Meta-epidemiological study: a step-by-step approach by using R.

The results of meta-analysis are usually influenced by characteristics of component trials. These characteristics include but are not limited to sample size, publication language, quality of study design and the number of participating centers. Because systematic review and meta-analysis is considered to be evidence of the highest quality in the era of evidence-based medicine, characteristics that may bias the results of meta-analysis should be fully investigated. Meta-epidemiological study aims to investigate the impact of study design characteristics on treatment effect. The article introduces how to perform meta-epidemiological study by using R software. There are several methodological approaches to perform meta-epidemiological study, depending on whether there is heterogeneity between meta-analyses and between trials. When there is no significant heterogeneity, these results would be similar.