The evolving role of stroke prediction schemes for patients with atrial fibrillation.

Our approach to managing patients with atrial fibrillation has changed substantially over the past 10 years, as a result of numerous high-quality observation studies and randomized trials. In this article, we will provide practical guidance for the use of oral anticoagulation therapy in patients with atrial fibrillation. We will review the evolution of stroke and bleeding risk prediction schemes and discuss their role in patient care. Initially, stroke prediction schemes were used to identify patients with atrial fibrillation at the highest risk of stroke, in whom the use of oral anticoagulant therapy was believed to be the most important. However; with the advent of new, safer, and more convenient oral anticoagulant drugs, the role of these schemes has shifted to the identification of the lowest risk patients, representing the minority of patients with atrial fibrillation, in whom oral anticoagulant therapy is not recommended. At the same time, schemes were developed to predict bleeding, the major risk of oral anticoagulant therapy. However; use of these schemes has been limited by their complexity and significant correlation with stroke schemes. In general, it is advisable to base the decision to use oral anticoagulation on the patient's stroke risk and use bleeding schemes to identify absolute contraindications or modifiable risk factors for bleeding. Prediction schemes have been useful clinical tools, invaluable in the design of clinical trials, and have assisted greatly in economic analyses. However, the nature and role of such schemes is now adapting to the current era of novel oral anticoagulant agents.