Study shows stealth corrections in scientific articles Some scientific articles are secretly changed

Silently altering scientific articles without transparently reporting what has been changed. It happens, as René Aquarius, researcher from Radboud university medical center, discovered. Together with international colleagues, he conducted research on these so-called 'stealth corrections'. They found 131 cases across different publishers. Now, the authors are advocating for better documentation of all changes in scientific publications. 

'Hey, these images are the same,' thought neuroscience researcher René Aquarius while conducting research for an animal experiment study on brain hemorrhages. In preparation for a new study, he delved into the literature and stumbled upon identical images that were associated with different groups. 'A mistake has been made here,' he thought. He made a report on the forum PubPeer, where scientific papers are discussed. The author of the article responded, thanked Aquarius, and said they would look into it. 

For a month, nothing happened, until Aquarius revisited the website of the publisher to look at the article again. To his surprise, the images had been altered, without any mention of it. On PubPeer, he found more researchers reporting changes to scientific publications that were not disclosed, known as stealth corrections. They united and spent months investigating unreported changes, by asking other ‘science detectives’ (or sleuths) and by using tools such as social media and prior PubPeer posts. 

Various publishers

In total, they found 131 articles published between 2005 and 2024 that received stealth corrections, spread across different publishers and disciplines. The results show that these changes occur across various publishers, including MDPI and Elsevier. A significant portion of the stealth corrections were found by researchers at an unknown publisher, BAKIS, with obscure journals. But Aquarius points out that this isn't the issue: 'These articles can still be found in PubMed, the online search system for scientific publications in the medical field. Scientists search PubMed and can easily find these articles to base their hypotheses on or to use in their grant applications.' 

Many of the stealth corrections involved substantive changes, such as modifications to figures or numbers in tables following a PubPeer report. There were also unexplained changes to the names of journal editors or authors. In some cases, articles that were published in a special issue were secretly moved out of it. The also found traces Large Language Models use, such as ChatGTP, that were quietly removed.  

Only a small number of articles eventually received an official correction. Aquarius believes this is unacceptable. 'It could suggest that publishers or editors are trying to cover up errors or changes. That is not in their best interest,' he says 'as that undermines the credibility of the authors, editors, publishers and science as a whole.' 

Moreover, readers are denied the opportunity to critically judge the article they are reading. 'With transparency and proper documentation, you safeguard the process of disseminating scientific knowledge. It’s always important to know if something has changed, because second-guessing every scientific article you read will hurt science as a whole.' 

Recommendations

Aquarius and his international colleagues offer three suggestions to prevent stealth corrections: 

• Develop and implement a mandatory system for publishers that publicly logs all changes that have been made to published works. 
• Create clear definitions and guidelines for all types of corrections. 
• Promote vigilance among scientists to publicly register stealth corrections. 

Aquarius' work is not finished. 'We don’t know how often this happens because it’s done secretly, and thus it is not visible. That makes this work extremely difficult to carry out in a comprehensive fashion. However, with our publication we have demonstrated that it does happen. That means we, as a scientific community, need to stay alert.' 

About this publication

This article was published in: The existence of stealth corrections in scientific literature – a threat to scientific integrity. Rene Aquarius, Floris Schoeters, Nick Wise, Alex Glynn, Guillaume Cabanac. DOI: https://doi.org/10.1002/leap.166.

• Want to know more about these subjects? Click on the buttons below for more news.

  News homepage news & media