A miscarriage is a heartbreakingly-common tragedy: perhaps one new mother in four will lose her unborn child.
Doctors would dearly like to prevent them. But how? One current theory is the ovaries of women who miscarry – especially those who miscarry multiple times – may not make enough of the hormone progesterone early in pregnancy. The hormone is important for preparing the womb for a baby.
That led doctors to propose a treatment: the blood thinner heparin, which, when injected, “thins” the blood, decreasing the tendency to form clots. At first, the results seemed exciting. A 2008 study by an Egyptian team, led by Dr. Ahmed Badawy, reported that, in a study of 340 women treated with heparin, the drug slashed miscarriage rates in half.
But recently, concerns were raised about the integrity of the trial. Several errors were noted in the way the data was analysed. The journal investigated – but the authors refused to provide their original data. In 2022, 3 years after the problems were flagged, the journal retracted it. The paper has been cited 184 times.
Then in 2023 a large, randomised controlled study has proven that heparin in most patients does nothing to prevent miscarriage.
“All these years we have told these women, ‘you should give yourself injections with heparin daily during pregnancy’, and it doesn’t work,” professor of obstetrics and gynaecology at Monash University, Ben Mol, tells Cosmos.
“There are two big problems here: one, there are areas in the world where researchers make up almost everything; and two, when this is pointed out, everyone looks away. Dr Ahmed Badawy has now had around 25 papers retracted.”
A bigger issue
When Mol was an editor of the European Journal of Obstetrics and Gynecology in 2015, he realised that two of the many research papers submitted for possible publication looked a little suspicious.
After careful scrutiny, he noticed the numbers in both papers did not make sense. So he wrote back to the authors, asking them a few critical questions – and heard nothing back. A few months later both papers popped up, this time in other scientific journals, with the issues Mol had raised corrected.
“They had changed all the things I asked, but never communicated with me again,” Mol says.
This wasn’t the first time he had come across allegedly fabricated medical data. In 2011, Mol discovered a retraction notice for a study on uterine fibroids and infertility published by a researcher in Egypt.
The journal which had published it was retracting the study because it contained identical numbers to those in an earlier Spanish study – except that that one had been on uterine polyps. The researcher had just copied parts of the polyp paper and changed the disease.
Since then, Mol and a group of sleuthing researchers have been determined to weed out fabricated research – especially in the field of women’s health.
Mol now believes as much as 30% of the knowledge-base that underpins Australia’s policies around pregnancy management, such as our understanding of how to treat and prevent recurrent miscarriages; excessive bleeding during delivery; and polycystic ovary syndrome (PCOS), is based on fabricated research.
He says the impetus for those in power to fix it is poor and mired in self-interest.
Some women’s health treatment in Australia under scrutiny
While Mol believes the scourge of fabricated research affects all scientific disciplines, so far he believes he has found concrete proof mostly within his own field. One of these relates to how we treat recurrent miscarriages (more than two).
There is a similar story with progesterone (but not quite identical – Mol cautions that we do not yet have as strong evidence for the use of progesterone as we do for heparin). In February 2022, the TGA approved the use of the micronised progesterone pessary (Utrogestan) in the management of women with unexplained first-trimester bleeding and a history of three or more miscarriages.
A subsequent study on Australian ground … found no benefit whatsoever.Professor Ben Mol
Utrogestan was already approved for women who require extra progesterone while undergoing fertility treatment, and for prevention of preterm birth in women with a short cervix, for which it was already listed on the PBS.
However, Mol argues that the evidence underpinning this decision comes from a series of fabricated research papers by researchers from Egypt and India, which were used as the basis for a 2019 Cochrane review by Dr David Haas from Indiana University and colleagues.
“Recent data shows that progesterone only prevents miscarriage in women who are bleeding in their pregnancies,” Mol adds. “Yet, we are using this drug widely.”
Another area needing investigation is the use of growth hormone (GH) as an adjunct treatment in ovulation induction and IVF.
