Science and Technology CommitteeJoint written evidence submitted by the Royal Society

1. The Royal Society has noted the Committee’s consultation on Clinical Trial data with interest. This is an area the Society has a touched upon in its report Science as an open enterprise published in June 2012. The report addressed the broader issues around making scientific data intelligently open. This letter will be addressing the Committee’s questions 3, 4 and 5.

2. The principal tenets of open data apply particularly to the topic of clinical trial data. This letter explains the reasons for this and why the UK should be opening up clinical trial data, the benefits and limitations of doing this, and the potential methods for doing it.

3. Open scientific data can benefit society in four main ways: it provides a layer of transparency that engenders public trust; it can quicken the pace of scientific discovery; it is a source of wealth creation; and it is a potential deterrent to scientific misconduct and fraud.

4. There are limitations to openness that need to be observed. In clinical trials where the data generated contains a large amount of personal information about individuals, limitations on the openness of that data are likely to be necessary for opening up the trial to outside parties. There are also occasions where the disclosure of data threatens security and safety and this is a legitimate reason for not opening up data, but must be explicitly stated. There are also legitimate reasons for embargoes of data for the purposes of commercial viability, but it is reasonable to expect the data to be made open within a specified time period.

5. Open clinical trial data would also lead to greater scrutiny of trial methodologies and chemical compounds. Clinical trials registries address the limits of what companies know about what is currently being (and has been) done, and would lead to more efficient use of resources. Clinical trial registries and compliance to them are vitally important.

6. There is also the concern, central to the movement to make clinical trial data open, that pharmaceutical companies conducting trials could selectively publish results, which may skew the understanding of how effective the clinical trials have been.1 There is evidence that this selective publication of clinical trial results can create favourable bias towards the uptake of a new drug.2 The partial reporting of clinical trials results distorts understanding and can be viewed as a form of scientific misconduct.3 Partial reporting of data may be due to innocent error, but can also be fraudulent if data is cherry-picked to demonstrate a relationship that would not be apparent if the full dataset were used and published.4 Non-reporting of trials slows down the progress of science and wastes public money.

7. In cases where it is essential for researchers to be able to access clinical trial data with personal information intact, there are ways in which this can be done safely and securely. “Safe Havens” are one method,5 where there is a fixed physical point of access to data, and the dataset is prepared for the specific, limited needs of the researcher. However, issues around the permission for access to this information need to be clear to all parties.

8. It should be noted that GlaxoSmithKline announced in October 2012 that they were embarking on a programme of making their clinical trial data open starting with existing treatments and most relevant advances, then opening up their archive of trials.6 They have also signed up to Ben Goldacre’s initiative AllTrials.net to register all clinical trials and disclose their clinical trial results and clinical study reports.7 , 8 This is a key development and GSK is to be commended.

9. The Royal Society continues to be interested in Open Data issues and awaits the outcome of the Committee’s inquiry.

February 2013

1 Hart B, Lundh A and Bero L (2012). Effect of reporting bias on meta-analyses of drug trials: reanalysis of meta-analyses. British Medical Journal, 344, d7202

2 Turner E H, Matthews A M, Linardatos E, Tell R A & Rosenthal R (2008). Selective Publication of Antidepressant Trials and its influence on apparent efficacy. The New England Journal of Medicine, 358, 252–260. Available at: http://www.nejm.org/doi/full/10.1056/NEJMsa065779

3 Chalmers I (1990). Under-reporting research is scientific misconduct. Journal of the American Medical Association, 263, 1405-1408.

4 Boulton, G, Rawlins, M, Vallance, P and Walport, M (2011). Science as a public enterprise: the case for open data. The Lancet, 377, May 14.

5 Thomas R & Walport M (2008). Data Sharing Review. Available at: http://www.justice.gov.uk/reviews/docs/data-sharing-review-report.pdf

6 http://www.guardian.co.uk/society/2012/oct/11/glaxosmithkline-clinical-trials-data?CMP=twt_fd

7 http://www.alltrials.net/

8 http://www.alltrials.net/supporters/gsk-statement/

Prepared 16th September 2013