Effect of Prophylactic Amiodarone on Clinical and Economic Outcomes After Cardiothoracic Surgery: A Meta-Analysis

BACKGROUND: Two previous meta-analyses of amiodarone for prevention of postoperative atrial fibrillation (POAF) after cardiothoracic surgery did not evaluate total hospital cost, concluded that data on stroke are incomplete, and did not evaluate the effect of clinical heterogeneity between trials. OBJECTIVE: To conduct a meta-analysis examining amiodarone's prophylactic impact on cardiothoracic surgery POAF, length of stay (LOS), stroke, and total costs. METHODS: Three reviewers conducted a systematic literature search of MEDLINE, EMBASE, CINAHL, and the Cochrane Library (1966-SEPTEMBER 2004). Studies were included if they met the following criteria: (1) randomized controlled trial versus placebo/routine treatment, (2) coronary artery bypass graft and/or valvular surgery, (3) Jadad score > or = 3, (4) reported data on incidence of POAF or stroke, LOS, or total costs, (5) used electrocardiographic/Holter monitoring, and (6) monitored subjects for > or = 2 days. A random-effects model was utilized. Subgroup and sensitivity analyses were conducted. RESULTS: Fifteen trials were identified, including 1512 and 1429 patients in the amiodarone and control groups, respectively. Amiodarone reduced POAF (OR 0.50; 95% CI 0.42 to 0.60) and decreased stroke (n = 8 studies), LOS (n = 10), and total costs (n = 6) (OR 0.47; 95% CI 0.23 to 0.96; -0.73 days, 95% CI -0.95 to -0.51; and -dollar 1619, 95% CI -3395 to 156, respectively). Surgery type, beta-blocker use, route of administration, use of a fixed-effects model, or exclusion of unblinded/unpublished studies did not affect the overall results. No statistical heterogeneity was observed for any endpoint evaluated (p > 0.22 for all comparisons). CONCLUSIONS: Prophylactic treatment with amiodarone decreases patients' risk of POAF and stroke while reducing LOS.