A fellow student of mine at SLAIS, Sue Bradley, recently won a prize and publication for a directed research project, entitled “Examination of the Clinical Queries and Systematic Review ‘hedges’ in EMBASE and MEDLINE.” The paper compares the translations of various search filters and “hedges” (some PubMed examples) used to narrow results in biomedical searches, and found some striking differences in the results across the different interfaces (OvidSP, EBSCO and PubMed).
The NLM Systematic Reviews subject subset hedge searches for high quality EBM material such as systematic reviews, clinical practice guidelines, and consensus development conferences. Although it has not undergone testing to determine sensitivity, specificity, precision, and accuracy, translations of it are available in OvidSP and EBSCO MEDLINE. A major problem was found with the OvidSP MEDLINE version of this, as well as with two other subject subset filters. Ovid was notified, and the problem was resolved in late June 2009. Thereafter, similar search results were found using the “equivalent” hedges in the three MEDLINE interfaces examined. The fact that Ovid MEDLINE subject subset filter search strings contain proprietary information that is not publicly available is alarming. Health science librarians need to be able to evaluate hedges to determine whether or not to use them. Without being able to examine the search string, it is difficult to evaluate the possible utility or limitations of a filter.
There has been a long thread recently–and long before, and surely to come again–on the MEDLIB listserv about various advantages and disadvantages (::cough::$$$::cough::) of OvidSP’s MEDLINE over the free PubMed interface, but a discussion of actual search results and filtering hasn’t seemed to surface in the discussion. Because narrowing the “clinically irrelevant” results out of the set is particularly important for health library reference, the topic of this paper is especially relevant.
- Bradley S. Examination of the Clinical Queries and Systematic Review “hedges” in EMBASE and MEDLINE. J Can Health Libr Assoc. 2010 Aug;31(2):27-37.