External validity of the ARISTOTLE trial in real-life atrial fibrillation patients

Our primary objective was to determine the proportion of patients with atrial fibrillation (AF) eligible for enrollment in a randomized controlled trial for a novel oral anticoagulant, the ARISTOTLE trial. A secondary objective was to describe the reasons for trial ineligibility.

We performed a cross-sectional study of an unselected population including 2274 patients in Skellefteå, Sweden with at least one verified episode of AF on or before December 31, 2010. Patients were classified as suitable or unsuitable for anticoagulant treatment according to current guidelines. The enrollment criteria from the ARISTOTLE trial were extracted from the original publication and applied to the population.

Among all patients with AF, 1579 were classified as suitable for anticoagulant treatment. Of these, only 658 patients (42%) were eligible for participation in the ARISTOTLE trial. Among the 921 patients ineligible for participation, 498 did not meet the ECG criteria, 272 had psychosocial problems, and in addition, 78 patients were excluded due to both of these criteria.

Our study shows that a majority of the patients in an unselected population with AF suitable for anticoagulant treatment were ineligible for participation in the ARISTOTLE trial. The applicability of the ARISTOTLE trial is therefore unknown for a considerable proportion of patients with AF in real life.