Risk in dementia care: searching for the evidence

Abstract

Effective and efficient search methods are required to retrieve robust evidence to inform the study of risk communication such as in dementia care. In this article, we draw on a study which appraised 12 bibliographic databases and one online search engine for this purpose by measuring their ability to identify relevant papers published up to December 2013 when applying a consistent search strategy on the topic of research on risk concepts and risk communication in dementia. We also searched reference lists of literature reviews. We retrieved 31 relevant articles. We took measures of sensitivity (ability to retrieve relevant papers) and precision (ability to avoid retrieving irrelevant papers), and we identified unique articles and the dispersion of relevant results when using database relevancy-sorting functions. We found that Cumulative Index to Nursing and Allied Health and PsycINFO had the highest levels of sensitivity; Social Services Abstracts, Social Care Online and Applied Social Sciences Index and Abstracts had the highest levels of precision. We rated Google Scholar (using the first 300 hits retrieved) third on sensitivity, seventh on precision, and found it had a more effective sort by relevancy function than any of the databases. We found that five databases and Google Scholar retrieved at least one study not identified by any other database. We found that none of the databases retrieved all of the relevant articles identified by that database within the first 25% of results when using the sort by relevancy function (where this was available). We concluded that it is necessary to use a number of databases for effective searching on this topic. The approaches we report in this article assist in creating a comprehensive search strategy and can be used by researchers to build social science risk knowledge methodically.

Keywords:

• databases
• bibliographic
• dementia
• evidence based practice
• information storage and retrieval
• review literature as topic
• risk
• risk communication
• systematic literature searching

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was supported by the Health and Social Care Research and Development Division Public Health Agency (HSC R&D, PHA) and Atlantic Philanthropies [grant number COM/4891/13]. This support is gratefully acknowledged.