3!

I discovered this years back and call it "Posterior Reperfusion T-waves".  We studied it and wrote this paper: Driver BE. Shroff GR.  Smith SW. Posterior reperfusion T-waves: Wellens’ syndrome of the posterior wall.  

Here are a few examples of posterior reperfusion T-waves: 

A middle aged man with ST depression and a narrow window of opportunity


MY Comment by KEN GRAUER, MD (9/21/2020):


I wish those cardiologists who continue to strictly adhere to STEMI millimeter criteria would begin reading Dr. Smith’s ECG Blog. IF they don’t — they will continue to overlook obvious OMIs that deserve to undergo prompt cardiac catheterization for optimal care.

  • The posterior wall of the LV is not directly viewed by any of the 12 leads on a standard ECG. Posterior leads (ie, leads V7, V8, V9) — have been suggested as a way to enhance ECG visualization of the posterior wall. With acute posterior MI — these posterior leads will sometimes manifest ST elevation not seen on the standard 12 leads.
  • CAVEAT: The amount of ST elevation you are likely to see with posterior leads in acute posterior MI is limited. As a result, the diagnostic utility of posterior leads is (in my experience) limited.

Suggested SOLUTION: Rather than taking the extra time to obtain another ECG with posterior leads (that at best — provides limited information) — GET GOOD at using the Mirror Test as an aid for recognizing acute posterior MI.

  • The Mirror Test is a simple visual aid: It helps the clinician recognize acute posterior infarction. It is based on the premise that the anterior leads provide a mirror image of electrical activity in the posterior wall. By simply inverting a standard 12-lead ECG, and then holding it up to the light — you can easily visualize the “mirror-image” of leads V1, V2, V3.
  • I’ve previously discussed clinical application of the Mirror Test on several occasions (SEE My Comment at the bottom of the page in both the September 13, 2020 post and the February 16, 2019 post in Dr. Smith’s ECG Blog).
  • As we will see momentarily — Use of the Mirror Test makes it difficult to understand HOW a cardiologist could look at the initial ECG in this case and fail to recognize acute OMI.

TODAY’s CASE: The patient in today’s case is a 50-something man who suddenly collapsed — and was found to be in VFib. The patient was defibrillated, and then taken to the nearest ED where ECG #1 was obtained (Figure-1).

  • As per Dr. Smith — there are obvious findings on this initial ECG that are diagnostic of acute OMI.

Figure-1: The initial ECG in the ED — as well as the 3rd tracing shown above, in which leads V4, V5 and V6 are replaced by posterior leads V7, V8 and V9. Mirror-image views of leads V1, V2 and V3 are shown to the right of each tracing (See text).

MY Thoughts on ECG #1: The rhythm is sinus tachycardia at 105-110/minute. The PR and QRS intervals are normal. The QTc looks somewhat prolonged — although this is difficult to assess given the increased rate. The frontal plane axis is normal at +70 degrees. There is no chamber enlargement. Regarding Q-R-S-T Changes for ECG #1:

  • There is a single small and narrow Q wave in lead III. This is of uncertain significance.
  • R wave progression shows normal transition, which occurs between leads V3-to-V4. Of note (and of significance) — the R wave becomes fairly tall (~7 mm) as early as in lead V2 (More on this in my discussion below).

Regarding ST-T Wave Changes:

  • The most remarkable finding in ECG #1 is the dramatic ST-T wave depression. This begins in lead V1 — and becomes marked in leads V2, V3 and V4.
  • Other ST-T wave changes in this tracing are more modest. There is slight ST elevation in lead III — reciprocal ST-T depression in leads I and aVL — and slight ST elevation in lead V6.

