Researchers face disciplinary action as dozens of their studies fall under scrutiny

A group of obstetrics researchers in the Middle East is facing disciplinary action after questions were raised about the validity of the data in dozens of their published studies. 

The tale — involving contaminated clinical trials, potentially fabricated PhDs, findings of misconduct that went ignored, accusations of terrorist sympathies and unresponsive journals — requires some unpacking, so bear with us. 

We begin with a study that appeared in April in the European Journal of Obstetrics & Gynecology and Reproductive Biology (EJOG). Esmée Bordewijk, a PhD student at the Center for Reproductive Medicine at Amsterdam University Medical Center, and her colleagues reported that they stumbled on the problems while conducting a literature review on ovulation induction for the venerable Cochrane Database: 

While performing data extraction, we noticed similarities between values in the baseline characteristics (i.e. age, Body Mass Index (BMI), parity) and outcome tables in different randomised controlled trials (RCTs) authored by Ahmed Badawy and Hatem Abu Hashim, from the Mansoura University. Following these concerns, we undertook a systematic assessment of the RCTs published by these two authors.

Using techniques similar to those of John Carlisle, the English anesthetist who helped expose the data deceptions of Yoshitaka Fujii and others, Bordewijk’s group looked more closely at 24 studies from Badawy  — a fellow of the Royal College of Obstetricians and Gynaecologists — and 11 from Abu Hashim, all published between 2006 and 2016. Out of those 35 papers, the researchers found 27 with problematic results.

For example: 

By analyzing the data we saw a pattern in the last digits for height, weight, and BMI in the trials of Badawy. In the Abu Hashim trials, we noticed a pattern for the outcomes pre-treatment endometrial thickness and number of miscarriages. The baseline characteristics of 17 of the 24 Badawy trials showed the same values after the decimal separator for height, weight and BMI. The distributions of the last digits deviated from their expected distributions. (Fig. 7, Supp. Fig. 1). Conspicuously, the digit seven never presented in these variables.

Bordewijk and her colleagues also showed that, in some cases, the probability that the Egyptian authors’ results were due to random sampling was exceedingly low: 

For 19 of 24 trials in which Badawy was first-author, the corresponding p-value for comparison with the uniform distribution was also less than 2.2*10^-16 (Fig. 9B, black dots). For five of 24 trials in which Badawy was not first-author, the observed distribution of two-group comparison p-values was consistent with the expected uniform distribution, with a p-value of 0.9192 (Fig. 9B, blue dots).

They observed similarly unlikely p values for Abu Hashim’s work: 

We have shown that the numeric results presented are very unlikely to either represent true biological values or to have been derived using the methods described by the authors. Apart from the findings presented, we also draw the reader’s attention to the number of recruits, and the number of distinct trials recruiting similar or even identical patient groups in the same hospital during overlapping periods of time …

The paper, however, was really the second shot. In February, the group, led by Ben Mol, an OB/GYN at Monash University, sent a letter to the editors of the journals involved alerting them to their investigation. In the letter, Mol — who earlier in the month had confronted Badawy in Cairo directly about his concerns — wrote: 

I herewith send you an analysis of 35 RCTs published by Badawy and Abu Hashim  In these 35 RCTs published between September 2006 and January 2016, we found a large number of similarities of baseline characteristics and outcomes between 27 studies. Also, analysis of the baseline characteristics of the RCTs indicated that their distribution was unlikely to be due to chance. The RCTs have been published in the following journals: Acta Obstet Gynecol Scand (5);  Arch Gynecol Obstet (4); Contraception (1); Eur J Obstet Gynecol Reprod Biol (1); Fertil Steril (10); Gynecol Endocrinol (2); J Minim Invasive Gynecol (1); J Obstet Gynaecol (4); J Obstet J. Gynaecol Res(2)Women’s Health (1) and Reprod Biomed Online (4). I have e-mailed to Dr. Badawy, but I did not receive a satisfactory reaction. 

I copy in the leadership of the RCOG and the ASRM. Dr. Badawy is a fellow of the RCOG and was the senior author on an ASRM 2018 prize paper.

I copy in the Cochrane editors from the Pregnancy and Childbirth Group and Gynecology/Fertility group, as many papers are in Cochrane reviews and in the Cochrane database for RCTs. I trust that the involved RCTs will be retracted soon.

From the studies of Gibreel, Ragab, El Refaeey and Marzouk we could not find direct evidence of data-fabrication, similar to the papers where Badawy or Abu Hashim are first authors. For these papers of Gibreel, Ragab, El Refaeey and Marzouk, I would suggest that the journals investigate, but that the burden of proof is with the authors. I have assessed the paper of El-Refaie earlier, including an assessment of the primary data, and I am convinced that that study is fabricated, too. I have longtime shared my concerns with the editor in chief, prof Olaf Ortmann, Arch Gynecol Obstet. For Shokeir 2011, I refer to another e-mail to follow this one. Dr. Shokeir has multiple papers where he copies data from articles of other authors; a different type of fraud.

