When we broke the story last week about a juicy retraction notice in the American Journal of Obstetrics & Gynecology (AJOG) — known by its readers as “the Gray Journal” — we wrote that there was more to it than we suspected. That’s an understatement.
As we reported, the AJOG retracted an article that it had published earlier this year because 1) the author, Laurence Cole, had failed to disclose a potential financial conflict of interest with a pregnancy test maker named Church & Dwight; and 2) the article lacked a “credible scientific reason given for conducting the study,” along with other flaws detailed in the notice. (As we wrote the other day, we wonder why those issues did not arise during the initial review of the manuscript — but more on that shortly.)
We’ve since learned that the journal’s move came after it received a sharply worded letter from a high-powered San Francisco lawyer demanding immediate retraction of the article on the grounds that it represented a “substantial” threat to the financial health of his client. That client? A maker of home pregnancy tests who is now in the process of suing the very firm that provided Cole with research funding he failed to disclose.
First, here’s what Cole, the hormone expert at the University of New Mexico whose paper the journal retracted, said about why he didn’t disclose that funding: He says that Church & Dwight — the New Jersey company that makes pregnancy tests as well as Arm & Hammer baking soda products and, maybe as a hedge against the pregnancy test business, Trojan condoms — had given him a “fairly substantial” grant several years ago to collect urine samples of pregnant women. The money funded a physician to run the project and recruit volunteers to participate. Although Cole wouldn’t disclose the exact amount involved, it obviously wasn’t trivial.
However, he said, the funding had nothing to do with the paper in question. Indeed, he said he approached the company about the manuscript when he was writing it and asked them if he should include them in the disclosure statement — but they demurred.
Cole said he accepted the decision because he didn’t see a direct conflict with the paper and the funding.
I don’t deny I have a relationship with this home pregnancy test manufacturer, but I don’t know what it has to do with this paper. I related to the journal that I was told by [Church & Dwight] not to cite them and I didn’t.
Of course, the journal editors and editorial board didn’t agree with Cole’s decision, as we reported. Now here’s one thing we found odd: The journal’s handling of the case diverged sharply from the way it has dealt with similar situations. Consider this May 2011 letter from the editors in the AJOG, again for failure to disclose a conflict of interest:
In response to questions from readers concerning a conflict of interest disclosure in a 2009 article,1 a letter to the editor,2 and the authors’ replies to letters to the editor,3, 4 the editorial board of the American Journal of Obstetrics and Gynecology wishes to inform readers that the following statement of conflict of interest was disclosed in the original submission of the article.
“Dr Witter did Phase III research on Intramuscular Ritodrine sponsored by Astra Pharmaceutical products in 1983 and also did phase II research of oral sustained release ritodrine sponsored by Kali-Duphar, Inc. in 1984-1985. Dr Zimmerman has reviewed legal cases of children with autism who were exposed to terbutaline and has been deposed on one case. Mr Reichmann was employed by Matria Healthcare (now Alere, Division of Iverness [sic] Medical Innovations, Inc) through February 2005. Dr Connors has reviewed one legal case involving terbutaline exposure and autism.”
And, in the November 2011 issue of the journal — the same one as the Cole retraction — appears this notice:
In response to a reader’s query about possible conflicts of interest relative to a Letter to the Editor published in the May issue of the Journal (O’Brien JM. 17-hydroxy progesterone caproate, progesterone, preterm birth prevention, and safety; who decides? Someone should. Am J Obstet Gynecol 2011;204:5.e16-7) Dr John O’Brien wishes to disclose the following:
(1)He was involved in studies of progesterone gel treatment for preterm birth prevention, financially supported by Columbia Laboratories, a maker of progesterone gel.
(2)He serves on Advisory Boards for Watson Pharmaceuticals, a company with a financial interest in marketing vaginal progesterone gel for preterm birth prophylaxis.
(3)He acknowledges that he and others are listed in the patent on the use of all progesterone compounds to prevent preterm birth (US Patent Number 7,884,093: Progesterone for the Treatment and Prevention of Spontaneous Preterm Birth).
We’re wondering why a simple post facto declaration was appropriate in these cases but not in Cole’s — particularly the second paper, which seems on its face to be a much stronger conflict.
The journal and publisher insisted that these two cases differ substantially from Cole’s, that the methodological problems in the latter aren’t present in the former. Again, that may be true, but it doesn’t square with what Moon Kim, one of the two top editors of the journal, told us about the flaws in Cole’s article:
When you look at more carefully, that’s not a reason why the retraction was made. The retraction was made because of this failure to disclose a major conflict of interest.
Again, we’re scratching our heads. The other two papers we note above didn’t require retractions. So what was the difference here?
Well, there was the letter from attorney Jeffrey Knowles, a specialist in false advertising and other corporate concerns at Coblentz, Patch, Duffy & Bass, to which we referred earlier. (He was co-lead attorney on the Napster case, representing the record industry.) We couldn’t get a copy of the entire letter, but here’s the money paragraph:
Publication of this flawed paper holds the potential to inflict substantial harm on SPD. We therefore demand that AJOG refrain from publishing the print version of the paper, that it cease publication in electronic form and that it publish a retraction. In the event that time constraints limit your ability to take any of these steps, we demand that AJOG accompany the publication of the paper with this letter. Any subsequent effort to remedy the reputational damage from this improvident publication is unlikely to be effective.
