Monday, June 5, 2023

A 40-something male with resolving chest pain and a "Normal ECG" by computer and cardiology overread

A 40-something male presented by ambulance with one hour of chest pain that was improving after sublingual nitroglycerine and 325 mg of aspirin, chewed.   


Here is his initial ED ECG:

What do you think?
Veritas algorithm.














I read this ECG without any history as reperfusing inferior and posterior OMI due to the Q-wave in lead III with minimal STE and reciprocal ST depression in V2-V4 (which should never be there).  The large upright T-wave in V2 is consistent with reperfusion.


I sent this ECG with no information to Pendell.  We send each other EKG by the dozens every day.  Most are OMI look alikes (mimics).  So when he sees OMI on the ECG, or reperfusing OMI, it is not because there is a high pretest probability.  It is because he is VERY good. 

He responded: "reperfusing inferior posterior OMI"

I sent it to the #PMCardio AI Bot (AKA Queen of Hearts) and she responded with "OMI mid confidence")

But the conventional algorithm read this as "Normal ECG"
And the overreading cardiologist (also with no clinical context) read this as "Normal ECG"



Now I need to tell you the whole story:

This patient actually had a prehospital ECG while his pain was intense.  Here it is:
Obvious Inferior Posterior STEMI (+) OMI.
The cath lab was activated prehospital

Zoll algorithm

But imagine if the patient had walked in.  Or had not had a prehospital ECG on the ambulance.  Then the ED doc would be dependent on that first ECG.

Initial troponin was:  3 ng/L  
We showed that the first troponin in acute STEMI is often negative in at least 27%.  

Here is the angiogram:

--Culprit is 100% stenosis in the proximal RCA. (It must have re-occluded between the ED and the cath lab)

--Lesion was stented.


This points out an important fact: the ECG and the angiogram are always recorded at different time points, and thrombi are dynamic, propagating and lysing.  


Aside on ECG Research: 20% of Definite diagnostic STEMI (Cox et al.) have perfect coronary flow by the time of angiogram.  Thus, in our ECG research, we need to find a surrogate outcome that reflects the state of the artery at the time of the ECG. Based on many studies of peak troponin (as a good measure of infarct size) in both STEMI and NonSTEMI, we have chosen certain cutpoints of for each of contemporary troponin I and T and high sensitivity I and T.


post PCI ECG

Reperfused inferior and posterior OMI

Veritas algorithm


Formal Bubble Contrast Echo

--Normal LV size, wall thickness, and systolic function with an estimated EF of 59%.

--Regional wall motion abnormality--basal to mid inferior and inferolateral hypokinesis.  (Echo segments of what we call the posterior wall are classified as "lateral")


Peak hs troponin I = 13,900 ng/L




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

MY Comment, by KEN GRAUER, MD (6/5/2023):

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


Today’s teaching points provide lessons worthy of emphasis.

  • Even more concerning than the fact that the the overreading cardiologist accepted the computer interpretation of this initial ECG as “normal” — is that fact that this “overreading cardiologist” clearly did not realize the NEED to STOP what he/she was doing, to inquire about the history (since as the "overreading cardiologist" — they are given the responsibility of overreading this ECG and clinically correlating it). This is because IF the history is recent CP — then this initial ED ECG has to be interpreted like Drs. Smith and Meyers did = "reperfusing inferopostero OMI until proven otherwise!"
  • My concern is that all-too-many cardiologists still fail to appreciate the basic physiology of acute coronary occlusion (ie, that it is an ongoing process — that may include one or more periods of spontaneous reopening and reocclusion, until a final disposition of the culprit artery is arrived at).
  • As per Dr. Smith — if the overreading cardiologist would have been the on-call cardiologist — and the prehospital ECG during CP (that showed an obvious STEMI) had not been obtained — that this OMI would have gone unrecognized, with probable disastrous consequence.

  • ECGs can not be interpreted in a vacuum. Clinical history is needed. Often serial ECGs are also needed (with repeating the ECG within as little as 10-20 minutes if the process appears to be active). This is because (as per Dr. Smith) — the 1st troponin will be normal in ≥27% of acute OMIs!

  • Serial tracings often show dynamic ST-T wave changes — which in a patient with CP, allow prompt definitive diagnosis that might not have been obvious with just a single tracing.

  • "Dynamic ST-T wave changes become much more meaningful — IF correlated to the presence and severity of CP (ie, reduced ST-T wave changes in association with reduction or disappearance of CP is diagnostic of spontaneous reopening — and provides clear indication for immediate cath lab activation in the hope of preventing spontaneous reclosure).
  • Current computer programs do not take into account the clinical history, nor serial ECG changes. Therefore — an ECG such as the initial ED tracing in today’s case will almost always be missed! The QOH AI program is already much more sensitive and specific for identification of acute OMI than previous computer programs — and — QOH will be able to incorporate clinical correlation with serial ECG changes in the not-too-distant future.


Finally — There are multiple abnormal ECG findings on the initial ED ECG, such that no overreading cardiologist should accept a computer reading of “normal” for this tracing. These abnormal ECG findings include:

  • More than just ST coving with slight elevation and disproportionately huge T wave inversion in lead III (relative to the tiny QRS amplitude in this lead) — is a wide initial Q wave, with fragmentation that produces a QrsR’ complex. This type of fragmentation in an inferior lead greatly increases the specificity for a definite inferior infarction over the simple presence of small inferior Q waves.
  • The fatter-than-it-should-be T wave in lead aVL is the mirror image opposite picture of the ST-T wave in lead III ==> reciprocal change that strongly suggests a recent (if not still ongoing acute) event.
  • The T wave in lead aVF is also inverted (albeit this shallow) — which together with what we see in leads III and aVF, suggests recent (depending on the history of CP) changes of inferior OMI, now most probably with reperfusion.

  • As per Dr. Smith — the disproportionately tall and fatter-at-its-peak-than-it-should-be T wave in lead V2, in the context of recent inferior OMI, indicates posterior reperfusion. Note that instead of the expected slight ST elevation in lead V2 — there is slight-but-real ST depression (positive “Mirror Test”), which is diagnostic of posterior OMI — which is further supported by the subtle-but-real ST depression in neighboring leads V3,V4.

  • The ST segments in leads V5,V6 are flat and slightly depressed — which is never “normal” in a patient with new CP.

  • The finding of a non-depressed ST segment in lead V1 — in association with acute infero-postero OMI — suggests that this is likely to be a proximal RCA occlusion, with a good chance of acute RV involvement.


Hopefully these lessons will be learned by many in the cardiology community who are still “stuck” in the STEMI paradigm.





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.