Search wide and dig deep: identifying ‘views’ research for systematic reviews

Article type
Authors
Kavanagh J1, Stansfield C2, Brunton G3, Rees R4, Thomas J4
1Social Science Research Unit, EPPI- Centre, London, UK
2Social Science Research Unit, Institute of Education, London, UK
3EPPI Centre, Social Science Research Unit, University of London, London, UK
4Institute of Education, EPPI-Centre, London, UK
Abstract
Background: Searching for views studies, (i.e. mostly qualitative research reporting people’s experiences, opinions and understandings about health issues), is acknowledged to be a challenging process. The EPPI-Centre conducted four systematic reviews of people’s views in the areas of obesity, walking and cycling and transition to motherhood. All reviews were used as evidence for UK evidence-informed decision making. Analysing where studies have been found is useful to inform future reviews, and to assess the utility of using a range of research sources (e.g. databases, websites, library catalogues, reference checking). Objectives: To assess the value of research sources used to identify research for four systematic reviews of research reporting people’s views about health related issues. Methods: A retrospective analysis was carried out. This included identifying from which search sources the included studies in each review were located, calculating the number of studies unique to each source and the precision of the results from searches of bibliographic database sources. Publication type was considered in the analyses. Additional comparisons were made of studies from two of the reviews, as these were within the same topic area. Results: Interim analysis of three reviews reveals that of a total of 118 included studies, N = 65 (55%) were identified through searches of ASSIA, Psycinfo, CINAHL and Pubmed. For one review, database precision ranged from 0.22–1.64%. A total of N = 95 (81%) studies were uniquely identified on one research source, of which N = 65 (55%) were not identified through searches of major bibliographic databases. Data will be presented for each type of search source and for individual reviews. Conclusions: Comprehensive searching for views involves screening many irrelevant studies. While literature is often found amongst a few sources, there is value in searching widely. Reasons why some sources provide more unique studies include publication type and relevancy of databases to review topics.