Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles.
We welcome the article by Krumholtz et al [1] on data sharing and agree with the concept that trial data should be accessible by all. This is important not only for unpublished trials, but also for those that are already published. Clinical trials provide a large amount of data that is important beyond the primary analysis. These are often used by the trial team to publish secondary analyses, but it is not uncommon for the study team to move on to the next project without fully exploiting the database. Making the trial dataset publically available allows other study teams to explore their own hypotheses, to conduct individual patient data meta-analysis, and obtain data in their preferred form. It also allows the primary analysis to be checked. Ultimately, transparency should be encouraged as it leads to higher quality research, better value for money and higher quality science.[2]
In 2010 we published a large placebo-controlled randomised trial of umbilical oxytocin for the treatment of retained placenta.[3] Stimulated by the debate on transparency and data publishing, we have recently published the study dataset on the 577 recruits on Dryad, the online data repository recommended by the BMJ[4]. We obtained ethical approval for this and, in line with recommendations to ensure confidentiality, we have published only 2 indirect identifiers in our database. Hrynaszkiewicz et al[5] suggest 3 or less indirect identifiers in the published dataset so as to ensure anonymity and patient confidentiality. Our dataset is published in SPSS format so as to provide detailed descriptions of the data. Users should refer to the original publication[3] for details of the study methods.
Kursheed Khan & Andrew D Weeks on behalf of the Release Study Team.
References
1) Krumholz HM. Why data sharing should be the expected norm. BMJ. 2015 Feb 5;350:h599.
2) Wellcome Trust. Sharing research data to improve public health: full joint statement by funders of health research. www.wellcome.ac.uk/About-us/Policy/Spotlight-issues/Data-sharing/Public-... - accessed 08Mar2015
3) Weeks AD, Alia G, Vernon G, Namayanja A, Gosakan R, Majeed T, Hart A, Jafri H, Nardin J, Carroli G, Fairlie F, Raashid Y, Mirembe F, Alfirevic Z. Umbilical vein oxytocin for the treatment of retained placenta (Release Study): a double-blind, randomised controlled trial. Lancet. 2010 Jan 9;375(9709):141-7.
4) Weeks AD, Alia G, Vernon G, Namayanja A, Gosakan R, Majeed T, Hart A, Jafri H, Nardin J, Carroli G, Fairlie F, Raashid Y, Mirembe F, Alfirevic Z. (2009) Data from : Umbilical vein oxytocin for the treatment of retained placenta (Release Study): a double-blind, randomised controlled trial. Dryad data repository. http://dx.doi.org/10.5061/dryad.g3gj1
5) Hrynaszkiewicz I, Norton ML, Vickers AJ, Altman DG. Preparing raw clinical data for publication: guidance for journal editors, authors, and peer reviewers. BMJ. 2010 Jan 28;340:c181.
Competing interests:
No competing interests
30 March 2015
Andrew D Weeks
Professor of International Maternal Health
Kursheed Khan and Andrew D Weeks on behalf of the Release Study Team
University of Liverpool
Department of Women's and Children's Health, Liverpool Women's Hospital, Crown Street, Liverpool L8 7SS
In an entirely altruistic world, everything would be shared. But when you've sweated to do an experiment, when you are the one who put it all together, and did the hard graft, why should you share your data with others who will gladly take the credit and the promotion? Yes of course, so that patients suffer less; but human beings are more complicated than that.
While what I state is simplistic - little medical research is done by single workers sweating overnight at their lab benches - there are complexities and difficulties lying behind this little sentence from the last paragraph of this editorial: "Academics must find ways to credit those who produced the data."
Retrospective dataset publication through Dryad: an example from a 2010 retained placenta trial
We welcome the article by Krumholtz et al [1] on data sharing and agree with the concept that trial data should be accessible by all. This is important not only for unpublished trials, but also for those that are already published. Clinical trials provide a large amount of data that is important beyond the primary analysis. These are often used by the trial team to publish secondary analyses, but it is not uncommon for the study team to move on to the next project without fully exploiting the database. Making the trial dataset publically available allows other study teams to explore their own hypotheses, to conduct individual patient data meta-analysis, and obtain data in their preferred form. It also allows the primary analysis to be checked. Ultimately, transparency should be encouraged as it leads to higher quality research, better value for money and higher quality science.[2]
In 2010 we published a large placebo-controlled randomised trial of umbilical oxytocin for the treatment of retained placenta.[3] Stimulated by the debate on transparency and data publishing, we have recently published the study dataset on the 577 recruits on Dryad, the online data repository recommended by the BMJ[4]. We obtained ethical approval for this and, in line with recommendations to ensure confidentiality, we have published only 2 indirect identifiers in our database. Hrynaszkiewicz et al[5] suggest 3 or less indirect identifiers in the published dataset so as to ensure anonymity and patient confidentiality. Our dataset is published in SPSS format so as to provide detailed descriptions of the data. Users should refer to the original publication[3] for details of the study methods.
Kursheed Khan & Andrew D Weeks on behalf of the Release Study Team.
References
1) Krumholz HM. Why data sharing should be the expected norm. BMJ. 2015 Feb 5;350:h599.
2) Wellcome Trust. Sharing research data to improve public health: full joint statement by funders of health research. www.wellcome.ac.uk/About-us/Policy/Spotlight-issues/Data-sharing/Public-... - accessed 08Mar2015
3) Weeks AD, Alia G, Vernon G, Namayanja A, Gosakan R, Majeed T, Hart A, Jafri H, Nardin J, Carroli G, Fairlie F, Raashid Y, Mirembe F, Alfirevic Z. Umbilical vein oxytocin for the treatment of retained placenta (Release Study): a double-blind, randomised controlled trial. Lancet. 2010 Jan 9;375(9709):141-7.
4) Weeks AD, Alia G, Vernon G, Namayanja A, Gosakan R, Majeed T, Hart A, Jafri H, Nardin J, Carroli G, Fairlie F, Raashid Y, Mirembe F, Alfirevic Z. (2009) Data from : Umbilical vein oxytocin for the treatment of retained placenta (Release Study): a double-blind, randomised controlled trial. Dryad data repository. http://dx.doi.org/10.5061/dryad.g3gj1
5) Hrynaszkiewicz I, Norton ML, Vickers AJ, Altman DG. Preparing raw clinical data for publication: guidance for journal editors, authors, and peer reviewers. BMJ. 2010 Jan 28;340:c181.
Competing interests: No competing interests