Table 2. Risk of bias for included studies.

Risk of bias assessment: each article was given a rating (low=1, medium=2, high=3) according to the NOS. Discrepancies were resolved by a third reviewer if necessary

NOS: Newcastle-Ottawa Scale

Author	Reviewer 1	Reviewer 2	Reviewer 3
Cai et al. [7]	2	3	2
Cao et al. [8]	3	3	N/A
Colaneri et al. [9]	3	3	N/A
Feng et al. [10]	3	3	N/A
Hu et al. [11]	3	3	N/A
Huang et al. [12]	3	2	2
Regina et al. [13]	3	3	N/A
Wan et al. [14]	3	3	N/A
Yan et al. [1]	3	3	N/A
Zhang et al. [15]	3	3	N/A
Zhao et al. [16]	3	2	3
Cao et al. [17]	3	3	N/A
Brill et al. [18]	2	2	N/A
Chan et al. [19]	3	3	N/A
Hirsch et al. [20]	3	3	N/A
Pei et al. [21]	3	3	N/A
Rubin et al. [22]	3	3	N/A
Wang et al. [23]	3	3	N/A
Zhang et al. [3]	2	3	2
Zhou et al. [2]	3	3	N/A