[Meta-analysis of therapeutic and preventive clinical trials]

Abstract

The paper surveyes the principles and the models of meta-analysis and gives the basic statistical formulae. For evaluation of data of clinical trials two illustrative examples are presented: 1. evaluation of association and homogeneity in fixed and random models of the common standardized treatment effects measured by continuous variable with known sample sizes, means and standard deviations; 2. evaluation of association and homogeneity in fixed model of: the common InRelative-Risk and Risk-Difference ot treatment effects determined by dichotomous variable categorized in 2 x 2 tables and completed with Stouffer and Fisher methods combining z and p values of percentage differences of the treatment effects. It is demonstrated that the presented two illustrative examples, as models of meta-analysis provide objective statistical methods for data summarization of clinical trials. Evidence based medicine applies also the methods of meta-analysis.