I “over-diagnosed” an ECG. Maybe you should too.

It’s important to be wrong now and then. Not just for the usual blather about being humble, understanding cognitive biases, or even nailing the Kobayashi-Maru test.

Don't do it...

Don’t do it Saavik…

No, it’s important to be wrong in the right sort of way, a willingness to be humble in the interest of patient care. Let me explain!

Case #1: I was sooo right.

EMS brought in a middle-aged male who was “found on the floor,” having been their for an unknown period of time. Their medical history and medications were also unknown, and his altered mental status didn’t help. Vital signs were okay, although the heart rate was unexpectedly low for someone who looked sick and dehydrated. While my resident was examining the patient, I talked with Sara, the paramedic, about the ECG.

EMS pseudo hyper k

“Huh,” I said to Sara, “funny it’s so slow, since he looks dry as dust. T-waves also look a bit funky – I wonder about hyperkalemia. Hey, don’t be afraid to empirically treat if the history and ECG make you suspicious.”

We grabbed our own ECG in the ED:

pseudo hyper k 2

ACUTE MI? Wat R U doin computer stahp

This was also supportive of hyperkalemia, and so I pushed calcium before waiting for the lab results. The potassium turned out to be 6.3.  #JediECGskillz   #AmIGoodOrWhat

Case #2:   #OrWhat

An elderly female with no prior ECGs, or records of any sort, was brought in by EMS with a report of “altered mental status.” She actually seemed mostly okay to me, but the veteran paramedic, Chris Lovell from Norwalk, showed me the ECG:

paced rhythm t slope

Now, I know what you’re thinking…” Yeah, it was paced, so some of you might say “you can’t tell anything from the ECG.” Probably should have listened to you! Two points in my defense however:

First, there are a number of case reports of hyperkalemia manifesting in paced rhythms. See here, here, and here, for example.

TL;DR? You might see loss of pacer capture, or significant QRS widening compared with an old ECG. Some of the better examples:


Second, did I mention I have Jedi-like skills in detecting hyperkalemia?


I proceeded to personally push 3 grams of calcium gluconate, and rechecked the ECG to document my “win.”


PRESS THIS NOW for the musical accompaniment.

No changes whatsoever. The potassium was completely normal.

Okay, so I was wrong. But was it a capital-F “Fail?” (Of course, my short answer is “No.”)

Here’s the longer answer why this was NOT a fail.

If an ER doctor tells you, with pride, that their accuracy in diagnosing STEMI is 100%, since they have never sent a “false-STEMI” to the cath lab, then they are either (best-case scenario) lying, or they are (worst-case scenario) very bad at their job.

If you never send a “false-STEMI” to the cath lab, it means that you are probably NOT picking up on a bunch of “true-STEMIs.” If you aren’t taking a chance on the small or subtle STEMIs, then you might be hurting patients.

The surgeons have understood this about appendicitis for generations. At least before CT scans started being used, a good surgeon was defined by the number of “negative” appendectomies you performed:

  • Too many meant you were too quick to cut, and had no sense of clinical judgment.
  • Too few meant you were missing “true cases,” letting them perforate & get septic.

We could view empiric treatment of hyperkalemia like that – if you aren’t overtreating at least sometimes, then you are probably missing critical chances to treat a potentially lethal condition. And unlike going to the OR or the cath lab, the EMS therapy (calcium chloride) is pretty benign.

So perhaps we should take pride in over-reacting (within your guidelines and protocols, of course). Maybe we could start to track our “negative-K” rates, and even start to brag about them!


  • Ryan says:

    Good job Doc. That last paced case looks like Sgarbossa B and C. In paced rhythms , Sgarbossa C is about 90% specific in identifying STEMI. There is C in Lead III and B in V2. I would have sent this pt. To cath as well or at least had cards consulted.

  • NICE post Brooks! I agree completely with the point you make — which is an important lesson for us all to learn.

Leave a Reply

Your email address will not be published. Required fields are marked *