[Reading Science] 'Zombie Papers' Are Rampant in the Medical Field

Widespread Use of Fake or Defective Data in Clinical Trials
Serious Medical Side Effects and Decline in Credibility
Origin of 'Anthelmintic Drug Effect' Rumors During COVID-19
"Urgent Need to Strengthen Research Ethics and Improve Verification Systems"

"Fake clinical trials and zombie papers are rampant in the medical field."

In drug development, which deals with human lives, precise clinical trials are essential to ensure safety and efficacy. First, animal testing (preclinical) is conducted to check for side effects, toxicity, and effectiveness. Then, clinical phase 1 trials confirm safety in healthy individuals, phase 2 trials verify efficacy in a small group of patients, and phase 3 trials examine safety during long-term administration and confirm drug effects in a large patient population. Through this thorough process, people generally trust the safety and effectiveness of drugs developed. However, in reality, many clinical trials are manipulated or have fatal flaws, and 'zombie papers' based on these trials are rampant in the medical community.

Stock photo. Not related to the article.

On the 18th (local time), the international journal Nature introduced Dr. John Carlisle, a British National Health Service anesthesiologist, as a representative figure. Carlisle, who is also the editor of the journal Anaesthesia, published a startling investigation result in October 2020. He reviewed over 500 papers published in his journal over three years from 2017. Among these, he examined individual participant data (IPD) from more than 150 clinical trials and found defects such as impossible statistics, inaccurate calculations, and duplicated numbers in as many as 44% of them. Furthermore, 26% of these papers were judged to be unreliable due to author incompetence or use of fake data. Carlisle named these clinical trials 'zombie' trials. They resemble real research but are fake or have fatal flaws. He said, "I expected about one in ten, but I was shocked by how many there were," adding, "They disguise themselves as trustworthy information but are actually just hollow shells."

Carlisle also raised the practical difficulty of accessing IPD. When the original data of clinical trials cannot be obtained, he verifies the trial results using only the statistical figures revealed in summaries. In this case, about 1% were 'zombie papers' using false data, and 2% had fatal flaws. This means that journal editors and editorial board members who review and publish papers containing clinical trial results cannot verify data manipulation or falsification without access to IPD. Accordingly, Carlisle pointed out, "Journals should assume that all papers potentially have defects and begin verification," and "Editors need to review individual patient data before publishing papers that include clinical results from randomized controlled trials."

This problem is not new. According to Nature, many experts have claimed in recent years that fake or unreliable clinical trials are surprisingly widespread. Various medical fields such as obstetrics and gynecology, pain medicine, anesthesia, orthopedics, and even COVID-19 research have published results using statistically impossible data, making them untrustworthy. Nature reported, "Some scientists believe that the claim that about a quarter of clinical trials are unreliable, based on their personal experience, may be an underestimate," and quoted Professor Ian Roberts of the London School of Hygiene & Tropical Medicine, who asserted that more than one-third of randomized clinical trials on specific topics might be manipulated."

In the past, fake clinical trials and fraudulent papers were considered problems at the journal level. The so-called paper mills?numerous commercial journals?produced and published papers indiscriminately to secure profits, causing this issue. However, as clinical trial results with false data and fatal flaws have spread widely, serious problems have arisen. Not only medical side effects but also statistical analyses and systematic verification processes, which are supposed to be thorough for actual clinical treatment, can be deceptively sanitized. Many medical guidelines often cite such flawed standards, and doctors refer to them in patient treatment.

Even clinical results reviewed and published by the authoritative medical research group and international nonprofit organization Cochrane Collaboration have been found to contain errors. Professor Roberts examined a 2005 study published by Cochrane Collaboration that claimed high-dose sugar injections reduce mortality after head injury. He found that the three key clinical results cited were all led by the same neurosurgeon in Brazil. However, the surgeon committed suicide before Roberts' investigation began, so it was not confirmed whether the results were manipulated.

The most recent case involves Japanese orthopedic researcher Sato Yoshihiro. Before his death in 2016, he was found to have fabricated dozens of clinical trial results on fracture prevention drugs and supplements. A total of 113 related papers were retracted. However, 27 of these had already been cited in 88 systematic reviews and clinical guidelines. Some were even used as the Japanese government’s recommended osteoporosis treatment.

The controversy over the antiparasitic drug ivermectin’s effectiveness during the COVID-19 pandemic is similar. At that time, clinical results claiming ivermectin was effective against COVID-19 emerged rapidly. However, many of these studies were criticized for data flaws, and in 2022, the Cochrane Collaboration concluded that over 40% of these clinical trials were unreliable.

Therefore, there are calls for thorough integrity management in research sites, adherence to research ethics, and the development and refinement of research methods to ensure precision, accuracy, completeness, and validity. Carlisle said, "The problem with the solution lies in fixing the materials (data). If this issue is not corrected, it will spread rapidly like an uncontrollable wildfire."