Article Image

The Elephant in the Lab Coat

Op-Med is a collection of original articles contributed by Doximity members.

It's been lumbering around for decades, this elephant in the lab coat of medical research. We've all felt its stride, a disquieting rumble beneath the surface of seemingly pristine publications and groundbreaking discoveries. But today the elephant — research fraud — has grown bolder, its presence now impossible to ignore.

In 2023, more than 10,000 papers were retracted from academic journals. That does not begin to capture how much research fraud actually occurs, with estimates suggesting that there should have been another approximately 190,000 retractions. Countless other studies harbor questionable methodologies, honest statistical errors, and lazy science. Now, the whispers of "misconduct" and "data manipulation" have reached the surface — and the eye of the public. High profile cases, like the July 2023 dismissal of Stanford President Marc Tessier-Lavigne and the report of 37 faulty studies at the Dana-Farber Cancer Institute have captured public attention and sparked a wider debate about higher education and its commitment to truth-seeking. Let us not be naive: much of the magnified focus on research fraud is the result of political partisans seeking to undermine academic institutions. But that does not mean they do not have a point.

The credibility of medical research isn't a political bargaining chip; it's the bedrock upon which physicians guide diagnoses, develop treatment plans, and ultimately, impact the lives of countless patients. Research is, in some ways, the foundation of trust in the doctor-patient relationship, as it makes our advice and expertise credible. This isn't just a matter of bruised egos or academic reputation; it has real-world consequences, hindering our ability to implement positive change in patients’ lives. As a medical student early on in my career, seeing research malpractice happen at institutions close to home hits a nerve, and heightens my concerns for the future of academia. It's time to confront the elephant. Let’s begin by defining the problem: what is research fraud?

Research Fraud: A Brief Definition and Case Study

At its core, research fraud encompasses any intentional fabrication or manipulation of data, methodologies, or conclusions. This includes knowingly falsifying results, plagiarizing existing work, misrepresenting statistical analyses, and deliberately obscuring negative findings. However, the murky waters of misconduct also extend beyond intentional malice. Honest mistakes, non-rigorous research practices, and improper data analyses, albeit unintentional, can have equally detrimental consequences and must be considered in this context. 

Intertwined in this problem is the existence of “super-productive researchers” who are now cranking out a paper every five days, on average. Disconcertingly, most of these “extremely productive researchers” are in clinical medicine, where research goes on to affect patient lives. An excellent case study of research fraud in medical practice comes from the height of the COVID-19 pandemic and the debate around the antiparasitic drug ivermectin. Not only was this one of the most analyzed and scrutinized research topics in recent history, it was also part of the aforementioned field of clinical medicine. One of the earliest studies of ivermectin for COVID-19 was the Elgazzar 2020 study from Egypt, which claimed that across 600 patients, ivermectin significantly reduced mortality. This study inspired other trials, and its massive effect sizes (two deaths in the ivermectin group as compared to 20 in the control, for example) even began to affect practice. As the smaller studies began to pile up, journals and independent media began to publish meta-analyses to determine the effectiveness of ivermectin with greater statistical confidence. In part thanks to the massive effect size of the Elgazzar study, it appeared that ivermectin reduced COVID-19 mortality.

We now know the Elgazzar study was entirely fraudulent. Some of the patients had died before the trial even began, several patients did not exist and were just copy-pasted data, and much of the data was highly edited. This only came to light because reporters, researchers, and volunteers were able to get access to the raw data by guessing the password of the protected file: 1234. Once this study was removed from the meta-analyses, much of their positive effect vanished. Still, some analyses claimed ivermectin reduced mortality — while others claimed it caused it. It seemed that the entire effect was determined by which studies you included, 10% to 15% of which were suspicious for fraud. A similar case unfolded politically, as supporters and detractors of ivermectin touted the studies which “proved” their point. With the benefit of hindsight, we now know that ivermectin did not significantly reduce mortality in patients without parasitic comorbidity. All this goes to show that research fraud is not an abstract, academic issue. In the case of ivermectin, it almost certainly harmed people’s lives.

A Swiss Cheese Model to Combat Research Fraud

Research fraud is a multifaceted problem; if one person could prevent it then peer review would be sufficient. Incentives abound for committing fraud: desire for prestige; the relentless pressure in academia to publish; the tendency for journals to favor studies with positive results, to name just a few. 

However, the incentives go beyond individual pressures. Academic institutions themselves are incentivized to turn a blind eye to questionable practices among their employees. Prestigious researchers attract coveted grants, boosting the institution's reputation and securing lucrative funding streams. The pressure to maintain this prestige can create a culture of silence, where concerns about misconduct are swept under the rug to protect valuable assets. 

Understanding the toxic brew of incentives fueling research misconduct is crucial, but the true test lies in devising effective solutions. Thus, I propose a proactive four-part approach inspired by the Swiss Cheese Model — a metaphor where multiple layers of defense work together to prevent harm — preventing fraud before it taints the scientific record. 

