Monday, November 5, 2012

ECG Computers Are Not Infallible

We are always told to read our own electrocardiograms. Though each ECG printout has a reading at the top as interpreted by a machine, we're advised to read it with our own critical eyes and make our own judgements. However, it is all too easy to quickly look at that computer interpretation while we are scrambling to get the next patient into the operating room instead of poring through every single lead on the page. Sure the computer interprets "antero-septal infarct, age undetermined" so often that we don't even bother to check if it is correct. As long as there a reasonable correlation between the reading and the the wave forms, we feel adequately informed.

The ECG above however is so egregiously off that I felt compelled to use it as Example A as to why we shouldn't rely on machines to do our readings for us and make clinical judgements based on them. As you can see, the computer read this ECG as "Sinus tachycardia with 2nd degree A-V block with 2:1 A-V conduction". That would be a pretty impressive ECG, almost like the kind they use to test us during ACLS. But even a cursory glance shows how off base the reading is. First of all, it's nowhere near being sinus tachycardia. As the printout shows a little to the left, the heart rate is only 61 beats per minutes. Then of course there is no evidence of a 2nd degree A-V block. These are all just overzealous interpretations made by the ECG computer. That is why an actual human interpretation is still essential in making clinical judgements in medicine. Even if it's just a disgruntled cardiologist making $20 for using his years of training and expertise to provide this critical life saving service.

No comments:

Post a Comment