Meta-analyses of drug efficacy and safety data obtained from clinical trails

Systematic reviews and meta-analyses of drug efficacy and safety data are essential elements of evidence based medicine. By promoting research activities in clinical practice, the number of clinical studies rapidly increases, and consequently increasing the number of studies with meta-analyses that compare the results of similar clinical trials.

In the lecture, the concept of meta-analysis and effect size will be presented. Additionally, statistical models/methods that have been developed in the field of meta-analysis will be described; namely, a fixed effect model, a random-effect model (DerSimonian and Laird method), an indirect comparison model (Bucher method) and mixed-treatment comparison model (or network meta-analysis). The importance of estimation of homogeneity of studies included in the meta-analysis will be shown. Graphical presentation of the results of meta-analyses such as forest plot and funnel plot will be described. In the end, the Bayesian approach to meta-analysis will be discussed.