A 2017 Australian review paper, ‘Growth hormone in IVF cycles: any hope’, which found a benefit for the use of the growth hormone, with less ovarian stimulation required, the collection of a greater number of eggs, and an improvement in pregnancy rates summarises mainly fake data, Mol says.
Nobody knows where their research is coming from or where it’s been done.Professor Ben Mol
“A subsequent study on Australian ground – and probably one of the few reliable studies in the field – found no benefit whatsoever,” Mol points out. “If you follow the review which includes the fake data and choose to use growth hormone, it will cost you between $1000 to $3000 per cycle in Australia.”
Yet another instance of untrustworthy data informing policy comes from Australia’s guidelines of PCOS, which have very recently been updated. This update, led by Australian researchers from Monash University, checked the integrity of the underlying papers for the first time, and found that nearly half of the studies in the field had to be discarded due to their unreliable data.
“We decided not to use 45% of the studies on medical treatment of PCOS because we don’t trust them,” Mol says.
“We have also written to all of these authors telling them this, and only three have responded. This means that 42 authors – and therefore, papers – are out there, and nobody knows where their research is coming from or where it’s been done – it is unthinkable”.
Lack of oversight of fabricated research in Australia
It’s not just Mol who thinks we have a problem with fabricated research – molecular biologist David Vaux, who has also worked on editorial boards of various journals such as The Journal of Cell Biology and Cell Death and Differentiation, is more than aware of the poison seeping into academic integrity.
In Australia, we have a self-regulation model, the same model that used to be used by the church when handling child abuse cases.Professor David Vaux
Vaux, a longtime advocate for more oversight into Australian research, believes that Australia does not take research misconduct seriously enough. This is in stark contrast to European countries, the US, Canada, and even China, who have national bodies that oversee academic integrity.
“In Australia, we have a self-regulation model, the same model that used to be used by the church when handling child abuse cases,” Vaux says. “We have colleagues having to make judgements on their own colleagues, and there’s bound to be conflicts of interests or perceptions of conflicts of interest here.”
Vaux suggests Australia needs to create “an independent body where people are experienced, don’t have conflicts of interest, where they can act rapidly, help whistleblowers, and help organisations carry out robust investigations”. This body should be overseen by either the Department of Prime Minister and Cabinet or the Attorney General to maintain full accountability.
“We don’t even know how big the problem is in Australia, because no one collects data, and the people who do the wrong thing try to hide it. We need to act as though it’s happening, because it probably is,” Vaux says.
There is some movement
Deputy director of Neuroscience Research Australia and one of Australia’s most senior scientists, Professor Simon Gandevia, founded Retractions Australia following strong concerns about the direction of the country’s research establishment.
Retractions Australia catalogues all articles which are retracted, and which have an Australian or New Zealand affiliated author.
The way science currently operates does not guarantee that when a study is published, it’s actually correct.Professor Simon Gandevia
Gandevia believes that retracted papers are likely the tip of an iceberg: “although only about 0.1% of all published papers are retracted, around 2% have dodgy images, and the real percentage with shonky work is likely higher,” he tells Cosmos.
“My motivation behind starting this, aside from wanting to see science done better, is to publicly show people that the way science currently operates does not guarantee that when a study is published, it’s actually correct.”
He points to a famous 2005 paper titled Why Most Published Research Findings Are False, which systematically looked at study biases, designs, solo, siloed investigators limited to small sample sizes and many other behind-the-scenes behaviours that impact on quality of research.
“As it stands, we may as well consider just tossing a coin on whether the data on most published research is right or wrong,” Gandevia adds.
It doesn’t help that many of the publishers and editors themselves have too much skin in the game to push for retractions or any accountability, Mol says. “Everyone just looks away – it’s not in their interest to retract studies, and even comes with legal risks.”
But, as Vaux says, this is not an unmanageable problem – we just need to get those in power to listen.
“Scientists are humans, and there’s always going to be some humans that do the wrong thing, and in any complex system you need a mechanism to handle cases when they arise,” he says.