Contributed Paper Abstract
The Evidence-Based Literature Search: Connecting Clinical Needs with Sound Searching Practices
Karl H. Woodworth, Grady Branch Librarian, Emory University, Atlanta GA.
Purpose: This paper will survey concepts of searching for journal articles having 'evidence-based medicine' content using MEDLINE and specialized databases such as the Cochrane Database of Systematic Reviews and the DARE.
Setting/Participants/Resources: The Grady Branch Library of the Health Sciences Library, Emory University School of Medicine, a branch medical library serving a major urban teaching hospital. Authors have worked closely with clinical faculty and housestaff actively requesting an evidence-based approach to the literature, and have maintained an evidence-based website for the past year from which locally produced search filters can be copied and pasted into MEDLINE. Although the primary literature source for Emory is the Ovid Technologies collection of medical databases, the author is experienced with PubMed and is prepared to discuss that environment also.
Brief Description: 'Evidence-based medicine', an effort occurring over the past decade to revolutionize medical practices to adhere more rigorously to objective scientific results, has created new demands about the kinds of citations that clinicians want to see in a literature search. Clinicians are teaching each other to search the literature using the 'therapy-diagnosis-prognosis-etiology' (TGPE) template. However, our work in the past several years to satisfy evidence-based search requests has revealed that clinicians have been asking for, and responding very favorably to, a different search 'template' that roughly corresponds to a 'levels of evidence' (LOE) classification scheme for medical literature. This paper discusses the growing layer of EBM-oriented databases that occur both as unique offerings and as subsets of MEDLINE, compares the TGPE search template with the LOE template, surveys the use of search filters to standardize effective EBM searching, and suggests more convenient ways to distribute search filters to the medical public.
Results/Outcome: This is an ongoing effort with very favorable input from local clinical staff.
Evaluation Method: Search result counts of filters for systematic reviews and randomized controlled trials compare closely with commercial products such as DARE. Anecdotal reports from local staff are very favorable.
Back to Contributed Papers
|
