Apologies for the length of this query:
I've just spent some time transferring the PubMed filters available through the Clinical Queries feature to save as filters that all our users can search when conducting their own Webspirs Medline searches. Having transferred the filters I then checked that I'd done the job correctly (initially I wasn't sure what "MH not exp" meant on PubMed; I assume its the equivalent of searching rather than exploding terms from the thesaurus on Webspirs.) In order to do this check I simply combined each filter with asthma on each version of Medline. The results were quite interesting:
Filter PubMed results Webspirs results
Therapy specific 4801 4967
Therapy sensitive 25176 25582
Diagnosis specific 1053 1081
Diagnosis sensitive 18260 15252
Aetiology specific 1321 1328
Aetiology sensitive 7485 7557
Prognosis specific 975 991
Prognosis sensitive 8608 8750
Questions
1 The results (with one exception) are quite close. However, I understood that Pubmed updated its records more frequently than Webspirs, and expected Pubmed to retrieve marginally more each time. The opposite was the case. Why?
2 What have I done wrong when transferring the sensitive diagnosis filter to Webspirs? I 'translated' the PubMed version
(sensitivity and specificity [MESH] OR sensitivity [WORD] OR (diagnosis [SH] OR diagnostic use [SH] OR specificity [WORD]))
into Webspirs as
(explode 'Sensitivity-and-Specificity' / all subheadings in MIME,MJME) or (sensitivity or specificity) or (sh=diagnosis) or (sh=diagnostic-use)
Is this not correct?
Any offers of help gratefully received.
John Addison
Deputy Librarian
Education Centre Library (0161 627 8463)
Royal Oldham Hospital (NHS) Trust
|