A recent study of antidepressant medication trials compared the drugs' published outcomes with all studies registered with the FDA.
A recent study of antidepressant medication trials compared the
drugs' published outcomes with all studies registered with the FDA.
The findings indicate that one-third (31%) of the registered studies
were unpublished. Interestingly, studies with negative or questionable
results (judged by FDA standards) were more likely to be either
unpublished or, when published, to convey false and misleadingly
positive outcomes. Publishing only positive results makes drugs appear
to be more effective than they are, when both published and unpublished
results are combined. For example, published literature on the
anti-depressant drugs reviewed indicated that 94% of the trials had
positive outcomes. Yet, the FDA analysis of all trials (published and
unpublished studies) showed just 51% of trials had positive results.
Such selective reporting and data manipulation poses great risks for
potential adverse effects for consumers, study participants, and
The New England Journal of Medicine, JANUARY 17, 2008