Selection of endpoints in atrial fibrillation studies

Atrial fibrillation (AF) is the tachyarrhythmia of the new millennium. There has been a dramatic increase in research on the management of this problem over the last 10 years. Presently, there is no clear consensus on the most appropriate endpoints to be used in studies of therapy for AF, particularly those concerning rhythm management itself. Endpoints for studies of rhythm management should be based firmly on the objectives of therapy for AF. Some objectives of therapy are obvious, but others, such as reduction of mortality, are not and are somewhat controversial. Clinically relevant endpoints are to be preferred but have been underutilized. Using clinical events as endpoints is complicated by the fact that event rates are low and large sample sizes are needed. Cost and cost-effectiveness are endpoints that are becoming increasingly important but also have been underutilized. Clinical classification of AF is an important factor to be considered in planning studies of AF rhythm management. Patient selection can have a profound effect on the outcome of certain surrogate endpoints. The main limitation of these endpoints is that they assume improvement in the surrogate measurement is closely correlated to a good clinical outcome. In fact, there is ample evidence that such a correlation is quite poor at times. Potential solutions to the problems discussed here include wider appreciation of the problem, use of carefully crafted composite clinical endpoints, and better calibration of surrogate endpoints against clinical endpoints. More research on these issues is urgently needed.