SCIE Report 13: Systematic searching on the AgeInfo database

By Brian J. Taylor, Emma Wylie, Martin Dempster and Michael Donnelly

Published June 2006

Context

The facility to undertake systematic retrieval of abstracts of articles from electronic bibliographic databases is an essential component of evidence-based practice. Systematic reviews of research require facilities for sophisticated searching that is both sensitive and precise.

Purpose

This report presents the results of an evaluation of the AgeInfo database, which is an information service provided by the Centre for Policy on Ageing (www.cpa.org.uk) and which is made available to the public at not cost through SCIE's own database Social Care Online (www.scie-socialcareonline.org.uk).

Audience

This report is intended for use by SCIE and the Centre for Policy on Ageing, and may also be useful to users of the AgeInfo database, such as researchers conducting systematic reviews.

Messages from the report

The AgeInfo database was searched using a clearly defined search question. Six other databases relevant to social work and healthcare were also searched to gauge the comparative advantages and disadvantages of AgeInfo. The modern development of Boolean algebra employed in electronic databases was used to create searches that were as sensitive and as precise as possible on each database, given their respective facilities.

Key findings

• AgeInfo had three of the four facilities available on most of the databases for refining search terms: suffixes could not be used, but it did use truncation symbols, had an index system and enabled proximity searching.
• AgeInfo could not filter for language or type of publication but could filter for date of publication.
• AgeInfo's selection of records was efficient and results could be emailed, although the search formula could not be printed with the results.
• AgeInfo's most significant weakness was that results could not be ordered alphabetically by author. AgeInfo may also have technical problems dealing with a large number of retrievals. It was beyond the scope of this study to compare the import of results into EndNote or Reference Manager.
• AgeInfo retrieved just over 13 per cent of the total number of relevant articles across all seven databases, and ranked fourth in terms of sensitivity.
• AgeInfo gave the highest level of precision of the databases used (76 per cent), which reflects the specialist focus of the database and its index terms.
• The search on AgeInfo retrieved 17 unique relevant articles; omitting to use AgeInfo would lead to a failure to retrieve about five per cent of the relevant articles retrieved across the seven databases.
• AgeInfo provides prompts to use 'older persons' for many words relating to older people, but not to the terms 'older men' and 'older women'.
• The records on AgeInfo have been indexed only intermittently with terms relating to age. It is strongly recommended that the Centre for Policy on Ageing develop a consistent policy on this, and that the thesaurus and guidance to users be amended accordingly.
• In order to retrieve the most articles on this topic using a more limited number of databases than those used for this study, a searcher would choose Social Science Citation Index and Medline as top priorities and AgeInfo and Cinahl as second priorities.
• In terms of the range and user-friendliness of search facilities, AgeInfo was comparable to the other databases, except for Medline, Cinahl and PsycINFO, which were decidedly superior.
• AgeInfo may also have uses for the broader purposes of evidence-based practice in terms of scoping the field and providing contextual material to underpin judgements on the application of evidence, but that was beyond the scope of this study.