I gave this Grand Rounds to UC Riverside a couple weeks ago.
It is my most up to date lecture on OMI NOMI and Use of AI, especially the Queen of Hearts, in the Diagnosis of Occlusion MI.
This was written by one of our really fine 3rd year residents, Katie Buerk.
This case was a 69 year old male, brought in by ambulance, after being found unresponsive in bed by family earlier that day. He had last been seen 3-7 days ago. For the medics, he was bradycardic in the 20-30s. They were unable to obtain a blood pressure.
On arrival to the ED, he was breathing spontaneously, had faint pulses, and was cool to the touch. His heart rate was in the low 20s and we were also unable to obtain a blood pressure. His temperature was 32.8 C. A quick POCUS which showed significantly reduced ejection fraction and trace B lines. He was given 50 mcg epinephrine with good response in both heart rate and blood pressure.
His rhythm on telemetry seemed to be sinus bradycardia vs junctional rhythm. Telemetry also showed possible ST elevation and peaked T waves, so he was given 2 g calcium gluconate empirically.
There is a regular wide complex rhythm without P waves at a rate of less than 100. Thus, this could be junctional, but is far more likely to be accelerated idioventricular rhythm (AIVR). AIVR is often a reperfusion rhythm. The AIVR has an RBBB configuration; thus, it is originating from the LV. There is also left axis deviation; thus it is originating in the inferior LV. The ST segments in AIVR can be analyzed similarly to LBBB (excessively discordant ST Elevation or Depression/concordant STE or STD).
In this case, there is excessively discordant STD in V2 and concordant STD in V3. Thus, it is diagnostic of Posterior OMI.
The AIVR suggests that it is a reperfused posterior OMI, and that is certainly possible, but I think one must assume it is active until proven otherwise.
The PMCardio Queen of Hearts AI Model thinks it is a reperfused OMI:
Case continued:
Initial EKG demonstrated ST depression in V1-V4, concerning for posterior OMI. Cardiology was consulted.
His heart rate had improved to the 80s after epinephrine administration. Subsequently, he became increasingly bradycardic and was noted to have myoclonic movements. At that time, he had no pulses and was in asystole on the monitor. CPR was initiated and he underwent 1 round of ACLS (CPR + 1 mg epi). During CPR, he started moving all four extremities spontaneously. During first pulse check, ROSC was achieved. We proceeded with intubation using ketamine and rocuronium. He was given 2 amps of bicarb empirically prior to intubation due to concern for profound acidosis. Norepinephrine started after intubation due to persistent hypotension.
Post intubation we obtained:
His initial labs returned notable for pH 7.01, bicarb 10, and pCO2 41. His lactate was 22. His blood gas demonstrates profound metabolic acidosis, driven by the lactate, with insufficient respiratory compensation. He was noted to have a glucose of 591 and an anion gap of 30. Potassium was 3.7.
His potassium was replaced, he was given 300 mg rectal ASA, and an insulin drip ordered. He was hyperventilated at a rate of 28 to provide some compensation for the metabolic acidosis.
He was persistently bradycardic, requiring 2 x 50 mcg epinephrine to maintain HR >60. Epinephrine drip was started and norepinephrine was discontinued. He was given vancomycin and cefepime empirically, though we overall had low concern for sepsis as the etiology of his presentation.
At this time, there had been multiple discussions with cardiology team, who were debating taking this patient to the cath lab. They felt that the asystolic arrest suggested a different etiology of cardiac arrest. With the severe acidosis and absence of ST elevation, they felt there was more likely to be a non-cardiac etiology of his presentation. Cath lab had not yet been activated, so he was taken to CT for a head and chest/abdomen/pelvis angiogram to rule out other causes of cardiac arrest.
It did not show pulmonary embolism or intra-abdominal pathology, but it did show this:
CT showed hypoperfusion of the right ventricular wall and the posterior wall, as well as significant calcifications of the LAD. There were no other radiographic findings to explain his presentation.
