BMJ Open Data Campaign
Posted on 6th June 2013 by Alice Buchan
As argued vehemently and elegantly by Ben Goldacre in his book Bad Pharma , doctors cannot make evidence-based decisions without the all the data available. Based on similar logic, the BMJ started their Open Data campaign . Since the start of this year, the BMJ have not published any trials in which the authors would not make anonymised patient data available on request.
On their Open Data campaign page, there are also links to four key examples, which I shall detail briefly below.
The main page for the BMJ’s information about open access to Relenza trials information is here. Essentially, it details correspondence between Cochrane reviewers and GSK regarding data that they would like to review, but that was not available to them.
This was the source of the BMJ’s first open data campaign, launched at the height of the swine flu pandemic, and tracks the release of data since Roche’s promise to do so in 2009. Found here, the Tamiflu page is possibly the most comprehensive campaign, with several interesting features. Such as links to BMJ editorial, feature, and other articles surrounding it, which all make an interesting read. There is also an excellent interactive timeline, starting with the first H5N1 “bird ‘flu” cases in Hong Kong in 1997, and including details of Cochrane reviews of treatments, dates of the approval of Tamiflu, and of publication of WHO guidelines, as well as BMJ articles relating to pandemic influenza and Tamiflu. Many of these include links to the relevant resources or articles.
There was also a 2007 feature on what lessons could be learnt from Vioxx. This looks at the trials and early warning signs that there might have been increased cardiovascular risk. It suggests some measures that could be taken by academics, industry, and academic journals to improve in future.
Similarly, this article looks at the failure of Avandia, and the lessons to be learnt from this. It highlights the need for sufficient evidence, and the possible long-term consequences of approving and prescribing drugs based on limited data.
Overall, there is a wealth of information available regarding the availability of unpublished data, and the steps being taken to make it available in the BMJ Open Data Campaign site, including a link to the web version of a 2012 special issue on missing data. Taken together, these articles, emails, and timelines should be of interest to anyone involved in medicine or evidence-based medicine, and really highlight the need for access to all the data in order to come to an evidence-based conclusion.
 Goldacre B. Bad Pharma. London. HarperCollins. 2012
 BMJ. Open Data. [accessed 4th June 2013] available at: http://www.bmj.com/open-data