Conflict of interest disclosure – the audience is not naive!

Parikh et al. have written on conflict of interest (COI) disclosure and interpretation in this issue of IJMS.^[1] We also read with interest a commentary on similar subject by Tannock IF in the February 2019 issue of Annals of Oncology.^[2]

Interest in disclosure of COI in the medical field is not new. Parikh et al. have shown that their Google search for the words “COI disclosure slide” had yielded 3,740,000 hits and a PubMed search (www.nlm.nih.gov) for “COI disclosure” yielded 5238 published articles.^[1] The year-wise distribution of these articles (from 2009 to 2019) is shown in Figure 1.

Close

Figure 1:

PubMed artilces “conflict of interest disclosure.”

Export to PPT

Several medical societies, including the Endocrine Society, already requires all speakers to show a disclosure slide in their presentation.^[2]

Ian Tannock’s paper also discusses COI slides during oral presentations. The methodology he has used is a limited part of what has been previously published.^[3] This article might be of interest because it has reported on a much larger data (video recording of 2128 presentation), better methodology (duration of display of the COI disclosure slide measured by digital stopwatch by two individuals), and analyses of additional information regarding all plausible relevant factors as well as the specified statistical methodology used.

Another, even earlier publication, compared the information regarding COI disclosure in PubMed published articles with that available in open payment databases (OPD). They reported the discordance between OPD- and self-disclosed COI to be as high as 65.0% by the article (P < 0.001).

The main contention of both these articles is that the display time of the COI disclosure slide is too short – 3 and 2.49 s, respectively.

We humbly suggest all of us give thought to the following points:

1. The audience is not naive: Tannock states, “disclosure of potential conflicts does not prevent or lessen bias.”^[1] We agree totally with this statement. It is also almost impossible to authenticate the validity and/or completeness of the disclosure information showed by speakers. That is also the reason we feel that any emphasis on showing COI disclosure slide for a longer time is meaningless. Of note is the fact that the audience at medical conferences is not stupid. It has been previously demonstrated that 73% of the audience is still alert for bias even when the speaker shows COI disclosure slide.^[2]

2. Why prevent participants from having the opportunity to grill the data and the presenter? Tannock suggests that “disclosures relevant to the trial being reported should indicate the amount of personal payments received in the prior 2 years, to the nearest 5000 Euros (or US dollars).” Most trail data are presented by one of the principal investigators. We ask, in which trial will the investigators’ fees be less than this amount? We are of the opinion that presenting data at conferences is not “purchasing silence.” It is, in fact, the exact opposite. This allows the audience to publicly question the data and grill the presenter.

3. Should we have a meaningless conference by shutting out the best work? Another solution touted is to “avoid having speakers at their meetings with financial COIs.”^[4] If this is implemented, it will exclude most (more than 90%) of existing speakers. This is because the brightest of minds, the most diligent researchers and the investigators with maximum patient load (remember investigators are selected from those who have the largest number of patients in their routine practice) will be automatically excluded from the study. In this case, why would the audience attend a meaningless or mediocre conference?

4. Value the deserving: It is a universal truth that anything provided free is not valued. Why should a key opinion leader sacrifice his personal time for any activity in which his contribution is not valued? Prohibiting receipt of personal payments for scientific insights and diligent effort would be a regressive step. Today the greatest innovation and advancement of science comes from a partnership between academia and businesses, where individuals are given commensurate financial rewards.^[5] In fact, teaching institutions and even government organizations are encouraging such an arrangement. As an example, the scale of private research and development (R&D) conducted at academic university facilities in Albany is double that of Boston, in relative as well as absolute terms.^[6] The Canadian Natural Sciences and Engineering Research Council, government’s largest source of support for R&D partnerships also wants greater collaboration and is striving to connect individuals having skills to businesses that require them. In fact, Massachusetts Institute of Technology awarded $2 million to 106 winners so that they could develop collaborations with outside businesses.^[7]

5. Does industry funding always influence credibility? We suggest that targeting honest and hard-working medical scientists and trial investigators with such “holier than thou” attitude needs to be condemned. The Dutch psychiatrists did a prospective study whose primary objective was to document credibility and clinical relevance of study results using a 10-point Likert scale.^[8] The presence of funding from industry did not influence credibility as perceived by the 395 participating psychiatrists (Mean Difference [MD] 0.12; 95% confidence interval [CI] –0.28–0.47). It also did not influence the clinical relevance scores (MD – 0.14; 95% CI –0.54–0.27). Policy decisions need to be balanced, reasonable and should be designed to actually meet stated objective without obstructing collaborative research.^[9]

6. Focus on the bigger picture: Do we need to focus on the measly 5000 dollars of payment, or should we be concentrating on the big fish? Why are we silent on the doctors who failed to disclose millions of dollars in payments from drug and health care companies in recent years, omitting disclosure of their financial ties from multitude of research manuscripts published in prestigious journals like The Lancet and New England Journal of Medicine?^[10]