Critical Appraisal of Systematic Review/Meta-analysis

Kim, Hyun Jung; Ahn, Hyeong Sik

Abstract

In recent years, the number of systematic review and meta-analysis literatures has markedly increased. Systematic reviews can give important information about clinical decision making when studies show different or even contradictory results. By utilizing systematic reviews, clinicians can get unbiased summaries of the estimates, which are a reliable source of clinical information. Contrary to the narrative review, the systematic review conducts a comprehensive research of relevant studies on a defined clinical question, and critically appraises the risk of bias in included studies. The systematic review usually includes meta-analysis which summarizes the quantitative estimates by using statistical methods. Most meta-analyses aggregate data from primary studies, but individual data are also commonly used. Explaining heterogeneity among included studies and subsequent subgroup analysis are often required. Systematic review and meta-analysis depend on the quality of included studies, and subsequently cannot overcome the limitations of primary studies. Also, meta-analysis is prone to publication biases and methodological flaws. Despite these limitations, systematic review has definitive strengths. Maximizing these strengths require reliable second-hand data and a comprehensive analysis.