Limitations:
Firstly, the included studies had different study protocols, including both randomized and nonrandomized trials. Secondly, the precise techniques of CBA for PVI and PWI might differ between different operators, while two of our studies used adjunct RFA, which might explain the high heterogeneity found in total ablation time. Thirdly, our sample size is small, too small to compare rare events such as AE fistula. Therefore, more randomized trials with larger populations are needed in the future.