Short report
Prohibiting or ‘managing’ conflict of interest? A review of policies and procedures in three European drug regulation agencies

https://doi.org/10.1016/j.socscimed.2009.09.002Get rights and content

Abstract

In light of debates about the relationship between interests and scientific expert judgments, and the potential for declarations of conflict of interest (COI) to minimize corporate bias, we reviewed the approach to COI in 3 European drug regulatory bodies. These bodies were the Irish Medicines Board, the Medicines and Healthcare products Regulatory Agency in the United Kingdom and the European Medicines Agency in the European Union. Official statements about COI laws and codes of practice in the 3 contexts suggest that COIs are prohibited. In practice, the approaches to COI in the 3 drug regulatory agencies presuppose and promote the ideas that COIs cannot and need not be eliminated as the risk of bias can be managed. Because the evidence about if and how COI affects micro-level decision-making in drug regulatory authorities is neither complete nor comprehensive, we advocate a precautionary principle model. Under this model COI would be prohibited on the grounds that it might influence the outcome of regulatory decisions.

Introduction

There have been a few classic studies of pharmaceutical regulation across Europe, but a systematic review of conflict of interest (COI) policies has been neglected, and is long overdue (Abraham and Lewis, 2000, Dukes, 1985, Mossialos et al., 2004). Disclosure of interests in pharmaceutical companies is now a normative requirement in many countries for scientific experts' participation in drug regulation decision-making. This has occurred despite protestations from scientists that the requirement to disclose COI ‘cuts at the epistemological foundation of scientific reasoning’ (Borgert, 2007, p. 4). Scientists have frequently voiced their belief that their interactions with industry do not influence their conclusions, ignoring strong psychological research evidence to the contrary (Kassirer, 2008). Another cut at the precepts of ‘pure science’ is the significant body of empirical sociological evidence of ‘corporate bias’ in the science of drug testing, bias that has resulted in the pharmaceutical industry being decisive in shaping regulatory policy (Abraham, 2007, Abraham, 2008). Furthermore, recent systematic reviews have shown strong evidence of ‘sponsorship bias’ in the production of knowledge about pharmaceuticals, evidence that pharmaceutical industry funding of clinical trials significantly enhances the chances of pro-industry results (Bero et al., 2007, Lexchin et al., 2003, Sismondo, 2008).

Researchers who have sought to trace and theorise bias in pharmaceutical regulatory science emphasise, however, that in micro-level decision-making about specific drugs, the relationship between interests and scientific judgments is not straightforward. Others too have highlighted that the evidence of the implications of COI for specific decisions that come from regulatory agencies is thin and equivocal (Lurie, Almeida, Stine, Stine, & Wolfe, 2006). John Abraham (2008, p. 877) argues that an ongoing challenge for sociology is to determine how interests may bias scientific knowledge-claims in drug regulation. Rather than determining particular decisions, Sergio Sismondo (2008, p. 1911) argues that COI is better understood as entailing complex and subtle influences. Close interaction between scientists and industry involving economic and other relationships, he argues, can cultivate a pro-industry disposition and alter scientists' habits of thought toward individual companies and their products. Moreover, Abraham and Sheppard (1999) show that, although European regulators often claim that they can detect the COI bias in drug development that occurs when regulators allow pharmaceutical companies to present their drugs in the most favourable light possible (a so-called ‘honest bias’) in practice they may fail to do so.

