Conducting a High-Quality Systematic Review.

Abstract

Systematic reviews (SRs) are a structured means of knowledge synthesis used by a variety of healthcare practitioners to aid in medical decision making. The SR, if conducted rigorously, is considered to be at the top of the hierarchy for research studies. In addition to synthesizing evidence, SRs identify research priorities, address questions that may not be answerable by individual studies, and identify gaps to be addressed in future primary research. There are several steps that need to be taken when developing SRs to provide the best available evidence-the most essential being the assessment of risk of bias (ROB). Several ROB tools have been developed for use according to study design. Increasingly used is the assessment of certainty of evidence using approaches such as those developed by the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) working group. Whereas ROB is assessed for individual studies, the certainty of evidence is assessed for each critical or important outcome across studies. Analysis can be quantitative (metaanalysis) or qualitative (narrative), with the former intended to develop estimates of the effect measure (ie, the statistic that compares collated data), with confidence limits around that estimate. This review will focus on the steps required to develop SRs, from registration of the review protocol to the conduct, analysis, and reporting, with a focus on the assessment of ROB and certainty of evidence to ensure the development of a methodological and rigorous process.