Managing or maintaining bias? Examining the conceptualisation of conflicts of interest in medical journal publishing
Item statusRestricted Access
Hendrick, Rachel A.
BACKGROUND: It has been claimed that the involvement of commercial companies in medical and health research poses risks relating to potential conflicts of interest. In response, many journals have developed conflict of interest policies, and there has been a proliferation of related guidance from publishers, professional associations and commercial companies, mostly centred on processes of voluntary disclosure. Studies and commentaries on these have raised concerns regarding the adequacy of such practices, but there has been limited analysis of the underlying context – how and why policies have been constructed in this way – or exploration of alternative approaches. AIM: This thesis examines how actors within medical journal publishing conceptualise conflicts of interest. It analyses their understandings of conflicts of interest: which types of interest are deemed most significant; which actor groups are seen as conflicted; and how conflicts are managed. Through doing so, it explores the barriers to, and possibilities of, change. METHODS: The study draws on two distinct sets of data. The first is a sample of conflict of interest policies and guidance. The second is 48 semi-structured interviews with actors working in a range of roles related to medical journal publishing. These data were thematically analysed to illustrate how medical journal publishing conceptualises and manages conflicts of interest, to identify perceived strengths and weaknesses of current approaches, and to identify potential opportunities for improvement. RESULTS: There appears to be an established discourse around conflicts of interest, which emphasises particular stakeholders, while others, who also have opportunities to influence journal content, are frequently absent from the debate. Financial interests are readily highlighted, while non-financial ones receive less attention and are thus often unregulated (Chapter 5). High levels of consistency characterise the ways in which actors discussed the management of conflicts of interest: for example, self-disclosure was regularly highlighted, despite the acknowledged weaknesses of this approach (Chapter 6). The existence of further mechanisms that offer the potential to assist in managing conflicts of interest were identified, though findings suggest that, in practice, these currently have limited uptake (Chapter 7). Interviewees’ suggestions of how conflicts of interest might be better managed (e.g. through greater data transparency) are also analysed. Overall, narrow interpretations of conflicts of interest and their management appear to have become institutionalised in ways that serve to limit the uptake of alternative approaches. DISCUSSION: Given the substantive importance that medical research can have on health policies and treatments, robust processes are required to protect the integrity and legitimacy of journals. This research shows that existing, institutionalised understandings of conflicts of interest have critical limitations, which leaves medical publishing open to potentially unethical practices that may be a source of bias in published evidence. This poses a significant threat to the desire to attain ethically robust, peer-reviewed medical/health research that can be used to inform policy and practice. Drawing on the interview data, the thesis explores some possible alternatives that may warrant further consideration.