External electrical and pharmacological cardioversion for atrial fibrillation, atrial flutter or atrial tachycardias: a network meta-analysis

Background: Atrial fibrillation (AF) is the most frequent sustained arrhythmia. Cardioversion is a rhythm control strategy to restore normal/sinus rhythm, and can be achieved through drugs (pharmacological) or a synchronised electric shock (electrical cardioversion). Objectives: To assess the efficacy and safety of pharmacological and electrical cardioversion for atrial fibrillation (AF), atrial flutter and atrial tachycardias. Search methods: We searched CENTRAL, MEDLINE, Embase, Conference Proceedings Citation Index-Science (CPCI-S) and three trials registers (ClinicalTrials.gov, WHO ICTRP and ISRCTN) on 14 February 2023. Selection criteria: We included randomised controlled trials (RCTs) at the individual patient level. Patient populations were aged ≥ 18 years with AF of any type and duration, atrial flutter or other sustained related atrial arrhythmias, not occurring as a result of reversible causes. Data collection and analysis: We used standard Cochrane methodology to collect data and performed a network meta-analysis using the standard frequentist graph-theoretical approach using the netmeta package in R. We used GRADE to assess the quality of the evidence which we presented in our summary of findings with a judgement on certainty. We calculated differences using risk ratios (RR) and 95% confidence intervals (CI) as well as ranking treatments using a P value. We assessed clinical and statistical heterogeneity and split the networks for the primary outcome and acute procedural success, due to concerns about violating the transitivity assumption. Main results: We included 112 RCTs (139 records), from which we pooled data from 15,968 patients. The average age ranged from 47 to 72 years and the proportion of male patients ranged from 38% to 92%. Seventy-nine trials were considered to be at high risk of bias for at least one domain, 32 had no high risk of bias domains, but had at least one domain classified as uncertain risk, and one study was considered at low risk for all domains. For paroxysmal AF (35 ...