Third EKG obtained 100 minutes after the first:
Initial troponin returned and was 42,747 ng/L.
This very high initial troponin tells us that the infarct is subacute and much, or most, of the damage is already done.
The patient went to the cath lab relatively quickly, at 160 minutes after arrival. He received an additional 300 mg rectal aspirin and 5000 unit heparin bolus in the ED.
Coronary Angiography
--First Diagonal occluded
--RCA: 95% distal disease with total occlusion of RPAV with the vessel small and no PCI option to RCA or RPAV RPDA severe diffuse disease, small vessel
--Ramus: Small vessel with severe diffuse disease
There was no possibility of PCI/stent
Echo:
Decreased left ventricular systolic performance mild to moderate.
The estimated left ventricular ejection fraction is 40-45%.
Probable anterior and anterolateral wall hypokinesis.
Right ventricular enlargement.
Decreased right ventricular systolic performance.
Assessment:
Cardiology thought this was cardiogenic shock from RV dysfunction.
Smith Comment: the RV was very ischemic on the CT scan and dysfunctional on echo, and this does explain the shock. But I'm not sure how to explain the RV ischemia based on either ECG or angiogram. There was no pulmonary embolism.
Case Continued
The post cath course was very complicated with cardiogenic shock and severe dysrhythmias that were eventually controlled, but he did not recover neurologically.
MY Comment, by KEN GRAUER, MD (4/6/2025):
-USE.png) |
| Figure-1: The initial 12-lead ECG obtained in the ED. |
-labeled-USE.png) |
| Figure-2: I've added an "X" to today's initial ECG. What does this "X" indicate? |
-USE.png) |
| Figure-3: Repeat ECG obtained after ROSC from an asystolic episode. |
-labeled-USE.png) |
| Figure-4: The 3rd ECG — obtained after return of a Troponin ~40,000. |
Written by Jesse McLaren
A 40 year old presented with 90 minutes of midsternal chest pressure, radiating to bilateral arms, with shortness of breath. Below is the first ECG, signed off by the over-reading cardiologist agreeing with the computer interpretation: “ST elevation, consider early repolarization, pericarditis, or injury”. What do you think?
There’s normal sinus rhythm with normal conduction, right axis and delayed R wave, and normal voltages. There’s ST elevation in V3-4 which meets STEMI criteria, which could be present in either early repolarization, pericarditis or injury. But there are also hyperacute T waves (HATW) in V4-5, which exclude early repolarization and pericarditis, leaving only LAD occlusion for this patient presenting with classic symptoms of ACS.
Here’s the PMCardio Queen of Hearts AI Model interpretation:
This is great example of how the Queen uses proportionality to identify hyperacute T waves: the T waves in V3 and V4 are almost identical, but in V3 they are proportional to its large QRS whereas in V4 they tower over its small QRS; the T wave in V5 is much smaller, but relative to its QRS it is large and bulky.
So using the OMI paradigm and its AI, this patient would have had immediate cath lab activation before the first troponin result. Let’s see what happens in the current STEMI paradigm.
Emergency physician: ‘STEMI neg’ but with elevated troponin = Non-STEMI
The first ECG was signed off. After only 90 minutes of chest pain, the first troponin was unsurprisingly in the normal range at 11ng/L (normal <26 in males and <16 in females), so the emergency physician waited for repeat troponin. After it rose to 150ng/L two hours later a repeat ECG was done:
There’s now a Q wave in V3 and a smaller T wave in V3-V4, proving this is LAD occlusion. Hyperacute T-waves remain in V3 and V4. There’s the same degree of ST elevation, but this time the computer calls it STEMI. But it was interpreted as no acute ischemia and the patient was referred to cardiology as Non-STEMI.
