Requiring anonymised raw data as a condition for publicationBMJ 2015; 350 doi: https://doi.org/10.1136/bmj.h437 (Published 28 January 2015) Cite this as: BMJ 2015;350:h437
- John H Noble Jr, emeritus professor, State University of New York at Buffalo1
The problem of conflict of interests for medical journals will largely go away when they adopt the policy of “publishing the reports of all and sundry, regardless of their commercial interests, subject to the pledge and actual delivery of anonymised raw data on which the report depends for internal and external validity.”1 2
By extension, authors of instructional materials and education reports could show their lack of bias by citing reports that have pledged and, better still, actually delivered on the promise of providing anonymised raw data. In the beginning, this will be a challenge for medical journals, which traditionally vet manuscripts for publication on the basis of imperfect and sometimes interest conflicted peer reviews.
The challenge now is for medical journals to stop going on about conflicts of interest and to adopt the robust policy of requiring, as a condition for publication, that authors provide access to the anonymised raw data underlying their reports.
Cite this as: BMJ 2015;350:h437
Competing interests: None declared.