Some stashed away studies suggest many antidepressants aren’t all they’re cracked up to be. From the NewScientist:
It's called the "file-drawer problem". A study fails to produce interesting results, so is filed away and forgotten - a practice that might mean antidepressants don't work as well as doctors think.
To get approval for the 12 antidepressants that went on the market between 1987 and 2004, drug firms registered over 70 clinical trials with the US Food and Drug Administration (FDA). But when Erick Turner of Oregon Health and Science University in Portland and his colleagues combed through medical journals, they found that 23 of these studies never made it into a journal. All but one of the unpublished studies concluded that the effect of the drugs was negative or questionable.