K. Dickersin et al., SYSTEMATIC REVIEWS - IDENTIFYING RELEVANT STUDIES FOR SYSTEMATIC REVIEWS, BMJ. British medical journal, 309(6964), 1994, pp. 1286-1291
Objective-To examine the sensitivity and precision of Medline searchin
g for randomised clinical trials. Design-Comparison of results of Medl
ine searches to a ''gold standard'' of known randomised clinical trial
s in ophthalmology published in 1988; systematic review (meta-analysis
) of results of similar, but separate, studies from many fields of med
icine. Populations-Randomised clinical trials published in 1988 in jou
rnals indexed in Medline, and those not indexed in Medline and identif
ied by hand search, comprised the gold standard. Gold standards for th
e other studies combined in the meta-analysis were based on: randomise
d clinical trials published in any journal, whether indexed in Medline
or not; those published in any journal indexed in Medline; or those p
ublished in a selected group of journals indexed in Medline. Main outc
ome measure-Sensitivity (proportion of the total number of known rando
mised clinical trials identified by the search) and precision (proport
ion of publications retrieved by Medline that were actually randomised
clinical trials) were calculated for each study and combined to obtai
n weighted means. Searches producing the ''best'' sensitivity were use
d for sensitivity and precision estimates when multiple searches were
performed. Results-The sensitivity of searching for ophthalmology rand
omised clinical trials published in 1988 was 82%, when the gold standa
rd was for any journal, 87% for any journal indexed in Medline, and 88
% for selected journals indexed in Medline. Weighted means for sensiti
vity across all studies were 51%, 77%, and 63%, respectively. The weig
hted mean for precision was 8% (median 32.5%). Most searchers seemed n
ot to use freetext subject terms and truncation of those terms. Conclu
sion-Although the indexing terms available for searching Medline for r
andomised clinical trials have improved, sensitivity still remains uns
atisfactory. A mechanism is needed to ''register'' known trials, prefe
rably by retrospective tagging of Medline entries, and incorporating t
rials published before 1966 and in journals not indexed by Medline int
o the system.