A theme issue in 2011 on unpublished evidenceBMJ 2011; 342 doi: https://doi.org/10.1136/bmj.d2627 (Published 28 April 2011) Cite this as: BMJ 2011;342:d2627
- Elizabeth Loder, associate editor
- 1BMJ, London WC1H 9JR
The BMJ has long championed efforts to ensure complete publication of clinical trial results. Studies published in the BMJ and elsewhere show that unpublished or selectively reported clinical research findings lead to distorted perceptions of treatment effects.1 2 The harm is compounded when the results of individual trials are combined in systematic reviews or meta-analyses. Doctors, patients, and policy makers rely on these evidence syntheses as authoritative summaries of knowledge about clinical questions. But when important evidence is unavailable the conclusions reached by these research summaries may be wrong.
Such evidence syntheses are widely considered to represent the highest form of evidence, but as we have asked in a previous editorial, “is this a tenable world view when so much evidence is missing?”3 The existing flawed clinical evidence base cannot simply be ignored because it forms the basis for future research. The current emphasis on comparative effectiveness research, in which new treatments are compared with established treatments rather than placebo, makes tackling the problem even more urgent. If the evidence base for established treatments is …