Six RCTs (n=693) were included in the review. Four RCTs had moderate-to-high quality, with summary scores of 9 or 10 points. Two studies scored fewer than 5 points and were considered poor quality. Three trials reported significant crossover from the control to the intervention arm.
There was a statistically significantly lower recurrence of atrial fibrillation at one year in the radiofrequency groups (risk ratio was 0.33, 95% CI: 0.21 to 0.51, p<0.001; six RCTs, n=693). The absolute risks of recurrence were 24 per cent in the radiofrequency ablation groups compared with 73 per cent in the control groups. There was statistically significant heterogeneity in the analysis (Χ2 p<0.001, I2=75.9%). Exclusion of the trial reported only as an abstract did not significantly alter the results of the analysis (risk ratio was 0.35, 95% CI: 0.21 to 0.57, p<0.001; five trials), which also showed significant heterogeneity. Further sensitivity analyses failed to explain the heterogeneity.