The credits for detection [of] this fraud go to a student whose name I will not disclose at this stage, and to members of our Monash-team who did the analyses.

I have put substantial effort into this topic over the last years. Unfortunately, this is only the tip of the iceberg. I am afraid that if we assess better, we will identify hundreds of problematic papers, of which the ‘results’ directly influence patient care. I leave it with you whether that is important or not. Personally, I would plead for a collaborative action of all of us, which results in a thorough investigation of all the RCTs from this country.

In April 2020, Mol’s group published a similar analysis in Fertility & Sterility for Shokeir, pointing to four papers: 

in which the author had copied-and-paste the exact same values from baseline and outcome tables from different author groups and that analysis of 4 other studies indicated that their distribution was unlikely to be due to chance.  

Meanwhile, in mid-February, Mol contacted Ashraf Hafez, the vice president for graduate studies and research at Mansoura to inform him of his group’s findings. He received an immediate reply, which he shared with us: 

Dear Prof. Mol,

We are thoroughly investigating your scientific concerns with manuscripts published by our Obs. & Gyn staff members.

We will get back to you ASAP with the conclusions of the investigation committee.

Mansoura University would not accept any breach of ethical research codes.

Thanks and do not hesitate to contact me for any further inquiries.

We emailed Hafez to find out the status of any investigation but have not heard back. However, Mol forwarded us a letter, dated October 7, in which Mansoura president Ashraf Bakr stated that Badawy, Abu Hashim and Shokeir are: 

under trial by the University Disciplinary Court to initiate an official hearing of the case to reach disciplinary actions, in accordance with the Egyptian Law of Higher Education. … the trial might take months to reach a final verdict.

The letter added that Mansoura has been in contact with Fertility & Sterility and the EJOG since June, and that the country’s Ministry of Higher Education is aware of the case. 

Mol, who has been working to expose the misconduct for more than three years, said he believes some 250 RCTs in his field from authors in the Middle East contain fabricated data: 

There is a combination of lack of governance and incentives for people to do this.  

The incentive is academic promotion. 

Mol said Abu Hashim and Badawy, both of whom hold PhDs, appear to have used fabricated data to earn their doctorates. He has contacted the institutions that awarded those degrees and they are investigating, he added. 

Abu Hashim had already been found guilty of misconduct in 2014, according to Mol and a member of the investigative committee at the time (we are protecting that person’s identity for safety reasons).The investigation found that Abu Hashim had committed misconduct — but the report was overruled by the administration of the university after the researcher complained that: 

the issue was political rather than scientific and that some of the committee members … were political opponents to the ruling military regime. 

Two of the authors of the report were subsequently fired, according to our source. 

Somewhat surprisingly, Mol’s new role as literature cop hasn’t earned him much praise from his peers: 

When I spoke out, individual colleagues said, ‘Ben, you have a nice career, be careful.’

But he is convinced the work is important enough to keep doing: 

I’ve said to myself, I’m doing much more good preventing 250 RCTS from being in the system than myself having 25 of my own. If this practice isn’t stopping, I don’t see an end to it. 

In a May 14 email titled “Truth about Ben Mol allegations,” Badawy defended his department’s research and accused his accusers of “real blackmail.”  Badawy contended that his facility could not produce any original data for the older papers because: 

We formally explained that we had our IRB established in 2016 and as a hospital policy we have all the paper records discarded every five years. If we are in hold of the data set for the trials you are entailing in your manuscript which is dating back to 15 years, we would be happy to reanalyze again using your questionable methods like Monte Carlo method.

He concluded: 

on behalf of myself and more than 30 staff members in different departments of Mansoura Faculty of Medicine involved in all the trials that:

1.       we are willing to cooperate with any formal investigation to reach a satisfactory conclusion.

2.       We will do all the effort to confirm the integrity of our data as this is crucial to our names, institution and country.

3.       We confirm our rights to take legal actions against yourself for the allegations and many false statements you make against us.

In a follow-up email, Badawy raised the legal stakes, accusing Mol of working with a terrorist in his inquiries: 

This is really a serious legal problem you put yourself in.

You included a message to you and cited manytimes of Mohamad E. Ghanem, allegations against Prof. Hatem Abo Hashem publication, so that everyone is aware of what has transpired.

