Steps in the undertaking of a systematic review in orthopaedic surgery.

In the last decades of the twentieth century it became obvious that modern medical care is replete with data and information, but in need of reliable evidence. This has led to an increased effort to systematically synthesise medical research and make it more useful for practitioners. Systematic reviews use an approach to research synthesis that minimises the risk of misinterpreting a body of evidence due to incomprehensive search or subjective opinion. Carrying out a systematic review is a rigorous procedure which corresponds to standard methodological steps in primary research studies. It involves posing a well-defined question, developing a robust search strategy, screening for relevant primary studies, critical appraisal of included studies, data extraction and processing, analysis and interpretation of results. In some, but not all systematic reviews it is appropriate to conduct a meta-analysis, which is a statistical procedure that integrates the results of several independent studies. Results of meta-analysis are graphically presented in forest plots, with pooled point estimate and its confidence interval represented as a rhombus, usually called a "diamond". Methodological quality of systematic reviews should not be judged by the quality of primary studies included, but by a distinct set of criteria specified in assessment tools such as AMSTAR. Systematic reviews and meta-analyses should be reported according to the PRISMA checklist. A major contribution to the development of methodological standards has been given by The Cochrane Collaboration, whose Handbook of Systematic Reviews of Interventions is the primary reference for all authors and referees of systematic reviews in health care.