Systematic reviews systematically evaluate and summarize current knowledge and have many advantages over narrative reviews. Meta-analyses provide a more reliable and enhanced precision of effect estimate than do individual studies. Systematic reviews are invaluable for defining the methods used in subsequent studies, but, as retrospective research projects, they are subject to bias. Rigorous research methods are essential, and the quality depends on the extent to which scientific review methods are used. Systematic reviews can be misleading, unhelpful, or even harmful when data are inappropriately handled; meta-analyses can be misused when the difference between a patient seen in the clinic and those included in the meta-analysis is not considered. Furthermore, systematic reviews cannot answer all clinically relevant questions, and their conclusions may be difficult to incorporate into practice. They should be reviewed on an ongoing basis. As clinicians, we need proper methodological training to perform good systematic reviews and must ask the appropriate questions before we can properly interpret such a review and apply its conclusions to our patients. This paper aims to assist in the reading of a systematic review.
Yuan et al. (2009) studied this question.