From Hydroxychloroquine to ACE Inhibitors: Understanding Retracted COVID-19 Studies and the Importance of Scientific Integrity
Scientific articles serve as critical cornerstones in shaping public health policies and implementing treatments for a myriad of diseases. Their influence extends to policy-making decisions and everyday life for individuals. Additionally, they play an instrumental role in enriching the knowledge base of healthcare professionals, and are considered fundamental building blocks of scientific research.
Reasons Articles May Be Retracted
Nevertheless, there are times when scientific articles need to be retracted, as they may contain errors or inaccuracies that impact their reliability and credibility. Retraction serves as a crucial component in a larger system of checks and balances within scientific research, aimed at preserving the integrity and veracity of published studies. There are several reasons why an article may be retracted, ranging from plagiarism and falsification of data to negligence or even an inadvertent mistake. In cases of an honest error, the policies and procedures involved in retraction may differ from those of deliberate misconduct.
A review of 2047 retracted articles with retraction notifications in PubMed from 1977 through 2012 stated that 21% of the retractions were attributable to error, the majority (67%) were attributable to fraud or suspected fraud, 10% were attributable to plagiarism. Duplicate publication and miscellaneous or unknown reasons accounted for the remaining retractions.
Examples of Retracted COVID-19 Articles
As of March 18, Retraction Watch listed 303 retracted COVID-19 articles. Perhaps the most famous example was The Lancet’s hydroxychloroquine study, which was published in May 2020. It quickly gained media attention for its information on the effectiveness of hydroxychloroquine. Several concerns were raised about the study, chief among them being the fact that the data came from a private company that did not allow for access to the raw data or an independent audit of the findings.
As a result, a group of scientists wrote a letter to The Lancet, asking that the data be reviewed. The Lancet then issued an expression of concern. In June 2020, The Lancet issued a retraction, informing the scientific community that "serious scientific questions have been brought to our attention." The authors of the study also issued a statement, saying that they could no longer vouch for the veracity of the data used in the study.
This retraction signaled a turning point in public opinion about treating COVID-19 with hydroxychloroquine. This retraction also emphasized the essential nature of scientific rigor and transparency in the midst of a global health crisis. In light of the retraction, The Lancet also retracted and replaced an editorial about the study.
Another example of a retracted COVID-19 article was published in the New England Journal of Medicine (NEJM) in March 2020 and was about the effects of angiotensin-converting-enzyme (ACE) inhibitors and angiotensin-receptor blockers (ARBs) on COVID-19. The article suggested that people taking these medications might be at higher risk of contracting COVID-19.
The scientific community questioned the findings, citing the very small sample size and that the study lacked sufficient statistical power to draw any conclusions. Experts also argued that ACE inhibitors and ARBs might actually help people with COVID-19 because the medications are known to have cardioprotective effects.
In April, a correction was issued, cautioning the public to interpret the results carefully and that more research was needed. But, in June, the article was retracted altogether by NEJM. There was concern about the validity of the data. This also highlighted the importance of scientific rigor and transparency during the COVID-19 pandemic.
Retraction and Replacement at the JAMA Network
I work at the JAMA Network, and there is a very clear policy about retractions in the AMA Manual of Style. The convention for retractions that qualify as honest mistakes follow the guideline of retraction and replacement. In the manual, it states that retractions should only be used in the case of pervasive errors. For typical errors, a correction should be issued. When a retraction is issued, the following must be included, according to the AMA Manual of Style:
Complete explanation for how the errors were discovered
An itemized listing of all errors and corrections
A marked-up copy of the original article identifying all errors and corrections (Note: A tracked-changes version of the original text of tables and figures is helpful)
Confirmation that there are no additional errors
Indication of whether the errors change the statistical direction of the results, interpretations, and conclusions
Letter of explanation summarizing all the above from all authors to be considered for publication
The journal editor then takes another look at the article, the author explanation, and the marked up copy. They identify the errors and corrections either on their own or (more likely) with the help of external peer and statistical review. If the editor, after doing all this, decides the scientific findings are still reliable and important, they will issue a retraction and replacement. A formal notice from the authors will be published in the journal.
According to the AMA Manual of Style, “The JAMA Network journals and The Lancet journals have published such notices and have replaced or republished the original article along with online supplements that include a version of the original retracted article showing the original errors and a version of the replacement article showing what was corrected. This option provides authors who have made inadvertent pervasive errors that have resulted in changes to results of published articles with a mechanism to retract the erroneous article without the “do not use” stigma associated with retractions that are reserved for acts of fabrication, falsification, and plagiarism. The ICMJE supports this option.”
Scientific Rigor and Transparency Are Key
Again, scientific articles are crucial to our understanding of the world around us. They effect everything from what medications we take when we get sick to our basic understanding of what is good for us and what is not. The system of checks and balances in the scientific community allows for science to be verified, and if it cannot be verified and vouched for, then an article may be retracted. Retractions allow for science to prevail and transparency to be promoted.
The reason scientific journals are peer-reviewed is to prevent errors that would result in a retraction. The lines of communication with journal staff and editors should remain open, and people should feel free to examine the veracity of data to the extent that scientific rigor and transparency are being promoted. As researchers, editors, and publishers, we have a responsibility to the public to inform them with accurate, reliable data that can be openly verified.