COVID-19 as a source of poor publications

It is not possible to discuss SARS-CoV-2 without acknowledging the fact that the pandemic has set the course for many discoveries. Owing to the competition among institutions, scientific advances are being made in a hurry. The dark side of COVID-19 research exists and has been addressed, chiefly by various publications.

1. Increased scientific output during the pandemic

During the pandemic, all scientific disciplines (biomedicine, economics, social sciences, mathematics, etc.) witnessed an increase in publications. There has been significant research output related to COVID-19 itself, as evidenced by over 350,000 articles (estimated at the end of 2021) and 65,000 preprints that were published [1]. Preprints were posted on repositories, and nearly two-thirds of them were posted on medRxiv, ResearchSquare, SSRN, RePEc, and bioRxiv (Fig. 1 ). Thanks to COVID-19, the field of health sciences (medRxiv) discovered preprints thirty years after the launch of arXiv for the fields of physics and mathematics, and seven years after the launch of bioRxiv, for biological sciences. These manuscripts are shared publicly through online repositories prior to undergoing peer review. Experts from the fields of media and journalism have not always considered the repositories” warnings such as medRxiv (https://www.medrxiv.org/): “Preprints are preliminary reports of work that have not been certified by peer review. They should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information”. The rise of preprints has changed science forever [2].

Fig. 1.

Cumulative COVID-19 preprints from the beginning of the pandemic (February 2020) to May 2022. Thanks to N Fraser and B Kramer [1] (figure under CC0 1.0 license).

There has been a noticeable peak in the English language publications by Chinese authors; prestigious journals are being adapted for Chinese audience. For instance, journals such as JAMA, NEJM and Lancet included abstracts and articles written in Chinese. Furthermore, Wellcome Trust has committed to making COVID-19 related research articles freely available and accessible in public repositories. Peer review process is being accelerated. There was an increased search for “hot papers” that will receive citations quickly after publication. There has also been a decline in publication standards that was discussed by the European Association of Science Editors: “We recognise that in times of crisis it may not always be possible to obtain all required data, and that reporting may – of necessity – be curtailed. To avoid misinterpretation, but also to facilitate the rapid sharing of information, we encourage editors to ensure that authors include a statement of limitations on their research. This will inform readers and strengthen the usefulness of any published research” [3]. Too many small clinical trials have had difficulty recruiting patients, while most have not been submitted to journals or have remained in the preprint form. Together with B. Plaud and E. Caumes, we have detailed the dysfunctions and shortcomings of publications, with examples related to the French publishing scenario [4].

In the biomedical field, ten prestigious journals, such as NEJM, Lancet, JAMA, Nature Medicine, BMJ, Annals of Internal Medicine, Lancet Global Health, Lancet Public Health, Lancet Infectious Disease, and Clinical Infectious Disease were analysed between 1 January 2019 and 1 January 2021. Some journals have increased their production, while others have had stable production owing to a decline in non-COVID-19 articles [5]. An 18% decline in the published non-COVID-19 related articles has been estimated. A decline in the number of original articles in favour of letters and clinical cases has also been observed. There were 47.9% original articles in publications related to COVID-19, while 71.3% original articles got published in non-COVID-19 related publications (P  < 0.001). Moreover, an increase in the number of authors per article has been noted. For clinical cases, the median was 9.0 authors for COVID-19 publications versus 4.0 for non-COVID-19 publications [5].

2. Discovering how scientific journals work

The functioning of scientific journals has become better known since the pandemic. Today, informed citizens, journalists, and the scientific community are aware of the dysfunctions of journals as they are being better described and publicised.

Author, editors, and publishers of prestigious journals are aware of the policy of retracting scientific articles. Prior to the pandemic, public and the media largely ignored the retraction policy of journals. According to RetractionWatch, up to April 24 2022, 221 COVID-19 articles were retracted from journals [6]. Evidently, many researchers engage in questionable research practices, therefore the actual number of retracted articles will be much higher. A survey pointed out that, among 6,813 academic researchers in the Netherlands, the prevalence of fabrication and falsification were 4.3% and 4.2%, respectively, and 51.3% of respondents engaged frequently in at least one questionable research practice [7]. A journal editor retracts an article when there is evidence of fraud or honest errors. Given below are two examples of virtuous retraction (The Lancet and NEJM) and a case of misconduct by a journal (International Journal of Antimicrobial Agents [IJAA]).

