Abstract
Little is known in clinical practice about antiarrhythmic drug (AAD) use in patients with atrial fibrillation (AF) (particularly younger ones) who do not have structural heart disease. Using the MarketScan database, we identified patients <65 years without known coronary artery disease or heart failure who had an AAD prescription claim (class Ic drug, amiodarone, sotalol, or dronedarone) after their first AF encounter. A multinomial logistic regression model was created to assess factors associated with using each available AAD compared with using class Ic drugs before and after dronedarone was marketed in the United States. Additionally, we used the Kaplan-Meier method to determine the rates of change in AAD use and discontinuation during the year after AAD initiation. Of 8,562 patients with AF, 35% received class Ic drugs, 34% amiodarone, 24% sotalol, and 7% dronedarone. The median patient age was 56 (interquartile range 49 to 61), and 34% were women. Both before and after dronedarone was marketed, there was a statistically significant lower likelihood of class Ic drug use versus other AAD use with increasing age, inpatient index AF encounter, and previous or concomitant anticoagulation therapy. During the 1 year after AAD initiation, the AAD change rate was 14% for class Ic drugs, 8% for amiodarone, 17% for sotalol, and 18% for dronedarone (p <0.001); the AAD discontinuation rate was 40% for class Ic drugs, 52% for amiodarone, 40% for sotalol, and 69% for dronedarone (p <0.001). In conclusion, we found extensive use of amiodarone that may be inconsistent with guideline recommendations and unexpectedly high rates of AAD discontinuation.