Figure 6.

Assessment of publication bias for mortality, disease progression analysis, lymph node metastasis and distant metastasis. In each funnel plot analysis, black dots represent each study's effect estimate (drawn on a log scale) plotted against its standard error. The outer dashed lines represent the 95% confidence limits around the summary effect estimate, within which 95% of studies are expected to lie in the absence of either biases or heterogeneity. P-values are from the results of Egger's test to assess publication bias.