Methods Statistik Metaanalyse

We calculated a pooled estimate of risk ratio (RR) with 95%-confidence interval (95%-CI) for each dichotomous outcome (XXX). Each trial-specific effect size was subsequently combined across studies in order to calculate summary estimates and presented as forest plot. We evaluated heterogeneity by the chi-squared test and calculated I². Significant heterogeneity was defined as an I² greater than 50% and chi-squared test p<0.1 [16].. The Mantel-Haenszel method was used for the random effects model. Heterogeneity was analyzed applying prediction intervals. The presence of publication bias was explored with a funnel plot. All statistical analyses were performed using R, version XXXX.