Rhythm vs. rate control of atrial fibrillation meta-analysed by number needed to treat.

Abstract

Whenever feasible, rhythm control of atrial fibrillation (AF) was generally preferred over rate control, in the belief that it offered better symptomatic relief and quality of life, and eliminated the need for anticoagulation. However, recent trials appear to challenge these assumptions. To explore the desirability of rhythm vs. rate control of AF by systematic review of pertinent, published, randomized controlled trials (RCTs) and a meta-analysis by number needed to treat (NNT) year(-1), with respect to diverse clinically important outcomes. RCTs of outcome primarily comparing rate vs. rhythm control in patients with spontaneous AF were identified. For each outcome and assuming rhythm control to be the active treatment, relative risk reduction (RRR) and NNT year(-1) were derived for individual trials together with an NNT year(-1) for all trials combined; corresponding 95% confidence intervals (CI) were also calculated. Adverse drug reaction (ADR) and quality of life reporting were also assessed. In all, data from five suitable RCTs (entailing 5239 patients) were analysed. For hospitalization, available RRRs and NNT year(-1) values were all clinically and statistically significant. Overall, one additional patient was hospitalized for every 35 assigned to rhythm control (95% CI 27, 48). For the endpoints of death, 'ischaemic' stroke and 'non-CNS' bleeding, there was no significant difference. ADRs were significantly more common in rhythm control patients, whereas quality of life assessments revealed no difference. Thromboembolism was associated with cessation of or subtherapeutic anticoagulation, irrespective of treatment assignment. Reduced risk of hospitalization and non-inferiority for other endpoints all favour rate control, the less costly strategy. If symptoms of AF are not a problem, treatment should target optimizing rate control and more widespread and effective prophylactic anticoagulation.