Late last year the journal Contraception retracted an article that had been published in August 2016 about the relationship between abortion laws and maternal mortality in Mexico. The retracted article, by Blair Darney of Oregon Health & Science University and colleagues, aimed to refute the findings of a study of the same subject published two years earlier in BMJ Open.
Retraction of a published article is about the worst thing that can happen to a scientist, so reluctance of the part of the authors, editors and publishers can be anticipated. The error/s have to be serious enough to invalidate the central finding of the study. In the case of Darney et al, a key result was turned on its head and used to discredit the earlier study and its authors.
Is maternal mortality related to abortion laws?
That study, “Abortion legislation, maternal healthcare, fertility, female literacy, sanitation, violence against women and maternal deaths: a natural experiment in 32 Mexican states” reported research by Elard Koch, an epidemiologist of the MELISA Institute in Concepción, Chile, and an international team who set out to investigate the alleged link between restrictive abortion laws and the maternal mortality rate (MMR) in Mexico and elsewhere.
Using official maternal mortality data from the 32 federal states of Mexico between 2002 and 2011, they found that there was a decline in MMR in the 31 states with restrictive abortion laws, but not in Mexico City – the only state with a permissive abortion law.
However, after controlling for a number of other factors known to impact maternal health — skilled attendance at birth; all-abortion hospitalisation ratio; low birth weight rate; contraceptive use; total fertility rates (TFR); clean water; sanitation; female literacy rate and intimate-partner violence – they found that abortion laws had no independent effect, either negative or positive. In other words, where mothers enjoy good healthcare and positive social and cultural conditions, they are unlikely to die from abortion — or anything else.
Dr Darney was not happy with these findings. They contradicted her belief that restricting abortion drove women to resort to what is known in family planning circles as “unsafe abortion”. She applied to the Society of Family Planning for a grant to replicate the Koch study and received US$250,000 from the abortion-advocacy organisation. (More about that later.)
An anti ‘anti-abortion’ study
Darney’s research proposal made it clear that she was not approaching the exercise with an open, scientific mind. She wrote that SFP “have failed to respond to anti-abortion junk science, which influences policy in the region [Latin America].” She proposed to “critically respond to recent anti-abortion publications,” clearly meaning the Koch study.
And yet the aim of the latter was not to argue for or against abortion legislation, but to examine its association with maternal mortality rates in Mexican states, using the available official data.
As Koch has pointed out, Darney therefore based her study in part on a false premise that
“reveals a pre-established conclusion or disconfirmation bias based on unscientific prejudices in regard to our article. The aim of the SFP research grant to Dr Darney therefore was apparently to discredit our BMJ Open article, regardless of whether our methods and results were scientifically sound or not.”
But truth will out. Despite her intention, Darney et al’s replication of the Koch study actually confirmed his main findings. Darney’s statistics show a 22.49-unit decrease in MMR in the 31 states that restricted abortion compared with Mexico City, where there are no restrictions. This is clearly shown in a table in their published article.
Incredibly, however, they assumed and stated the exact opposite in their written report.
In their conclusion, for example, they say that “our multivariable regression (Table 2) suggests that, accounting for time trends and common state-level sociodemographic and health systems factors, Mexico City (the only state with access to abortion on demand) is associated with a 22.5-unit decrease in MMR compared with the 31 states with restricted access.”
Darney seems to have been so eager to expose “junk science” that she and her colleagues produced some themselves – or at least a “junk” commentary on it. They “actively used the misrepresented result throughout their article, confirming its pivotal role to discredit our BMJ Open study,” says Koch, who points to an ad hominem tone in references to his own team’s study throughout the Contraception article.
But what about the editor of the journal? And the peer reviewers? Did they not see anything amiss? This seems inexplicable unless one understands the relationships between the author, the journal and the funding organisation.
Internal relationships – conflict of interests
Contraception is the house journal of the Society of Family Planning, published by Elsevier. According to SFP’s website when visited by Koch in October 2016, the editor-in-chief, Carolyn Westhoff, is a prestigious member of the board of directors, a founding fellow and immediate past president of SFP (the latter information is no longer on the website).
This means that Dr Westhoff was simultaneously a director of the funding source for Darney’s paper and director of the editorial board for the Elsevier journal at the time of its submission and publication (though Westhoff has denied that the board has anything to do with such funding decisions).
