Exploring heterogeneity in meta-analysis: Meta-regression analysis.

Meta-regression is simply the extension of the regression analysis. Whereas in the regression analysis, the outcome and the covariates are measured for each subject, in meta-regression, the outcome and the covariates are measured on the level of the study. Because studies differ at least in the sample size, we need to assign a weight in each study under a specific model (fixed-effect or random-effects model) before implementing a meta-regression analysis. Weight assignment is not addressed in a regression analysis, because we study subjects. The differences between the regression and the meta-regression analysis are obvious, but the goal is the same: to assess the association between an outcome and one or more covariates