Incidence and predictors of myocardial infarction among patients with atrial fibrillation.

OBJECTIVES We sought to evaluate the utility of excluding myocardial infarction (MI) in patients presenting to the emergency department (ED) with atrial fibrillation (AF) and to identify predictors of MI in this group. BACKGROUND Patients with AF are frequently admitted to the hospital, in part, to exclude an associated MI. There are no prospective data on unselected patients to support this common practice. METHODS We conducted a prospective cohort study of all patients who presented to a single-center ED with the primary diagnosis of AF. RESULTS Of a total of 255 patients, 190 (75%) were admitted to the hospital, and 109 of them underwent a standard "rule-out MI" protocol. Of these 109 patients, six (5.5%) were identified as having an acute MI at the time of admission. Chest pain was present in 39% of patients, with a sensitivity and specificity for the occurrence of MI of 100% and 65%, respectively. ST segment elevation or depression was present in 43% of patients, with a sensitivity and specificity of 100% and 51%. The presence of either major ST segment depression (>2 mm) or elevation on the admission electrocardiogram (ECG) was present in 6%, with a sensitivity of 100% and a specificity of 99%. The resulting positive and negative predictive values were 86% (95% confidence interval [CI] 42% to 99%) and 100% (95% CI 96% to 100%), respectively. Use of this criterion would have reduced the number of rule-out MIs in our study group by 94%, with no loss of sensitivity. CONCLUSIONS Chest pain and ST segment depression are extremely common findings in patients presenting to the ED with AF and have limited power to predict MI. In contrast, ECG evidence of ST segment elevation or depression >2 mm appears to be a reliable discriminator of which patients are at risk for MI. Patients without significant ST segment changes are at very low risk for MI and may not require performance of the rule-out MI protocol or hospital admission if clinically stable.