Meta-Analysis of Non-Comparative Binary Outcomes and Its Solution by Stata

Objective To introduce the method of meta-analysis for non-comparative binary outcomes and its realization in Stata. Methods We first introduced principles and methods of meta-analysis for three types of non-comparative binary outcomes, and then replicated results of three published meta-analyses in Stata. Results The key point of doing these meta-analyses was to choose the effect size indices which were of normal distribution or could be transformed into normal distribution. The replicated results were consistent with the original literatures. Conclusions Meta-analyses for three types of binary outcomes, including prevalence, incidence density, and odds, could be done in Stata conveniently.