Editor—We agree with Guyatt et al that providing evidence based answers to clinical questions requires intensive study and application if the process is seen in the context of a systematic review of the original literature.1 The main stumbling block remains the difficulty of constructing complex searches appropriate for a range of potential sources.
We suggest that many typical clinical questions can be answered by using a limited range of extracting, evaluating databases, which can be interrogated with simple two step or three step search formulations. Typically these sources contain several thousand references, as opposed to several millions on large databases such as Medline or EMBASE with their unfavourable signal to noise ratio.
The three sources that in our experience have a high yield of material related to evidence based health care are the clinical queries option in PubMed2; the Cochrane Library with its four sections (systematic reviews, the CRD (Centre for Reviews and Dissemination) database of reviews of effectiveness, the register of controlled trials, and the NHS economic evaluation database)3; and the TRIP (turning research into practice) database from the Centre for Research Support, Cardiff.4
These three databases typically retrieve fewer than 10 references provided that two or at most three relevant and discriminating terms are selected for a simple search. We often suggest to trainees that they should formulate their searches as if they were sending a telegram: which two or three words would you transmit to a colleague to ensure that he or she can imagine the clinical question? Thus the question “How efficient is a single dose of a steroid for outpatient croup?” suggests the search “croup and outpatient,” which identifies small sets (<10) containing a target reference5 on any of the three databases mentioned above.
This pragmatic approach, although no substitute for systematic reviews for those undertaking more extensive searches, is influenced by William of Occam's principle of “if in doubt keep it simple” and is a valid option for busy clinicians.
References
- 1.Guyatt GH, Meade MO, Jaeschke RZ, Cook DJ, Haynes RB. Practitioners of evidence based care. BMJ. 2000;320:954–955. doi: 10.1136/bmj.320.7240.954. . (8 April.) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pubmed. www.ncbi.nlm.nih.gov:80/entrez/query/static/clinical.html
- 3.www.update-software.com/clibhome/clib.htm [password required].
- 4.www.ceres.uwcm.ac.uk/frameset.cfm?section=Trip.
- 5.Geelhoed GC, Turner J, Macdonald WBG. Efficacy of a small single dose of oral dexamethasone for outpatient croup: a double blind placebo controlled clinical trial. BMJ. 1996;313:140–142. doi: 10.1136/bmj.313.7050.140. [DOI] [PMC free article] [PubMed] [Google Scholar]