A hands-on practical tutorial on performing meta-analysis with Stata

Statistical synthesis of research findings via meta-analysis is widely used to assess the relative effectiveness of competing interventions. A series of three papers aimed at familiarising mental health scientists with the key statistical concepts and problems in meta-analysis was recently published in this journal. One paper focused on the selection and interpretation of the appropriate model to synthesise results (fixed effect or random effects model) whereas the other two papers focused on two major threats that compromise the validity of meta-analysis results, namely publication bias and missing outcome data. In this paper we provide guidance on how to undertake meta-analysis using Stata, one of the most commonly used software packages for meta-analysis. We address the three topics covered in the previous issues of the journal, focusing on their implementation in Stata using a working example from mental health research.