Saturday, March 24, 2018

You have two hours to save this patient's life

Written by Pendell Meyers, edits by Steve Smith

A female in her 60s with history of CAD s/p PCI and CABG, alcohol abuse, and recurrent pancreatitis presented at 14:55 complaining of sudden onset epigastric pain. Initial vital signs were heart rate 44 bpm, respiratory rate 16, BP 143/67, SpO2 96% on room air. On initial exam she was in mild distress and complaining of severe nausea.

Here is her initial ECG:
What do you think?

There is decreased ECG quality due to baseline movement. Despite this, there are clearly hyperacute T-waves in lead III with reciprocal negative hyperacute T-waves in aVL (and lead I) with likely a small amount of STD in aVL. This is diagnostic of acute transmural inferior acute MI, with the most likely etiology being acute coronary occlusion. There is clear STD in V2 indicative of posterior involvement. Sinus bradycardia despite critical illness also points toward inferior acute coronary occlusion (as the SA and AV node are generally supplied by the same vessel as the inferior wall).

Here are two immediate repeat ECGs performed in efforts to get less baseline wander and artifact:

The same findings are evident throughout, but with baseline movement obscuring various sections. There is even a very small amount of STE present in III on the third ECG.

The emergency physicians activated the cath lab. Apparently this decision was made only on perceived STE in lead III on the last of the ECGs above.

This happened to have been the 6th emergent cath lab activation called by the ED within the past 4 hours (!). The cardiologist came to bedside and deactivated the cath lab. His opinion was that the ECGs did not meet STEMI criteria, that she was probably not suffering from ACS given her pain was epigastric and he interpreted the physical exam as showing epigastric tenderness to palpation.

Initial labs returned showing a first troponin negative.

At approximately 15:30, the nurse called for a physician to bedside because the patients blood pressure dropped to 89 systolic and she became lightheaded.

Here is the repeat ECG at that time:
Massive STD V1-V5, maximal in V2-V3, diagnostic of posterior acute MI. Interestingly, the inferior hyperacute T-waves are no longer present. There is obligatory STE in aVR generated by the large amounts of STD elsewhere.  

ST Elevation in aVR in the setting of inferior MI should also tip you off to possible right ventricular MI (RV MI) and to the need for a right sided ECG. In fact, it is a good idea to do a right sided ECG in every inferior MI, as there is no finding that is 100% sensitive for RV MI on the standard 12-lead ECG.

A possible explanation is that the thrombus in the proximal segment of the vessel supplying the inferior wall may have embolized distally and occluded the portion supplying only the posterior wall. The problem with this hypothesis is that occlusion of the proximal part of the vessel should also have caused posterior ischemia at that time. It is possible that the first ECGs were too early to detect this posterior ischemia. This theory is not provable without cath, but is an anatomic possibility. We do not have records to indicate if her native RCA is patent, or what graft supplies the inferior wall if applicable.

Based on this ECG, the cath lab was reactivated with concern for posterior acute coronary occlusion.

The cardiologist came back and this time took the patient immediately to the cath lab. However, the cath lab was occupied by at least one of the prior 5 "STEMI activations" from earlier in the day, thus there was a delay in which the patient was kept in the cath holding area.

At approximately 1600, the patient acutely decompensated even further. She was rushed into the cath lab, where she lost her pulses just as she was placed on the cath lab table, suffering PEA arrest. ACLS was initiated. The notes state that a transvenous pacemaker was attempted but "never entered her heart" for unclear reasons. No left heart cath was performed. I cannot find an explanation as to why this patient was not an ECMO candidate. She was pronounced around 1700.

It is overwhelmingly likely that this patient died from acute coronary occlusion, but I do not have an angiogram to prove it. Autopsy was not performed.

Learning Points:

1) You must advocate for your patients, as "subtle" findings such as hyperacute T-waves are simply not yet recognized by a wide range of providers as an early finding of acute coronary occlusion (the event for which "STEMI" is a very poor surrogate term)

2) Repeat ECGs are important.

3) Bradycardia in the setting of any critical illness must prompt immediate consideration of hyperkalemia, acute coronary occlusion (usually inferior), and primary rhythm disturbance, among other less common and less reversible causes.

4) This patient, and many others with similar initial ECGs, is an unfortunate victim of a broken paradigm for acute MI management: STEMI vs. NSTEMI. This is a false dichotomy that inspires ignorance to anything except the ST segments. The "STEMI criteria" seem to have been one of the barriers to emergent catheterization which was not overcome in this case, ultimately leading to the patient's death. It is very likely that this patient could have been saved in the two hours from initial ECG to cardiac arrest.


  1. thanks for these impressive ECG's, there very helpfull, i'll share them on my own (dutch) blog with your permission

  2. This ECG with the obs and patient presentation immediately said to me this patient was having an inferiopoInferioposterior MI especially with the STE in III and mirror image STD in AVL . STD in V2 significant for posterior involvement but the obs and patient presentation really tell a story beyond the alcohol issues they are confirming the ECG.

