2.5 Statistical analysis
We performed meta-analysis of included studies using a random-effects model and the generic inverse-variance method of Der Simonian and Laird to calculate pooled HR [18]. We extracted from these studies the freedom from AT rates and complications rates. For the analysis of pooled recurrent rate, if the study does not provide HR, we manually calculate HR with methods by Tierney et al. [16]. For the analysis of adverse events, we calculate pooled OR, or OR if outcome was available from one study only. The heterogeneity of effect size estimates was assessed using forest plots to detect non-overlapping confidence interval (CI), and then was calculated using the Q statistic and I2 statistic. For the Q statistic, substantial heterogeneity was defined as p< 0.10. The I2 statistic ranges in value from 0 to 100% (I2< 25%, low heterogeneity; I2= 25%–50%, moderate heterogeneity; and I2> 50%, substantial heterogeneity). A sensitivity analysis was performed to assess the influence of the individual studies on the overall results by omitting one study at a time. Publication bias was assessed using funnel plot and Egger’s regression tests [19] (p< 0.05 was considered significant). All statistical tests were performed using the STATA 14.2 software (College Station, TX).