A Practical Guide to Conducting Meta-Analysis in Stata 17 for Public Health Research: Applications to Binary, Continuous, and Prevalence Data

Abstract

Meta-analysis is a quantitative statistical method used to synthesise findings from multiple independent studies addressing the same research question. It improves statistical power, increases precision, and supports evidence-based decision-making. This manuscript provides a practical, step-by-step guide for conducting meta-analysis using Stata version 17. Three common applications are demonstrated: (1) meta-analysis of binary outcomes using risk ratios, (2) meta-analysis of continuous outcomes using standardised mean differences (Hedges’ g), and (3) meta-analysis of prevalence data using the Freeman–Tukey double arcsine transformation. For each example, detailed instructions are provided for data preparation, computation of effect sizes, assessment of heterogeneity (Q statistic, I², τ²), generation of forest and funnel plots, and evaluation of publication bias using Egger’s test. Results from the worked examples illustrate key concepts including interpretation of pooled estimates, weighting of studies, and implications of heterogeneity. The guide is designed for clinicians, public health researchers, and graduate students seeking a clear and reproducible framework for performing meta-analysis in Stata. By combining conceptual explanations with practical commands, this manuscript serves as a comprehensive introductory resource for applied meta-analysis.