Skip to main content
. 2019 Mar;24(2):e217–e230. doi: 10.4317/medoral.22786

Figure 2.

Figure 2

a) Funnel plot for all publications. Four publications (3,10 [including results prior to corticosteroid therapy in “active” oLP only],11,13) were found to reveal too extreme rate estimates for implant failure. b) Funnel plot for all publications following exclusion of outliers (3,10 [including results prior to corticosteroid therapy in “active” oLP only],11,13), revealing no risk of bias.