Provide feedbackProvide feedback

« Back

CARISMA researchers use ILR to document arrhythmias in post-MI patients with low ejection fraction, identify high-degree AV block as independent predictor of cardiac death

Thursday, April 10 2008 | Comments
Evidence Grade 0 What's This?
By Courtneay Parsons

Results from the CARISMA study suggest new-onset atrial fibrillation and potentially serious bradyarrhythmias and ventricular tachyarrhythmias occur in a substantial number of patients who have an ejection fraction <=0.40 following an acute myocardial infarction. Study data also suggest high-degree atrioventricular (AV) block is an independent predictor of cardiac death in these patients.

In the current analysis, CARISMA researchers documented the incidence and prognostic significance of cardiac arrhythmias assessed by an implantable electrocardiogram loop recorder (ILR). The study included 297 patients who received an ILR <=21 days following an acute MI and had an ejection fraction <=0.40.

Arrhythmias of interest included sinus bradycardia (<=30 bpm, >=8 sec), sinus arrest (>=5 sec), AV block (<=30 bpm, >=8 sec), nonsustained ventricular tachycardia (VT; >=125 bpm, >=16 beats), sustained VT (>=125 bpm, >=30 sec), and AF (>=125 bpm).

During a mean follow-up of 1.9 years, 137 patients (46%) exhibited one of the prespecified arrhythmias (new-onset AF, 27%; high-degree AV block or sinus bradycardia, 17%; nonsustained VT or VT/ventricular fibrillation, 17%). Eighty-six percent were asymptomatic.

There were 25 cases of cardiac death during the trial. In univariate analyses, predictors of cardiac death were AV block <30 bpm (HR, 7.0; P=.0004), sinus bradycardia <30 bpm (HR, 5.8; P=.004), and nonsustained VT (HR, 3.4; P=.025).

However, in multivariate analyses that included the prespecified arrhythmias as time-dependent covariates, high-degree AV block was the only independent predictor of cardiac death (HR, 4.8; 95% CI, 2.0-11.5; P<.001). (Thomsen PE. Presentation 412-11.)

Print  |  E-mail


Be the first to write a comment for this article!

You must be logged in to post a comment.