Evidence-based research is used to generate, summarise, and understand the best available practices to inform decision-making. Systematic reviews (SR) and meta-analysis (MA) have become a valuable tool for these goals in public health, medicine and pharmaceutical research. MA is a statistical procedure for combining the results of multiple studies investigating a common intervention or issue to produce a pooled effect size and evaluate interventions' efficacy across studies. This article outlines the usefulness of systematic reviews and meta-analysis in medicine and public health. Steps in undertaking the systematic reviews and meta-analysis include forming a team, identifying and refining the research question, determining the inclusion and exclusion criteria, registering the SR and MA protocol, searching for the studies, selection of the studies, data extraction, data analysis and presenting the results. The review also outlines the issues that can impact meaningful meta-analysis. The heterogeneity in the included studies, conditions studied, interventions, and end-point measures is one of the major issues encountered in meta-analysis. Quantification of the heterogeneity can be done by I2 statistics and prediction intervals. Sub-group analysis, outlier detection followed by sensitivity analysis, and meta-regression can be applied to explore and reduce heterogeneity. Random effects model, Knapp-Hartung, likelihood estimates, and Bayesian models can be applied in a highly heterogenous meta-analysis.
Gandhi et al. (Tue,) studied this question.