3.4 Risk of bias and sensitivity analysis
Visual inspection of the funnel plot showed no evidence of a significant
publication bias(Fig.7 ), confirmed by Egger’s test for
publication bias (P = 0.9008) (Supplementary Fig.5 ).
Leave one out sensitivity
analysis was performed to re-estimate the pooled incidence rate of
adverse events. The results revealed that no single study had a
significant impact on the overall effect (Supplementary Fig.6 ).