Medicine & Ethics

The Papers Your Doctor Trusts Might Already Be Wrong

By Jackson Bomongcag, YouthCare Connect 10/3/2026

In January 2026, the journal Nature retracted a paper that had helped scientists of the last two or three years think about treating lung cancer. The paper had been cited 192 times. It had 49 authors. 48 of them said they had no idea that anything was wrong with it.

An investigation by the Francis Crick Institute found that a PhD student had created the fraudulent data that the paper published. For almost three years, numerous medical researchers and doctors had used this data in their work. Some of them built upon the ideas that the paper created. Only when it was found during an internal investigation of Nature did the fraud come to light.

Here’s the part that should bother you more than the fraud itself: This is not an anomaly anymore.

The number that should worry you

In the year 2000, scientific journals retracted around 140 papers. That was all of the scientific literature published that year.

By 2023, there had been over 10,000 retractions of scientific literature in that same year alone. The retraction rate for scientific journals, which is the number of papers that get retracted out of all the papers that are published, went from being around 0.02% in 2016 to 0.2% in 2025. That’s a tenfold increase in less than a decade.

Out of all the retractions of scientific journals, as of April 2026, 60% of those papers have been found to contain actual misconduct. That means that the scientists behind these retractions had fabricated data, ideas, results, and research reports. This is not due to typos, miscalculations, or other forms of human error. These authors had faked their research.

This is not an issue with a few bad apples among scientific researchers. Rather, it is an issue with a system that permits more fraud in scientific publishing than in previous decades. It could be that there are more instances of fraud entering the system, or perhaps that there are more investigations into those instances of fraud than before – or perhaps both.

The business of fake science

Some scientific fraud is not committed by individuals alone. It is becoming an industry.

Paper mills are companies that take money from scientific authors and journals and publish fabricated research. These published papers contain no actual research by the authors listed on the paper. It is a business with revenue, with goals, and with a means of earning that revenue from the scientific community.

Between 2016 and 2025, 20% of retractions of health and life science journals came from papers published by paper mills.

One of the most prominent examples of this business creating fraudulent health and life science publications is the journal publisher Hindawi, which is owned by the scientific journal publishing company Wiley. In 2023, Hindawi retracted over 8,000 articles alone – more than any journal has ever retracted in a single year. Many of those articles were published by paper mills. Wiley lost between $35 and $40 million in revenue as a result of the fraudulent publications. Eventually, the company retired the Hindawi journal publisher.

When it reaches a patient

Fraud in an obscure journal is bad. But fraud that impacts medical practitioners and patients is much worse.

In 2020, one of the world’s most prominent medical journals, The Lancet, published a research paper that claimed that using the medicine hydroxychloroquine increased the death rate of patients with COVID-19. The research came from a company called Surgisphere, which claimed to have medical records of 96,000 patients treated in 671 hospitals around the world. The revelation of this paper caused the World Health Organization to suspend its COVID-19 drug trial for hydroxychloroquine.

When people asked the company Surgisphere for the data of the patients that they had treated, the company refused. Independent researchers were not able to verify the data that The Lancet published. Both The Lancet and the New England Journal of Medicine retracted the papers that were based upon the data that Surgisphere published. A real-time medical trial was altered due to data that might never have existed.

There is another well-known field in which fraud has occurred and continued for years – the field of Alzheimer’s. In 2006, a research paper published in the field of Alzheimer’s disease claimed that a specific protein structure was the cause of memory loss in mice with the disease. This paper was cited over 2,300 times. It became one of the most-cited scientific papers to be retracted.

In 2024, Nature published a retraction notice for the paper, stating that the research contained “excessive manipulation, including splicing, duplication, and the use of an eraser tool.” The fraudulent alterations to the research had been published since 2006. For over two decades, other scientific researchers had relied upon this data and published images of the brains of the mice as part of their own research on Alzheimer’s.

The part almost nobody talks about

Here’s a detail about the retractions of scientific journals that most people don’t talk about: Retracting a paper does not remove its influence on the scientific world.

When researchers investigate the citations of retractions of scientific research, over 90% of the time, they find that the papers that cited the retractions did so as if the retractions had not happened. There were no caveats to the citations, no changes to the references, no discussion of the reasons for the retraction.

According to the research director at Retraction Watch, a well-known researcher of the retractions of scientific journals, papers that are retracted are cited “as if it hadn’t been retracted” more than nine times out of ten.

The retraction process is intended to be the time when a fraud is acknowledged and the researcher is punished. Yet, instead, the fraud is acknowledged by a few individuals, and almost everyone else continues to use it as if it were genuine.

Why this keeps happening

The number of scientific publications that an individual researcher has is one of the main indicators of their success in the field. More publications mean more research funding, higher ranks in scientific institutions, and advancement towards tenure. This incentive encourages scientists to take shortcuts to increase their publication rates. An even greater incentive is provided for the researchers who are willing to take even more shortcuts to increase the number of their publications.

The Office of Research Integrity in the United States investigates misconduct in federally-funded scientific research. However, they only investigated misconduct in scientific research during 2025 in exactly two cases. That’s the lowest number of investigations into fraud in scientific research since the Office of Research Integrity began reporting such investigations.

Finally, artificial intelligence has become another reason for the increase in scientific fraud. Scientific journals use artificial intelligence tools to identify manipulated scientific images faster than a researcher could possibly inspect every publication. However, the same AI tools can also create fake microscopy images that are indistinguishable from real scientific images. This tool, meant to investigate and catch scientific fraud, has also become the tool to commit the fraud.

What actually helps

The most useful development in the fight against scientific fraud is the availability of a database of fraud.

Retraction Watch, the database that has been tracking scientific retractions since 2010, is now free to the public after being purchased by Crossref in 2023. It contains over 60,000 publications that have been retracted from scientific journals.

Publishers of clinical trials are also introducing statistical tools to detect fraudulent data before the trials are published. However, rather than being a solution to the size of the problem of scientific fraud, it is the first such change in years to the system.

The honest way to think about scientific fraud in medical research is that it has never been zero, and never will be. But in the past decade, there have been significant changes to the number of cases of fraud that are being caught and investigated. Right now, though, the number of individuals who are investigating fraud is increasing at a rate that suggests that many of them are still too far along in their research projects – too many have published their papers, too long, and past too many researchers who failed to investigate the origins of the fraud in the publications that they cited.