It used to be easy to find systematic reviews. The Cochrane Library in combination with the Database of Abstracts of Reviews of Effects (DARE) in theory together covered all systematic reviews of effectiveness. The UK Department of Health and the National Institute for Health Research (NIHR) initially funded the production of DARE in 1994. Unfortunately, funding has ceased and new records have no longer been added to DARE since March 2015. An archive version of DARE can be accessed via the Centre for Reviews and Dissemination's website.1 The Database of Abstracts of Reviews of Effects included systematic reviews that evaluate the effects and delivery of health and social care interventions.
What now? Of course we still have MEDLINE, Embase and other databases where one can search for systematic reviews. However, this is not easy and very laborious, so there is a need to fill the gap left by the absence of DARE. Individual collections of systematic reviews are still available, such as those from the Joanna Briggs Institute or Cochrane. Fortunately we also have some alternatives that aim to identify all systematic reviews, not only covering interventions, but also other topics such as etiology, diagnosis or prognosis.
Epistemonikos states on its home page that it “combines the best of Evidence-Based Health Care, information technologies and a network of experts to provide a unique tool for people making decisions concerning clinical or health-policy questions.”2 (para.1) Its aim is to gather scientific information (i.e. evidence) that might be relevant for health decision-making, and to provide rapid and reliable access to the best available evidence for real-life questions. The target users are health professionals, researchers and health decision-makers. The Epistemonikos Foundation and Cochrane have established a partnership aimed at improving the knowledge base for making decisions in healthcare and global health policy.
Epistemonikos was established in 2009, and is a health evidence database that focuses on systematic reviews and related publications, including overviews of reviews, primary studies and structured summaries. These publications are sourced by a team of collaborators who regularly look through a range of resources such as PubMed, Cochrane and many more. It is multilingual and it is free to use. Signing up allows users to save searches and documents. The search interface offers a basic search option which supports nine different languages or a more advanced search option limited to English which enables searches by author, title and/or abstract. The advanced search option also allows the use of Boolean operators. Results are displayed by relevance and can be filtered by year or publication type. Links to free full-text articles are provided if these are available. The information provided includes the citation and the abstract of the paper. A useful feature is the links provided to evidence related with the article found, e.g. links to the papers included in a systematic review. However, there is no critical appraisal of evidence.
KSR Evidence has been developed by Kleijnen Systematic Reviews Ltd with the aim of helping users make sense of the ever-increasing volume and complexity of evidence, and provides a user-friendly, time-efficient way to access systematic reviews in healthcare.3 The database aims to include all systematic reviews and meta-analyses of all health topic areas published since 2015 (some topics date back from 2010), and for many reviews provides a short, accessible bottom line which gives the key clinically relevant implications of the results reported in the review and includes information about the reliability of these results. KSR Evidence is not free; users need to have a subscription.
In KSR Evidence, sensitive and advanced search strategies have been developed to identify systematic reviews and meta-analyses. These searches encompass a large number of databases including MEDLINE, Embase and the Cochrane Library. The results are checked by a team of information specialists, and any record that may be a systematic review is then assessed based on the full paper by a systematic reviewer. Articles are included if they claim to be either a systematic review or a meta-analysis, or if the methodology used in the paper is that of a systematic review.
Only 29% of all systematic reviews have been rated at low risk of bias in KSR Evidence. Systematic reviews are critically appraised using criteria based on the Risk Of Bias In Systematic Reviews Tool (ROBIS).4 An adapted version of this tool is used to provide the necessary summary information which is useful for busy healthcare professionals, researchers and other end users. The risk of bias tool looks at four domains: study eligibility criteria; identification and selection of studies; data collection and study appraisal; and synthesis and findings. From these domains an overall summary of the risk of bias and a bottom line are provided. The bottom line will inform users if the review has been conducted in line with accepted methods of undertaking systematic reviews and if the evidence is reliable.
Users can search the database with basic or advanced search options, and filter results according to risk of bias assessment, year, subject area, author or journal. All critical appraisals are written by one reviewer and checked by a second reviewer. Given that 19,000+ systematic reviews are currently published per annum, critical appraisals are continually added, but it is not possible to do this for all of them. If no assessment is available, KSR Evidence gives subscribers the option to request a fast track assessment of any review for a nominal administrative fee. As of April 2018, there are over 70,000 systematic reviews in KSR Evidence and over 6300 critical appraisals.
Comparing Epistemonikos and KSR Evidence, both aim to provide access to all systematic reviews, and both can be the first place/port of call for end users – clinicians, researchers, consumers and policy makers looking for systematic reviews for whatever purpose. The main differences are that the former is free to use, whereas the latter requires a subscription; the former provides links to evidence related to articles found which can be really useful for a systematic reviewer, while the latter provides bottom line summaries and assessments of risk of bias which can be really useful to health care professionals, decision makers and guidance or guideline developers.