Intended for healthcare professionals

Editorials

The rise of ambiguous competing interest declarations

BMJ 2018; 361 doi: https://doi.org/10.1136/bmj.k1464 (Published 10 April 2018) Cite this as: BMJ 2018;361:k1464
  1. Barbara Mintzes, senior lecturer,
  2. Quinn Grundy, postdoctoral fellow
  1. Faculty of Pharmacy and Charles Perkins Centre, University of Sydney, Australia
  1. barbara.mintzes{at}sydney.edu.au

Terms such as “unpaid consultancy” should be consigned to the dustbin

A recent analysis in the Journal of General Internal Medicine reports on a new trend: many more authors of medical journal articles are disclosing that they are “unpaid consultants” to pharmaceutical, biotech, or medical device companies. Before 2000, there were one or two such disclosures a year; this increased to over 150 a year from 2012 to 2014.1

This form of disclosure may reflect greater openness, or it may disguise industry sponsorship of medical research with new terminology. Menkes et al suggest that declarations of “unpaid” consultancies “may do more to conceal than illuminate.” Although it implies that cash was not transferred, the definition of “unpaid” does not preclude other financial and material advantages. These consultancies may include payments for conferences, travel, or hospitality, for example. Such gifts can be more than incidental: Australian industry reports of sponsored events for health professionals included “a $176 000 (£95 875; €144 000) junket to Vancouver for nine dermatologists.”2 Attendees received only food and drink at 65% of the 116 845 industry sponsored events for health professionals held in Australia from 2011 to 2015, making them all “unpaid” and yet in receipt of industry largesse.3

There is solid evidence that industry sponsorship is associated with research outcomes that are more favourable to the sponsor. This association holds for both direct sponsorship4 and more indirectly through the financial links of principal investigators leading clinical trials and the outcomes of those trials.5 Doctors who receive even inexpensive meals from drug companies are more likely to prescribe their brands, with a dose response seen by the numbers and price of meals received within four drug classes.6

A random sample of biomedical journal articles published in 2016 found the highest prevalence of reported conflicts of interest among authors of commentaries, editorials, and drug related studies.7 Bias in commentaries is especially concerning because of their role in summarising evidence and providing practical advice for care. After the publication of the Women’s Health Initiative, for example, commentaries on hormone replacement therapy by conflicted authors were more likely to support its use, despite the evidence that serious harm outweighed benefit.8

There are no analyses of whether similar bias results from unpaid consultancies, nor other “non-financial” conflicts of interest. However, the potential is clear, and the increasingly widespread nature of these declarations suggests that such research is needed.

Disclosure does not mitigate risks of bias associated with industry sponsorship, but it does let readers consider the possible contribution of such biases when interpreting outcomes.9 This purpose is subverted by unclear or incomplete disclosure. Despite efforts by the International Committee of Medical Journal Editors to develop standard reporting criteria guiding authors’ conflicts of interest, including non-financial support, reporting is often incomplete and inconsistent with other information sources.

For example, an analysis of 404 articles with authors who had been described as having financial relationships with manufacturers in US whistleblower legal cases found adequate disclosure in only 15% of the articles.10 Databases of industry payments to health professionals under national Sunshine Acts, such as the US Open Payments database, enable comparisons of companies’ reports with clinician authors’ self reports of industry funding. Over half of authors’ disclosures were incomplete as compared with Open Payments reports in a sample of articles in three surgical and two medical specialties.11

There is growing interest in the impact of researchers’ non-financial interests. The label “unpaid” suggests these consultancies could fall into this category, although this classification is debatable. This confusion adds to an already unclear reporting category. Bero and Grundy note that no consistent definition for non-financial interest exists, and the inclusion of opinion, expertise, and even religious upbringing or political persuasion has created confusion between the role of human agency in scientific research and conflicting interests.12 The addition of the category of “unpaid consultancy” to this mix is unlikely to provide clarity.

Without a definition that allows readers to understand what it means, the reporting category of unpaid consultancy should be consigned to the dustbin. If it includes gifts such as food, drink, or expenses, or funding for a research centre, such payments should be reported as financial links to companies. The US National Sunshine Act provides clear guidance to pharmaceutical companies to report all payments above $10 in value given to healthcare professionals, including non-cash gifts. Given the advances in transparency required of industry, medical journals should introduce similar unambiguous reporting standards.

Footnotes

  • Competing interests: We have read and understood BMJ’s policy on declaration of interests and declare the following: BM was an expert witness on behalf of plaintiffs in a Canadian class action suit on cardiovascular risks of a testosterone gel. QG has no competing interests to declare.

  • Provenance and peer review: Commissioned; not externally peer reviewed.

References

View Abstract