The talk focuses on how peer reviewed journals provenly publish the positive test results, and not the negative results. He gives the example of Tamiflu, where the 50/50 split between actual positive/negative tests was not reflected in the journals at all - anyone reading the journals would get a remarkably positive view of the drug.
www.ted.com/talks/ben_goldacre_what_doctors_don_t_know_about_the_drugs_they_prescribe.html
The talk focuses on how peer reviewed journals provenly publish the positive test results, and not the negative results. He gives the example of Tamiflu, where the 50/50 split between actual positive/negative tests was not reflected in the journals at all - anyone reading the journals would get a remarkably positive view of the drug.