Article Text

Download PDFPDF

Conflict of interest disclosure and the polarisation of scientific communities
  1. Thomas Ploug1,
  2. Søren Holm2,3,4
  1. 1Centre for Applied Ethics and Philosophy of Science, Aalborg University Copenhagen, København S, Denmark
  2. 2University of Manchester, Manchester, UK
  3. 3Center for Medical Ethics, Faculty of Medicine, University of Oslo, Norway
  4. 4Department of Health Science and Technology, Aalborg University, Denmark
  1. Correspondence to Professor Thomas Ploug, Centre for Applied Ethics and Philosophy of Science, Aalborg University Copenhagen, AC Meyers Vænge 15, København S 2450, Denmark; ploug{at}hum.aau.dk

Abstract

In this article, we introduce and define the notion of polarised scientific communities. We show how polarisation may generate genuine conflicts of interest that require specific efforts to be identified and reported, and suggest a simple heuristic for the identification and reporting of the polarisation of research. We use examples from the debates about breast cancer screening and the prescription of statins to people at low risk of heart disease.

  • Ethics
  • Publication Ethics

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

In recent years, conflicts of interest have received significant scholarly attention and the disclosure of potential conflicts of interest has become a widespread requirement in relation to the publication of research. The purpose is to ensure the disclosure of interests that may unduly influence research so that users of research can take these conflicts into account when weighing up the evidence. Recently, these have been recognised to include financial and non-financial interests and the International Committee of Medical Journal Editors (ICMJE) has in its most recent guidelines greatly expanded the range of reportable conflicts.1–4 There is, however, a common type of conflict of interest which is not covered by these guidelines. This type of conflict of interest is generated by the polarisation of scientific communities and we will suggest ways in which it can be identified and reported.

Polarisation as a conflict of interest

What is polarisation? In the present context, polarisation of a scientific community can be defined as a situation in which reputable scientists hold radically opposed views leading to the segregation of the scientific community into groups in part constituted by their opposition to other groups in the field. Polarisation goes beyond mere disagreement. It occurs when researchers begin (1) to self-identify as proponents of a particular position that needs to be strongly defended beyond what is supported by the data and (2) to discount arguments and data that would normally be taken as important in a scientific debate. Two types of polarisation can be distinguished: (1) Polarisation of the scientific community in relation to scientific findings and methods and (2) Polarisation of the scientific community in relation to policy advice provided on the basis of the scientific findings. Both types of polarisation can be seen in debates about breast cancer screening programmes. On the one hand, the community is polarised as a result of the controversy over the exact reduction of the mortality rate among those screened, the extent of overdiagnosis and the number of false positives.5 This, at least in part, reflects underlying methodological disagreements over how to study the effects of cancer screening programmes, for example, the choice between cohort studies and randomised trials, and between different kinds of randomised trials.6 ,7 On the other hand, the community is polarised as a result of differences in the advised policy on breast cancer screening partly independent of the scientific methods and findings.6 ,7 Thus, even if disagreements over scientific methods and findings are set aside, there still is controversy over the justifiability of running screening programmes, and not least over how to run screening programmes, for example, opt-out versus opt-in enrolment and so on.8 ,9 Different groups of researchers consistently publish research supporting one or the other of these positions. Polarisation is not uncommon as is perhaps best illustrated by the perennial popularity of the confrontational ‘X should be introduced Yes/No’ debates in the general medical journals.

The polarisation of a scientific community may generate genuine conflicts of interest. That is, conflicts that threaten the objectivity of science, and may in turn bias public debate and political decision-making.10–12 More specifically, the threat is that a polarised group may nourish an interest in advancing the position and views of the group, and that this interest may come to be a main criterion and goal for the choice of methods, the reporting of findings and the provision of policy advice. This interest may well be based on an honest conviction that one is right and thus not be in any way morally reprehensible, but even honestly held convictions can introduce potential biases in research and reporting. One could perhaps argue that the threat to the objectivity is weaker and more tractable than in conflicts of interest based upon financial interest. After all, there seems to be a greater degree of transparency in the case of polarisation of a scientific community; many researchers will be able to point out the fields in which scientific controversy have resulted in polarisation, and can identify the polarised groups and the individual researchers within these groups. Hence, it is likely that the effects of polarisation are attenuated to the same extent and by the same mechanisms as conflicts of interest based on financial interests are diminished by the disclosure of these interests in the publication process. But this is not always the case. First, it is not clear that researchers outside a polarised field will always be able to identify the polarisation, or the polarised groups and individual researchers within them. This will often require a certain familiarity with the relevant field. Second, researchers within a polarised group in a polarised field may not themselves be able to identify the field as polarised or see themselves as belonging to a polarised group, and thereby potentially biased. Rather, it seems likely that many researchers within a polarised field would find it difficult to identify themselves as being polarised; they are after all ‘only after the truth’ and the only people seen as polarised may well be ‘the others’. Third, and most important, it is unlikely that the public, media and political decision-makers will always be able to see that the research at hand is produced by research groups in a polarised field. This entails that there is a conflict of interest that should be taken into account in the process of disseminating scientific information to the public or in the process of collecting evidence for decision-making in the health field, for example, in preparing healthcare technology assessments.

