Figure 3.
Random-effects meta-analyses using Sidik-Jonkman estimation method demonstrated no statistically significant differences in infection with the two treatment modalities.