Nurse notes: the silent scream of the heart
The emergency nursing notes document the patient complaining of chest pain refractory to nitro, with a rising trop:
2200: ECG shown to ED MD
0020: repeat ECG shown to ED MD, patient complain of midsternal chest pain
0520: nitro x 3. Chest pain still persists. Cardiology aware. Repeat troponin ordered
0630: lab called for high troponin 3900. Paged cardiology
0800: patient complains of chest pain. Repeat blood work and ECG
0845: repeat trop over 7000. Cardio aware
1030: repeat trop sent, no change in chest pain
1100: heparin drip started
1130: transfer for cath, still complains of chest pain
Here’s the ECG repeated at 0800
There’s now Q waves V3-4. T waves have deflated and inverted in V3-4 suggesting some degree of reperfusion, but the patient still complained of ongoing chest pain. Troponin rose to 12,000 before cath.
Cardiology: delayed cath = Non-STEMI
Here’s the interventional cardiology note describing the ECGs, management and outcome:
“He has had transitory peak T-waves, ST-segment elevation, and biphasic T-waves during recurrences of pain. This morning, he was also found to have a rising troponin. He was thus referred for emergent invasive assessment. Code STEMI was called…Mid LAD had serial 70 and 60 percent lesions and was occluded in the distal segment…An excellent result was achieved. The total occlusion was recanalized and stented from 100 to 0%.”
Here's the discharge ECG, with ongoing Q waves and reperfusion T wave inversion:
Peak troponin was 47,000 ng/L and echo showed an akinetic apex with EF 45%.
What should the discharge diagnosis be, and why does this matter?
The patient had an ECG with ST elevation (and hyperacute T waves), activation of code STEMI, a 100% LAD occlusion on angiogram, a massive peak troponin, and an akinetic wall. The discharge diagnosis should reflect the underlying pathology of Occlusion MI, regardless of whether the ECG was interpreted to show STEMI criteria, and regardless of time to reperfusion. But because of the delayed reperfusion, the discharge was Non-STEMI.
Thus, a case with more than 12 hours of delay for reperfusion will not be flagged for review.
_______
Smith: we have an article under review that shows that the variable most closely associated with the final diagnosis of "STEMI" vs. "Non-STEMI" was a door to balloon time less than, vs. greater than, 90 minutes. Not whether there was or was not Acute Coronary Occlusion. Not whether the ST segments met STEMI millimeter criteria.
_______
The problem is not just for this patient, and simply changing the discharge diagnosis in this case from "Non-STEMI" to "STEMI" is not the solution. This is just an example of the broader problems of the STEMI paradigm for research and quality improvement.
For research, 12 hours to PCI is deemed ‘early intervention’ for Non-STEMI. If two of the same patients were part of Non-STEMI trials like TIMACS, and randomized into ‘early’ (16 hours) vs ‘delayed’ intervention (52 hours), there would appear to be no benefit to the ‘early’ intervention, because the damage was already done. But clearly this 'Non-STEMI' patient with OMI would have benefited from immediate cath lab activation on arrival, when their first troponin was 11ng/L, rather than after after it rose to 12,000ng/L after 12 hours of refractory ischemia.
For quality improvement, the discharge diagnosis also matters. Classifying as STEMI vs Non-STEMI can be more reflective of reperfusion time than ECG findings or patient outcomes, which allows cases like these to be normalized. Instead, if patients are classified by the actual outcome of OMI vs NOMI, then this patient clearly had a missed OMI. This is not to assign blame but to identify multiple opportunities for improvement:
1. ECG: using OMI signs and AI, to activate the cath lab on arrival, before waiting for the troponin
2. POCUS for complementary regional wall motion abnormalities for subtle OMI
3. Clinical: patient alerts for refractory ischemia (refractory chest pain), and empowering nurses to advocate for patients
4. Troponin: troponin is a rear-view mirror that shows damage that has already happened, so the first troponin is unreliable with acute symptoms and serial troponin will lag behind the damage of ongoing occlusion. But refractory ischemia with rising troponin is an indication for cath lab activation regardless of the ECG.
-USE.png) |
| Figure-1: Today's initial ECG. |
