What instructions are available to health researchers for writing lay summaries? A scoping review

Protocol and registration

We conducted a scoping review using methods outlined in the PRISMA extension for scoping reviews(13) and information in the Joanna Briggs Institute Reviewers’ Manual for scoping reviews.(14) A protocol for this study was completed prior to data analysis and is on Open Science Framework.(15)

Data sources

We searched the following data sources: journals, global health organisations, professional medical organisations and multi-disciplinary associations, consumer advocacy groups and health research funding bodies.

Eligibility criteria

Eligibility criteria were developed for each data source (Table 1). Within these data sources, we only included those which included research based on the top ten non-communicable diseases (NCDs) responsible for the greatest global disability burden,(16) most of which are chronic health conditions. We chose this focus because people with chronic health conditions are frequent users of these health information sources.(17,18) For professional medical organisations and multi-disciplinary associations, consumer advocacy groups and health research funding bodies, we included those from Australia, USA, UK, Canada and New Zealand. We chose these five countries as they are largely English speaking and have ready access to these sources. Examples include The Cardiac Society of Australia and Diabetes UK.

Search

The full search strategy and date(s) for each data source is detailed in Table 1. One reviewer conducted all Google searches based on the search strings in Appendix A. The relevant web pages were saved to PDF and the cache and browsing data cleared between each search.

Selection of data sources

Two reviewers independently viewed all Google search results and selected relevant websites. When selected websites were agreed upon, both reviewers visited each website and extracted all data items. Discussion resolved all disagreements.

Data charting process

A data charting form was developed by one researcher containing the key study characteristics to extract. Two reviewers searched the website of each identified data source to determine if they require, suggest, or refer to lay summaries. They then extracted lay summary writing instructions from the data sources. These instructions formed the primary data for this study. We also collected information on several categories of observations and recorded the data items according to these category components.

For journals, both reviewers viewed two original articles in each journal and noted whether a lay summary was included. If so, the article was saved as a PDF. We then consulted the author instructions and searched for the lay summary writing instructions. We saved these instructions for data analysis.

For all other data sources, both reviewers searched the website of each data source for any mention of lay summaries. Where available, reviewers used the search function, searching common terms for a lay summary such as “lay summary”, “plain English summary”, “plain language summary” and “summary”. We also searched for lay summaries under each tab on the website home page. When lay summaries and instructions for writing them were located, we followed the same process we followed when searching journals.

The most concise and widely used definition of a lay summary is that developed by the INVOLVE group, which was established in 1996 as part of the National Institute for Health to facilitate public involvement in research.(19) According to jargon buster produced by INVOLVE, a lay summary is “a brief summary of a research project or a research proposal that has been written for members of the public, rather than researchers or professionals. It should be written in plain English, avoid the use of jargon and explain any technical terms that have to be included”.(19) We took a broader definition of ‘lay summary’ for this review, including any summary or distillation of a research study that was not the abstract. This broader definition helped ensure we captured all available lay summaries. During data charting, we located lay summaries and instructions that suggested the target audience for the summary was not a non-expert lay public. We identified three clear target audiences for lay summaries. (Table 2).

Data items

Using an inductive approach, we identified characteristics of lay summaries and lay summary writing instructions and extracted data on these characteristics. (Table 2).We located three formats for lay summaries: written, graphical and video summaries. Written lay summaries are the most well-known format, however journal publishers now produce graphical and video summaries, referred to by the labels graphical abstract and video abstract. In this review, we refer to all three formats as lay summaries for consistency.

Synthesis of results

We collated results for the lay summaries and the lay summary writing instructions in the review. We summarised the number of websites that mentioned or required a lay summary and those that provided lay summary writing instructions, grouped by data source. Data was also extracted for; format of the lay summary, the audience for the lay summary, requirement for a lay summary (optional or mandatory), the party expected to write the lay summary, and all labels used to refer to the lay summaries.

Main findings

Figure 1 shows the process of website selection. Overall, 556 websites were identified across the five data sources, of which 526 were searched for the study. We excluded 30 websites because of duplication, broken website links or inability to access necessary text. Remaining websites included 179 journals, 41 global health organisations, 147 professional medical associations and multidisciplinary organisations, 154 consumer advocacy groups and five funding bodies. From these 526 websites, 124 (23.6%) published or mentioned lay summaries and 108 (20.5%) provided lay summary writing instructions. (Table 3). Of note, none of the data sources we identified detailed how their instructions were developed or mentioned end-user involvement.

