Pharmacologic and non-pharmacologic interventions for preventing postoperative atrial fibrillation in cardiac surgery: systematic review and network meta-analysis

Objective Comparing pharmacologic and non-pharmacologic interventions for preventing postoperative atrial fibrillation in open cardiac surgery with control. Design Systematic review and network meta-analysis. Data sources PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials. Methods Following a pre-registered protocol, screening, data extraction, and risk-of-bias assessments were performed. Certainty of evidence was evaluated using the CINeMA framework. A Bayesian network meta-analysis was constructed to compare pharmacological and non-pharmacological interventions, with robustness checked via sensitivity and heterogeneity analyses. Secondary outcomes included mortality, hospital length of stay, and ICU duration. Results 392 randomized controlled trials with 830 arms involving 82490 participants undergoing open cardiac surgery were included. Pharmacological and non-pharmacological interventions were evaluated in 319 and 73 studies, respectively. In pharmacological intervention, compared with controls, the most robust prophylactic effects were observed for the combination of amiodarone plus beta-blockers (OR 0.23, 95% CrI 0.13 to 0.42; NNT 4.85) and magnesium plus beta-blockers (OR 0.26, 95% CrI 0.16 to 0.42; NNT 5.13). Beta-blocker monotherapy was found to be slightly superior to amiodarone alone. Across various beta-blocker types, sotalol demonstrated higher robustness in the prevention of POAF. Among non-pharmacological approaches, the combination of amiodarone and pacing was shown to be more effective (OR 0.23, 95% CrI 0.09 to 0.60; NNT 4.77). Within the pacing modalities, the greatest efficacy was achieved with biatrial pacing (OR 0.29, 95% CrI 0.19 to 0.45; NNT 5.33). No significant impact on secondary outcomes was identified for the studied interventions. Conclusions Combination therapies were found to be superior to monotherapy for the prevention of POAF. Among single-agent interventions, amiodarone and beta-blockers remained the most robust options, demonstrating comparable prophylactic efficacy. Regarding non-pharmacological strategies, the most favorable outcomes were atrial pacing combined with amiodarone and biatrial pacing alone.