Data sleuth flags 30 randomized clinical trials from researcher in Egypt

Thirty randomized clinical trials involving a researcher in Egypt who has already had six papers retracted show signs of research misconduct and data fabrication, according to the authors of a recent preprint. 

Ben Mol, one of the authors of the preprint and a professor of obstetrics and gynecology at Monash University in Australia, has spent several years investigating the work of Sherief Abd-Elsalam, a hepatologist and gastroenterologist at Tanta University in Egypt. Abd-Elsalam denies that his research is false or fabricated. 

Mol has been exposing research misconduct in his own field for years. His work revealed dozens of dodgy obstetrics papers by Ahmed Badawy and Hatem Abu Hashim of Mansoura University, in Egypt, as well as serious problems with clinical trials led by Ahmed Maged at Cairo University, research about c-sections also from Cairo University, and urology research by Iranian researcher Mohammad Reza Safarinejad. 

Abd-Elsalam said in an email that he disagrees with the allegations in the preprint. “Where is the problem? We don’t know,” he wrote. 

Mol began to question the integrity of Abd-Elsalam’s work after noticing a number of suspicious randomized controlled trials (RCTs) he helped conduct on controversial COVID-19 treatments including ivermectin, hydroxychloroquine and the antiviral favipiravir. 

Four of the seven COVID-19 papers that list Abd-Elsalam as an author have been retracted so far, as we reported earlier. The retraction notices cite data overlaps, inadequate data validation, and doubts about the randomization process as reasons for the retractions. 

Abd-Elsalam is the first author in three of the four retracted papers:

• Hydroxychloroquine in the Treatment of COVID-19: A Multicenter Randomized Controlled Study
• Efficacy of favipiravir in COVID-19 treatment: a multi-center randomized study
• Remdesivir Efficacy in COVID-19 Treatment: A Randomized Controlled Trial
• Efficacy of combined Sofosbuvir and Daclatasvir in the treatment of COVID-19 patients with pneumonia: a multicenter Egyptian study

Abd-Elsalam has also earned a retraction for a non-RCT paper because of data fabrication; he disagreed with the retraction. 

• Efficacy and safety of sofosbuvir plus ribavirin for treatment of cirrhotic patients with genotype 4 hepatitis C virus in real-life clinical practice

Mol’s preprint had originally included a double-up of one of Abd-Elsalam’s papers in the appendix listing his retracted work. The duplication was amended following our reporting.  

Seeing the extent of the problematic COVID-19 RCTs led Mol to investigate Abd-Elsalam’s other work. The preprint, which was posted in March, details all 30 RCTs Mol and his co-authors found searching PubMed, Google Scholar and Scopus for papers describing clinical trials and listing Abd-Elsalam as an author. Out of the 30, seven are the RCTs on COVID-19 and the rest are in the field of gastroenterology. 

The publications flagged by the authors have been cited more than 1,200 times. To date, just one of the gastroenterology papers has been retracted. As a consequence, clinicians might be giving misinformed advice to patients, according to Mol. 

The preprint includes a series of allegations about data fabrication, inconsistencies and plagiarism in the 30 RCTs. For example, Mol and his colleagues showed that in nine of the gastroenterology studies the recruitment start date in the trial registration and the published paper were inconsistent. 

Another detail the authors found striking among 15 of the RCTs is that they reported no drop-outs, despite some having long follow-up periods and patients with terminal illnesses. 

The instances of data fabrication allegedly include a number of highly unlikely P-values among the papers. The authors decided to recalculate them by performing a comparison of means and concluded they were incorrect in nine of the RCTs. 

The authors also noticed unusual data in tables reporting patient characteristics. For example, one paper published in 2019 had 40 binary variable categories and 36 of them were even – a result which has a 0.00000929% probability of happening by chance, according to the preprint. 

“It goes on and on and on,” Mol said. “It is full of these mistakes.” 

In his emailed response to the allegations, Abd-Elsalam wrote: “I am not sure what I did that …led them to conclude that there are concerns regarding the trials I participated in.” 

He added that the non-COVID-19 articles were mostly thesis papers, and claimed that he was listed as an author because of his writing skills.

In October 2022 and February 2023, Mol contacted Abd-Elsalam to share his findings. 

