Service providers and search interfaces
It is obvious that searching different databases will give different results (for example searching CINAHL versus searching Medline). However, the interface selected for searching each database also has important implications. Searching the same database via different interfaces can result in different results. For instance searching Embase via Embase.com or via OVID, or searching MEDLINE via Web of Science, EBSCO, OVID, or even via PubMed.
The differences in the results of searching the same database via different interfaces can be a result of different search functionality and search syntax as well as the currency of the database in different interfaces and even the content of the database via different interfaces. For example, there are more duplicate records on OVID Embase than on Embase.com. By accessing MEDLINE via the Web of Science – citation data can be obtained. The differences between MEDLINE via OVID and via PubMed are much discussed and not only include different ways of searching but the content differs as PubMed contains more than MEDLINE with extra content such as PubMedCentral and articles before they receive full MeSH tagging (in-process citations) and books.
Most papers comparing different search interfaces for the same database are solely descriptive in nature with no evaluation of the impact on the results and thus the implications of searching via one interface as opposed to another are not formally measured. There are, however, a number of published evaluations that have gone beyond a descriptive comparison. The majority of these have compared Medline and PubMed (1, 2), although other comparisons are for CINAHL (3), The Cochrane Library (4) and AMED (5). Three studies found that the same search in the same database can vary in different interfaces, both when doing freetext searching, and also searching for keywords (3,4,5).
In relation to PubMed – MEDLINE comparisons - PubMed contains more than MEDLINE, and may have a higher sensitivity in some cases (1,2).
One problem that all such research in this area suffers from is that of currency. Database interfaces are constantly being developed and improved, more so than the databases themselves. Also, there are few studies that evaluate the performances in real life search strategies, comparing recall and precision, relevant unique results and also providing in-depth explanations of the differences.
Before searching, one should learn how each interface works (truncation, mapping to subject headings, proximity operators). For example, adjacency/proximity operators can work with different meanings in some interfaces such as Ovid and Embase.com, adj3 or NEAR/3 represents two or fewer characters between the words combined and in other interfaces, such as, Web of Science and EBSCO the number represents three or fewer characters between the words combined. It may also be beneficial to search the same database via more than one interface. This may be particularly be the case for MEDLINE and PubMed where PubMed offers more content but other interfaces of MEDLINE offer more controlled searching.
- (1) Irvin, E. Does is matter which version of MEDLINE you search? 12th Cochrane Colloquium: Bridging the Gaps. Oct 2-6 2004. Ottawa, Ontario, Canada. [Further reference details] [Publication appraisal] [Free full text]
- (2) Katchamart, W., Faulkner A, Feldman B,Tomlinson G, Bombardier C. PubMed had a higher sensitivity than Ovid-MEDLINE in the search for systematic reviews. J Clin Epidemiol 2011;64:805-807. [Further reference details] [Publication appraisal] [Free full text]
- (3) Allison MM. Comparison of CINAHL via EBSCOhost, Ovid, and ProQuest. J Electron Res Med Lib. 2006;3:31-50. [Further reference details] [Publication appraisal] [Free full text]
- (4) Craven J, Jefferies J, Kendrick J, Nicholls D, Boynton J, Frankish R. A comparison of searching the Cochrane library databases via CRD, Ovid and Wiley: implications for systematic searching and information services. Health Info Libr J. 2014;31:54-63. [Further reference details] [Publication appraisal] [Free full text]
- (5) Younger P, Boddy K.When is a search not a search? A comparison of searching the AMED complementary health database via EBSCOhost, OVID and DIALOG. Health Info Libr J 2009:26:126-135. [Further reference details] [Publication appraisal] [Free full text]