Sharing our trial results, regardless of whether they reflect positively or negatively on our medicines is the right thing to do.
It’s essential if we are to advance scientific understanding and build trust. It allows us to acknowledge the great contribution made by those who take part in our clinical research. These patients volunteer their time with the hope not just that they might ultimately benefit from an investigative treatment themselves, but that they might further the understanding of a disease and its treatment through their contribution.
Publishing the results of our research also provides healthcare professionals with critical information to help them apply trial results to daily clinical practice, so they are able to make the most appropriate treatment decisions for their patients. That’s why we post information about all our clinical studies on publicly accessible registers before they start, and update these with results summaries after the studies are finished. We also seek publication of the results of all clinical research we conduct on our medicines in peer-reviewed scientific journals, regardless of whether they reflect positively or negatively on our medicines.
To evaluate how we are measuring up to our transparency commitments, GSK recently conducted a six-year retrospective review to determine whether there was any bias in the publication of our study results. We explored whether the outcome of a study influenced our submission to peer-reviewed journals, and also whether the acceptance rates from those journals differed according to whether a study had a positive or negative result. Our review examined more than 1,000 studies carried out during that time.
The study’s results, published recently in the BMJ, found no evidence to suggest that we favour submitting positive studies over negative ones for publication. Indeed, submission and publication rates were higher in our review for studies with a negative outcome (92% submitted, 77% published) than those with positive outcomes (79% submitted, 66% published).
Patrick Vallance, President, R&D commented: “The results of this analysis allow us to challenge perceptions that pharmaceutical companies cherry-pick what is submitted for publication according to favourability towards their medicines. It’s clear that GSK does not do that and the medical journals did not apply any bias either. These results reinforce our strong belief that the scientific community can derive as much, if not more, insights from negative data as positive data, and we must continue to ensure access to all of our clinical study data involving patients, with the aim of ultimately benefitting patients”.
An in-depth look
Looking at the study in more detail provides some broader measures supporting GSK’s progressive approach to data transparency.
of GSK’s studies had results posted to one or more public registers
At the time of analysis cut off, 85% of GSK’s studies had been submitted as a full manuscript. An additional 7% had been disclosed as congress abstracts (not included in submission or publication rates). In addition, 98% of GSK’s studies had results posted to one or more public registers, including 100% of studies subject to FDA requirements for posting to www.clinicaltrials.gov. The study also shows our continuous efforts to re-submit manuscripts after initial rejection, in some cases three or more times, to get our research results out to the wider scientific community.
Continuously exploring and improving
We want to enable researchers to learn and build on our research so we’re always exploring other vehicles to increase access to clinical research – such as our publicly accessible study register and an online portal for researchers to access the detailed data behind our research.
Conducting analyses of our efforts to submit and publish our research ensures we continue to hold ourselves accountable for fulfilling our obligation to patients and physicians - while correcting misperceptions regarding publication bias that can erode the credibility of our research.