Abstract
EDITOR, - After Harrington's criticism of the use of unpublished data in drug advertisements1 and Humphreys's response2 I surveyed advertisements in the BMJ during 1993 to discover more about where drug studies are published. I analysed 48 advertisements, which contained 175 references. Fifteen advertisements (9%) referred to “data on file.” This figure is in line with results of an earlier study, which found that 11% of references …