Finding What You Seek: Analyzing Therapies for Nonvalvular Atrial Fibrillation

Nonvalvular atrial fibrillation is common and can be associated with few hemodynamic symptoms. However, one of its consequences, systemic embolization, produces substantial morbidity and mortality. Over the past decade, the management of patients with nonvalvular atrial fibrillation has been the subject of many clinical trials, numerous conferences, and more papers than any clinician has time to digest. New drugs and therapeutic options abound. Clinicians managing these patients are awash with data and, perhaps, conflicting interpretations of those data. In this issue, Catherwood and colleagues (1) extend a simpler model published 5 years ago (2) to offer a new cost-effectiveness analysis of therapies for nonvalvular atrial fibrillation. They conclude that patients with nonvalvular atrial fibrillation are best and most cost-effectively served by cardioversion followed by antithrombotic therapy with aspirin. When atrial fibrillation recurs, as it will eventually in most of these patients (3, 4), the authors suggest a second attempt at cardioversion, continuation of aspirin therapy (or initiation of warfarin therapy in patients who remain in atrial fibrillation), and initiation of amiodarone therapy in patients with a moderate or high risk for systemic emboli.

Because cost-effectiveness analyses are explicit, they allow us to explore diverse assumptions and their implications. But decision models always simplify the real world: by limiting the number of strategies considered, by ignoring crossovers among therapies, by constraining the potential outcomes examined, or by considering a limited time horizon. Various recently published analyses of nonvalvular atrial fibrillation, although more similar than different, have simplified the decision in different ways. Two considered only antithrombotic …