• Download figure
• Open in new tab

Figure 1.

Flow diagram of included data sources

Lay summaries

Journals contributed the largest number of lay summaries and lay summary writing instructions, with only half being mandatory. Most lay summaries were written by the author of the published paper.

Lay summary writing Instructions

For each format of lay summary, we identified key elements in the instructions such as content, purpose and structure (Table 4). The most common elements for written lay summaries referred to: structure (84.2%), content (84.2%) and word count/length (77.9%). The least common elements were readability (3.2%), use of jargon, acronyms and abbreviations (25.3%), and wording (30.5%). All instructions for graphical lay summaries referred to content and technical specifications, with (69.2%) including purpose. Most instructions for video abstracts included technical specifications (92.3%).

The instructions for writing lay summaries varied in content, intended audience and level of detail, ranging from a single sentence to several paragraphs. For 15 (12.1%) lay summaries, no instructions were provided at all. Large commercial publishing companies such as Elsevier often had standard instructions that applied to each of their journals.

Instructions for how to structure the lay summary most commonly referred to the use of bullet points and/or the use of specific headings such as ‘The Known, The New, The Implications’ (Medical Journal of Australia - journal). Instructions for content varied in length and level of detail, often as questions to address or headings to use, for example, ‘What is already known about this subject?, What does this study add?, How might this impact on clinical practice?’, (Heart - journal). Instructions for word count/length were usually expressed as a maximum number of words or sentences per bullet point. Word counts ranged from 10 to 500 words. Bullet points varied from 3 to 6 bullet points of 85 characters or 15 words each. Technical specifications for both graphical and video abstracts were detailed and precise, specifying file size, format and resolution.

Summary of evidence

In this scoping review of journals, global health organisations, professional medical organisations and multi-disciplinary associations, consumer advocacy groups and funding bodies from Australia, USA, UK, Canada and New Zealand, we found that only 23.6% published or mentioned lay summaries, with just over 100 providing instructions for writing lay summaries of health research. Very few provided instructions related to readability, use of jargon, acronyms and abbreviations, and wording. Some instructions provided structured formats via subheadings or questions to guide content, but not all. Only half mandated the use of lay summaries.

Strengths and limitations

This review represents the most comprehensive review of lay summary writing instructions, including data sources across 526 websites. This is the first review to synthesise information on the style, content and technical specifications of instructions for producing lay summaries. However, our results were limited to information that was freely available and accessible on each website. Health and medical publishing is dominated by a few large publishing companies. Therefore, many of the instructions for writing lay summaries we identified were common across multiple journals, which may influence the proportion of elements in the instructions. However, we selected journals across seven categories from the journal of citation ranking reports. The journals included in this review represent high quality, widely read journals.

Comparison to the literature

This work builds on the existing evidence-base for lay summaries to date. A review of the biomedical literature showed the labels commonly used to describe a lay summary are ‘lay summary’, ‘plain-language summary’ and ‘plain-English summary’.(1,3-8,20-22) The INVOLVE definition of lay summaries(2) also uses the label ‘plain-English summary’. These labels, however, are contrary to the findings of this review, which showed much greater variation in labels used for lay summaries and identified only 18 (11.0%) data sources that used these common labels. This is concerning given that the use of varied and imprecise labels for lay summaries may cause these summaries to not reach their target audience; the identification of a lay summary by the reader often begins with its label, and an unfamiliar or confusing label may deter readers. A universally accepted, clear and well-defined label may increase appeal of the lay summary and engagement by readers. Use of labels more easily accepted or recognised would also enable search engines to more easily locate lay summaries for users, making lay summaries more useful.(21)