Using the Mirror Test in ECG #1:

  • To facilitate recognition of the clinical significance of the ST-T wave depression in leads V1-thru-V3 of ECG #1 — I’ve placed the mirror-image of these leads to the right of ECG #1 (Figure-1). Keep in mind that the mirror-image of these anterior leads gives us a picture of the electrical activity occurring in the posterior wall of the LV.
  • ISN’T the mirror-image view of these anterior leads for ECG #1 all but screaming out, “I’m having an acute posterior OMI” (ie, large Q waves with hyperacute and markedly elevated ST segments)?
  • Although I did not make a mirror-image of lead V4 from ECG #1 — it should be apparent that an inverted lead V4 would look virtually the same as inverted leads V2 and V3 do in the mirror-image view.
  • Returning to my comment above regarding R wave progression in ECG #1: — The rapid development of a surprisingly tall R wave as early as in lead V2 “becomes” a Q wave in the mirror-image view.
  • LEARNING Point: Maximal ST depression in leads V2-thru-V4 (especially when the ST-T waves are shaped as they are in ECG #1) in a patient with new chest pain (or sudden cardiac arrest, as in today’s case) — is diagnostic of acute Posterior OMI until proven otherwise!

Continuing with TODAY’s Case: The 1st tracing in this case to look at posterior leads was ECG #3, obtained 2 hours after ECG #1. As per Dr. Smith: For posterior leads to enlighten interpretation of the initial ECG — “they must be recorded at the same time”That said — several important points can still be made about ECG #3:

  • Note how small the amplitude of QRS complexes and ST-T waves in leads V8 and V9 of ECG #3 are. Even though this tracing was obtained 2 hours after ECG #1 — it would seem almost certain that the amplitude of ST-T wave changes in these posterior leads would not be nearly as large as we see for the amplitude of ST-T wave changes in the mirror-image of leads V1, V2, V3 for ECG #1. Posterior leads are not needed to make the diagnosis of acute posterior OMI in ECG #1.
  • Dr. Smith emphasized that the tall, upright T waves in leads V2, and especially in lead V3 of ECG #3 reflect reperfusion changes. AREN’T these reperfusion changes in ECG #3 much EASIER to visualize (and conceptualize) from the mirror-image view of leads V1, V2 and V3 seen to the right of ECG #3?
  • Dr. Smith also emphasized that the T wave inversion in leads V8 and V9 of ECG #3 provides another indication of acute reperfusion. Comparing the depth of the T wave inversion of leads V8 and V9 in ECG #3 — with the depth of the T wave inversion in the mirror-image view of leads V1, V2, V3 for ECG #3 — ISN’T it much EASIER to visualize (and conceptualize) these ECG changes of acute reperfusion from the mirror-image view? ST-T wave changes that are seen in posterior leads are generally much smaller in amplitude than ST-T wave changes seen in the mirror-image view for leads V1, V2 and V3.

BOTTOM Line from Today’s Case: It is hard to justify the decision by cardiology not to cath the patient in today's case because the ECG “did not show a STEMI”.


  1. We have learned through practice to be humble before the menace that's called ACS. So I will be not the first to cast a stone.Some observations. To the best of my knowledge a nstemi preceded by ventricular fibrillation ,in a patient with a history of high probability for CAD, is an absolute indication for cardiac catheterization . I am not sure how "no stemi , no cath" can stand up to scrutiny and is probably a moment of weakness and bad judgement happened to everyone.
    Also did you not obtain, serial "troponins"? Was an echocardiogram performed or not?
    Plus it was a stemi. 30 minutes after the first ecg the second still showed changes in v1 to v3. Who can decide if an automatic reperfusion is enough in such a patient? And exactly what's the rationale for calling it a nonstemi? Because st changes were not as expressed 30 mins late! I don't think that counts. Furthermore , even if we assume this for a nonstemi not eligible for emergency catheterization ( i think it is eligible for emergency cath), since when we don't "cath" nstemis not today not tomorrow not ever? This event is highly concerning about the way modern hospitals operate.

    1. @ Prodromos — THANK YOU for your comments! I have lived my professional life trying my utmost “not to cast a stone” at others. I taught Family Medicine for 30 years — and primary care physicians are among those WIDE OPEN for potential hindsight judgments. But my role in contributing to this ECG Blog is to “Call a spade a spade” — ALL in the hope of providing constructive feedback so that future practice improves. It’s a TOUGH line to walk trying to do so without calling into question certain decision-making.

