Abstract
It's common knowledge that published medical findings tell only part of the story, the other part consisting of unpublished records of negative trials, sometimes of the same agents. No one, of course, knows precisely how much lies below the waterline, but informed estimates suggest that as many as half of all trials undertaken remain unreported —a percentage sufficient to cast a shadow over the entire corpus of published medical research. The effect is incalculable. That negative findings of trials of cancer chemotherapies would be withheld from publication seems inconceivable, but a paper documented that practice, and its slanting effect on the published literature, some 30 years ago....