In addition to the labels given to lay summaries, this review also builds on the literature related to the content of instructions for writing such summaries. Only 3% of instructions identified in this review recommended the use of a readability tool, and fewer than one-third of instructions gave direction regarding type of language and/or jargon. This finding is inconsistent with the INVOLVE definition of lay summaries which states that lay summaries “should be written clearly and simply, without jargon and with an explanation of any technical terms”.(2) Specifying readability is likely to be important because previous research has showed that most lay summaries are written at a high reading level, which may make them difficult for consumers to understand.(5,6) Poor readability can hinder the ability of a non-expert lay audience to improve their knowledge and understanding of their health condition.(23) This can lead to misinformation and poor self-management, potentially resulting in reduced health status.(23) The solution is to provide health information, such as lay summaries that are “accurate, accessible, and readable”.(18)

Implications for research and practice

Our review highlights the need for instructions to specify the intended audience of the lay summary. Some labels reflect their intended audience (e.g. “clinical perspective”, “clinical significance” for health practitioners/clinicians). Other labels, however, were ambiguous and may confuse readers, particularly the lay public (e.g. “Context”, “What’s New”, “Highlights” and “Summary at a Glance”). If the target audience and label is unclear, it may not be obvious that it is referring to a lay summary of the research article. Future work should explore the labelling preferences of each target audience, and work should be done to facilitate greater consistency across data sources.

Related to this, we took a definition of lay summaries broader in scope than that of INVOLVE(2) to ensure this was a comprehensive review. Despite this broader definition, we did not expect to locate as many as a quarter of lay summaries specifically targeted at an expert audience (medical practitioners/health professionals and researchers). The eLife review(10) noted two journals also include research summaries aimed at an expert audience e.g. The Proceedings of the National Academy of Sciences. In 2016, the journal Nature trialled the use of summaries written specifically for other researchers.(10) With the presence of article abstracts, it is unclear why some publishers of health research focus on delivering research summaries to an expert audience at the expense of conveying health research to a non-expert lay public. Given the technical nature of most research articles, there is a need for lay summaries aimed at a lay audience to satisfy their demand for trustworthy health information.(20) The public, especially those with chronic illnesses represented by the top 10 NCDs(16), consider journals a primary source of health information.(20) Lay summaries are also a tool that a non-expert lay public can use with their treating clinicians to aid in shared-decision making.(20)

Unanswered questions and future research

This review has highlighted two primary issues with lay summary writing instructions. First, inconsistency of guidance for authors writing lay summaries. Second, many don’t appear to appreciate the needs of a non-expert lay audience; our findings suggest that distributors of research evidence may prioritise the structure, content and length of the lay summary over accessibility for the target audience. One solution to these issues would be the development of a framework for writing lay summaries that is universally accepted by publishers of research evidence. This framework could include the easy identification as a lay summary by its label, consideration for the target audience (e.g. grade reading level and identifying the target audience at the top of the lay summary) and describe requirements of consumer collaboration. Given that most lay summaries (96%) are written by the authors of the research article who are unlikely to have had formal training on writing for a non-expert lay audience,(23) providing authors with training along with a standard framework could help them write more effective lay summaries. The skills and techniques used to write a research article are different to those required to communicate in lay language.(23)

None of the data sources detailed how their instructions were developed or mentioned end-user involvement. Involving consumers in the co-design process could produce lay summaries that better meet their needs.(24) Few guidelines exist to help authors work with stakeholders such as consumers on the development of lay summaries, however the protocol developed for researchers in the AGE-WELL network is a promising foundation.(1) This protocol outlines six steps in co-creating lay summaries with stakeholders. These steps begin with an investigation of the principles for writing a lay summary, and the target audience and key stakeholders. This is followed by involving stakeholders in writing lay summaries and conducting workshops to provide them with information and support.(1) Guidelines and lay summary instructions, applied correctly, can lead to high-quality research for dissemination to a wide audience. To assist their ongoing implementation, collaboration between authors, publishers and other stakeholders would be beneficial.(1)

A final area for future research relates to graphic abstracts. Graphical abstracts can attract attention to an article and are an ideal format for sharing on social media, which could improve their reach to a wider audience.(25) However, they need to be well designed with an emphasis on the main findings, which takes time and effort.(25) In 89% of graphical and 75% of video lay summaries respectively, we found the purpose was clearly outlined as attracting reader attention and increasing engagement. An evaluation of these two formats for lay summaries would be beneficial to gauge their effectiveness in meeting this objective. If effective, these formats for lay summaries offer an important alternative to written lay summaries for readers for whom a more visual medium is easier to use.