An article reported the ineffectiveness of hydroxychloroquine (HCQ) based on 96,000 administrative records [8]. It was retracted within 13 days owing to the pressure from the HCQ lobbies. The arguments were admissible, although the peer review was correctly carried out in a hurry (personal contacts). The reason for retraction was not fraud but the impossibility of accessing the source data. An NEJM article evaluating cardiovascular drugs used for COVID-19, with data extracted from the same database as the Lancet article, was retracted on the same day for the same reason (no access to source data) [9]. Investigation could not be carried out on the site of the company that published the source data; hence we will never know if these articles were valid or fraudulent. Most articles do not share source data [10]. If we retract articles for which source data are not available, we would have to retract more than 50% of the original articles.

The first article on a COVID-19 treatment published by the Infectious Diseases Marseille IHU (Institut Hospitalo-Universitaire), France, was cited more than 5500 times (according to Scholar Google), without any clinical data to support the efficacy of the treatment. Later, in the same journal, a Dutch epidemiologist opined that: “As outlined below, this study suffers from major methodological shortcomings which make it nearly if not completely uninformative. Hence, the tone of the report, in presenting this as evidence of an effect of hydroxychloroquine and even recommending its use, is not only unfounded, but, given the desperate demand for a treatment for Covid-19, coupled with the potentially serious side-effects of hydroxychloroquine, fully irresponsible”. [11] The editor of IJAA, co-author of the article and an employee of the IHU in Marseille, France, failed to retract this article. Research has evidenced the manipulation of journals, based on the conduct of the Marseille IHU, dubbing them “self-promotional journals” [12]. Three criteria qualify these journals as exploitative:

• •when a well acquainted group of authors regularly contribute a large proportion of articles to a journal and cite each other's articles very often;
• •these authors maintain a good relationship with the editors of journals, oftentimes these editors happen to be authors or even reviewers of their articles;
• •publication of poor quality research. Journals, such as IJAA and New Microbes New Infections aim to serve a group of researchers and rapidly publish articles in a complacent manner. Most often, these are articles that are rejected by legitimate journals. These are known as parochial newsletters. This long-suspected practice now has been described in detail.

Scientists and media introduced the concept of authors with dirty hands [13] to refer to the paternalistic and authoritarian behaviour of researchers who impose their names on all the publications of their collaborators. They exist in all disciplines.

3. Ignorance of the principles of responsible research conduct

The principles of the Singapore Declaration on Research Integrity were released in 2010 [14]. Ignorance of the principles of responsible research conduct is common. Here, I have quoted three of the 14 responsibilities:

• •researchers should employ appropriate research methods, base conclusions on critical analysis of the evidence and report findings and interpretations fully and objectively;
• •researchers should keep clear, accurate records of all research in ways that will allow verification and replication of their work by others;
• •researchers should limit professional comments to their recognized expertise when engaged in public discussions about the application and importance of research findings and clearly distinguish professional comments from opinions based on personal views.

The principles of open science have been long ignored because data sharing remains weak. This is not going to change any time soon. A study shows that, out of the 924 COVID-19 trials conducted, only 17.3% of researchers have the intention to share their individual patient-level data [15]. Research has highlighted poor communication, numerous misconducts, and offered interpretations related to the publish-or-perish system [16], [17]. Time and again, surveys have indicated the key role of publications in highlighting the lack of responsible research conduct. Even if the researchers are given ample time for research work, the misconduct related to COVID-19 research will persist.

4. Misinformation: absence of a culture of critical thinking

Many experts have spoken outside their fields of expertise appealing to the public through messages and at times relying on preprints or articles published in predatory journals. From a young age, children should be informed that online reading is akin to reading graffiti on a wall. Things would have been different if informed citizens, journalists, and researchers had a critical mind and knew how to differentiate between facts and opinions. For instance:

• •the Marseille IHU would have had a small voice in the media, if the doubts regarding the director's production had been described and acknowledged since 2006 [18];
• •hypotheses about the possible effectiveness of HCQ would not have been listened to, if the biologists would have simply explained why it was wrong;
• •and shameful publications would not have been considered by the media if predatory journals had not published articles that assumed the efficacy of HCQ, even after they were turned down by legitimate journals.

Let us bear in mind that we live in a period in which the best and worst have been observed. The arrival of preprints in the field of medicine and data sharing will change our research behaviour for a long time to come. Misleading or fraudulent publications will leave a lasting mark on the public at the risk of increasing mistrust of science.

Disclosure of interest

Editor of a non-for-profit blog Revues et Intégrité.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.