Results of clinical trials are often not published. We were interested in finding out how this comes about, and what past trial participants and others think about it.
It is estimated that about 4 in every 10 randomized controlled trials are not published in academic journals or only published after a long delay. Clinical studies with unfavourable results for the tested drug are less likely to be published—a phenomenon known as publication bias. When trial results are not published, this makes it more difficult to understand the risks and benefits of drugs. This has led to less informed patient care and unnecessary harm to patients.
We conducted a research study called “Inside trials: an interview study of clinical trials and clinical trial reporting.” The aim of this qualitative study was to examine factors that contribute to nonpublication of clinical trials in Canada and related ethical issues.
This University of British Columbia study was based on interviews with 34 participants, including 10 clinical trial participants, 17 clinical trial investigators, 1 clinical research coordinator, 3 research administrators, and 3 research ethics board members. Interviews for the study were conducted between March 2019 and April 2021.
Trialists who participated were from a range of medical fields, including cardiovascular medicine, endocrinology, hepatology, infectious diseases, oncology, psychiatry, and rheumatology. Trial participants had taken part in trials of treatments for cardiovascular disease, C. difficile infections, chronic pain, diabetes, eye disorders, and multiple sclerosis. Participants were from the Canadian provinces of British Columbia, Alberta, and Ontario.
Reporting our study findings
This research study produced findings on three main issues:
- Industry sponsor influence on whether the results of clinical trials are reported
- Factors, other than industry influence, relating to whether results of clinical trials are reported
- Reporting of clinical trial results as an ethical responsibility to research participants.
Findings on these issues will be reported in separate academic journal articles. As the results are published, key findings will also be reported on this website and links to abstracts or full articles will also be reported.
To date, we have reported our findings on the first two of these issues:
- An article on industry sponsor influence in clinical trial reporting is published in the journal Clinical Therapeutics. An abstract for this article is available on the journal’s website. While the final article is only available to the journal’s subscribers at this time, the accepted manuscript and an appendix with additional information may be downloaded from this site. Clinical Therapeutics also published an editorial accompanying our article in which it pledged to address scientific bias in its publications.
- An article on other factors relating to whether results of clinical trials are reported is published in the British Journal of Clinical Pharmacology. An abstract for this article is available on the journal’s website. While the final article is only available to the journal’s subscribers at this time, the accepted manuscript and an appendix with additional information may be downloaded from this site.
The results of our study, covering all of the issues listed above, are reported in the main researcher’s PhD dissertation, “Nonpublication and publication bias in clinical trials in Canada: a qualitative interview study,” which is available for download from University of British Columbia’s cIRcle repository.
Industry sponsor influence in clinical trial reporting: key findings
We found that industry sponsors of clinical trials have a weaker incentive to report the results of certain trials, including those with unfavourable results for the tested drug and those for drugs they have decided not to market.
Interviews with trial investigators and others highlighted mechanisms through which sponsors may influence whether a trial is reported:
- First, sponsors may influence reporting by stopping a trial early and not proceeding to publish the findings.
- Second, sponsors typically own and may control access to the key data from a trial, which may be an obstacle to investigators reporting trial results if the sponsor does not support publication.
- Third, sponsors may negotiate clinical trial agreements in multicentre trials which do not protect the ability of site investigators to publish based on all of the data from a trial if the sponsors and trial leaders do not proceed with publication.
Early phase internal company trials of investigational drugs are an additional source of unpublished trials. Importantly, dependence of researchers and research centres on funding from industry sponsors may weaken their ability to negotiate terms with industry sponsors that fully protect the right of investigators to publish findings.
Factors contributing to nonpublication and publication bias: key findings
While a range of factors contribute to nonpublication and publication bias, our study suggests powerful incentives within the research system, which provide greater rewards for reporting positive findings than negative findings, shape clinical trial reporting practices. Key findings on incentives:
- Investigator experiences or perceptions of the difficulty of publishing negative findings in journals may discourage investigators from submitting trials with negative results for publication.
- The accounts of interview participants suggested positive findings may be more likely to lead to industry and nonindustry funding for trial research.
- Research institutions may contribute to incentives of researchers to focus on positive findings through hiring and promotion decisions which reward researchers for attracting funding and publishing in prestigious journals. Positive findings also tend to receive greater recognition.
Policies to encourage full reporting of clinical trials have not been strong enough to counterbalance these incentives. Until recently, Canada’s main public funder of clinical trial research, the Canadian Institutes of Health Research, did not require reporting of trial results within a specific timeframe. Also, Health Canada has not introduced mandatory requirements for investigators or sponsors to report trial findings within a clinical trial registry, as has been done in the United States and European Union.
We would like to express our gratitude to everyone who was willing to take part in this study and to discuss their views and experiences, without whom this research would not have been possible. In addition, we would like to thank the clinical research coordinators who helped with recruitment for the study by seeking consent from past trial participants for us to contact them about the study, and their colleagues who were supportive of this research.
Additional information about the study:
This study was conducted by Richard Morrow as his PhD research under the guidance of his co-supervisors Colin R. Dormuth (University of British Columbia) and Barbara Mintzes (University of Sydney) and PhD committee members Garry Gray (University of Victoria) and Michael R. Law (University of British Columbia). Scott Garrison (University of Alberta) also provided critical feedback on the study.
The Principal Investigator of the study was Colin Dormuth, Associate Professor, Department of Anesthesiology, Pharmacology & Therapeutics, University of British Columbia.