Systematic reviews are used by health care providers, policy makers and managers to inform healthcare decisions.1,2 Systematic reviews can also inform future research, and guidance from organizations specializing in the conduct of systematic review states that authors of reviews should provide clear recommendations for research in the systematic review report.3,4 Explicit recommendations for research are also advised in the PRISMA statement.1 However, the guidance is not as clear-cut when it comes to including recommendations for practice in systematic reviews. The focus of this editorial is to discuss whether recommendations for practice should be included in systematic reviews.

The PRISMA statement currently suggests that "a general interpretation of the results in the context of other evidence"1 is required in the conclusions section of a systematic review, which often takes the form of an "Implications for Practice" section.1 However, the statement does not require systematic review authors to make recommendations within this section. The Cochrane Collaboration and the GRADE working group guidance goes one step further, and specifically states that systematic reviews should not include recommendations for practice.4,5 It is the view of these organizations that making recommendations for practice is beyond the scope of a systematic review. As we know, recommendations should take into account more than just the findings of research, and consider the balance between desirable and undesirable effects, the quality of the evidence, values and preferences, and costs.6,7 Therefore it is the view of these organizations that systematic review authors themselves are not in the ideal position to create such recommendations, as they are "not in a position to weigh the trade-offs between the desirable and undesirable consequences of adhering to a recommendation."5(p403) As such, recommendations are seen as the purvey of clinical practice guideline developers only.4

Clinical guidelines can be defined as "statements that include recommendations intended to optimize patient care that are informed by a systematic review of evidence and an assessment of the benefits and harms of alternative care options."8 We know that not all clinical guidelines are created equally, and that only a percentage of those touted as evidence-based guidelines can be "trusted". Characteristics of "trustworthy" clinical guidelines include the use of a rigorous development methodology, clear reporting of recommendations linked to the evidence, include systematic reviews in their development, and are conducted using a transparent process including extensive external reviews.9 There are many examples of such guidelines, with the amount of guidelines developed increasing significantly in recent years.10

We agree with the view that trustworthy clinical guidelines, following a rigorous methodology and taking into account the evidence, the balance between benefit and harms, patient values and preferences and resources, are the ideal source of recommendations for practice. We would also suggest that in the ideal situation, clinical guidelines would include in their development not only systematic reviews on the effectiveness of interventions for certain conditions, but also the feasibility, appropriateness and meaningfulness of healthcare practices. However, the development of a "trustworthy" clinical guideline is a drawn out process, requiring adherence to strict methodology and significant investment of resources. As such, for many conditions and practice areas there is a distinct lack of trustworthy clinical guidelines, particularly in nursing and the allied health fields.

Given the lack of high quality clinical guidelines for all areas of healthcare, and the difficulty in creating these, we suggest that recommendations can and should be included in systematic reviews as well. The Joanna Briggs Institute have a long history of pragmatism,11 evidenced by our long history of advocating for the best available evidence to be used to inform practice. For example, although we know that the ideal systematic review investigating therapeutic interventions would contain only randomized controlled trials, given the lack of these study types in some fields of health, since our inception we have allowed authors to include "lower level" evidence in systematic reviews being submitted to this journal. Similarly, although we know that clinical practice guideline developers and their panels are the best placed to make recommendations for practice, given the lack of high quality guidelines for all areas of practice, we suggest that systematic reviewers who have conducted a comprehensive search and rigorous evaluation of the literature are also able to provide recommendations for care.3

In reality, this approach is not too dissimilar from other organizations conducting systematic reviews. Joanna Briggs Institute reviews have an "Implications for Practice" section similar to systematic reviews produced through other organizations. The main difference is that we recommend providing useful, actionable and explicit recommendations for practice in this section, rather than the indefinite, ambiguous and implicit implications seen in other reviews. Of course, the recommendations made in JBI systematic reviews should be interpreted carefully by the reader, and should be seen as either indicatory and/or interim recommendations rather than absolute, as in most cases they would not have been developed following the processes recommended for trustworthy clinical guidelines. However, these recommendations can be used as a basis for the development of more qualified recommendations in clinical guidelines or, in their absence, to inform (but not necessarily direct) clinical practice.

Systematic reviews are the pillar of evidence-based healthcare, playing a key role in informing guideline development and clinical decision making. We argue that systematic reviews can (and should) provide recommendations for practice that take into account the feasibility, appropriateness, meaningfulness and effectiveness of health care practices.

For guidance on the JBI approach to creating recommendations for practice, view the information and supporting documents available at:http://joannabriggs.org/jbi-approach.html#tabbed-nav=Grades-of-Recommendation

Director, Transfer Science, The Joanna Briggs Institute

Conflicts of interest:

Dr Munn is the convener of the Joanna Briggs Institute Levels of Evidence and Grades of Recommendation methodological group and current chair of the Scientific Committee. He is also a member of the GRADE working group and the Guidelines International Network Tech group and Australia and New Zealand regional group.

References