Background: Post-operative atrial fibrillation is a common complication of cardiac surgery and has been associated with increased incidence of other complications including post-operative stroke, increased hospital length of stay and increased cost of hospitalisation. Prevention of atrial fibrillation is a reasonable clinical goal and, consequently, many randomised trials have evaluated the effectiveness of pharmacological and non-pharmacological interventions. We systematically reviewed the literature and prepared meta-analyses to better understand the role and effects of various prophylactic therapies against post-operative atrial fibrillation.
Objectives: To assess the effects of pharmacological and non-pharmacological interventions for preventing post-cardiac surgery atrial fibrillation.
Search strategy: We searched CENTRAL, MEDLINE, EMBASE and CINAHL from earliest achievable date to June 2003. We hand searched references from reports and earlier reviews. We searched abstract books and CD-ROMs from annual scientific meetings of American College of Cardiology, American Heart Association, North American Society of Pacing and Electrophysiology and European Heart Organization between 1997-2003. No language restrictions were applied.
Selection criteria: Randomised controlled trials comparing pharmacological interventions or non-pharmacological interventions with control treatment, placebo or usual care for the prevention of post-operative atrial fibrillation in post-coronary artery bypass grafting or combined CABG and valvular surgery.
Data collection and analysis: Two reviewers assessed trial quality and extracted data. Study authors were contacted for additional information.
Main results: Fifty eight studies were included with a total of 8565 participants. Interventions included were amiodarone, beta blockers, solatol and pacing. Results favoured treatment for post-operative atrial fibrillation. The data for stroke favoured treatment by a non-significant effect size of 0.81, 95% confidence interval 0.51 to 1.28. Similarly, a positive indication for length of stay was derived but it too was not significant with a weighted mean difference of -0.66, 95% confidence interval -0.95 to -0.37. A positive result for cost of hospitalisation in favour of treatment was achieved, but the statistic is not significant due to low power and large standard deviations: a weighted mean difference of -2717, 95% confidence interval 7518 to 2084. Beta-blockers had the greatest magnitude of effect across 28 trials (4074 patients) with an odds ratio (random) of 0.35, 95% confidence interval 0.26 to 0.49. Across all treatment, the odds ratio favoured treatment with a ratio (random) of 0.43, 95% confidence interval 0.37 to 0.51.
Reviewers' conclusions: Intervention is favoured across the three pharmacological interventions studied and the one non-pharmacological intervention, pacing. The length of stay data favoured treatment (-0.66, 95% confidence interval -0.95 to -0.37).