The ongoing COVID-19 pandemic has brought unprecedented disruption to almost every area of life, including science and scientific publishing. Many researchers have seen their projects interrupted, while others have re-invented themselves as coronavirus researchers.

The general public are taking an unprecedented interest in the latest research findings. Complicated aspects of epidemiology, diagnostics, pathogenesis or therapy that are normally left to experts operating outside the limelight, now trigger controversial discussions on social media platforms.

At the same time, public health policymakers urgently need scientific findings. Measures to control the pandemic or to mitigate its consequences must be implemented in real-time to respond to a rapidly unfolding situation for which no-one has a “recipe book”, as the world has never before faced a pandemic caused by an infectious agent with the characteristics of SARS-CoV-2.

The enormous need for relevant data, a flood of COVID-19-related funding opportunities and offers of expedited editorial and review processes by scientific journals, have resulted in over 40,000 papers on COVID-19 within just nine months!

The resulting sense of urgency, and the unfortunately common politicisation of public health measures like universal wearing of non-medical (cloth) masks, re-opening of schools etc., are re-shaping the manner in which scientific knowledge is shared.

While there are many pressures and incentives to be quick, speed comes at the expense of thoroughness. It has become routine to disseminate scientific data before they have been peer-reviewed by experts in the field. The peer-review process is one of the cornerstones of the scientific process – despite its many flaws, it is still regarded as the gold standard as manuscripts are subjected to the scrutiny of experts (who may reject it altogether, or ask for more or less profound modifications and improvements) before they become public.

A deluge of manuscripts related to the pandemic is being uploaded onto pre-print servers for health sciences and biology  ̶  as of 30 September 2020, 9178 articles on COVID-19 and SARS-CoV-2 are on medRxiv and bioRxiv alone. While some stimulate helpful discussions among experts, many are picked up by journalists and others and widely reported, oblivious to the fact that they still need to pass through the “safety net” designed to ensure the quality and integrity of science. Premature application of alleged yet unsubstantiated findings can be harmful.

Even worse – some papers are accepted and published by reputable journals despite serious shortcomings. Certain drugs have been touted as beneficial in COVID-19 patients despite a lack of proper evidence. Preliminary observations may be useful if subsequently put to test in proper randomised controlled trials. If they however become political currency, in pre-print and later published form, this can cause damage, as when President Trump praised chloroquine as a remedy against COVID-19.

Subsequently, the prestigious journal Lancet published a paper claiming the opposite: an analysis of a global hospital patient registry had shown that not only did chloroquine not help but that it led to higher mortality among COVID-19 patients. However, this “study” is equally invalid: There are serious doubts about quality or even existence of the underlying database, leading to the eventual retraction of this and a related scientific paper.

These are examples of failures of two of the components of scientific quality assurance, namely editorial oversight and peer review. The subsequent layers, critical reception by peers and replication by other studies, exposed the major flaws of these papers. This proves that fundamentally, the scientific process and its self-correcting nature are intact. However, even temporary aberrations can cause harm – by leading to poor decisions (for example, suboptimal treatment of patients or insufficient prevention of infection) or “only” by distracting from real issues and better alternatives.

The ongoing pandemic places an enormous burden on editors, reviewers and just about everyone else in the scientific community. Those best placed to provide meaningful peer review on submitted manuscripts are probably the same people who are themselves trying to obtain funding and ethics approvals, conducting trials, analysing data and writing manuscripts and thus do not have time to undertake peer reviews.

On the other hand, “informal” peer review may yield unexpected benefits, even before a paper is submitted to a journal. An example from the early 2020 is the “disappearance” of a pre-print paper falsely claiming that the SARS-CoV-2 genome contained elements from the genome of HIV. Vigorous “open” peer review, taking place on science blogs and Twitter, seems to have prompted the withdrawal of the manuscript.

A rational approach will go a long way, especially if supported by some knowledge of the field in question. An example: Most antiviral drugs are the end product of painstaking research conducted over decades, starting with basic virology and culminating in clinical trials. Does that not make it highly unlikely that a decades-old antimalarial like chloroquine would have major, hitherto unrecognised antiviral activity? Or major toxicities not seen in decades?

Fortunately, the scientific endeavour remains largely intact. Even during the pandemic poor (or even fraudulent) papers will be called out in the end. But to maintain or restore public trust in science, scholarly publishing needs to ensure high quality peer review and be prepared to sacrifice high scores on the public relations side in the interest of maintaining the integrity of science.

Prof Wolfgang Preiser heads the Division of Medical Virology in the Faculty of Medicine & Health Sciences at Stellenbosch University (SU). Dr Rika Preiser is a Senior Researcher at the Centre for Complex Systems in Transition at SU. This article is an abridged version of their recent paper Academic publishing in pandemic times in the South African Journal of Science.