Systematic reviews bring together research evidence, helping people who need to make decisions by providing them with a summary that minimises bias and avoids too much emphasis on the findings of individual studies. Systematic reviews are scientific, research activities. They are increasingly common, not least because of the work of The Cochrane Collaboration (http://www.cochrane.org). Several thousand new reviews are published annually across all areas of health, in hundreds of journals. A few hundred of these new reviews each year are Cochrane reviews (http://www.thecochranelibrary.com). Almost all of the 3700 full Cochrane systematic reviews currently available investigate the effects of healthcare interventions. But, there are also Cochrane reviews of research into methodological issues in healthcare evaluation and diagnostic test accuracy. More generally, most systematic reviews are of treatments for illness but reviews have also been done in other areas, including risk factors, prognosis, and genetic predisposition to disease.
The starting point for any systematic review should be a clear question to be answered. This determines which types of study should be included: randomised trials for comparing the effects of interventions, case control or cohort studies for risk factors, etc. A focused question helps when planning the searches that will be needed to find research evidence for the review. This searching should be wide ranging, to minimise the impact of publication bias which can limit the availability of some types of results, usually those that do not favour the new intervention. Another advantage of a wide search may be that it boosts the amount of data available for the review. This makes it more likely that the review will be able to detect small differences or associations, through increased statistical power.
Traditional, non-systematic review articles can be misleading. The authors of such articles, who might be recognised experts in the topic, may provide more of an opinion than an assessment of the existing evidence. They might not have the time to identify and bring together all relevant studies, or they might selectively discuss and combine research that confirms their views and prejudices. Systematic reviews seek to avoid these problems by using a predefined, explicit methodology. The methods used include steps to minimise bias in the identification of relevant research, the selection of studies for inclusion and the collection, analysis and interpretation of their findings.
Systematic reviews do not always contain a statistical synthesis, or meta-analysis, of the results from the included studies. This might not be sensible if the designs of the studies are too different for an average of their results to be meaningful or it might not be possible if the data available from the studies are incompatible. Some systematic reviews are “scoping” or “mapping out” reviews, in which the existing research is identified, appraised and summarised, with the primary aim being to learn from what has already been done, in order to carry out better studies in the future.
The factors that are important when conducting a systematic review are also the ones to remember when reading one. Questions to ask as you read a review include: How were studies identified? Is all the relevant research included? Are the analyses appropriate?
This section is based on material prepared within TENALEA, a European project to make available an online system for the registration and randomisation of participants into research studies in health care. TENALEA has been developed by the National Cancer Institute in Amsterdam in The Netherlands, working with partners in some of the largest cancer research groups in Europe. It is now being deployed, with funding from the European Union (contract C029334). More information, including details of how to use the system, is available at http://www.tenalea.com.