Publishing only successful trials?

That's the question one has to ask when a study by British scientists found that almost one in three (29%) large clinical trials in the United States remain unpublished five years after they are finished.  Interestingly, trials funded by industry were more likely not to have been published (32%) than others (18%).

We do require drug companies to report the results of their trials in a web-based database, Clinicaltrials.gov.  Yet, the scientists concluded that the legislation to report is ignored.