Interactive evidence maps help support the use of research evidence in practice and policy-making through their visual appearance, to show what is available and where gaps exist. Design is a key element. What are the options and issues to consider when creating one? Claire Stansfield and Helen Burchett reflect on their experiences from developing two maps on digital interventions for alcohol and drug misuse (1).
Why create an interactive evidence map?
Interactive evidence maps provide a visual overview of a collection of evidence and enable users to explore the content at varying levels of detail. They are useful to describe the nature of research or interventions (such as population focus, study design, setting), and by doing so also illustrate gaps. They can be particularly useful for seeing at a glance, what exists or is lacking, across a broad topic area. Digital alcohol and drug interventions are an example of such a broad area. For our project we produced two maps, which are contrasting examples of the purposes and the level of details of maps.
One map contains 1,250 primary research citation records of research evaluations and gives a visual overview of what intervention evaluations had been conducted, with options for filtering the map view, but providing only the citation details for each individual record. In contrast, our second map of intervention practice and systematic reviews, includes just 58 records, and fewer options for filtering the map view, but it contains a detailed descriptions within each record. Although evidence maps normally focus only on research evidence, this map describes 40 interventions available for use in England, alongside 18 internationally-focused systematic reviews. We highlight three key considerations in developing the maps.
1) Developing an organising framework
Developing a structure for organising the map was of fundamental importance to categorise the records and guide users. A street-map shows transport networks, road systems, footpaths, and there are conventions and expectations around using and navigating this type of map. Interactive evidence maps take a variety of forms and users may not be familiar with using them. A map needs at least two elements (one for rows and one for columns), and a third element is an option for segmenting records at each row and column intersection. Each element should be defined carefully and applied consistently to ensure integrity of the maps.
The final organising framework for our maps was derived from discussions with the maps’ commissioners, the former Public Health England. The rows in the maps distinguish between interventions directed at drug and alcohol misuse. A third row contains both interventions directed at drugs and alcohol and those where it was unclear as they were generically targeted to ‘substance misuse’. The columns separate the intervention purpose of prevention, treatment or recovery from alcohol or drug misuse. Though other options considered were the underlying intervention strategy used (e.g. tracking or feedback, cognitive therapy), and the digital nature of the intervention (e.g. smartphone app, videoconference call). For the second map, interventions in practice and systematic reviews were separated and the systematic reviews were segmented into three quality ratings.
2) Striking a balance between a broad view and detail
The level of detail in the maps was another consideration. For example, should drugs be grouped as one, or should there be separate rows for each drug? We grouped the drugs together, as separating them seemed to create a level of detail that would prevent the user from gaining an ‘at-a-glance’ sense of the landscape. However the primary research map allows filtering to show only those that relate to specific drugs, and both maps allow searching within records for a drug name. The map of primary studies also has options to filter the visual overview on study measures (outcomes, process, cost, other), publication year and geographical location, which helps in navigating a diverse collection of records.
For the map of intervention practice and systematic reviews, we produced detailed descriptions about each record. For the systematic reviews we extracted pertinent findings and also summarised their population and focus.
Copyright issues prevented displaying abstracts in the primary research map; they are displayed in the second map where permitted.
3) Enhancing usability
The appearance and usability of the map is fundamental to its appeal and use. Although the maps initially requires a little familiarisation, we tried to make them as straightforward as we could. We piloted them with stakeholders and colleagues, which provided useful feedback.
The labels describing each element need to be clear and concise to the user, and an accompanying glossary helps explain the labels.
We used the EPPI-Mapper tool, which integrates with EPPI-Reviewer review management software via a json file type (it is also possible to create a simple map without an EPPI-Reviewer account using an EndNote RIS file) (2). While underlying functions are fixed by the tool, there was flexibility to adapt the maps to our own needs.
We created tabs along the top of the maps’ webpages to summarise the methods used to produce the maps, display the glossary, and provide guidance on using the map.
Final reflections
The process of developing the map involved multiple iterations, ensuring that both the content and appearance are credible, consistent and usable. It was particularly time-consuming to prepare and finalise the bespoke descriptions for the second map. EPPI-Mapper has acquired additional functionalities since we produced these maps, including improved functionality for adding summaries.
Going forward
There are different approaches to developing maps to visually represent research in broad topic areas. Our two maps show two distinct but complementary uses of the mapping tool. When developing maps, decisions are needed on organising the content, the level of detail to convey and how it is provided (e.g. filters or segments). Usability needs to be considered through choice of labels, colours, supplementary content and user instructions.
You are welcome to share your thoughts on using and creating maps via the comments option below.
About the authors
Claire Stansfield is a Senior Research Fellow at EPPI Centre, UCL Social Research Institute, who focuses on information science for systematic reviews and evidence use support.
Helen Burchett is an Associate Professor at the London School of Hygiene and Tropical Medicine. She has been working on public health/health policy systematic reviews for over 20 years.
Bibliography
- Digital interventions in alcohol and drug prevention, treatment and recovery: Systematic maps of research and available interventions [Internet]. Available from: https://eppi.ioe.ac.uk/cms/Default.aspx?tabid=3879
- Digital Solution Foundry, EPPI Centre. EPPI-Mapper [Internet]. EPPI Centre, UCL Social Research Institute, University College London, UK; 2023. Available from: https://eppi.ioe.ac.uk/cms/Default.aspx?tabid=3790