“Maybe I see it wrong, maybe there’s an explanation, maybe he also regrets it and decides to retract his own work,” Mol said. “But he basically denied it all and didn’t respond very seriously, so here we are.”

After getting in touch with several editors, some were swifter than others in responding to Mol’s concerns. 

The study, “Efficacy of favipiravir in COVID-19 treatment: a multi-center randomized study,” published in Archives of Virology in January 2021, received an editor’s note citing an investigation into the article in August 2021, and a subsequent retraction in November 2021. The paper has been cited 146 times. 

Tim Skern is the journal’s editor-in-chief and also a co-author on the preprint analyzing the 30 RCTs. According to an email exchange shared by Mol, Skern wrote to Anna Lockhart, a senior editor of medicine and life science journals at Springer, expressing his concern about a second paper published in Archives of Virology in 2020. 

The study, “High success rates for the use of sofosbuvir/ombitasvir/paritaprevir/ritonavir + ribavirin and sofosbuvir/simeprevir/daclatasvir + ribavirin in retreatment of chronic hepatitis C infection after unsuccessful sofosbuvir/daclatasvir therapy: a real-life experience,” has been cited eight times. 

We have since learned that Skern’s original email referred to the wrong paper. In a subsequent email we were not originally sent, Skern requested an editor’s note on “Efficacy and safety of sofosbuvir plus ribavirin for treatment of cirrhotic patients with genotype 4 hepatitis C virus in real-life clinical practice.” As noted above, that paper was retracted in September 2022.

In an April 2022 email to Lockhart, Skern wrote:

It is clear from the analyses of two other expert statisticians that the following paper is not trustworthy (real terms fabricated) and must be retracted …

He also asked her to attach an editor’s note to the paper saying that it’s under investigation. The paper does not currently have the note. Anna Lockhart did not respond to our request for comment. 

In April 2023, Mol shared his findings with the editors of the European Journal of Gastroenterology & Hepatology, a Wolters Kluwer title, which published six of the RCTs included in the analysis. 

The lead publisher for medical journals at Wolters Kluwer, Phil Daly, responded with follow-up queries but to date none of the six papers has been retracted or received an expression of concern. Daly did not respond to our request for comment. 

Mol and his co-authors have written to the editors of the journals which have published 10 of the 30 RCTs, not including those of the retracted papers. The only response they’ve received is the one from Wolters Kluwer. 

“The journals don’t all respond, that is unbelievable for me,” he said. “I increasingly find that everybody is looking away from it and we don’t want to speak up. I feel myself a little bit the buffalo separate from the herd.”

Of the retracted COVID-19 papers co-authored by Abd-Elsalam, the journals took between 63 and 414 days to retract following an email from Mol and his colleagues. 

“It’s complicated to get the data out of these publications,” Mol said. “And it’s extremely uncomfortable to say to people well, maybe the work you were doing is not honest.”

In his email, Abd-Elsalam linked to a preprint he says he co-wrote (he is not listed as an author). The paper questions the credibility of a checklist developed by Mol to assess RCTs and uses 16 of Mol’s own RCTs in its analysis. 

The authors of the preprint conclude: “Using the TRACT (Trustworthiness in Randomized Controlled Trials) checklist, 16 RCTs co-authored by its developer (Mol) showed red flags that need to be investigated further.” 

Abd-Elsalam says he and his colleagues have also made several comments on Mol’s work on PubPeer, some of which he claims have been removed by the site. 

Responding to Abd-Elsalam’s concern about his work, Mol wrote in an email: 

The retraction was made at our own request because of an error. Otherwise I had 16 of my papers investigated without any concern raised.

One problem is that there is not a collective response from the academic community to people like Abd-Elsalam.

In August 2022, the Journal of Exploratory Research in Pharmacology announced on Twitter that Abd-Elsalam had been appointed to its editorial board. He is currently coordinating 74 clinical trials, with 35 trials marked as recruiting, according to clinicaltrials.gov. 

Mol and his co-authors are calling for all papers on which Abd-Elsalam is an author to be marked with an expression of concern.

Update, 6/23/23, 1900 UTC: After learning that we had not been provided with the entire relevant email exchange, we added a paragraph noting Skern originally referred to the wrong paper, and struck through of a passage referring to the wrong paper.

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].