Clinical Trial Data Can Be Supplemented by Real-World Evidence: An Example for an Oral Anticoagulant in Patients With Nonvalvular Atrial Fibrillation at Risk of Stroke