      That said — we would be remiss if we didn’t call attention to the mistakes in this case that should not be made. Referral to the interventionalist should not have been refused in this patient who presented post-cardiac arrest with an initial ECG showing the dramatic ST-T wave depression that we see in ECG #1. Though “not a stemi” (by the millimeter-definition) — all cardiologists (in my opinion) should recognize that dramatic ST-T depression localized to leads V2-through V4 (as in ECG #1) in a patient with either new chest pain or an episode of VFib have a STEMI-Equivalent pattern corresponding to acute Posterior MI. This HAS To BE assumed an acute OMI ( = acute coronary occlusion) until proven otherwise. And as you say — even if this wasn’t a stemi-equivalent — every cardiologist in 2020 should by now appreciate (and accept) that there ARE other reasons that serve as a clear indication for prompt cath, even when there no stemi is present.

    2. So, we agree on this. I am a "general cardiologist" from Greece. I would like to express my deepest gratitude to you. Being a content creator, and of this qualiity and caliber , it's huge. I have been enjoying your content for some time now. But it is with this I have a chance to say thank you.

  2. Excellent case vignette. I must admit I always believed (as illustrated in your post 2 weeks back)that posterior wall mi is characterized by R/S>1,st depression and tall upright t waves in v2, v3.but if I understand you correctly,the interpretation(both on direct and mirror image)depends on St segment. Also, in the previous case,t wave was positive in both acute and reperfusion that case, how do you interpret reperfusion?

    1. Hi Subhasish — IF I understand your comment correctly, you are referring to our Sept. 13, 2020 post ( ). If this is not the post — please SPECIFY which post you are referring to.

      The diagnosis of acute Posterior MI is made on ECG by a number of factors — not the least of which is potentially consistent timing with respect to the clinical history. Although the finding of a narrow complex, but tall R wave in lead V1 (or taller-than-expected R wave in leads V2,V3) can be helpful — this is neither that sensitive nor specific a sign for acute posterior MI. It may assist as a supportive finding.

      Most posterior MIs are accompanied by ST-T wave changes in other lead areas (often in lateral chest leads with LCx occlusion — or inferior leads with RCA occlusion). But you CAN have “isolated” (or relatively isolated) acute posterior MIs.

      You are correct that in general, it is the amount, shape and distribution of anterior ST depression (as illustrated in the mirror-test images I’ve made on various posts) that helps most. ST depression tends to be maximal in leads V2,V3 — extending to V4 — which distinguishes posterior MI ST-T change distribution from diffuse subendocardial ischemia that is more generalized without maximal ST depression in V2,3,4. Depending on timing, interventions undertaken, and ongoing patient history — positive T waves that are often peaked suggest reperfusion changes. As I showed in the mirror-image views for ECG #3 in today’s case — the “shape” of these reperfusion T waves in anterior leads looks ischemic (symmetric) when these leads are inverted.

      I hope the above addresses your question — :)

  3. Hello. Great series here.

    How would you differentiate a posterior MI from a NSTEMI in anteroseptal leads? Wouldn't both have ST depressions in v1-v3?

    1. @ Imran — THANK YOU for your Question! The KEY is that with a posterior OMI ( = Occlusion-based MI) — there is LOCALIZED ST depression (usually in leads V2-thru-V4 — though may begin in V1) — vs with a NSTEMI — in which case the ST depression will not be as localized. Today’s case is a beautiful example of a posterior OMI with localized ST depression in leads V1-thru-V4 in ECG #1 (maximal in V2,V3,V4) + the positive “Mirror” Test. Note in ECG #3 (in the Figure-1 of My Comment, above) how the reperfusion ST-T wave changes (in this case, T wave peaking) are also localized. You won’t see that in a NSTEMI. (NOTE: Many “NSTEMIs” were really OMIs that were MISSED by clinicians “stuck” in the NSTEMI paradigm (See My Comment in the July 31, 2020 post = — as well as MANY other of our posts that show clinical examples of this) — :)


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