Corruption of the medical literature is impossible to prevent

Medicine improves regularly. Our tests, our surgeries, our medicines are better now than they were 10 years ago. The main driver of this improvement is the scientific method. Using this, the worse is discarded and the better retained. Doctors have to work hard to keep up with the progress. They rely on impartial and accurate information transmitted by articles, reviews, and CME material in order to advise and treat their patients.

To help assure its readers that presentations are unbiased, journals insist on full disclosure from its authors. Authors have to state whether they actually wrote the article. CME presentations require its readers to state if the presentation appeared biased and if so in what way.

This article argues that there can be no certainty that information is unbiased, fair and accurate, because the individual physician will not be able to detect two common deceptions.

The first is to omit key facts deliberately. The doctor won’t know what was left out. You can’t know what you aren’t told. One example of that is Bombardier et al.’s rofecoxib (Vioxx) study published in the NEJM. Merck and its ghostwriters left out the fact that some patients had had heart attacks. When the editors looked at the CD containing the article, they found it was in Microsoft Word, which contains prior versions, as well as the finished article. Prior versions showed that the sentence mentioning the heart attacks had been deleted. Now, pharmaceutical companies and their ghostwriters are smart enough to send articles without including prior versions, so that discovery of this deception, which was embarrassing and costly for Merck, will not occur again.

While the first deception assumes the person transmitting information is corrupt, the second assumes that the data are corrupt. The person transmitting the information may be above reproach. How can this happen? It can occur several ways: (1) Negative studies can be discarded, with only the positive ones published; (2) Studies can be done with comparison to placebo, not to best known treatment; (3) Studies can be ghostwritten by companies paid by the study sponsor; (4) Editors can be biased by ad revenue from drug companies, the main revenue source for most journals; and (5) An editor may be a paid investigator for drug company. This is probably not an exhaustive list.

This type of deception is a major problem. It can even affect formal guidelines based on an evidence-based review of the literature, because the literature has been corrupted by companies who are quite rationally, from their point of view, trying to portray their products in the best light.

Because the companies have an incentive to portray their products favorably, these deceptions will continue. They are difficult to detect. In fact, once they are uncovered, the rational response of the company is to invent new ways to avoid detection. Finally, as far as I know, corruption of the medical literature is impossible to prevent.

Bradley Evans is a neurologist. 

Submit a guest post and be heard on social media’s leading physician voice.

View 4 Comments >

Most Popular