Trust a computer read of "Normal ECG" at your peril!

This case was contributed by Brooks Walsh, an emergency physician in Connecticut.

Don't trust the "Normal ECG"

It's important to periodically reemphasize that the computerized ECG interpretation  can miss critical findings.  

Some authors have suggested that ECGs interpreted as "normal" by the computer are "unlikely" to be significant. However, "unlikely" is doing a lot of the heavy lifting in that conclusion!

The challenge of emergency medicine, after all, is to churn though the sea of benign presentations, looking for that "unlikely" crucial diagnosis. If you "outsource" this vigilance to the computer, you might miss that critical presentation.

The Case

An older person came to the emergency department with chest discomfort. It was not severe, but it was uncomfortable, and had not improved after an hour.

An ECG was obtained immediately at triage:

 

The computer program read this as "Normal Sinus Rhythm, Normal ECG." Can we trust this?

Probably not in this case.

First of all, there's some pretty suggestive ST elevation in the anterior leads...

... along with reciprocal ST depression in lead III. 

 

Seems pretty clearly to be an anterior occlusion myocardial infarction (OMI), but some folks might worry it "doesn't meet STEMI criteria" since the STE in V2 doesn't clearly exceed 2mm.

 

Further evidence, though, is provided by the subtle-STEMI calculation in MDCalc. Using the QTc of 421 ms provided by the computer... 

 

... we get a score of 20, quite above the threshold of 18.2.

You can read all about this formula here:

12 Example Cases of Use of 3- and 4-variable formulas to differentiate normal STE from subtle LAD occlusion

So we're concerned about the ECG - How did the patient do? Well, although the cardiac team was immediately activated, there were issues with moving to the cath lab directly. 

 

Interestingly, the patient reported that the chest pain had almost totally resolved. About 15 minutes after the first ECG was recorded, a second ECG showed what the computer (again) reported as a normal ECG.

 

Despite the computer's optimistic interpretation, however, subtle anterior STE is still seen, and subtle ST depression persists in lead III.

Fortunately, the patient was able to soon proceed to the cath lab, now only reporting 1/10 chest pain. Given the improved symptoms and a "normal" ECG (followed by a "normal-er" ECG), one might expect the angiography to find only a minor or partial obstruction.

Nope.

Angiography showed a TIMI 0 lesion in the mid-LAD. Following placement of 2 stents, TIMI 3 flow was restored.

  

On the ECG subsequent to the PCI, the ST-T can finally be interpreted as normal (but not the QRS -- there are little Q-waves in V1 and V2)!

 

Lesson: 

Even if the ECG shows only subtle signs of of acute coronary occlusion, and even if the computer calls the ECG "normal," and even if the symptoms are seemingly improving, a significant portion of the myocardium can be at risk!