Is transparency really a panacea?

It is often remarked that sunshine is the best disinfectant. This view of disclosure as a powerful accountability tool has become an unquestioned orthodoxy in many governance circles. It stands behind and informs the recent US Physician Sunshine Act (PSA).¹ This federal legislation uses a national public registry that discloses physicians’ financial relations with the pharmaceutical and medical device industry to tackle the problem of commercial conflicts of interest (COIs). The PSA stress on disclosure is seen to empower the public to take greater control over health decisions and to mitigate the influence of bias rooted in commercial COIs.² And it has inspired a call in the UK for a similar approach to tackling commercial COIs.³

But how effective is disclosure as a stand-alone measure in mitigating bias? What does the evidence show? Greater attention to disclosure’s performances and governance character suggests that the limits of this accountability tool have yet to be fully appreciated.⁴

Consider the use of disclosure to manage commercial COIs on Wall Street. In the wake of a long list of profiled scandals involving tainted auditing due to consulting contracts with corporate clients, disclosure of the commercial COIs became the governance option of choice on Wall Street. The measure failed to prevent further scandals involving commercial COI and kept the biases in tact.⁵ It sustained the status quo. A recent study sees a similar pattern in psychiatric medicine and has led to the call to go beyond transparency to combat bias.⁶ The study dovetails with research that suggests that local state Sunshine laws in the US have not had a discernible effect on physicians’ prescribing habits.⁷ Given disclosure’s poor track, it is not clear how a federal registry would better empower the public to mitigate bias?

Indeed Jerome Kassirer, former chief editor of the New England Journal of Medicine, has argued that discussions about commercial COIs have been captivated by a powerful mythology involving transparency.⁸ In medicine – and the science that underpins it – isolating a bias and assessing its implications normally involves specialised training, knowledge and practice. But the public lacks the required training and background knowledge and a registry of commercial COI does not provide it. The need for these tools and how to use them is especially important given that bias often functions in a self-serving unconscious way.⁵ A registry simply draws attention to a commercial conflict and flags attention to the possibility that medical information that is being presented may or may not be biased. But it does nothing more. It does not inform the public whether the person with a commercial COI is the grip of bias or not. Nor does it provide the means to assess the implications of a bias. And to suggest that these tasks can be accomplished simply through disclosure is to endow this accountability tool with extra ordinary super governance powers. It is not surprising then that disclosure has had a poor record on combating bias on Wall Street and in psychiatric medicine. It has been mythologised as a cure-all.

What lessons can be drawn from assembled reminders of the limits of transparency that might be useful for further discussion in the UK on tackling commercial COIs in medicine?

First, transparency is clearly not a panacea or magic bullet. The point is perhaps obvious and it has been made before. But it bears repeating given a strong tendency to elevate transparency to the governance status of an unassailable self-evident philosophical principle. This step into mythologising can be resisted by keeping our sights focused on the work that disclosure can and can not do as an accountability tool. And what it cannot do on its own – as the evidence shows – is effectively combat bias. Hopefully, this reminder will help prevent a historical amnesia on disclosure’s limits from becoming to firmly embedded in public policy.

Second, we also need to keep our sights fixed on the actual problem. In addition to diverting us down the wrong governance path, another consequence of being bewitched by the PSA approach to bias is that it can blind us from seeing the real problem that remains to be solved. The debate over mitigating the negative influences of bias has been framed mainly around the need to disclose commercial COIs. And as a result, our attention has been diverted from fixing our governance sites on the commercial conflicts themselves and their corrupting influence.⁸ This is the problem that needs to be foremost addressed and a registry of commercial COIs on its own does not solve it.

Third, it might also be tempting to reject transparency as an accountability tool given its flaws. But this would be a colossal mistake. What needs to be rejected is simply a mythological view of transparency that is embodied in the PSA. One lesson that emerged from the evidence on Wall Street and in psychiatric medicine is that disclosure is insufficient to ensure the independent assessment of the integrity of data. A national registry that discloses commercial COIs could play an important role here. From such a registry, experts who are free of commercial COI could serve on practice guideline and oversight committees that serve a public gate keeping role in assessing the integrity of medical data. And members of such committees should have unfettered access to raw clinical trial data, past, present and future in order to better ensure public trust in a healthcare system that is rife with commercial COI.

As the popular adage goes, there is no need to throw the proverbial baby out with its dirty bathwater – even if the most popular cleansing agent cannot get the job done on its own.

Declarations

Competing interests

None declared

Funding

None declared

Guarantor

MW

Ethical approval

Not applicable

Contributorship

MW is responsible for all aspects of the text.

Acknowledgements

This paper is dedicated to the loving memory of Rita Doreen Wilson – a kind and generous soul.

Provenance

Invited contribution