Managing conflicts of interest: who, and how?

A very public slanging match has arisen between The Lancet and one of Australia’s most high-profile medical opinion leaders in psychiatry. The Lancet’s editor, Dr Richard Horton, has queried whether or not Professor Ian Hickie, of the University of Sydney, adequately disclosed his involvement with the pharma company Servier when writing a review for The Lancet on the role of new melatonin-based agents in treating major depression. Hickie has been accused of bias, selective referencing and inadequate declarations of conflicts of interest (COIs). These claims were made in The Lancet’s letters pages, sustained by Horton’s subsequent trickle of tweeting — claims that Hickie strenuously denies.

In this context, the publication in this issue of “Policies and practices on competing interests of academic staff in Australian universities” (doi: 10.5694/mja11.11224) is timely. Chapman and colleagues investigated the state of disclosure policies in Australian universities and the level of public access to disclosure information. Their findings are concerning. A third of universities declined to participate, for unclear reasons. Only eight universities maintained a central register of COIs for all staff, and two universities had policies stating that information regarding COIs would remain confidential. Without regulations in place at the university-wide level, medical faculty attempts to manage COIs will become toothless.

Transparency is vital for managing medical COIs. Institutions must provide the right environment and policies for their staff, and the institution itself, to encourage “clean” research and practice, leading to valid knowledge and useful treatments. The assessment of medical faculty and hospital campus regulatory efforts have been informed by these principles in the United States (Acad Med 2011; 86: 293-299) and, more recently, in Australia (Med J Aust 2011; 194: 121-125). However, these studies revealed significant policy variation and gaps, incomplete efforts, and failures to report on regulations governing COIs. Australia’s efforts lag considerably behind those of the US.

The issues are magnified for those who are leaders in the profession, for whom industry involvement is a two-edged sword. Dr Bernard Lo, of the University of California, San Francisco, summarises the issues (N Engl J Med 2010; 362: 669-671): relationships with pharma are trade-offs between new treatments and knowledge translation and the industry’s desire for new products, markets and profits — trade-offs that are more acute for key opinion leaders. However, any relationships must be at arm’s length. Nor should they be seen as a professional status symbol with significant personal perks. Without effective COI management, our leaders will turn into what would essentially be sophisticated sales staff, and their publications and any guidelines they present would become advertorials.

The effects of COIs extend beyond the individual involved. Our students and doctors-in-training develop values and knowledge from working with their mentors. They, and the community, need to feel sure that they and their teachers are working primarily in their patients’ interests rather than those of pharma. Should an academic be teaching medical students or mentoring junior doctors while on pharma’s books as a key opinion leader?

But still the question remains: after public declaration of COIs, what then? Academic institutions must investigate and intervene where there are concerns to ensure that research and teaching is not biased or coerced. This protects the academic and the institution. Who should do the investigating? Hopefully, when the University of Sydney looks at the Hickie issue that was raised in The Lancet — as they must — it will be done at arm’s length and the investigators will have at their disposal a policy that enforces accountability.

Medicine’s involvement with industry will continue to be a reality — for good or bad. Pharma clearly needs us. We should be using the power that that gives us to rethink our industry relationships on our terms and in the interests of the community. And, hopefully, pharma will also realise that an independent medical profession is in its interests too.