Cross-sectional analysis of UK research studies in 2015: results from a scoping project with the UK Health Research Authority. Issue 10 (17th October 2018)

• Record Type: Journal Article
• Authors: Clark, Tim; Wicentowski, Richard H; Sydes, Matthew R
• Is Part Of: BMJ open. Volume 8:Issue 10(2018)
• Publication Date: 2018-10-17
• Subjects: clinical trials; Medicine -- Research -- Periodicals; 610.72
• DOI: 10.1136/bmjopen-2018-022340
• Languages: English
• ISSNs: 2044-6055

Abstract

Abstract : Objectives: To determine whether data on research studies held by the UK Health Research Authority (HRA) could be summarised automatically with minimal manual intervention. There are numerous initiatives to reduce research waste by improving the design, conduct, analysis and reporting of clinical studies. However, quantitative data on the characteristics of clinical studies and the impact of the various initiatives are limited. Design: Feasibility study, using 1 year of data. Setting: We worked with the HRA on a pilot study using research applications submitted for UK-wide ethical review. We extracted into a single dataset, information held in anonymised XML files by the Integrated Research Application System (IRAS) and the HRA Assessment Review Portal (HARP). Research applications from 2014 to 2016 were provided. We used standard text extraction methods to assess information held in free-text fields. We use simple, descriptive methods to summarise the research activities that we extracted. Participants: Not applicable—records-based study Interventions: Not applicable. Primary and secondary outcome measures: Feasibility of extraction and processing. Results: We successfully imported 1775 non-duplicate research applications from the XML files into a single database. Of these, 963 were randomised controlled trials and 812 were other studies. Most studies received a favourable opinion. There was limited patient and public involvement in the studies. Most, but not all, studies were planned for publication of results. Novel study designs (eg, adaptive and Bayesian designs) were infrequently reported. Conclusions: We have demonstrated that the data submitted from IRAS to the HRA and its HARP system are accessible and can be queried for information. We strongly encourage the development of fully resourced collaborative projects to further this work. This would aid understanding of how study characteristics change over time and across therapeutic areas, as well as the progress of initiatives to improve the quality and relevance of research studies.