SPD is short for Swiss Precision Diagnostics, a joint venture of Procter & Gamble and Alere — formerly Inverness. Cole’s paper, the letter states, “appears calculated to contradict” research that would be favorable to the company’s position in the market for home pregnancy tests.
Cole said the friction between him and Alere stems from work he did in the mid-aughts showing that the only way to accurately determine early pregnancy through a urine test — we’re talking the first three weeks — is to measure a molecule called hyperglycosylated human chorionic gonadotropin (hCG). He’s has published numerous papers trashing various home pregnancy tests for being unable to perform as advertised. Church & Dwight made a home test that could detect hCG; Inverness, the market leader at the time, did not.
As a result of his finding, Inverness’ share of the home testing market fell, Cole said. But the company eventually began producing tests that could detect hyperglycosylated hCG, and Alere is now suing Church & Dwight to stop the latter manufacturer from claiming that its tests are the only ones that can identify the substance. (Inverness also launched a suit in 2010 against Church & Dwight for patent infringement involving First Response.)
Cole said he learned that Alere lodged the complaint with the journal when he was deposed a few months ago in that lawsuit. Jonathan Gothorpe, an in-house attorney for Alere, confirmed that Swiss Precision Diagnostics is suing Church & Dwight for false claims, and that Cole’s “name certainly has come up.”
Alere, we should also note, advertises in AJOG and is a major sponsor of the annual meeting of the Society for Maternal-Fetal Medicine, which the journal represents.
All of that seems like important context. But AJOG co-editor in chief Kim said the letter did not influence the journal’s handling of the matter. Perhaps not. But we’ll note that this is the second time in recent memory that Elsevier, which owns AJOG, has backed down in the face of legal threats.
Kim said that had Cole initially disclosed the relationship with Church & Dwight, his article probably would not have been retracted. But, he added, the paper
started out as a marginally acceptable [publication]
and that in combination with the methodological issues unearthed by the second review, the article could not stand.
Yet Cole rejects the notion that the paper was without scientific merit. He said he received a thorough pre-publication review by two readers, who had many comments, but not those that arose in the post-publication review.
They accepted the article. Why did they accept the article if they didn’t see a single bit of significance? There’s major significance here in trying to explain why hCG levels vary so dramatically…
Cole — whom, we should note, was forced to withdraw a 2009 paper in Fertility & Sterility for what he called a computer error during the submission process, but said he has not had any other articles retracted — also disputed the more subtle points of the retraction notice, including the use of certain instruments and the calibration of a urine test. He provided us an email exchange with the editors in which he addressed those points, as well as a rebuttal email he sent to the editors upon reading the published retraction notice.
We don’t know the science here, so we asked an hCG expert to review the article
— which has been completely removed from the journal’s site — and the retraction notice and give us his thoughts on the merits of each. We’ll update with that feedback.
But in a sense it doesn’t really matter. One thing is perfectly clear: Despite its efforts to discredit Cole’s paper from top to bottom — and despite the useful appearance of an undisclosed conflict of interest to assist in the deconstruction — the journal at one point thought the article was worth publishing. Two reviewers and an editorial staff felt it passed muster.
The editors and the publisher now insist the paper was “borderline” from the start. So how many of the approximately 25% of submitted manuscripts (from “domestic” authors) the editors accept are that weak?
A second question we had involved Cole’s status at the University of New Mexico. We’d assumed that he had gone on leave as a result of the retraction — or at least, because of circumstances related to it — but that’s not so. According to Cole, he has been on disability leave for several months after his department head requested that he seek a neuropsychiatric evaluation for memory lapses. Cole said he submitted for the testing locally, and that the results showed signs of what he called “major” memory problems.
However, Cole said he did not believe the results and decided to get a second opinion from clinicians at Baylor in Houston. He did so, he said, and was declared in sound mind.
All this isn’t irrelevant to the retraction. Cole initially blamed his failure to acknowledge the conflict with Church & Dwight on his bad memory. He said he eventually remembered receiving a request from the company not to disclose the relationship, but by then it was too late.
Cole’s chairman William Rayburn said it would be “totally inappropriate” for him to comment on the matter and referred us to a university lawyer who has not responded to a request for an interview.
Meanwhile, however, Cole has been in limbo — a situation he is attempting to restore with the help of a lawyer, who has said that either the university restore Cole to his position or pay him a $500,000 severance. Whatever the outcome, he said he plans to resume stewardship of the hCG laboratory.
Cole’s lawyer, Donald Gilpin, declined to discuss the details of the case. But he did confirm the basics of Cole’s claims. Gilpin added that the university has other beefs against the researcher but would not specify them.
Oh, and about that vague bit in the retraction notice alluding to a punishment levied against Cole? He has received a two-year ban from publishing in the AJOG.
Update, 6:30 p.m. Eastern, 11/10/11: Commenter Neuroskeptic points out that a PDF of the paper is in fact available on the site. It didn’t seem to be when we checked earlier, but we’re glad to correct this post and the previous one with strikethroughs.