The impact of COVID-19 on the publication practices of researchers and journals

  • Hervé Maisonneuve
  • Univadis Medical News
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.

While confined in the Alps (at -10º C tonight), my first thought is of all the professionals involved in the fight against SARS-CoV-2. I've had calls from colleagues, including ICU physicians, describing situations I've never seen before. Some are rightly upset with those who take the time to write manuscripts and submit to journals: there is no sense of urgency with these endeavors.

The role of scientific journals is to disseminate validated data to the scientific community, journalists and informed citizens. The journals analyze articles submitted by researchers before deciding whether or not to accept them. This evaluation is done by colleagues known as reviewers: the commonly-spoken ‘peer-review’ which generally takes 4 to 6 weeks to be done properly.

A pandemic is normally accompanied by the rapid appearance of scientific journal articles. This was the case with HIV, Ebola, SARS. Is the dissemination of information about SARS-CoV-2 different from that of other pandemics? We will know later, because journals have changed a lot in the past 10 years due to technology and new economic models. The race to publish continues during a pandemic.

How authors have been behaving:

  • Since January 2020, in less than three months, several thousand articles have been published. And take care with these numbers, because literature published in Chinese may elude us.
  • In a simple, less qualitative way, with a high margin of error (March 26, 2020), the single search term 'Coronavirus' for the year 2020 returns 1,483 hits on PubMed and 7,790 on Google Scholar...Searches with the words 'COVID-19' or 'SARS-CoV-2' give other results and numbers.
  • Preprints (articles posted on a platform in an open archive) are numerous, despite the resistance to using such means in medicine, generally. These are manuscripts that have not been peer-reviewed, pending publication. As of March 26, 2020, there are 765 of them, 574 on medRxiv and 191 on bioRxiv). These manuscripts will probably be published later; some are already published, others will remain as preprints.
  • Many articles have only Chinese authors. In other pandemics, Chinese authors often signed in collaboration with other international teams.
  • Articles published in Chinese circulate only when they have an English summary, which is too simple a view of the work.

The practices of scientific journals:

  • Prestigious journals have all created spaces dedicated to COVID-19 articles, and in general, these articles are all open access. Some journals have podcasts including interviews with experts. It is not necessary to be a subscriber or to pay for temporary access (paywall).
  • Two basic research journals (Nature and Science) and three medical journals (JAMA, NEJM, Lancet) seem to be leading the way. Some prestigious journals have added abstracts in Chinese to their articles.
  • Urgency seems to take precedence over quality. Competing journals are looking for “hot” papers. The “hot” paper, whatever its contents, attracts citations. It increases the impact factor that measures the journal’s notoriety. The journal’s editors assess the risk of publishing a not very good article: it will be cited, especially by those who say that the articles are bad.
  • Journals promise a quick peer-review, sometimes even within 48 hours (or less). Some journals have issued calls for reviewers. What is the quality of this peer-review, and how do authors who are in a hurry respond to requests from reviewers? This accelerated peer-review is called fast-track, and it is usually used for exceptional articles.
  • The NEJM says it receives 40 manuscripts a day, and accepts 2% of them. Prestigious journals generally accept 8 to 11% of the submitted manuscripts. The editors of JAMA have discovered the same patients described in different manuscripts, a practice that does not meet the ethical standards of the publication.
  • Groups such as The Lancet have a COVID-19 Resource Centre with the compilation of COVID-19 articles from the group's 19 journals. Similarly, the Cell Press Coronavirus Resource Hub has a Chinese version in addition to the English version, and compiles articles from about 50 journals.
  • Specialty journals, often under the aegis of scholarly associations, also seek to attract “hot” papers. They have a few articles. Later, they will receive higher quality papers that the prestigious journals will no longer accept because the trends have changed.  

A resource that categorizes articles by research area

An English site analyses the COVID-19 literature overall and regularly updates its observations. This effort by the University College of London is great, except that the inclusion criteria are not explicit enough. As of March 26, 2020, there are 1,761 articles for which titles, names and affiliations of authors are cited with or without abstracts. Links to the articles are given. This shows that 1,149 articles do not contain primary data (n=987) or are concerned with other viruses (n=162)! These are diverse opinions, editorials and reflections...even literature reviews or meta-analyses already published! Among the 612 articles containing data, it is not surprising to find 142 clinical cases, 130 articles on transmission and risks, 76 articles on diagnosis, and 70 articles on genetics/biology. I randomly browsed through a few articles and found that the conclusions were often of the type: “These preliminary findings need further research”, or “Complications are seen mostly in the elderly”, or “No one is allowed to go out”.

It will take time to understand the changes impacting the work of scientific authors and journal editors. That some things are wrong is the only certainty I have when analyzing the dissemination of scientific data. I am not convinced that speed and haste in publishing is an indicator of the quality of publications.