Interpretation of clinical studies in electrophysiology: statistical considerations for the clinician.

Abstract

Atrial fibrillation (AF) outcome studies play an essential role in the development of clinical evidence to improve the management of AF patients. Understanding the statistical considerations involved in study design and interpretation is crucial if electrophysiologists are to change practice. In this review, with the guidance of a medical statistician and a clinical trialist we provide an overview of important statistical issues for the clinician, with a focus on clinical studies in AF ablation. Various types of study designs including randomized controlled trials, superiority, and non-inferiority studies are described, along with their implications and limitations. Appropriate sample size calculation is fundamental to ensure statistical power and efficient resource use. Multiplicity in study endpoints is useful to encapsulate the varied effects of an intervention/treatment, although statistical adjustments are required to account for this. Finally, we discuss the limitations with the current primary endpoint used in AF ablation studies, namely, freedom from atrial tachyarrhythmia of >30 seconds, and propose AF burden as a more relevant primary endpoint, based on findings from recent clinical studies. However, technical challenges need to be overcome before AF burden can be routinely adopted, especially the need for non-invasive, long-term monitoring. The emergence of newer technologies, particularly wearable technology, offers significant promise in filling this gap.