In February this year, Elisabeth Bik, a well-known "academic anti-counter", broke out the news and named more than 400 problem papers at one time. Almost all of these came from dozens of hospitals in China. They were highly similar in strips, graphic layouts and titles, and seemed to be produced from the same "paper factory". Bick even believes that there may be thousands of such articles from paper factories.
For a time, the "paper factory" incident attracted high attention and discussion among domestic people.
After just 3 months, Bick once again exposed a batch of suspected fake papers. The 8 papers listed this time all came from different hospitals and different authors. The diseases studied in the paper are also completely different types of cancer, but the results are surprisingly consistent, and even the charts and data in the papers have large-scale similarities.
published such a blatant academic fraud paper in international academic journals, which can be said to have refreshed the lower limit of academic fraud, simple and rough.
What is sad is that, like the previous question paper incidents, all the relevant personnel exposed this time are still from China.
A practicing physician and health blogger on Weibo said that this is "crazy fraud". Three of the papers were actually noted in the article, and the project was funded by the National Natural Science Foundation.
This time, the people involved are from , affiliated to hospital, thoracic cardiovascular surgery, , Fujian Medical University affiliated to general surgery, , Zhengzhou University 1 affiliated to respiratory department, Liaoning Provincial Cancer Hospital, Liaoning Provincial Cancer Hospital, Colorectal Surgery Department, Shandong University Qilu Hospital Obstetrics and Gynecology Department, and , Qingdao University Affiliated to general surgery department.
Figure | 8 papers (Source: PubPeer)
Among these similar papers, most of the papers were from 2017-2018, of which, the earliest was published in 2017 and the latest was published in April 2020. These 8 papers were published in the following 5 journals, and the impact factor is between 3-6.
International Immunopharmacology, impact factor 3.361, 1 article;
Cellular Physiology and Biochemistry, impact factor 5.5, 1 article;
journal of cellular and molecular medicine, impact factor 4.658, 1 article;
Open Biology, impact factor 3.89, 2 articles
Oncotarget (deleted from MEDLINE in 2017), 3 Article
"Break through all my understanding of academic fraud. Judging from the signature of the article, from front-line doctors to chief deputy chief physicians and deputy directors of the hospital, many articles are funded by the National Natural Science Foundation of China." A Weibo user said.
DeepTech verified the situation with corresponding author of 8 papers by email. As of the time of publication, no reply was received. Team members have not responded to PubPeer so far.
There may be problems with the paper itself, and we still need to wait for further response from the team.
8 papers are highly similar to
, and various signs indicate that this batch of problematic papers is likely to be the hands of the "paper factory".
(Source: Twitter)
At the same time, Bick posted a post on the academic exchange platform PubPeer, comparing two of the papers, and found that the two have amazing similarities. Among them, the data of one paper came from 286 patients with gastric cancer at Zhangzhou Hospital affiliated to Fujian Medical University; the data of the other paper came from of China Medical University and 384 patients with non-small cell lung cancer at Liaoning Cancer Hospital.
Bike found that both of these papers, published in 2018, are based on completely different research projects, with similar Kaplan-Meier curves, similar line graphs, similar western blots, similar bar graphs and similar statistics.
diagram | Similar Kaplan-Meier curves and line graphs
diagram | Similar western blots and bar charts
diagram | Similar experimental data
BIK even found that the similarity in text is also very obvious.
(Source: Twitter)
Bik further compared the Kaplan-Meier curves for all 8 papers, and although there are no black lines in the middle in several papers, the shapes of the red and green curves are surprisingly similar.
Figure｜8 papers all have similar Kaplan-Meier curves (made by DeepTech)
She asked on Twitter, "If there are 2 different patient groups, 2 different cancer types and 2 different research institutions, what is the chance that the results will be exactly the same?" Then, Bick gave a positive answer, "apparently zero."
(Source: Twitter)
The corresponding authors of the two papers used the same email address: [email protected], which also conforms to the previous inference by German independent scientific journalist Leonid Schneider, "The controller of these accounts is definitely not a doctor listed as the author of the paper, but the operator of the paper factory."
Figure | Bick speculates that problem papers are not simply plagiarism
In this series of papers, Bick keenly discovered in comparison that there were large overlaps in the paper texts. More importantly, the two papers just happened to "collide" in the reception/publishing time.
information shows that one of the articles was accepted by the journal in May 2018 and published on the website in August of the same year; the second paper manuscript was first received in July 2018 and finally published in October of the same year. The detail of
shows that before the first paper was published, a suspected similar paper manuscript was completed and received, which means that this will not be a simple paper plagiarism.
Bike speculated on Twitter that one of the possible reasons is that the first paper was "stealed" during the peer review stage; the second is that the two papers may have been from the same paper factory, and the reason for the similarity is that the paper factory "sold the same paper".
DeepTech contacted the anti-counterfeiting team of the paper about this matter. Team members confirmed that there are enough reasons to speculate that these are more like organized writing than personal plagiarism.
team members also revealed that in addition to the 8 papers exposed this time, the team discovered other similar problem papers a few years ago. This is not the first time this situation has happened. The
team shared a blog with DeepTech, recording a batch of suspected fake papers discovered previously. The method of similar pictures recurring in many papers is exactly the same as this one, and the author of the paper is also from China.
Figure | Through the review, the team found that there was cross-reuse of multiple paper charts, which were called "paper multiples"
shocking academic fraud.
just happened 3 months ago 3 months ago. This is not the only "paper factory" fraud incident that happened around us recently.
