Monday, June 17, 2019

More precise interpretation of the results of the 4-variable formula.

If you haven't heard of the 4-variable formula for differentiating Normal Variant ST Elevation in V2-V4 from Ischemic ST Elevation due to LAD occlusion, then go here:

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

I have often said that the closer the formula value is to the cutpoint of 18.2, the more accurate.

But how wide is the variation?

Here I plot the graphs:

Specificity
Specificity is 97% at a cutpoint of  20.7

Sensitivity
Sensitivity is 97% at a cutpoint of 17.0

Accuracy

===================================
Comment by KEN GRAUER, MD (6/17/2019):
===================================
Today’s brief Blog post illustrates insightful sensitivity, specificity and accuracy data for use of Dr. Smith's 4-Variable Formula for differentiating between normal variant ST elevation in leads V2-thru-V4 from ischemic ST elevation due to LAD occlusion.
• Near the top of the page is a LINK to 12 Example cases. I’m repeating the link to this page — because it provides an excellent opportunity for REVIEW! (LINK to 12 EXAMPLE CASES — from Dr. Smith’s November 3, 2017 Blog).
• For each of these 12 cases — Dr. Smith provides Formula values and specific commentary on the ECG abnormalities present — followed by “the Answer” (ie, if the ECG turned out to be a case of acute LAD occlusion or a repolarization variant ). It is worth spending a few minutes going through these 12 cases to solidify your understanding and to “hone” your acute ECG diagnostic ability.
MTHOUGHTS: I think Dr. Smith said it BEST in answer to one of the commenters on this November 3, 2017 blog (Figure-1):
 Figure-1: Comment by Dr. Smith on November 6, 2017 in answer to one of the commenters (See text).
================

• Dr. Smith’s formulas serve as an AID. The amount of assistance they provide may vary, depending on the experience of the interpreter. You don’t need the formula — if you recognize that the ECG is clearly abnormal and suggests acute LAD occlusion. Even if the formula were to be “negative” in such a case — this should not dissuade you from acting as if there is acute LAD occlusion.
• In participating in an ECG study with Dr. Smith — I had the opportunity to interpret 1,000 tracings, in which I was immediately provided with the Formula calculation. To keep myself “honest” — I always interpreted each of these ECGs first — and only then looked at the Formula value. I found knowing the Formula value to be very comforting in my interpretations! When I thought the ECG was negative for acute occlusion — knowing that the Formula value was clearly below the “cutpoint threshold” was exceedingly reassuring, and allowed me much more quickly to move on to the next tracing. On the other hand — learning that the Formula value was near (or above) the cutpoint alerted me to much more closely go back and review those occasional tracings that I otherwise would have thought were negative. As per Dr. Smith in his Comment above (Figure-1) — I too was occasionally surprised by the Formula value! — and on those occasions, awareness of a higher-than-anticipated Formula value was extremely helpful to me!
In my opinion — each of the 12 Example Cases at the above link that were positive manifest morphologic features that should suggest possible acute LAD occlusion even without use of the Formula. We have discussed these features on numerous occasions in this ECG Blog — but it may be helpful to repeat them here: KEY ECG Features that may suggest Possible LAD Occlusion include the following:
• One or more of the chest leads show T waves that are disproportionately tall (compared to the R wave in the same lead) — as well as being fatter-than-they-should-be at their peak and/or wider-than-they-should-be at their base.
• Loss of the normal upward concavity of the ST segment upslope (being replaced by straightening of the ST segment takeoff — or by coving of the ST segment).
• Poor R wave progression (ie, loss of anterior R wave amplitude, compared to what should be expected).
• Unexpected ST elevation in lead V1.
• Reciprocal ST-T wave changes in the inferior leads (including unexpected ST-T wave flattening in these leads).
• Unexpected ST elevation in lead aVL.
• NOTE #1: Sometimes the above ECG findings in one or more leads may be subtle. I always look first for those one or two leads in which there is NO doubt about the abnormality. It often becomes easier to identify more subtle abnormalities in neighboring leads after identifying one or two leads that are clearly abnormal.
• NOTE #2: The more leads showing abnormal ST-T wave findings — the greater the likelihood of acute ischemia. In each of the 12 Example Cases at the above link — there were at least 2 definitely abnormal leads plus several additional leads with more subtle changes.
BOTTOM LINE  For any provider, ready availability of Dr. Smith’s 4-Variable Formula value can be an invaluable aid for teaching, expediting clinical decision-making, and improving accuracy. While I personally prefer to interpret ECGs before I am told the Formula value — others may prefer to use the Formula value sooner, as a “heads up” to alert them to a high likelihood of an abnormal ECG.
• P.S. — Be aware of EXCLUSIONS for using the Formula.