Bizarre (Hyperacute??) T-waves

Thanks to one our great HCMC nurses, Ryan Burch.  He figured this one out.

A dialysis patient presented with dyspnea.  He was a bit fluid overloaded and not hyperkalemic.

 This ECG was recorded:

This was sent to me in a text that woke me from sleep, but not simultaneous with patient care.
Truly bizarre T-waves in I, aVL, III, aVF, aVR
Lead II is unremarkable, and leads V3-V6 are also slightly bizarre.

What do you think?

My answer, as I looked with bleary eyes at my phone: "I have to say I've never seen this one before."

Later, I looked into the chart and found an ECG from a few days before:

I texted back:
"Those T-waves were gone 5 minutes later.  Artifact!"

Ryan Burch, RN, was the nurse caring for the patient, later sent me the same ECG, stating the following:

"This ECG had people stumped and concerned but I read an article in www.ecgmedicaltraining.com (see below) about an artifact a few weeks prior which I thought looked similar and the suggestion was that a lead had been placed over an artery.  I switched lead placement and this ECG was recorded 5 minutes later:"

He found that the left arm electrode had been placed near the patient's left arm dialysis fistula, which was pulsating with a palpable thrill.

This resulted in the following:

Wandering Artifact only

Explanation

All leads are derived from 3 bipolar electrodes and one unipolar electrode.
Leads I, II, and III depend on bipolar leads voltage differences:
--Lead I uses the right and left arm
--Lead II uses the right arm and the leg
--Lead III uses the left arm and the leg.
--The Wilson (or Goldberger) Central Terminal is used to produce the augmented (a) leads:
aVR, aVL, aVF.

The voltages are calculated as follows (Thanks to Ken Grauer for sending these):

• I = L - R
• II = F - R
• III = F - L
• aVR = R - (L + F/2)
• aVL = L - (R + F/2)
• aVF = F - (R + L/2)

As you can see, the only lead that does not use the left arm electrode is lead II.  Since lead II is the only normal lead in this ECG, the left arm electrode must be the affected electrode.  Indeed, the patients dialysis fistula was on the left arm and was pulsating with each heart beat, moving the electrode and causing artifact.

Arterial pulse tapping artifact

https://www.aclsmedicaltraining.com/blog/guide-to-understanding-ecg-artifact/

This online article references the article below by Emre Aslanger, a great guy who occasionally corresponds with me about ECGs.

Aslanger E, Yalin K. Electromechanical association: a subtle electrocardiogram artifact. Journal of Electrocardiology. 2012;45(1):15-17. doi:10.1016/j.jelectrocard.2010.12.162.

Incredibly, this case was just published in Circulation on January 22, 2018 (thanks to Brooks Walsh for finding this!) 
Asymptomatic ST-Segment–Elevation ECG in Patient With Kidney Failure.   https://doi.org/10.1161/CIRCULATIONAHA.117.032657.  Circulation. 2018;137:402-404.Originally published January 22, 2018

Here is a case from Circulation year 2000 that was misdiagnosed as due to pancreatitis.  But you can tell from the normal lead III that this was a right arm electrode problem:
http://circ.ahajournals.org/content/101/25/2989.full

It is full text!! 

Why is there also artifact in precordial leads?
Aslanger explains:

“[O]ne may expect that the leads not connected to the electrode affected by the source of disturbance would be free of distortion; but this is not the case. When one of the limb electrodes is affected by a source of disturbance, it distorts not only the corresponding derivation but also [the others] which are all calculated by mathematical equations…”

“…precordial leads [are also affected] because the Wilson central terminal, which constitutes the negative pole of the unipolar leads, is produced by connecting 3 limb electrodes via a simple, resistive network to give an average potential across the body.”