“Mohamad E Ghanem”, his full name “Mohamad El-Saeed Mohamad Hassan Ghanem” is NO longer a staff member in our department. In fact, he has been segregated from the Mansoura University many years ago (Mansoura University President Decision No. 6529, October 2017) after being accused by the Supreme State Security Court of Egypt for terrorism. “Mohamad E Ghanem” is one of the heads of the “Muslim Brotherhood terroristic group” in Egypt with accusations of planning assassinations and plots. Being segregated from our University, he got used to attacking Mansoura University, its staff members, and even the Egyptian government.  Attached 2 official files (in the Arabic language) confirming that “Mohamad E Ghanem” Full name “Mohamad El-Saeed Mohamad Hassan Ghanem” is accused several times being one of the heads of the “Muslim Brotherhood terrorist group” in Egypt.  …

Badawy did not respond to a request for comment. 

Abu Hashim, in an email to us, pointed to a September 2020 article in EJOG in which he defended himself and his colleagues, and stated: 

[the] Borderwijk–Mol’s article is heavily infested by several major flaws that undermine its validity and render all their conclusions meaningless. 

Regarding the 2014 misconduct findings, he added: 

it is absolutely incorrect. These words have no basis in truth … Be vigilant, and avoid defamation so that you do not fall under the legal issues 

Journals have been aware of the problematic papers for roughly eight months — Mol’s group notified editors before they published their article in EJOG — but only a handful have been retracted to date. 

Badawy has lost four articles in Fertility & Sterility stemming from the investigation. According to the retraction notice for the 2007 article “Clomiphene citrate or aromatase inhibitors for superovulation in women with unexplained infertility undergoing intrauterine insemination: a prospective randomized trial:”  

This article has been retracted at the request of the ASRM [American Society for Reproductive Medicine] Publications Committee.

Several reports of randomized clinical trials were recently the subject of an investigation conducted by the publications committee of the ASRM. The committee reviewed concerns related to the validity of data reported in those reports. The committee noted significant duplication of data reported in this manuscript with data reported in another report (PMID 17686479; https://doi.org/10.1016/j.fertnstert.2007.05.010) pertaining to a different study group that could not be explained. As we cannot vouch for the validity of the data, we have issued a retraction of this paper.

The referenced paper, “Anastrozole or letrozole for ovulation induction in clomiphene-resistant women with polycystic ovarian syndrome: a prospective randomized trial,” also has been pulled. 

The two other retractions involve “Clomiphene citrate or letrozole for ovulation induction in women with polycystic ovarian syndrome: a prospective randomized trial,” and “Letrozole versus combined metformin and clomiphene citrate for ovulation induction in clomiphene-resistant women with polycystic ovary syndrome: a randomized controlled trial.”

The journal also has issued an expression of concern for three other papers, all of which it published in 2009: 

• “Extended letrozole therapy for ovulation induction in clomiphene-resistant women with polycystic ovary syndrome: a novel protocol;” 
• “Clomiphene citrate or anastrozole for ovulation induction in women with polycystic ovary syndrome? A prospective controlled trial;” and
• “Luteal phase clomiphene citrate for ovulation induction in women with polycystic ovary syndrome: a novel protocol.”

As for Shokeir, we see one retraction, from Fertility & Sterility, and two expressions of concern for a pair articles in the Journal of Minimally Invasive Gynacology, one of which he wrote with Abu Hashim.

We reached out to the editor of the journal for comment, but received a response from the ASRM asking us not to contact journal staff directly. They then sent us a statement reiterating the retraction notices and expressions of concern, but which did not offer any other information.

We have contacted the editors of several of the journals involved, but have yet to receive answers to our questions.  

We did get a reply from a spokesperson for the Archives of Gynecology and Obstetrics, who told us: 

For confidentiality reasons, we are unable to comment on articles that may or may not be under consideration for retraction.

We are in the process of investigating an article by Badawy and others after concerns were raised. These issues are often complex and can take time to fully unravel. While we are looking into this, we have placed an Editorial Note on the article to make readers aware that concerns have been raised. As stated in the note, there is an investigation by the University of Mansoura, the results of which we have been awaiting for several months.

We will take the appropriate action once our investigation is concluded and we will then update the scientific record accordingly.

We also asked about whether the journal or the publisher — Springer Nature, in this case — had received any legal threats from the Egyptian authors. We were told: 

we treat any correspondence with authors as confidential and cannot comment further on this.

Although some of the papers have been incorporated into meta-analyses and reviews, Mol said their clinical impact is generally modest. Still, he said, some of what he has seen has been dismaying:  

One of these guys has faked a paper on how to treat women with fertility problems after cancer.

How can you make up data on that? That’s the most vulnerable group of patients. 

Update, 1445 UTC, 10/12/20: Contraception editor Carolyn Westhoff tells us:

Sorry to put you off.  

As you are surely aware, the pandemic has led to record-breaking submissions to many journals including Contraception.    I have been swamped with keeping up with the immediate workload, and thus have simply not been able to find time and energy to investigate a ten-year-old paper.   

This is not to belittle the seriousness of the specific problem (or the more widespread problems of this type), but only to say that my capacity is quite overwhelmed.  Figuring out how to reach a more sustainable workload is beyond me at present.  

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].