To test the robustness of results, an analysis of the interstudy variability
was carried out by comparing interstudy variance with intrastudy
variance. It was recommended to determine, before the analysis,
what size of estimated variance attributable to study would be considered
negligible