Objective: Over 900 journals are registered with the Cochrane Collaboration as being handsearched for reports of controlled trials. As retrospective handsearching is gradually completed, the question remains: Should resources continue to be devoted to handsearching journals prospectively? We are comparing reports of potentially eligible trials identified from handsearching with those identified from a search of the same journals, using the Cochrane highly sensitive search strategy for MEDLINE.

Methods: Each issue of 12 UK-published general health care journals indexed in MEDLINE is being handsearched for all reports of potentially eligible studies published in 1996. The Cochrane search strategy, limited by journal title and publication year, is being run on MEDLINE. Reports of trials retrieved from handsearching will be compared with those retrieved from the MEDLINE search to identify and explore the differences in retrieval between the two techniques.

Results/Discussion: Because of the delay between publication and the appearance of a corresponding MEDLINE record, our analysis of 1996 publications will be carried out in June 1997. Our study should help to inform decisions concerning the need to handsearch journals prospectively.