[Systematic reviews in practice. VIII. Searching and assessing systematic reviews]

In a systematic review (SR) the information from the original studies is searched, selected, extracted, appraised and summarised in a transparent, valid and reproducible manner. SRs are therefore an efficient and reliable source of information for the clinician, health policy maker and researcher. SRs can be identified in various sources. The search strategy is determined by the desired comprehensiveness, the availability of various sources and the need for additional comments on the SR by a referee. In Medline a sensitive filter is available for searching SRs. An SR should be based on a carefully formulated clinical question that can be answered on the basis of data from the original studies. SRs should be thoroughly assessed as they can differ substantially in terms of quality. In such an appraisal the following aspects should be considered: question, search strategy, selection of studies, quality assessment, data extraction and data presentation, (statistical) summary of the data, statistical and clinical heterogeneity, results and conclusion. Various assessment lists are available. Differences in conclusions between SRs on the same subject can be analysed in a systematic manner and can often be resolved.