Regulatory agencies' COI policies can be read as official recognition of the risk of bias, or at least of public perceptions of such risk, and variations in these policies as differing approaches to ‘taming conflict of interests’ (Epstein, 2006). As will be seen below, in some contexts taming the risk of bias in pharmaceutical regulation involves techniques of ‘risk assessment’. Although frequently advocated as key to this project, COI disclosure has been criticised for offering a significantly weaker means of minimizing the risk of bias than transparency of the scientific process, and, even when rigorously implemented, for being too piecemeal a response and an empty ritual (Borgert, 2007, De Vries and Bosk, 2004, Sismondo, 2008). Sheldon Krimsky's (2003) analysis of the increasing entanglement of public science with private interests in US universities reveals a new ethos of academic commercialism in which COI is framed as inevitable, but manageable. His work demonstrates how, within this ethos, COI has been redefined without pejorative connotations and become the behavioural norm. While applicable to the situation in the US, Krimsky's analysis cannot be generalized to Europe where this type of ethos may have deeper roots. However, recent years have witnessed an intensification of regulatory commercialism as many drug regulatory agencies now compete with each other for business and are subject to the ‘tests of the market’ (Abraham, 2008, p. 875). This is part of a broader transformation of the state in the era of capitalist globalization and the emergence of ‘competition states’ which prioritise meeting the demands of global economic competitiveness (Kirby, 2005). With these trends and debates in mind, in this research we investigated the normative character of the COI policies and practices of 3 drug regulatory agencies in Europe, the understandings of COI that they presuppose and promote including ideas about what should and should not be allowed.

Section snippets

Methods

We reviewed the approach to COI in the Irish Medicines Board (IMB), the Medicines and Healthcare products Regulatory Agency (MHRA) in the United Kingdom and the European Medicines Agency (EMEA) in the EU. We chose these 3 agencies to examine differences and similarities between a relatively small national agency (IMB), a large national agency (MHRA) and a transnational agency (EMEA). Specifically we looked at 3 issues. What are the policies and procedures within each agency for dealing with

Disclosure of COI policies

The IMB, MHRA and EMEA operate within diverse regulatory frameworks, but shared features are the requirement that COI have to be declared, and an official rationale that this undertaking provides public reassurance that the agencies' decisions are guided by public health rather than commercial interests. In keeping with the 1995 Ethics in Public Office Act, the 1995 Irish Medicines Board Act requires the disclosure of ‘pecuniary and other beneficial’ interests by IMB board members, employees

Discussion

The general thrust of official pronouncements about COI in the contexts considered here suggest that the 3 agencies' view of COI is that it is prohibited and has to be ‘resolved’ in favour of the public interest. In practice, however, notwithstanding significant variations in policy and procedure, the approach to COI in the IMB, MHRA and EMEA is one of management rather than prohibition. In all 3 agencies close interaction between experts and the pharmaceutical industry is taken-for-granted and

References (27)

  • C. Borgert

    Conflict of interest or contravention of science?

    Regulatory Toxicology and Pharmacology

    (2007)
  • S. Sismondo

    How pharmaceutical industry funding affects trial outcomes: causal structures and responses

    Social Science & Medicine

    (2008)
  • J. Abraham

    Science, politics and the pharmaceutical industry

    (1995)
  • J. Abraham

    From evidence to theory – neo-liberal corporate bias as a framework for understanding UK pharmaceuticals regulation

    Social Theory and Health

    (2007)
  • J. Abraham

    Sociology of pharmaceutical development and regulation: a realist empirical research programme

    Sociology of Health and Illness

    (2008)
  • J. Abraham et al.

    Regulating medicines in Europe: Competition, expertise and public health

    (2000)
  • J. Abraham et al.

    Trading risks for market: the international harmonisation of pharmaceuticals regulation

    Health, Risk & Society

    (2001)
  • J. Abraham et al.

    The therapeutic nightmare: The battle over the world's most controversial sleeping pill

    (1999)
  • L. Bero et al.

    Factors association with findings of published trials of drug-drug comparisons: why some statins appear more efficacious than others

    PLoS Medicine

    (2007)
  • R. De Vries et al.

    The bioethics of business: rethinking the relationship between bioethics consultants and corporate clients

    Hastings Centre Report

    (2004)
  • M.N.G. Dukes

    The effects of drug regulation: A survey based on the European studies of drug regulation

    (1985)
  • R. Epstein

    Overdose. How excessive government regulation stifles pharmaceutical innovation

    (2006)
  • European Medicines Agency

    EMEA procedure on the handling of conflicts of interests for EMEA scientific committees members and EMEA experts

  • Cited by (47)

    View all citing articles on Scopus

    We would like to thank the anonymous reviewers of this paper for their helpful suggestions.

    View full text