About
 Introduction
 Search Concepts
 PubMed
 Ovid
spacer.gif (43 bytes) spacer.gif (43 bytes)

backspacer.gif (43 bytes)next

Backspacer.gif (43 bytes)Next
spacer.gif (43 bytes)
backspacer.gif (43 bytes)next
Backspacer.gif (43 bytes)Next

 spacer.gif (43 bytes)

 spacer.gif (43 bytes)concepts backspacer.gif (43 bytes)next

3. Constructing the Systematic Search

Constructing the systematic search is probably the more difficult to perform, but fortunately, because of the limited number of searches of this type there has been considerable world-wide effort to devise effective search strategies for all of the common clinical questions. If you can find the time it is worth reading the article 'Developing Optimal Search Strategies for Detecting Clinically Sound Studies in Medline'  Haynes et al. Journal of the American Medical Informatics Association 1994 1 (6);447-58.

I have presented the search strategies from that paper below, for both Ovid and PubMed, together with their quoted sensitivities and specificities. You can simply copy and paste them into the relevant search box and run them. You will note that there are actually two searches for each category, one of which is optimised for sensitivity, and one for specificity.

Table 1. Systematic Searches
Category
sens/spec		 BMA OVID MedlinePlus Results PubMed Results
Therapy sensitivity
99%/74%		 randomized controlled trial.pt. or dt.fs. or tu.fs. or random$.tw. 1205397 "randomized controlled trial" [PTYP] OR "drug therapy" [SH] OR "therapeutic use" [SH:NOEXP] OR "random*" [WORD] 1196311
Therapy specificity
57%/97%		 (double and blind$).tw. or placebo.tw. 82076 (double [WORD] AND blind* [WORD]) OR placebo [WORD] 98649
Diagnosis sensitivity
92%/73%		 exp "sensitivity and specificity"/ or sensitivity.tw. or specificity.tw. or di.xs. or du.fs. 2425797 "sensitivity and specificity" [MESH] OR "sensitivity" [WORD] OR "diagnosis" [SH] OR "diagnostic use" [SH] OR "specificity" [WORD] 2640389
Diagnosis specificity
55%/98%		 exp "sensitivity and specificity"/ or (predictive and value$).tw. 101585 "sensitivity and specificity" [MESH] OR ( "predictive" [WORD] AND "value*" [WORD]) 102645
Aetiology sensitivity
82%/70%		 exp cohort studies/ or exp risk/ or (odds and ratio$).tw. or (relative and risk).tw. or (case and control$).tw. 597971 "cohort studies" [MESH] OR "risk" [MESH] OR ("odds" [WORD] AND "ratio*" [WORD]) OR ("relative" [WORD] AND "risk" [WORD]) OR ("case" [WORD] AND "control*" [WORD]) 628689
Aetiology specificity
40%/98%		 case-control studies/ or cohort studies/ 54668 "case-control studies" [MH:NOEXP] OR "cohort studies" [MH:NOEXP] 56309
Prognosis sensitivity
92%/73%		 incidence/ or exp mortality/ or follow-up studies/ or mo.fs. or prognos$.tw. or predict$.tw. or course.tw. 909271 "incidence" [MESH] OR "mortality" [MESH] OR "follow-up studies" [MESH] OR "mortality" [SH] OR prognos* [WORD] OR predict* [WORD] OR course [WORD] 1004204
Prognosis specificity
49%/97%		 prognosis/ or survival analysis/ 175832 prognosis [MH:NOEXP] OR "survival analysis" [MH:NOEXP] 178159
Sensitivities and specificities quoted from Haynes RB, Wilczynski N, McKibbon KA, Walker CJ, Sinclair JC. Developing optimal search strategies for detecting clinically sound studies in MEDLINE. J Am Med Inform Assoc 1994 Nov-Dec; 1(6): 447-58.