PLOS’s continued commitment to transparency in trial publication

As of today, PLOS is retiring its two “Hubs”, which have respectively highlighted and brought together open-access content in two specific areas: Biodiversity and Clinical Trials. These sites were initially developed to experiment with how the open-access literature might be reorganised and filtered across journals, providing ways of enabling readers in particular fields to access material of interest to them. The Hub:Clinical Trials was dedicated to highlighting clinical trials research across the PLOS journals, as well as maintaining an archive of the original work previously published in PLOS Clinical Trials. Although the Hub will no longer exist, all unique articles originally published in PLOS Clinical Trials will continue to be openly available at http://www.plosclinicaltrials.org, as well as in PubMed Central. Trials research published in other PLOS journals will be available as usual on those journals’ sites. The decision in no way diminishes PLOS’s commitment to publishing all correctly conducted and reported clinical trials and we also intend to explore new ways of aggregating content in future. PLOS has always been committed to raising awareness about the effects of publication biases in relation to clinical trial data, and to providing ways of addressing those biases. This commitment now continues through a number of ongoing initiatives which PLOS journals continue to support.

Firstly, PLOS ONE continues its commitment to publishing the results of clinical trials research, irrespective of trial outcome. The journal’s publication criteria have always emphasised that for a paper to be published, following rigorous peer review, that work is technically sound and properly reported. In relation to clinical trials research, PLOS ONE (and other PLOS journals) requires authors to use CONSORT guidelines in reporting their study; to have prospectively registered their trial in a public, international registry such as clinicaltrials.gov; and to provide a copy of the original trial protocol as supporting information for editors, reviewers and readers to evaluate against the published report. The journal therefore provides a unique mechanism to rectify publication bias in the trials literature – a bias that has been described by the Alltrials Campaign as potentially “leading to bad treatment decisions, missed opportunities for good medicine, and trials being repeated”. Since January 2011, PLOS ONE has accepted for publication 406 articles classified as trials. Included amongst the papers published are frequent examples of well-conducted studies which fail to find “positive” effects of experimental interventions – such as this study looking at a complex intervention aimed at decreasing x-ray referrals for patients with acute low back pain; and this study evaluating the effects of an advocacy initiative aimed at reducing the risk of pedestrian injury.

Secondly, PLOS staff members continue to be involved in external initiatives which aim to address publication bias and to increase transparency in trial publication. Both Ginny Barbour (Editorial Director for Medicine at PLOS and Chief Editor, PLOS Medicine) and I have been involved in the European Medicines Agency’s Advisory Groups regarding its development of a new policy for making publicly available the original datasets  for clinical trials which underpin drug regulatory decision making in Europe. The Agency has now released its draft policy which is open for public comment. The policy makes some important statements. For example, EMA does not consider clinical trial data to be commercially confidential, and the draft policy states that the interests of public health outweigh the considerations of commercial confidentiality. The draft policy would make all summary clinical trial data and documents (where there are no concerns about patient privacy) open access, at the time of the EMA’s report of a positive, negative or withdrawal decision on a drug indication.

PLOS is an initiating organisation behind the Alltrials initiative, which it endorses. Alltrials calls for “all trials to be registered and reported”. PLOS’s declaration of support for Alltrials can be read here. As a result of recent campaigning, some forward-thinking drug companies – such as GSK and Medtronic  – have supported or developed initiatives to freely share individual patient level trial data with academic researchers for independent scrutiny.

Finally, PLOS journals (specifically PLOS Medicine and PLOS ONE) have recently declared support for a creative new approach for getting unreported trial results into the published literature. Both journals have stated that they will support the RIAT initiative, a proposal whereby authors independent of the original triallists or sponsors could revive unreported trials and independently publish their findings based on documents obtained through public-domain documents, such as via freedom of information requests.

These and other initiatives are beginning to change the trials landscape, but there is still a very long way to go before the public can be reassured that the findings of all trials will be made immediately available in an unbiased way to inform clinical decision making. PLOS editors will continue to support efforts aimed at ensuring the full and honest reporting of the results of all trials.

My competing interests: I have contributed to one of the advisory groups of the European Medicines Agency in relation to publication and access to clinical trial data. I have also acted as a contributor to meetings of groups aiming to develop consensus statements for improved reporting of research, such as the development of the CONSORT guideline for patient-reported outcomes (I’ve received subsidy for travel expenses for such meetings). I am on the advisory group for Current Controlled Trials.

This entry was posted in Peer review and tagged , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>