1) Building a culture of integrity

This begins with our students. Research for students has begun to resemble international paper mills, where students are offered recommendation letters and dozens of publications with minimal impact. We must encourage good research practices in medical school and reduce the pressure to publish. Residency committees should stop being naive to the reality that allows students to publish dozens of papers.

One organization attempting to teach students how to do research properly and with integrity is the International Research Olympiad (IRO). Full disclosure: I am on the advisory board of the IRO. But I joined the board because I genuinely believe that teaching students how to critically evaluate and perform methodologically sound research at an early age will pay dividends in the future.

2) Expanding the toolkit for fraud detection

Research fraud is not a new problem. Brilliant minds have proposed tools and techniques to detect fraud using mathematical and statistical techniques. Looking out for p-hacking, ensuring pre-registration of predictions, and ensuring that the sample sizes are even possible to achieve are great tools with a strong track record of identifying fraud. The REAPPRAISED Checklist, proposed in 2020 after its inventors spent a decade attempting to get journals to retract a clear-cut fraud case, offers a strong starting point. 

However, we can expand this toolkit further. In the ivermectin case study, where 10% to 15% of published studies were suspicious for fraud, techniques like the Carlisle-Stouffer-Fisher Method helped to find implausible data. Journals and academic institutions must enforce the use of such techniques before allowing any research paper to be published. 

Notably, with over 5 million papers published each year, it is unfeasible to rely only on human checklists. With artificial intelligence (AI), we may have a new solution to automatically review papers for fraud at scale. Modern AI such as GPT-4 and Gemini have finally advanced to be able to deal with complex unstructured data. Such tools could flag all submitted research for potential methodological errors and/or duplicated or photoshopped images, and could perform statistical tests. Review at this scale could have prevented the recent high-profile fraud cases.

3) Reforming academic institutions for oversight

Academic institutions ought to create independent roles to review research output before publishing. At present, there are too many perverse incentives for in-house reviewers to ignore misconduct until it becomes public knowledge. 

Thankfully, there are already groups that do astounding work in reviewing research and acting as “data-sleuths,” such as the volunteer-maintained PubPeer. Perhaps if academic institutions pooled funds and were required to fund anonymous, independent, third party groups like PubPeer to hire experts and review institutional research output, oversight would be strong enough to significantly limit the amount and degree of research fraud.

In addition, institutions should reform career advancement pathways to ensure doctors not interested in research can still advance. In an era of physician shortages, we should allow doctors to achieve career advancement through spending time in the clinic helping patients. That way, only those truly motivated to further science would pursue research. 

Finally, institutions must reform how they fund research. Too often, institutions that do not truly understand new technology award grants to researchers who show unrealistic output. Our higher education institutes should ensure that research funding is utilized by those who produce impactful research of high quality — and potentially utilize new tools to identify such researchers early.

4) Enforcing higher standards in academia

In the ivermectin case study, the statistical techniques to identify fraud are helpful, but nothing can beat the power of access to raw data. It is a real shame that we still allow research to be published without the data being publicly available, especially in clinical medicine, where there is concern for bias. Even with papers which claim to have available data online, in my experience it's a coin flip whether the link even works or the data has been maintained. We need to appropriately penalize journals when they do not adhere to such standards or when they have many retractions, to elevate journals which publish good research and to encourage investment in fraud prevention techniques.

A Call to Action: Before We Lose Faith in the Very Fabric of Research

Confronting the issue of research fraud is undoubtedly uncomfortable, especially for those of us deeply invested in academia. It shakes the very foundation of our trust in research as a source of truth and progress. However, ignoring this elephant in the lab coat is not an option. We are teetering on the precipice of a dangerous reality where health care professionals might be forced to assume the research they rely on is inherently flawed, as some have already suggested. With 20% of published clinical trial research not occurring or not being honestly reported, this assumption is not a radical one.

The prevalence of research fraud is not just an issue for institutions or researchers; it concerns us all. Patients deserve to know that the medical decisions impacting their lives are based on sound science, not fabricated data. The public deserves transparent and trustworthy research that guides critical policies and advancements. For physicians in academia, advocacy can start at home. There is no one-size-fits-all approach to fraud, but the proposed Swiss Cheese Model may allow us to openly discuss the elephant in the lab coat.

How is your institution working to prevent research fraud? Share in the comments.

Aditya Jain is a third-year medical student at Harvard Medical School. His previous works include "The Future is STEM" and medical fiction shorts for In Vivo Magazine. He is a published researcher on the applications of artificial intelligence in medicine. When he's not busy with rotations, he enjoys playing guitar, reading sci-fi, and nature hiking. He tweets @adityajain_42. Aditya is a 2023–2024 Doximity Op-Med Fellow.

Illustration by Jennifer Bogartz

All opinions published on Op-Med are the author’s and do not reflect the official position of Doximity or its editors. Op-Med is a safe space for free expression and diverse perspectives. For more information, or to submit your own opinion, please see our submission guidelines or email

More from Op-Med