Another Cardiac Arrest - Is it OMI this time? Use your skills from the previous post!

Written by Pendell Meyers, with edits by Steve Smith:

A 49 year old male with history of HTN suddenly complained of shortness of breath, and walked down the street to the fire department, where he walked in the door and collapsed. The fire department crew on scene immediately recognized cardiac arrest and began CPR within seconds. Initial rhythm was VF, shocked into pulseless VT. En route, paramedics shocked him out of VF/VT 5 times, with an intermittent wide complex rhythm between episodes of VF/VT. He arrived in our ED with a perfusing rhythm, intubated and without any spontaneous movements.

Here is his initial 12 lead ECG, approximately 10 minutes after his most recent ROSC:

What do you think?

The computer wrongly diagnosed left bundle branch block, and had no comments concerning ischemia. The computerized QRS duration is stated as 138ms.

The interpretation of this ECG hinges completely on correct understanding of the QRS complex and location of the J-point. Your eyes are drawn to V2-3 and you are tempted to believe that the entire shape represents the QRS complex. If that were true, the QRS complex in V3 would be almost 300ms long! This is not possible, or so unlikely that we must consider it incorrect until proven otherwise.

In reality, the computer is actually correct about the QRS duration of ~138 ms. If you search carefully for the J-point (see annotated ECG below), you'll see there is actually a massive amount of STD in V1-V5, and there is STE in V6, I, and aVL.

Like the case from yesterday, the end of the QRS is obliterated by massive ST segment deviation, mimicking LBBB in the precordial leads.

Should you consider hyperkalemia? Yes of course, as it is wide and bizarre to the less trained eye. So give some calcium if you can't tell, while your resuscitating and sorting it out. If it's hyperkalemia, calcium will usually produce rapid change in the ECG.

With the correctly identified J-point seen above, there is obvious and massive STD in V1-V5, with massive STE in I and aVL and reciprocal STD in II, III, and aVF. This indicates acute and total ischemia of the posterolateral walls.

Remember, this is an immediately post-ROSC ECG. This means that you must consider whether these changes are due to the immediate post-ROSC state, or not.

In my experience, the magnitude and anatomically logical distribution of these ST segment changes makes it less likely to be simply post-ROSC abnormalities, and much more likely to represent true full thickness focal ischemia due to an occluded artery (OMI).

There are still two possibilities:

1) A focal fixed stenotic lesion, combined with the low flow state of cardiac arrest, creating such poor flow through the stenosis that there is effectively an occlusion during the down time.

2) An acute coronary occlusion (OMI).

Obviously the clinical history of this case points toward true OMI. But in many cases it will be more confusing. If you are unsure, get a repeat ECG 20-30 minutes after the initial post-ROSC ECG. If it was #1 (combo of fixed lesion with temporary low flow during down time), then the changes will improve or disappear. If it was #2 (OMI) and the artery is still occluded, the changes will persist.

Here is the ECG 23 minutes after the initial ECG:

Practice finding the J-point and the ST segment deviations before looking at the annotated ECG below.

Annotated ECG showing persistent massive ST deviations, including anterior STD and high lateral STE indicative of persistent posterolateral OMI. 
Conveniently for our education, there are 4 PVCs spaced perfectly to show us that we can see analogous ST changes within the PVCs in all 12 leads, including proportionally excessive discordant ST segment deviations in V5-6, I and aVL, II, III, and aVF, and aVR. There is concordant STD present in the PVCs in leads V3 and V4. The J point is inappropriately isoelectric in V2.

The ECGs are diagnostic of posterolateral OMI.

Unfortunately, the ED physicians and cardiologists all believed these ECGs represented LBBB. It is even documented that "no Sgarbossa criteria are present." (Regardless of the fact that the Sgarbossa and modified Sgarbossa criteria are inapplicable in this case, they are indeed both present). We believe from experience that the modified sgarbossa criteria can be applied to PVCs. This case supports our claim, because almost all 12 leads have a PVC with either concordance or excessive discordance.

Luckily for the patient, the team decided to take the patient to the cath lab because he had complained of SOB before collapsing in front of the firemen and suffering a VF cardiac arrest. The initial positive troponin T of 0.22 (normal less than 0.01 ng/mL) also helped push the patient appropriately to cath lab.

In the cath lab they found severe chronic 3 vessel disease, including chronic appearing stenoses of the left main (40%), proximal LAD (90%), "chronic total occlusion of all diagonals", proximal LCX (75%), mid RCA (80%), with an acute thrombotic 100% occlusion of the mid-distal LCX (the culprit, obviously). A stent was applied to the LCX culprit lesion, with TIMI 0 flow improved to TIMI 3 flow after intervention.

Here are the angiogram images:

The RCA is overall patent but has a lesion at the level of the mid-RCA.

Annotated, with arrow showing the mid-RCA lesion.

This appears to be a diffusely diseased but overall currently patent LAD. There are few to none visible diagonals.

The left main can be seen to bifurcate into the LAD (running horizontally across the top of the image) and the  LCX (running vertically down the middle of the image). The LCX shows several points of stenosis and then is abruptly occluded. See next image for annotation.

Arrow showing the site of the acute thrombotic LCX occlusion.

The wire has crossed the occlusion, preparing to perform the intervention.

After intervention, TIMI 0 flow has been restored to TIMI 3 flow, exposing the long course of the LCX which was previously occluded.

Another view post-intervention.

Here is his ECG about 18 hours later:




His troponin T peaked at 11.78 at around 14 hours after presentation. This indicates a very large MI and portends much worse prognosis.

Next day ECG:



Unfortunately, the patient suffered anoxic brain injury. I do not have the details to explain why this occurred in a patient with short down time, immediate bystander CPR, and appropriately timed reperfusion therapy. The patient ultimately died 2 weeks later.


The ECG pitfalls in this case are common, especially among the sickest patients with OMI. The STEMI vs. NSTEMI paradigm does not prepare us for cases like this. The STEMI vs. NSTEMI paradigm fails even in those with normal-appearing QRS complex, and fails even more as the QRS complex becomes more and more abnormal. Only advanced ECG interpretation and training will help these patients in the cases when the clinical history does not already provide the correct management.


Learning Points:

Finding the end of the QRS complex and the J-point is a basic but absolutely crucial skill.

If you see what appears to be an extremely wide QRS complex (greater than 200ms), then consider hyperkalemia and/or incorrect measurement of the end of the QRS (which is most importantly caused by obliteration of the QRS complex by ST segment deviation)

You must be able to recognize posterior STEMI/OMI, despite the fact that it is downplayed by the STEMI vs. NSTEMI paradigm, and even more downplayed in actual practice by some physicians (EM and cardiology).

Look for supplementary evidence of OMI within PVCs.