plainlanguage

A plain language summary is a summary of evidence that a general audience, without a medical or research background, would understand and be able to use. Plain language summaries of systematic reviews of the effects of treatments include background information about the health condition, the population, and the treatment (any action intended to improve health or relieve suffering) and information about the main findings of the review. It also includes a comment on adverse effects and on any limitations of the review. An example can be seen here.

We have tested different formats for plain language summaries, and are continuing to improve how we present the key findings of systematic reviews in plain language, to enable people everywhere to use this information when making decisions. Researchers from several Cochrane entities have joined forces to develop and test a common format for Cochrane Plain Language Summaries with funding from the Cochrane Collaboration [Glenton 2010]. This work has been based on earlier efforts to develop Summary of Findings tables [Rosenbaum 2010a, Rosenbaum 2010b], summaries of evidence for policy makers [Rosenbaum 2011] and for consumers [Glenton 2002, Glenton 2006a, Glenton 2006b, Santesso 2006]. In addition, this work was influenced by a growing body of evidence regarding the presentation of research results to patients. This evidence suggests the following:

• Patients’ perceptions of risk appear to be more accurate when they are presented with numbers rather than words [Trevena 2006]

• Patients’ perceptions of risk appear to be more accurate when they are presented with absolute rather than relative risk formats [Akl 2011]

• Tables may be preferred and understood by consumers better than narratives [Schünemann 2004, Schwartz 2009]

• Consistency in the numeric formats used and the avoidance of comparing ‘‘apples and oranges’’ is increases comprehension [Lipkus 2007]

However, research is still lacking for other aspects of how to present research findings. For instance, we know little about how to convey uncertainty around the magnitude of effect [Politi 2007] or how to convey the certainty of the evidence.