It provides the first hard data on a practice known as selective reporting in which the good news about a drug is made public but the bad new is not.
Experts say it's not that antidepressants don't work but that their effectiveness has been exaggerated. They say this might tip the scales against prescribing the drugs in borderline cases.