How to discover and report polarisation

The current guidelines on reporting potential conflicts of interest do not include any specific requirement of reporting conflicts of interest that may arise due to membership of polarised scientific communities. The conflict of interest form and the guidelines from the ICMJE recognise that non-financial interests may influence scientific work by requiring authors to reveal non-financial relationships and activities, conditions and circumstances that may be perceived to have influenced the submitted work. But none of the examples of non-financial conflicts of interest mentioned in the guidelines are close to polarisation. The guidelines mention ‘intellectual passion’ as a source of potential bias,2 but polarisation is more than just individual intellectual passion; polarisation is, at least partly, a group phenomenon. We believe that it is both possible and desirable to aid researchers in identifying and reporting polarisation as a type of conflict of interest and that it will not be difficult to expand current guidelines to achieve this.

A researcher may detect or identify that he or she belongs to a polarised group through reflection on some simple questions. As suggested above, polarisation may be easier to identify from outside the polarised field than from within. It seems especially difficult to identify oneself as belonging to a polarised group, although it is of course not impossible. Many researchers will be aware of their own positioning, but some will not. However, given the nature of polarisation, it will show itself in the evaluation of the research of others. Belonging to a polarised group thus seems to be intimately interwoven with the view that other researchers are one-sided in their research, that is, that they persistently favour one particular conclusion in their scientific work. A researcher may therefore identify that he or she is part of a polarised field and group by reflecting on the following two questions:

  1. If the results of your current (well planned and well conducted) project point in the opposite direction of the results of your previous research on this topic, would your first reaction be to reanalyse the data and reconsider your methods, or to reconsider your previous conclusions?

  2. If your findings were the exact same as the opposing researchers in this field of research, would your policy recommendations be any different from the recommendations of the opposing group?

The first question aims at identifying type 1 polarisation. The second question aims at identifying type 2 polarisation. Although these questions are not in themselves sufficient to decisively identify polarisation, a ‘Yes’ to both of them strongly indicates its presence.

The reporting of the polarisation of a field could easily be implemented into existing forms for disclosure of conflicts of interest. The specific content of the reporting depend on the kind of research, but let us just give one example related to intervention-research. If polarisation occurs, a polarisation statement could look like this:

  1. This article reports research in a polarised field.

  2. The research group I/we belong to generally believe that the intervention we have researched should/should not be introduced in healthcare.

The first part of the statement identifies the field as polarised. The second part indicates that the researcher (or group of researchers) generally assumes a certain position in the controversy central to the field. However, if a researcher (or group of researchers) upon reflection genuinely believes that he or she is researching in polarised field but remain neutral, then he or she should be given the opportunity to report this.

Journal editors have important roles to play here. They are in a position to require the disclosure of polarisation as a potential conflict of interest, and should do so. This will enable them to take account of polarisation in their editorial processes in the many cases where they are not themselves able to identify polarisation because they are in a similar position to the average user of scientific evidence. If editors are aware that a field is polarised, they also need to take account of this in their editorial processes and decision-making, for example, when choosing referees or when scrutinising the quality of the science or the policy recommendations. The recent controversy concerning the publication in the BMJ of two articles on the prescription of statins to people at low risk of heart disease exemplifies what can go wrong when an editor ignores known polarisation in a scientific/policy area. In her editorial summarising of the view of BMJ on the controversy, the editor in chief explicitly says about one of the groups involved that ‘… they will be considered by their critics to be too partisan …’, and about another group that ‘… they too may be considered by their critics to be insufficiently independent …’.13 The polarisation of the statin field was well known in the general medical community prior to the publication of the two articles.

We have in this brief analysis shown that polarised scientific communities generate genuine conflicts of interest that require specific efforts to be identified and reported in the conflicts of interest form. Our analysis has focused on publication but is also relevant to other areas like grant review where biases generated by polarisation have to be taken into account. Many fields of science are—and should be—characterised by disagreement and debate. Our concern here is when the zealous quest for truth and knowledge moves beyond productive disagreement and scientific communities stiffen in polarised groups threatening the objectivity of science and the openness of public debate and decision-making.

References

Footnotes

  • Contributor Both authors contributed equally to the design of the study, analysis and writing of the draft paper. Both authors are guarantors.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Response
    Brian D Earp

Other content recommended for you