medical researcher’s first port of call when looking for quality articles and summaries of clinical research. This is indeed the case, and the Cochrane Library now boasts several hundred systematic reviews and hundreds of thousands of peer reviewed summaries of randomised controlled trials. The story behind the Cochrane project is worth telling. In 1972, epidemiologist Archie Cochrane called for the establishment of a central international register of clinical trials. (It was Cochrane who, as a rebellious young medical student, marched through the streets of London in 1938 bearing a placard which stated, “All effective treatments should be free”. His book Effectiveness and efficiency15 caused little reaction at the time but captures the essence of today’s evidence based medicine movement.) Though he never lived to see the eponym, Archie Cochrane’s vision of a 100% accurate medical database, the Cochrane Controlled Trials Register, is approaching reality. The Cochrane Library also includes two “metadatabases” (the Cochrane Database of Systematic Reviews and the Database of Abstracts of Reviews of Effectiveness) and a fourth database on the science of research synthesis (the Cochrane Review Methodology Database). This entire library is available on CD-ROM from the BMA bookshop. Published articles are entered on to the Cochrane databases by members of the Cochrane Collaboration,16 an international network of (mostly) medically qualified volunteers who each take on the handsearching of a particular clinical journal back to the very first issue. Using strict methodological criteria, the handsearchers classify each article according to publication type (randomised trial, other controlled clinical trial, epidemiological survey, and so on), and prepare structured abstracts in house style. The Collaboration has already identified around 60 000 trials that had not been appropriately tagged in Medline. All the Cochrane databases are in user friendly Windows style format with a search facility very similar to that used in the common Medline packages. Numerical data in overviews are presented in a standardised graphics way to allow busy clinicians to assess their relevance quickly and objectively. In 1997 some of the founder members of the Cochrane Collaboration published a compilation of articles reflecting on Cochrane’s original vision and the projects that have emerged from it. Despite its uninspiring title,
Non-random reflections. . . is a fascinating account of one of medicine’s most important collaborative initiatives in the 20th century.17 Finally, if you are interested in becoming involved with the Cochrane Library projects, contact the Cochrane Library Users Group on http://www.york.ac.uk/inst/crd/clug.htm.