Identifying digital intervention research: Using bibliometric analyses to inform searching and mapping

Article type
Authors
Stansfield C1, Burchett H2
1EPPI-Centre, University College London Institute of Education
2London School of Hygiene & Tropical Medicine (LSHTM)
Abstract
Background Bibliometric analyses, such as word co-occurrence networks, topic modelling and research field networks can provide an overview of trends and patterns of a research area. Digital technology for promoting health behaviour change is a growing interdisciplinary field. Interactive digital interventions for health promotion and treatment use a range of modes such as smartphone apps, websites, videoconferencing, and websites. Systematically identifying research evidence within this area is challenging owing to the variety of digital modes and the heterogeneous terminology used to describe them. As part of undertaking a systematic map of research on digital interventions for prevention, treatment and recovery of alcohol and drug misuse, a readily available set of primary studies and systematic reviews provided an opportunity to explore how bibliometric tools could describe these studies in order to inform search term development and mapping.

Objectives To explore ways that digital interventions have been described in primary studies and systematic reviews and their references and citation relations. To compare differences between them in order to explore the utility of highlighting the different conceptual areas within this domain to inform search strategies and mapping.

Methods The sets of citations were analysed using text analysis, topic modelling and text visualisation tools for keyword co-occurrence: 1241 primary studies and 23 systematic reviews and their references. The distribution of research areas of 22 systematic reviews and references were mapped using Web of Science research fields.

Results
We will present the trends and differences across the sets of citations and an overlay map of research areas. Key terms and phrases will be grouped into broad categories to distinguish different facets of the digital components of the intervention: nature of technology (digital, electronic, voice recognition, automated); hardware (e.g. computer, smartphone); technology interface (internet, app, text-message); strategy of the technology intervention (eCBT, e-health).

Conclusions This exploratory study informs the potential of applying bibliometric tools for identifying research described by diffuse terminology. It contributes to increasing the breadth of research evidence available to users of reviews, including healthcare consumers.