  3. Tenderness in the epigastrium
    Can you tell more about this entity in ACS patients?

    1. This is most likely a consultant finding something that does not exist to confirm his bias.

  4. Thanks for the "courage" in publishing this unfortunate case


  5. Tragic but highly insightful case … I’ll add the following thoughts to the comments by Drs. Pendell Meyers & Stephen Smith. i) The ECG reflects the net effect of events. We are not told if serum electrolytes were normal in this 60yo woman with a history of alcohol abuse, who presented with nausea and abdominal pain. This is relevant, since if serum K+/Mg++ were reduced — that might account for the less-than-expected ST-T wave changes that we see in initial ECGs (which perhaps, might in part be attenuated by the effect of electrolyte abnormality on ST-T waves …). ii) Virtually all leads on the initial ECG contribute to a cohesive story of likely acute RCA occlusion. That is, not only lead III — but also inferior leads II and aVF show hyperacute-looking T waves that are clearly more prominent-than-they-should be given abrupt angulation at the point of onset of the T wave, following a relatively flat ST segment. There is mirror-image ST-T wave depression in leads I and aVL to what is seen in lead III. In the chest leads — acute posterior involvement is supported by the appearance of lead V2, which shows abrupt development of a taller-than-expected R wave, with ST-T depression that is also the mirror-image of the ST-T wave in lead III. In light of lead V2 — the flat ST segment in lead V1 may herald acute RV involvement. ST-T waves in leads V4,V5 are distinctly flat — and in this light, especially considering the small amplitude of the R wave in lead V6 — this V6 T wave also looks hyperacute. iii) I’ll propose an alternative explanation for the reason why inferior hyperacute changes may have been “lost” in the last ECG — namely, point i) above = the ECG reflects the net effect of events. I interpret this last ECG as suggestive of diffuse subendocardial ischemia — since compared to initial ECGs, there has been a GENERALIZED increment in J-point ST depression in virtually all leads, with exception of increased ST elevation in lead aVR. What would have been ST-T wave depression in the inferior leads and lead V6 has been attenuated by hyperacute T waves that had been present in these leads. Perhaps acute decompensation was precipitated by shock from associated acute RV involvement? In any case, humbling Lessons to be learned in the Learning Points highlighted by Drs. Meyers & Smith.

  6. Wow! Stephen Smith and Ken Grauer in one blog. Who needs cardiologists.

  7. I think your description of the circumstances surrounding the case is probably highly pertinent to the decision not to go to the cath lab and the confirmation bias which seems to have informed it. That first ECG is just so barn door for 'acute, severe badness happening in coronary arteries'.

    1. David,
      I'm not sure what you are trying to say exactly.
      Steve Smith

  8. Steve...

    Excellent cases and great contributions by those commenting. I would like to suggest that one be very careful assuming that - in cases of acute inferior epicardial ischemia - ST depression in Leads I and aVL are BOTH reciprocal to the inferior STE. aVL is 150 degrees from Lead III and you can't get more reciprocal than that in the frontal plane. However, the Lead I axis is perpendicular to aVF and really bears very little to no real reciprocity at all to Leads II or III. So, ST depression in Lead I - although reciprocal - is usually not reciprocal to the inferior changes but rather to something else that is closer to 180 degrees away: the right ventricle. Also, there is marked ST depression in V2 - V4 but relatively little STD in V1. That was most likely due to the cancellation of forces caused by the competition of STE caused by a RV infarction with the reciprocal changes to the posterolateral wall infarction that would have been seen in V7 - V9. Check other ECGs with acute inferior MIs (and no right ventricular MI) and see how many have reciprocal changes in Lead I. Typically, you will just see the reciprocity in aVL.

    1. Jerry,
      Lead I is 120 degrees from lead III and frequently there is reciprocal ST depression in lead I in inferior MI that is NOT due to RV MI. It is correlated with RCA vs. circ.
      I have heard the contention that lead I STD is due to RV MI and disagree with it. And so I have therefore gone ahead to study it (manuscript is under consideration).
      We found that, in 150 consecutive inferior MI due to RCA as the infarct artery, lead I ST depression had ZERO correlation with occlusion proximal to, vs. distal to, the RV marginal branch.
      We also studied the value of STE, vs. none, in lead V1. We hypothesized that ST elevation in lead V1 would be specific for RV MI, and that it would also be sensitive for RVMI if there was no ST depression in V2. However, although the specificity was indeed good, the sensitivity was not so good.
      So we now have data with which to conduct this conversation.
      I'm not sure when it will be accepted for publication.

    2. Steve...

      I was not aware of your research. I look forward to the article.


  9. Hi Steve, long time follower from Slovenia, this stuff is as usual invaluable, bravo.
    May i suggest your team in future post ECMO blog case, some indications, criteria, what should be tried before it, etc

    1. dr. john,
      Good timing, because we are about to start doing ECPR for out of hospital cardiac arrest. Patients who have witnessed arrest, no signif comorbidities, shockable rhythm, not responding to 3 shocks, Epi, and amio, will have ECMO placed immediately upon arrival in the ED and then go to angiogram.
      When we have one, I'll post it. But there won't be much in the way of 12-lead ECGs, as the patient will still be in V Fib.


Recommended Resources