The Danger of Scientific Fraud

A recent article in The Economist highlighted the increasing level of fraud found in scientific literature.¹ It detailed the concerns of Ben Mol, MD, PhD, a former editor of the European Journal of Obstetrics and Gynecology. He and a group of colleagues recently reviewed the literature in their areas of expertise and found 750 previously published papers with significant flaws. Of that group, only 80 have been retracted.

To add to the problem, a number of these papers had already been included in systematic reviews, thus perpetuating the use of their misinformation. An example cited by Mol was a systematic review on the efficacy of steroid injections given to women who had elective Cesarean sections. These injections were said to prevent breathing problems in newborns. A drawback of this technique was that it could cause infant brain damage. Unfortunately, before the flaw in the original paper was discovered, it was included in a review published in 2018.

When Mol’s group evaluated the review paper, they found it included three studies they deemed unreliable. A revised review on the subject published in 2021 that excluded these three articles found the benefits of this approach to be questionable.

Mol and his colleagues are not the only ones who have found problems in papers published in refereed journals. The Economist article reported that an online database, Retraction Watch, found almost 19,000 papers that had been retracted. The reasons ranged from suspected fraud to plagiarism as well as conflict of interest. Other editors and reviewers have found an increasing number of questionable studies. Unfortunately, some papers are never retracted and those that are can take months or years to be rescinded. The article estimated that one in a 1,000 papers is retracted. The editors of Retraction Watch suggest it is closer to one in 50.

The Economist article suggests that, in some cases, these fraudulent papers have resulted in direct harm to patients. An example was the use of beta blockers before surgery, which was intended to reduce the number of heart attacks and strokes. This practice was a result of a single study published in 2009. It was eventually determined to have been based at least in part on fabricated data. One estimate was this approach had caused at least 10,000 deaths a year in Britain alone.

According to The Economist, another significant problem was found in a 2006 article published in Nature concerning the relationship of Alzheimer disease to amyloid plaques. This one article resulted in leading research in the wrong direction by indicating that a particular form of the amyloid beta was the plaque-forming protein. In July 2022, Nature published an expression of concern while it investigates these data.

It is always important to read the scientific literature with a somewhat jaundiced eye. Paying attention to conflicts of interest and to the accuracy of the bibliography is strongly suggested. With the advent of artificial intelligence, we can assume that falsification of data will increase.

Thomas G. Wilson Jr., DDS
Editor in Chief
twilson@belmontbusinessmedia.com

Reference

1. There is a worrying amount of fraud in medical research. The Economist. Available at: economist.com/science-and-technology/2023/02/22/there-is-a-worrying-amount-of-fraud-in-medical-research. Accessed May 9, 2023.

From Decisions in Dentistry. June 2023;9(6):6.