What should journals do when peer reviewers do not disclose potential conflicts?

Story image for Journals from Retraction Watch (blog)

Peer reviewers, like authors, are supposed to declare any potential conflicts of interest. But what happens when they don’t?

Take this case: In a court transcript from Feb. 23, 2017, Bryan Hardin testified that he was a peer reviewer on a 2016 paper in Critical Reviews in Toxicology, which found that asbestos does not increase the risk of cancer. In the deposition, Hardin—who works at the consulting firm Veritox—also said that he has testified in asbestos litigation on behalf of automakers, such as Ford, General Motors, and Chrysler, but said he had not disclosed these relationships to the journal.

Last year, the first author of the 2016 review withdrew a paper from another journal (by the same publisher) which concluded asbestos roofing products are safe, following several criticisms — including not disclosing the approving editor’s ties to the asbestos industry. In this latest case, the journal told us it believes the review process for the paper was up to snuff, but two outside experts we consulted said they believed Hardin’s relationships — and failure to disclose them — should give the journal pause.

We obtained a copy of the transcript from Christian Hartley, who was representing a man suing a mining company because the man developed cancer after being exposed to asbestos at work. When Hartley asked Hardin whether he had told the journal about testifying for companies involved in asbestos litigation, Hardin responded:

No. If — if that’s a new expectation, I’m not aware that as a peer reviewer you’re supposed to disclose that sort of thing, but I — I don’t recall that I did.

When we asked Hardin to confirm whether he had reviewed the 2016 paper and had disclosed his industry relationships to the journal, Hardin responded “I have been a peer reviewer on more than one asbestos-related paper” and “I have been retained by ‘several’ companies involved in asbestos litigation.”

Hardin also testified that he has never published on asbestos in the peer-reviewed or non-peer-reviewed literature.

A spokesperson for Taylor & Francis, which publishes Critical Reviews in Toxicology, told us:

Our in-house editorial team thoroughly examined the review process on this article earlier this year, to ensure it had followed Committee on Publication Ethics guidelines (the standards to which we expect all our journals to adhere to). The Editor-in-Chief made his assessment of the manuscript based on four individual review reports, at a ratio of 3:1 from academia and industry. These reviewers were all selected independently.

The spokesperson added:

On completing the investigation in this case, we were satisfied that the submission and review process had followed all the standards we would expect.

“An updated evaluation of reported no-observed adverse effect levels for chrysotile asbestos for lung cancer and mesothelioma” has been cited once, according to Clarivate Analytics’ Web of Science.

Should Hardin’s ties to the asbestos industry be considered a conflict of interest?

According to Elizabeth Wager, a publications consultant at Sideview and a member of the board of directors of The Center For Scientific Integrity, our parent non-profit organization:

I’d definitely regard giving expert testimony as a conflict of interest that should be declared by a potential reviewer (or an author). This is also in line with the views of many journals, such as PLOS which specifically notes that “acting as an expert witness” constitutes a non-financial competing interest. In other words, it doesn’t matter whether or not you were paid for such services, just testifying suggests a strong allegiance or interest which editors deserve to know about.

Trudo Lemmens, a professor of Law & Bioethics at the University of Toronto in Canada who was not involved in this case, told us:

In my opinion, a peer reviewer who has as an expert been working for a particular industry on a specific issue should at least disclose to the journal his ties to the industry whose interests can be affected by the publication of a paper on that topic.  And journal editors should exclude such peer reviewers from reviewing a paper on that or a related topic, or at least ensure that there are several other more independent reviews of the paper. If for one reason or another they think it is important to get a review by an expert with such a conflict of interest, they should assess that review much more critically. They should then also provide an opportunity to the authors to respond to the peer-review before making a publication decision.

Lemmens added:

But the more cautious approach would be to exclude reviewers with such a clear COI.

Wager noted that:

It’s up to the editor to decide whether a conflict of interest is so great as to disqualify the reviewer, but, whatever the editor decides, s/he obviously needs to be made aware of it by the invited reviewer…If a reviewer’s failure to disclose a relevant interest comes to light after publication the journal should look at the review comments again.

Hardin told us he believes that editors approach him as a peer reviewer “because they are aware of my prior employment with the government and my current consulting relationships.” (According to Hardin’s bio, he was deputy director of the National Institute for Occupational Safety and Health and an assistant surgeon general in the U.S. Public Health Service).

Other relationships

The authors of the 2016 paper all work for Cardno ChemRisk  — a firm that has testified for companies involved in asbestos litigation. In the paper’s “Declaration of interest” section, two of the authors—Jennifer S. Pierce and Brent L. Finley—disclosed that they “have served and may serve again as experts in asbestos litigation:”

Cardno ChemRisk has been engaged by numerous companies involved in asbestos litigation, and two of the authors (Drs Pierce and Finley) have served and may serve again as experts in asbestos litigation.

Pierce, the first author on the 2016 review, is also the last author of a 2016 study in the Journal of Occupational and Environmental Hygiene, which concluded that asbestos roofing products are safe. In September 2016, the journal told us it planned to retract the paper, after critics of the paper said it provided misleading information, contradicted key aspects of the original research that it references, and did not disclose the approving editor’s ties to the asbestos industry. The authors ultimately decided to withdraw the paper before the editors could formally retract it. (Although we consider retractions and withdrawals equivalent, the journal said it distinguishes between the two — a withdrawal is author-initiated, while a retraction is publisher-initiated.)

Finley, the last author on the 2016 paper, is also last author on a 2012 paper published in Critical Reviews in Toxicology, which again concluded that asbestos does not increase the risk of cancer. The paper received an erratum in 2013, which provides the authors’ previously undisclosed conflicts of interest:

…Six of the published papers referenced in this article were evaluated