The science of research synthesis is rapidly evolving; hence the methods employed in the conduct of Cochrane reviews have developed over time. The aim of the Cochrane Handbook for Systematic Reviews of Interventions (the Handbook) is to help Cochrane review authors make appropriate decisions about the methods they use, rather than to dictate arbitrary standards. Wherever possible, recommendations are informed by empirical evidence. The guidance provided here is intended to help review authors to be systematic, informed and explicit (but not mechanistic) about the questions they pose and how they derive answers to those questions. Interpretation and implementation of this guidance requires judgement and should be done in conjunction with editorial bases of CRGs.
This Handbook focuses on systematic reviews of the effects of interventions. Most of the advice contained within it is oriented to the synthesis of clinical trials, and of randomized trials in particular because they provide more reliable evidence than other study designs on the relative effects of healthcare interventions (Kunz 2007). Some chapters, however, provide advice on including other types of evidence, particularly in forms of care where randomized trials may not be possible or appropriate and in considerations of safety or adverse effects. In 2003, The Cochrane Collaboration expanded its scope to include Cochrane Diagnostic test accuracy reviews. Guidance for the conduct of these reviews is contained in a separate document: the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy.
This Handbook has 22 chapters organized into three parts. Part 1 introduces Cochrane reviews, covering their planning and preparation, and their maintenance and updating, and ends with a guide to the contents of a Cochrane review or protocol. Part 2 provides general methodological guidance relevant to all Cochrane reviews, covering question development, eligibility criteria, searching, collecting data, within-study bias, analysing data, reporting bias, presenting and interpreting results. Part 3 addresses special topics that will be relevant to some, but not all, Cochrane reviews, including particular considerations in addressing adverse effects, meta-analysis with non-standard study designs and using individual patient data. This part has chapters on incorporating economic evaluations, non-randomized studies, qualitative research, patient-reported outcomes in reviews, prospective meta-analysis and reviews in health promotion and public health. A final chapter describes the new review type, Overviews of reviews.
Each chapter contains a list of key points to summarize the information and draw out the main messages for review authors.
The Handbook is largely prepared by The Cochrane Collaboration’s Methods Groups, whose members conduct much of the methodological and empirical research that informs the guidance.
Although the main intended audience for the Handbook is authors of Cochrane Intervention reviews, many of the principles and methods are applicable to systematic reviews applied to other types of research and to systematic reviews of interventions undertaken by others (Moher 2007).