Fig. 3

Graphics from MA1: a Forest plot displaying included studies and calculated treatment effects (SMD). b Funnel plot displaying publication bias. c Regression model of moderator analysis 1 showing the relationship between the treatment effect and the moderator “animal type”. d Pore-sizes of all the studies included in MA1 (weighted with the inverse of their variance for their size, represented with a dotted line in the case of the reported pore-size range) vs. their corresponding calculated TE. e Coplot displaying pore-sizes of all the studies included in the rabbit subgroup analysis vs. their corresponding TE (SMD) (below), but allocated to three graphs, corresponding to three clusters of porosity values (above)