The system doesn’t work perfectly and it can’t cull every false positive finding, but a 2015 analysis showed that after the launch of ClinicalTrials.gov in 2000, the proportion of trials that found a positive benefit plummeted.9 Prior to 2000, for example, 57 percent of large randomized, controlled trials sponsored by the National Heart, Lung, and Blood Institute showed that the treatment worked. After 2000, that number dropped to 8 percent.
