Quality assessment and publication bias
According to the Cochrane method of evaluating RCTs, trials of acceptable to good quality were found (Supplementary Table 5). The funnel plots indicated that publication bias had no impact on the quantitative results (Supplementary Figure 1).