3 months ago, Bick and her team discovered more than 400 papers from different institutions, all of which seemed to be produced in the same "paper factory". Almost all of these papers came from dozens of hospitals in China, including several Grade A hospitals. Bick mentioned the following 6 Grade A hospitals:
Jining First People's Hospital (Grade A): 101 papers (almost no duplicate authors, involving 15 pediatrics, 6 cardiology, 6 endocrinology, 6 nephrology, 5 vascular surgery and other different departments);
Jilin University Sino-Japanese Friendship Hospital (Grade A): 59 papers;
Qingdao University Affiliated Hospital (Grade A): 23 papers;
Linyi Central Hospital (Grade A): 16 papers;
Zhengzhou University First Affiliated Hospital (Grade A: Henan's largest Grade A): 16 papers;
Jining Medical College Affiliated Hospital (Grade A): 12 papers.
According to statistics at the time, the earliest of these similar papers was published in 2016, and most of them were from 2018-2020. The journals involved are not all so-called "water-filled journals". The overall quality is pretty good, which is also the reason why it caused a big sensation at that time.More than 400 papers listed in
are mainly published in 6 journals, with the influence of these journals ranging from 2 to 5. Among them, Cellular Physiology and Biochemistry were kicked out of SCI in January 2019. It also involves many publishing houses such as Wiley and Elsevier.
Pick's discovery comes from clues provided by anonymous user Indigofera Tanganyikensis, who pointed out on PubPeer that the backgrounds between the Western imprint detection plates of some papers were very similar. Later, they found that not only the backgrounds of the same paper were similar, but also unexpectedly similar between the detection plates of different papers.
After noticing this clue, Bick followed up. She found that in more than 400 papers, the spacing between western blot bands was very regular, with dumbbell or tadpole-like appearance without common stains. All strips are placed on a similar background, indicating that they were copied and pasted from other sources, or were computer-synthesised.
In addition to the data graph itself, the bar charts of these papers are also very regular: mostly gray tones, with black bars on the left, and each picture is a bilateral error bar. It is not ruled out that such a layout may be a relatively popular standard image style, but the layout designs of hundreds of papers in different institutions are the same, which is indeed rare.
Bike even summarized the title templates of these "paper factories", namely "insert a molecular name, choose a verb (usual present form, third-person singular form), choose one to two cell transformation processes, choose a cancer or cell type, choose a conjunction, choose a verb (current progress form), insert the name of miRNA or neural pathway".
In this latest event, the same rule was found in the titles of these 8 papers. For example,
Interleukin-37 mediates the antitumor activity in colon cancer through β-catenin suppression;
Interleukin-36 Receptor Antagonist Is Associated With the Progression of Renal Cell Carcinoma;
Interleukin-35 Expression in Non-Small Cell Lung Cancer Is Associated With Tumor Progression.
As of this time of publication, up to 453 papers have been on the list. If this phenomenon continues to rampant, this list will be longer.
At that time, TigerBB8, one of the anti-counterfeiting team members, said that in the process of exposing the fraud of papers, he had received complaints from Chinese paper buyers, "During the day, I have outpatient surgery; after get off work, I have to take care of my children. After 10 o'clock in the evening, I only have a little time, but this is far from enough, because scientific research requires a lot of time."
In response to this latest incident, some Weibo netizens also said that part of the reason for such incidents is that there are problems in the doctor's professional title promotion system.
Academic fraud is too rampant
In recent years, the phenomenon of "paper fraud" has become common.
The previous "paper factory" incident was also very common. At this stage, the academic publishing industry still lacks a systematic way to identify the problem of image fraud.
mentioned in a previous report by DeepTech that the Bick team believes that analyzing images and data is an important way to uncover the falsification of papers, and this is also one of the most important concerns of the Bick team in the process of fighting papers.
For a long time, cracking down on deliberate paper fraud has not been the focus of academic publishing. For example, after receiving the paper, reviewers of the peer review process generally do not doubt the authenticity of the data and pictures, but focus on whether the experimental design in the paper is reasonable and whether there are improper aspects.
Previously, according to Nature, so far, most academic journals have not used targeted means to identify whether the pictures in papers overlap. The reason is nothing more than that they are too expensive or time-consuming, and there is no software that can screen papers on a large scale.
However, in recent years, Chinese scholars have become increasingly influential in international academic journals, and the frequent occurrence of plagiarism and malicious fraud has made this issue pay more attention.
Recently, international academic publishing giants led by Elsevier and John Wiley are taking action to explore how to automatically identify and mark images in scientific research papers that have been tampered with or reused.
They have proposed the first formal cross-industry initiative ever, and have also set up a new working group to set standards for software that screens out problematic paper images.
This working group is led by the Standards and Technical Committee, a member of the International Association of Science, Technology and Medical Publishers (STM). Headquartered in Oxford, UK, STM Association is an organization that mainly protects the interests of academic, scientific, technical, medical and professional publishers.
Working Group is chaired by IJsbrand Jan Aalbersberg, head of research integrity at Elsbrand Jan Aalbersberg. In addition to Elsbrand and John Wiley, members also include representatives from publishers such as Springer Nature and Taylor & Francis, covering almost all mainstream academic journals.
"The ultimate goal is to create an environment that can help us automatically recognize image tampering," said Aalbersberg.
But so far, this response to image fraud is still in its early stages. Behind the fraud of papers, if various systematic problems from the academic and medical circles have not been resolved, "falsification" and "anti-falsification" will inevitably be a long-term tug-of-war. After the
incident, domestic Weibo netizens commented that this phenomenon "has caused real scientific researchers who are suffering from thinking, exhausted, and stay up late to write papers."
On the other hand, in the originally relatively pure academic community, a large number of malicious fake articles will bring irreparable pollution problems to scientific papers.
