Bayesian Network Meta-analysis of Randomized Controlled Trials on the Efficacy of Antiarrhythmics in the Pharmacological Cardioversion of Paroxysmal Atrial Fibrillation.

Abstract

Since atrial fibrillation (AF) is one of the major arrhythmias managed in hospitals worldwide, it has a major impact on public health. The guidelines agree on the desirability of cardioverting paroxysmal AF episodes. This meta-analysis aims to answer the question of which antiarrhythmic agent is most effective in cardioverting a paroxysmal AF. A systematic review and Bayesian network meta-analysis, searching MEDLINE, Embase, and CINAHL, were performed, including randomized controlled trials (RCTs) enrolling a population of unselected adult patients with a paroxysmal AF that compared at least two pharmacological regimes to restore the sinus rhythm or a cardioversion agent against a placebo. The main outcome was efficacy in restoring sinus rhythm. Sixty-one RCTs (7988 patients) were included in the quantitative analysis [deviance information criterion (DIC) 272.57; I2 = 3%]. Compared with the placebo, the association verapamil-quinidine shows the highest SUCRA rank score (87%), followed by antazoline (86%), vernakalant (85%), tedisamil at high dose (i.e., 0.6 mg/kg; 80%), amiodarone-ranolazine (80%), lidocaine (78%), dofetilide (77%), and intravenous flecainide (71%). Taking into account the degree of evidence of each individual comparison between pharmacological agents, we have drawn up a ranking of pharmacological agents from the most effective to the least effective. In comparing the antiarrhythmic agents used to restore sinus rhythm in the case of paroxysmal AF, vernakalant, amiodarone-ranolazine, flecainide, and ibutilide are the most effective medications. The verapamil-quinidine combination seems promising, though few RCTs have studied it. The incidence of side effects must be taken into account in the choice of antiarrhythmic in clinical practice. PROSPERO: International prospective register of systematic reviews, 2022, CRD42022369433 (Available from: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022369433 ).