An introduction to systematic reviews and meta-analyses

Abstract

A high quality systematic review with meta-analysis is considered to provide a high level of evidence in health-care research. It is important to follow rigorous methodology while conducting a systematic review and reporting its findings. To minimize the chances of error and bias, at least two (preferably 3–4) authors should be involved at all stages of the conduct of a systematic review. Registering the protocol on open websites such as PROSPERO is essential to enable transparency and scientific rigor. Important steps are writing the protocol, conducting detailed literature search through multiple databases, identifying studies that meet strict inclusion criteria, abstracting the data, contacting authors to obtain additional information, assessing the risk of bias of included studies, synthesizing evidence qualitatively, conducting meta-analysis using appropriate statistical models, assessing publication bias and heterogeneity, exploring heterogeneity through sensitivity and subgroup analyses, grading the certainty of evidence following the GRADE framework, reporting using the preferred reporting items for systematic reviews and meta-analyses guidelines, describing the limitations and providing implications for clinical practice and future research. Clinicians should evaluate the credibility of systematic reviews prior to applying their results in clinical practice. This article provides a brief introduction to the methodology of systematic reviews and meta-analyses and also provides links to some high-quality free resources for prospective systematic reviewers.