Screen2Go: a pilot smartphone app for citation screening

Article type
Authors
Huckvale K1, van Velthoven M1, Car J1
1Imperial College London, UK
Abstract
Background: Although Cochrane’s RevMan software supports authors during the latter stages of a review, citation searching and abstract screening are not part of the current or planned feature set. Screening up to several thousand citations can be a daunting task for a researcher with multiple demands on their time.

Objectives: To develop a pilot smartphone and tablet app to explore the feasibility and acceptability of a convenience approach to citation screening. This would allow a user to screen batches of citations during otherwise unproductive periods of the day, for example when commuting to work.

Methods: We used a rapid application development approach to build a prototype using the Microsoft Visual Studio SilverLight software development environment for Windows 7 Phone.

Results: The prototype app was developed to support a simple workflow. Citations exported from bibliographic software, such as Endnote, can be uploaded to a secure website and are automatically downloaded to the phone for screening óffline’. When a mobile data connection becomes available, screening choices are sent back to the website where the user can then view and export batches of citations grouped by decision. The app proved highly acceptable to a small group of test users. It appeared most effective for rapidly excluding citations of obvious irrelevance. Citations with missing information had to be marked as ’unsure’ and triaged in a separate step using a desktop computer. Applied to a real-world screening of some 12,000 articles, there was no obvious deficit in the precision of screening decisions made using this workflow.

Conclusions: A smartphone app may be an effective way to support the citation screening process by allowing users to review smaller batches of citations at convenient times. A future randomised crossover study will formally compare the performance of screeners using the tool against usual practice.