She was also the one who handled a paper sent to Contraception by Koch et al with the results of their replication of Darney’s study, and their critique of – and complaints about – Darney’s representation of it. The reviewers for her study are not public and involvement by the board cannot be ruled out.
In addition, there are — or were at the time of publication – other close relationships between SFP, the publisher Elsevier, the journal, and Blair Darney. For example, her father, Dr Philip Darney, is the founder and a major donor of SFP and was a member of the editorial board of the journal at the time her paper was submitted and published. Darney herself replaced her father on the board in January 2018, a year after her paper was published.
All these internal relationships did not augur well for a retraction, let alone the apology Koch and his colleagues believe they are still owed.
The battle for a retraction
Koch’s team set to work to replicate Darney’s study and found that her key statistical finding (the 22.49 decline in MMR in states with some restrictions on abortion) was basically correct. It was clear that it was a not a typographical error. But there certainly was gross misinterpretation of the results and other methodological flaws and omissions. In May 2018 Koch et al submitted a paper to Contraception describing their replication and noting:
“Overall, we found evidence supporting a potential case of research misconduct. The ad hominem accusations of a lack of transparency and false conclusions they stated for our study were unsupported and untrue.”
On August 10 Westhoff emailed Koch that there was in fact “an error” in the Darney paper and that she would ask the author to submit an erratum. “This erratum will acknowledge that this error was identified by a careful reader.”
A careful reader! A mere acknowledgement of one (“small”) error! The authors of the BMJ article were not having that. What happened subsequently is reported in some detail in an article published on Retraction Watch last month. The following is a summary.
Given the fundamental nature of the error, the tone of Darney’s paper and the potential damage to the reputation of the researchers and their institutions, Koch and colleagues replied saying nothing less than a retraction would be adequate. This should be accompanied by “an editorial comment pointing out the large issues leading to retraction.” They noted the conflict of interests involved, and raised the prospect of an investigation by the Office of Research Integrity and perhaps by COPE (Committee on Publication Ethics). However, they made it clear that they would prefer an immediate retraction based on “the assumption that honest mistakes were made.”
A couple of weeks later they heard from someone at Elsevier saying that the author “feels this correction does not change the overall conclusion of her paper.” The publishers were seeking an independent review of Koch et al’s “allegations”. Koch and co-authors responded detailing their concerns. Then on October 31, Westhoff emailed Koch to say that the journal had decided that Darney’s paper required retraction. Some time after that a retraction notice appeared online. It read:
This article has been retracted at the request of the Editor-in-Chief and Authors.
The authors recently discovered an error that affected the results in their article on the relationship between state-level maternal mortality in Mexico and state-level abortion legislation. In Table 2 the beta-coefficient for abortion legislation was calculated as -22.49 and erroneously interpreted as +22.49. This error affects several of the paper’s conclusions, and thus the editor and authors have jointly made the decision to retract the paper.
The authors would like to express their sincere regret at the errors in their initial report.
Not surprisingly this wording did not satisfy Koch and his co-authors. They hired a Washington D.C. lawyer, Paul Thaler, who has represented researchers accused of misconduct, and Thaler wrote to Contraception suggesting a different and much fuller retraction notice, one that explained who actually “discovered” the “error”, which was in fact a serious misrepresentation, and ended with the statement: “The retraction of the article removes the basis the authors relied on for criticising the BMJ Open study.”
Elsevier’s lawyer responded on December 21 last rejecting that suggestion and saying among other things that the retraction notice already posted included a “clear apology from the authors”.
But Darney et al’s “sincere regret at the errors in their initial report” is clearly not the same thing as regret for trying to rubbish a perfectly good study and blackening the reputations of its authors while you were about it.
Meanwhile, in response to questions from Retraction Watch, Westhoff has made excuses for herself and tried to shrug off the episode by saying that nobody much read either of the articles anyway!
It is likely, however, that many scientists have read the Retraction Watch article. Dr Koch told MercatorNet that he and his colleagues are very happy that the full story has been brought to light by this specialised scientific news site.
The Darney-Contraception episode, he says, has been one of the most difficult of his career.
“The article has put at risk not only my reputation and that of my colleagues, but also proper recognition of our research. That it has ended satisfactorily for us is a relief. It allows MELISA to continue being a serious, reliable and rigorous research centre, as well as to put our energy into our current research focus, next-generation gene sequencing, which is relevant to maternal and fetal medicine.”
Carolyn Moynihan is deputy editor of MercatorNet.