Sunday, March 20, 2022

I was reading EKGs on the system and came across this one

I was reading through EKGs on the system and came across this one:


Computer interpretation:


Preliminary result by Interface

SINUS RHYTHM
NONSPECIFIC T-WAVE ABNORMALITY
BORDERLINE ECG


What do you think?







Look in Lead V1 and you will see fully upright atrial waves.  They are regular and are exactly associated with the QRS in a 3:1 ratio.  It is very slow atrial flutter, at a rate of around 200.  

True sinus P-waves are not upright in V1; they are biphasic up-down.  

Atrial flutter waves are usually upright in V1.  This ECG shows atrial flutter with 3:1 conduction.  The 3:1 conduction implies AV conduction defect, which could be intrinsic or due to medications or hyperkalemia.  

I looked into the chart and it was from a patient who was admitted with sepsis.  K was 4.6 and she was not on an AV nodal blocker; therefore, she likely has some AV node disease.

I called the physician caring for the patient.  She ordered another ECG.  By the time I saw the repeat (2nd), it had already been overread by a physician.  Here it is:

Now, look in lead I.  The same pattern is present.  
However, this time the QRS is irregularly irregular.  
So although lead I suggests atrial flutter with 3:1 conduction, this is probably atrial fibrillation. 
 
The overreading physician was influenced by the computer interpretation to diagnose sinus rhythm.

Computer interpretations are proven to badly influence the overreading physician.

That is why I always recommend reading the ECG first, THEN looking at the computer interpretation to see if it found something that you missed.

Whether it is atrial flutter or fibrillation is not of great importance.  They require the same management: rhythm control, assessment of AV node function, and evaluation for embolic risk and prophylaxis with an anticoagulant.


Later, this was recorded:

Atrial fibrillation


So the patient got a new diagnosis of atrial fibrillation.  An important diagnosis that should not be missed.


See this full text article I wrote in 2019: 

Smith SW et al. A deep neural network for 12-lead electrocardiogram interpretation outperforms a conventional algorithm, and its physician overread, in the diagnosis of atrial fibrillationInternational Journal of Cardiology.

See this statement: "When the conventional algorithm (Veritas®) was incorrect, accuracy of Veritas® + physician was only 62% (CI 52–71)"  In other words, the physician overread was adversely impacted by the computer interpretation.  

ABSTRACT

Background

Automated electrocardiogram (ECG) interpretations may be erroneous, and lead to erroneous overreads, including for atrial fibrillation (AF). We compared the accuracy of the first version of a new deep neural network 12-Lead ECG algorithm (Cardiologs®) to the conventional Veritas algorithm in interpretation of AF.

Methods

24,123 consecutive 12-lead ECGs recorded over 6 months were interpreted by 1) the Veritas® algorithm, 2) physicians who overread Veritas® (Veritas® + physician), and 3) Cardiologs® algorithm. We randomly selected 500 out of 858 ECGs with a diagnosis of AF according to either algorithm, then compared the algorithms' interpretations, and Veritas® + physician, with expert interpretation. To assess sensitivity for AF, we analyzed a separate database of 1473 randomly selected ECGs interpreted by both algorithms and by blinded experts.

Results

Among the 500 ECGs selected, 399 had a final classification of AF; 101 (20.2%) had ≥1 false positive automated interpretation. Accuracy of Cardiologs® (91.2%; CI: 82.4–94.4) was higher than Veritas® (80.2%; CI: 76.5–83.5) (p < 0.0001), and equal to Veritas® + physician (90.0%, CI:87.1–92.3) (p = 0.12). When Veritas® was incorrect, accuracy of Veritas® + physician was only 62% (CI 52–71); among those ECGs, Cardiologs® accuracy was 90% (CI: 82–94; p < 0.0001). The second database had 39 AF cases; sensitivity was 92% vs. 87% (p = 0.46) and specificity was 99.5% vs. 98.7% (p = 0.03) for Cardiologs® and Veritas® respectively.

Conclusion

Cardiologs® 12-lead ECG algorithm improves the interpretation of atrial fibrillation.





===================================

MY Comment, by KEN GRAUER, MD (3/20/2022):

===================================


Insightful case presented by Dr. Smith that conveys a number of important teaching lessons. I'll add the following points.
  • The ECG diagnosis of AFib (Atrial Fibrillation) — can at times be deceptively challenging. AFib is an easy diagnosis to make when the rhythm is obviously irregularly irregular — without P waves — and, with coarse "fib waves". That said — this is not always what we see.
  • On occasion — I have found me asking myself, "Is this rhythm truly AFib?" — because the rhythm is not all that irregular, coarse "fib waves" are absent, artifact may be present, and deflections that almost look like P waves may be seen. Such is the case with the 2nd ECG in today's case (ECG #2 in Figure-1) — in which except for the relative pause at the beginning of the rhythm and early beat #5 — the rhythm is not all that irregular.

 

Figure-1: For clarity — I've put the first 2 tracings in today's case together (See text).


