Fig. 4

Assessment for publication bias of included studiesThe funnel plot shows the relationship between the effect size and the sample size of the studies included in the systematic review. Studies with high variability and effect sizes near 0 are missing and there are a number of studies which lie outside the expected funnel. In particular it is clear that small studies that have been published are those with relatively large effect sizes (those points on the lower right of the plot). This funnel plot asymmetry suggests publication bias.