Hempel

Appraisal of: Hempel S, Rubenstein LV, Shanman RM, Foy R, Golder S, Danz M, Shekelle PG. Identifying quality improvement intervention publications--a comparison of
electronic search strategies. Implement Sci. 2011 Aug 1;6:85.

Reviewer(s):

Corinne Holubowich

Full Reference:

Hempel S, Rubenstein LV, Shanman RM, Foy R, Golder S, Danz M, Shekelle PG. Identifying quality improvement intervention publications--a comparison of electronic search strategies. Implement Sci. 2011 Aug 1;6:85.

Short description:

Authors compared several search strategies to identify quality improvement intervention publications. 10 strategies were developed or applied and evaluated in MEDLINE/PubMed. Three sets of publication collections were derived from experts in the field at the Agency for Healthcare Research and Quality (AHRQ), Standards for Quality Improvement Reporting Excellence (SQUIRE) and the Cochrane Effective practice and Organization of Care Group (EPOC). Search strategies were developed and then tested against these collections for recall. The authors strived to balance yield, recall, recall-to-yield ratio, precision and face validity. Of the 10 strategies tested, the mean recall was between 5% and 53%. The most successful strategy identified based on yield was a simple text word strategy: (‘quality’ AND ‘improve*’ AND ‘intervention*’). Authors noted that quality improvement tends to be a difficult discipline to search since a specific MeSH term does not yet exist.

Limitations stated by the author(s):

Authors noted that the recall rate of their tested strategies were low, which may in part be explained by the reference sets being somewhat low in number. The chosen search terms were not derived in an automated manner, therefore making them somewhat subjective. Boolean and adjacency were used to combine terms, rather than testing individual terms. Rather than focusing on maximizing the retrieval rates, the authors hoped to develop generalizable search strategies. Recall rates would have likely improved if high search yields were not a concern. Authors reported that their results were only moderately successful.

Limitations stated by the reviewer(s):

None

Study Type:

Single study

Related Chapters:

Value of using different search approaches

Tags:

Search Filters, Organizational