A matter of trust

Stephen Bradley

Maintaining trust is essential for the long-term viability of the pharmaceutical industry. Greater openness, including responsible stewardship of research data, is the best way to build that trust.

The pharmaceutical industry has a problem with trust. To some extent, this is unfair – publicly funded research has been dogged by the same issues. But trust matters. Patients should be able to trust the pharmaceutical industry to participate in clinical trials, while taxpayers and regulators need to be able to believe that the claims made are grounded in verifiable evidence. Any further deterioration in trust could well lead to further regulatory impositions on the pharmaceutical industry and fuel calls for the state to take over more aspects of drug discovery and manufacture. Greater openness is the most effective way to build trust.

The good news for pharma is that standards are so inconsistent that industry could easily outflank academia and lead the way in transparency. In fact, pharma’s strengths in terms of processes and quality assurance may give it a competitive advantage over academia in adopting transparent practices. Commercial sponsors of research already achieve greater compliance in reporting clinical trials than their academic counterparts, while the success of some companies in achieving greater transparency suggests that such practices could become much more widespread.

As part of a collaboration, we have proposed that three basic standards of transparency should be widely adopted. This blog post outlines some of the benefits and challenges to pharma in supporting such basic measures before considering the particular challenges and responsibilities with regards to data sharing.    

Disclosing interests

The intersection of industry and academia has been dogged by scandals around conflicts of interest that have undermined trust for decades. Pharma has much to gain by ensuring that it meets society’s basic expectations by helping to achieve mandatory disclosure of interests. The industry’s establishment of a register of interests is to be applauded, but voluntary disclosure is patchy, and ongoing opacity around conflicts of interest continues to threaten the industry with reputational harm.

By lobbying for a mandatory register of interests, the pharmaceutical industry could help quell perceptions that it is complicit in underhand practices and demonstrate that it is sincerely committed to tackling this issue. It has little to fear from full disclosure – Sunshine Acts have been introduced through much of the world applying to clinical researchers. The UK pharmaceutical industry can go one better and help grant the public a right to find out about potentially competing interests for all researchers, both clinical and non-medically qualified. Ahead of legislation to require disclosure by law, the industry can do much to normalise transparency by encouraging comprehensive and up-to-date disclosures on existing platforms, for example, through staff appraisals and before signing off on academic and commercial partnerships.

Tackling publication bias with Registered Reports

Industry can also take the lead in embedding open science practices in publishing research. Pharma can readily build on high rates of trial reporting by preregistering all kinds of research on public platforms. Meanwhile, private sector employers can help recognize the scientific contributions of their researchers and improve job satisfaction by encouraging staff to get work published, whatever the results. Given a tendency for ‘negative’ findings to languish in the file drawer, the industry should lobby journals to accept the registered reports publication format. Registered reports are accepted by peer reviewers on the basis of methods and provide a route to publication regardless of results, providing agreed study plan is implemented. Given that only a tiny proportion of biomedical research is published as registered reports, industry could very readily outperform publicly funded institutions by implementing this mechanism to overcome publication bias. Such practical action would do much to demonstrate the sincerity of industry in tackling an issue that has done much to tarnish its reputation.

Sharing study documentation and responsible data stewardship

Thorough appraisal of published research often requires access to original documents, such as study protocols. We have called for such documents to be available for all publicly funded studies through an open repository, but again industry is probably particularly well placed to institute sharing of such basic documentation as standard. 

Sharing of research data, such as individual participant data (IPD), presents different challenges since there are legitimate concerns about risks to privacy, even for apparently anonymized information. So, it is not appropriate to demand that all IPD be made publicly available. But industry and researchers need to start constructing the norms and processes required to support data sharing and engaging with the public on these issues. Given the additional discoveries that can be leveraged by sharing data and the necessity of opening research findings to scrutiny, it is ethically problematic, as well as inefficient, to hoard data without also investing in the means for the data to be used to accomplish the maximum benefits for all.  

Although sharing data comes with risks and the public has appropriate concerns, ruling out data sharing feels like a paternalistic cop-out. The onus is on researchers and industry to engage with the public and to help explain that the risks to privacy can be mitigated and that these need to be weighed against the costs and harms of hoarding data. It is unduly pessimistic to assume that barriers cannot be overcome. Already, many patients are calling for their data to be shared.

Hopefully, the recently announced Goldacre Review will help establish rational principles that reference the need to protect privacy and to obtain the maximum public benefit from patients’ data. But there are already opportunities for the pharmaceutical industry to think imaginatively in designing new trials. For example, consent processes might involve allowing participants to see examples of the kind of anonymized data that might be shared about them, with different variables colour-coded to demonstrate the different tiers of detail that would be made available to different agencies and applicants. Transitioning to a system that facilitates responsible data sharing will not happen overnight, but those who control this valuable resource have a responsibility to ensure it is utilized for the public good, on behalf of those who donated it.

Deciding what data can be shared and with whom is a daunting challenge that requires specialized expertise. Initiatives such as the Yale Open Data Access Project (YODA) can support these processes by offering third-party arbitration of requests to access study data. Data that are shared through services such as YODA are also likely to be more persuasive and less vulnerable to suspicion than released data that have been cherry-picked to show a product in its most favourable light.

Negative perceptions about pharma are a liability for the sector, but there are clear opportunities for the industry to demonstrate it takes its corporate social responsibility seriously. Supporting basic measures to improve transparency is a prudent and inexpensive way for the industry to safeguard its long-term interests.

Stephen Bradley is a steering group member of the Declaration to Improve Health Research (improvehealthresearch.com), which calls for three basic measures to be widely adopted in biomedical research. Stephen Bradley would like to thank Nicholas DeVito for providing comments on a draft of this blog post. The views presented here are those of the author alone and do not necessarily reflect those of signatories of the Declaration to Improve Health Research or employers or funders. A statement of interests for Stephen Bradley can be found here