15-30% of strokes are cardioembolic and 60-90% of stroke patients present with ECG abnormalities.Â (Source)
So let me ask you some questions. Is it important to obtain a 12 lead ECG during a suspected acute stroke in the prehospital setting? Why? Why not? Does itÂ reallyÂ matter either way? Before we discuss the answerâ€¦
Hereâ€™s the abnormalities youâ€™reÂ likelyÂ to encounter in acute stroke:
1.Â Long QTc & Ischemic ST changes.Â Found in 76% of hemorrhagic strokes & 90% in ischemic strokes. (Source)
2.Â Afib. Detected in up to 25% of new ischemic strokes. (Source)
Hereâ€™s what youâ€™reÂ unlikelyÂ to encounter in acute stroke:
1.Â STEMI. Outside of a fewÂ case studiesÂ (and good anecdote) thereâ€™s little in the literature of concomitant STEMI & stroke however each disease tends to follow one another within days or months. 2-6% of stroke patients die of cardiac causes in the first 90 days after stroke. (Source)
2.Â Changes indicative of acute PE.Â Again the literature was scarce.
Now letâ€™s look at some literature:
â€œIn patients with ischemic stroke and intracerebral hemorrhage, these ECG abnormalities (and QT prolongation) most often represent preexisting coronary artery disease. The specificity of ECG changes to diagnose acute myocardial infarction is low in the acute phase of stroke.â€
The authors note that the ST & QTc changes most likely were due toÂ arteriosclerosis and it was extremely difficult to look at an acute stroke patientâ€™s ECG and correctly say that the patient was not having an MI. In other words the physicians were unable to distinguish that the patient was not suffering an MI from that patient suffering an MI.
â€œThe six-month mortality rate in the patients with ECG changes was 38.9% whereas it was 15.2% in those with normal ECGâ€.
The authors note that observing an abnormal ECG in an acute stroke patient more than doubled their mortality rate at 6 months (likely due to the severity of the underlying cardiovascular disease).
â€œBoth disturbances in rhythm and conduction and â€œischemicâ€ ST-T alterations were detected and the frequency of the former exceeded that of the latter. The ECG alterations were transient in 32 patients and permanent in four. New electrocardiographical abnormalities in patients without evidence of heart disease prior to the stroke were associated with poorer prognosis.â€
The authors noted a high incidence of transient abnormalities vs. permanent ones.Â (remember paroxysmal Afib was one of the potential causes of cryptogenic stroke we covered inÂ Tales from the Cryptogenic, pt. 1)
â€œPatients with a prolonged QTc interval were more likely to die within 90 days compared with patients without a prolonged intervalâ€¦The estimated survival at 90 days was 70.5% and 87.1%, respectivelyâ€¦The identified threshold cutoffs for increased risk of death at 90 days were 440 milliseconds for women and 438 milliseconds for men.â€
â€œOf the 150 patients with stroke, 138 (92%) showed ECG abnormalities. The most common abnormalities were also changes from prior tracings: QT prolongation (68 patients, 45%), ischemic changes (59, 35%), U waves (42, 28%), tachycardia (42, 28%), and arrhythmias (41, 27%). Patients with cerebral embolus had a significantly increased frequency of atrial fibrillation (9 patients, 47%); and with subarachnoid hemorrhage an increased frequency of QT prolongation (20, 71%) and sinus arrhythmia (5, 18%)â€¦Stroke patients had an increased frequency of pathologic Q waves (30 patients, 20%) and left ventricle hypertrophy (39, 26%), but these were not new findings at the time of the stroke.â€
An old paper from 1979 but a good read nonetheless.
â€œThe most common ECG abnormalities associated with stroke were T-wave abnormalities, prolonged QTc interval and arrhythmias, which were respectively found in 39.9%, 32.4%, and 27.1% of the stroke patients and 28.9%, 30.7%, and 16.2 of the patients with no primary cardiac disease. We observed that other ECG changes comprising pathologic Q- wave, ST-segment depression, ST-segment elevation, and prominent U wave may also occur in selected or non-selected stroke patients; thereby simulate an acute myocardial injury.â€
Similar to #5 but published in 2013 and conducted in a different population furthering the validity of the changes as they correlate to stroke.
This should lay a foundation for us to further explore the relationship of the ECG to acute stroke. Back to answer the question from earlier regarding prehospital 12 leadsâ€¦.wellâ€¦presently thereâ€™s simply not a hard and fast answer. A case can be made either way but the real goal is as always:Â recognition andÂ activation of a stroke alert and rapid transport to an appropriate facility without delay. If you can snag a 12 lead without compromising this dictum then why not? If it slows you down, or if a protocol is broadly applied to standard 911 transport paramedics of unequal skill and efficiency then we are doing a disservice when one can be obtained in the ED. Prehospital 12 leads in acute stroke neednâ€™t become distracting novelties.
Thus far there is consensus on in-hospital 12 leads but a lack of scientific consensus on prehospital 12 leads. Let the discussion ensueâ€¦
P.S., as a historical asideâ€¦theÂ first paper ever published on the link between cardiac changes and strokeÂ was in 1947.
Now on toÂ part 2.