In today's case — the correct rhythm diagnosis was initially overlooked for both of the tracings shown in Figure-1. In each case — the computerized report said "sinus rhythm" — which apparently influenced the treating clinicians.
  • It is hard to believe that the computer is calling the rhythm sinus for both of the tracings in Figure-1. This just goes to show How Wrong the computer can be! But we should never be depending on the computerized interpretation to begin with!
  • There are Pros & Cons to use of the computerized report (Please see Figure-2 and Figure-3 below from my suggested approach on this topic).

  • The BASICS: Never look at the computerized interpretation until after you have made your own independent assessment! In general — the computer is excellent for calculating rates, axis, and intervals. In my experience — it is terrible for assessing any rhythm other than sinus! The computer usually does OK for assessing normal tracings — but it may overlook certain findings (such as subtle ST elevation and hyperacute T waves, to name just a few). 
  • BOTTOM Line: The computerized report may help by suggesting certain findings that you may not have thought of — BUT — you must always overread what the computer says. IF you disagree with what the computer says — then cross out those parts in the computer report that you disagree with!

QUESTION: 
  • WHY was the rhythm misdiagnosed for both tracings in Figure-1?


ANSWER:
A systematic approach was not followed.


There are many systems for 12-lead ECG interpretation. Regardless of the system you choose to use — the KEY is to always begin by assessing Rate and Rhythm.
  • The easiest way to avoid ever again calling a rhythm sinus when sinus rhythm is not present — is to always spend the first 3-to-5 seconds of your interpretation of every 12-lead ECG you see by finding the long lead II rhythm strip — and looking in front of each and every QRS complex. IF an upright P wave with constant PR interval is not present in lead II — then unless there is dextrocardia or lead reversal, the rhythm is not sinus.
  • In ECG #1 — there is no clear upright P wave in lead II. Therefore the rhythm in ECG #1 is not sinus!
  • After lead II — the 2nd-best lead to look for P waves in is lead V1. Unlike lead II — polarity of the P wave in lead V1 does not help in determining if the rhythm is sinus, or arising from another atrial focus (because sinus P waves can be positive, negative or biphasic with sinus rhythm in lead V1). 
  • IF we don't see atrial activity in either lead II or lead V1 — then look in the other 10 leads before concluding that atrial activity is absent.

  • As per Dr. Smith — there is regular atrial activity in lead V1 of ECG #1 (slanted BLUE lines in lead V1 of Figure-1). The rate of this atrial activity is ~200/minute (it takes 3 large boxes to record 2 P waves) — and, this atrial activity does appear to be conducting (because the PR interval preceding beats #7 and 8 in lead V1 of ECG #1 is constant). I do not see P waves in other leads.

  • My Impression: The rhythm in ECG #1 is not sinus. Instead — there are low-amplitude-but-regular P waves at ~200/minute that are conducting with a 3:1 AV ratio. The rhythm is either atypical AFlutter (Atrial Flutter) or ATach (Atrial Tachycardia). I favor ATach as the rhythm — because the rate of ~200/minute is considerably slower than the usual 250-350 atrial rate for untreated AFlutter — the 3:1 AV conduction ratio is very uncommon for untreated AFlutter (which most often manifests 2:1 or 4:1 AV conduction) — and because of the isoelectric baseline between P waves.
  • That said — Clinically, it does not really matter IF the rhythm in ECG #1 is ATach or atypical AFlutter — because initial management in the ED is virtually the same. Practically speaking — distinction between these 2 entities is often impossible outside of the EP lab, because there is lots of overlap in ECG features of these 2 arrhythmias (See Figure-4 below).

What about the Rhythm in ECG #2 of Figure-1?
Beginning our assessment by focusing on the long lead II rhythm strip in ECG #2:
  • Although parts of the rhythm in ECG #2 look regular — caliper measurement confirms that QRS complexes are irregularly irregular throughout the tracing.
  • Although several deflections in this long lead II rhythm strip almost look like P waves — the shape and distance of these deflections varies with respect to neighboring QRS complexes. Therefore — there are no definite P waves.
  • The presence of irregular irregularity of QRS complexes without P waves defines the rhythm in ECG #2 as AFib, here with a controlled ventricular response. AFib is a diagnosis that can be accurately recognized by clinicians attentive to the principles described above.

  • Beat #5 occurs earlier than other beats on this tracing. The reason for the different QRS morphology of this early beat #5 — is the result of aberrant conduction.  


Figure-2: "My Take" on the Pros & Cons of Computerized ECG Interpretations (From Grauer, K: ECG-2014-ePub, KG/EKG Press).





Figure-3: Example of hyperacute T waves missed by the Computer Report (Continuation of Figure-2).




Figure-4: Clinical considerations for distinguishing AFlutter from ATach with Block (From Grauer K: ACLS-2013-ePub, KG/EKG Press).




No comments:

Post a Comment

DEAR READER: I have loved receiving your comments, but I am no longer able to moderate them. Since the vast majority are SPAM, I need to moderate them all. Therefore, comments will rarely be published any more. So Sorry.