The pandemic of sars coronavirus was introduced to the public at large the system of peer review of scientific articles, which is in theory a guarantee of quality, but it can also be a source of misunderstandings, which make the popularization of science is sometimes difficult. This system used long before the introduction of computer technology is experiencing an important development since the advent of the Internet, which has experienced an abrupt acceleration with the COVID-19, due to the urgent need of scientific results immediate. The scientific community is currently exploring methods for review of articles by peers, based on the new technologies.
Algorithms to match scientific articles and reviewers
The publisher of scientific open-access Frontiers has developed a tool for artificial intelligence (AI) to help funders, who fund the studies, to identify the right specialists, to suggest them to revise their publication. The experts are linked to articles based on keywords or semantic analysis in-depth of their publications.
This helps speed up the process of revision in the event of urgent situation, such as that related to the COVID-19. In normal circumstances, the review process of research funding usually occurs in the reading committee and may take several months. However, since the outbreak of COVID-19, the experts are less available and the emergency requires the shortest possible time. Frontiers has, therefore, put this tool in place to deal with the little time available to evaluate the proposals.
A research infrastructure is more open
Heather Joseph, executive director of the Scholarly Publishing and Academic Resources Coalition, an organization of university libraries who makes great efforts to ensure a research infrastructure that is more open, sees the pandemic as an opportunity to open the search in a way more permanent. Ijad Madisch, a virologist German, founder and ceo of ResearchGate, a social network for researchers that has seen an increase in activity and collaboration around Covid-19, considers that a restriction on publication of rejected articles due to their aspects “sensitive” or controversial might be an alternative to censorship, real or supposed.
lbert-László Barabási, for example, an eminent scientist of the network at the Northeastern University of Boston, has written an article on “a network of medicines, to identify opportunities for reuse of drugs”, which has been rejected by the scientific journal, BioRxiv. He was finally released on the platform, the ArXiv instead, but wondered on Twitter whether it would not be more logical to BioRxiv to create a list restricted to research scientists Covid-19 potentially sensitive. Madisch offers a configuration “where the research community could give its opinion prior to its publication to the public”.
How the pandemic has passed the peer review on Twitter
In the United States, the example is clear with the case of a journal article, MedRxiv the most debated up to the present, “COVID-19 Seroprevalence of antibodies in Santa Clara county, California”. The journal has reported on the results of screening tests for antibody to coronavirus at 3300 county residents recruited by advertisements on Facebook, of which 1.5% were positive. The authors then performed a number of statistical adjustments that have increased their estimate of the percentage of county residents who had been infected by the coronavirus, from 2.49% to 4.16%, which was 50 to 80 times the number of confirmed cases at the time, and it involved a mortality Rate of Covid-19 of only 0.12 to 0.2 percent – not very different from the rate usually reported for seasonal influenza (although the actual ratio of deaths by influenza infections is likely to be lower).
Some experts in infectious diseases, as the scientific become influencer during the pandemic, Trevor Bedford, turned to Twitter to criticize the skeptics. They have denounced the article as “a fake” and demanded an apology from the authors. No excuse of this type has been presented, but the authors have replaced the document on MedRxiv by a revised version, which has changed the estimate of the percentage of county residents infected with the virus by a range of 1.3% to 4.7%. They also expressed their appreciation to the many criticisms that the document had received, concluding that: “We believe that our experience offers an excellent example on how the pre-publications can be a good way to provide a crowdsourcing massive useful reviews and constructive comments by the scientific community in the broad sense in real-time for issues that are important and timely.”
The crowdsourcing of comments may be good for the science
Other scientists are not so sure that the discussion brutal – and public – around a document is so positive. In an opinion article for the news site, scientific Stat, two teachers consider that scientists should not take the comments seriously.
But, a peer review is more formal, is it really more qualitative? “For me, there is no doubt that more eyes on something means that, in the end, a better judgment can be worn,” explains Sever of MedRxiv, a molecular biologist with a long experience in writing scientific journals.
Finally, a start-up French offers an interesting initiative, a visual tool allowing researchers to find and explore scientific publications more easily than on the existing platforms, Connected Papers. The Coronavirus will not only put the light on the world of research, but will also be a witness to the initiatives and digital tools to boost search and make it more accessible.