Pharmaceutical companies running clinical trials are increasingly being required to register them in one or more databases. Once the trials are finished, the results ought to be published in medical journals, but there is little incentive to publish negative data, for example from trials that were terminated early due to poor drug performance or dangerous side effects. These results may quietly disappear, leading to unrealistic estimates of the effectiveness of drugs and poorer medical decisions. This talk introduces a project headed by Ben Goldacre, currently at the prototype stage, aiming to encourage publication of clinical trial data by combining information from the different trial repositories in one publicly accessible website, crosschecked with publications. This would allow the identification of trials with no published results, making these absent data points publicly visible.
http://github.com/crowbot/trials
Developer for mySociety. Interested in medical science, civil society, climate change.
Nerd cheerleader, Bad Science person, stats geek, research fellow in epidemiology, procrastinator. TED here: http://bit.ly/qUcrmf And email: ben@badscience.net bio from Twitter
Sign in to add slides, notes or videos to this session