Anna Obenauf had never posted her results to a preprint server, but she decided to make the jump in April. She was racing against another team to get findings on a rare skin cancer out quickly, so she uploaded her manuscript to bioRxiv — just like thousands of COVID-19 researchers have been doing during this pandemic. It was a turning point for Obenauf, a cancer biologist at the Research Institute of Molecular Pathology in Vienna, who particularly liked the quick feedback she received (L. Leiendecker et al. Preprint at bioRxiv http://doi.org/dw3f; 2020). She says she will probably continue to post some of her team’s work on preprint servers in the future.
The COVID-19 crisis has underlined just how fast and open science publishing can be — when scientists want it that way. Researchers working on the pandemic are sharing preliminary results on preprint servers and institutional websites at unprecedented rates, embracing the kind of early, public sharing that physicists and mathematicians have practised for decades. Journals have whisked manuscripts through to formal publication in record time, aided by researchers who have rapidly peer-reviewed the studies. And dozens of publishers and journals, including Elsevier, Springer Nature and the New England Journal of Medicine, have made coronavirus research — new and old — free to read. They have pledged to continue doing so for the duration of the outbreak, and have encouraged or, in some instances, required researchers to post their manuscripts on preprint servers.
Even before COVID-19 spread around the globe, momentum was growing to share results early online and to make work open access. The coronavirus publishing frenzy has underlined the worth of these objectives, says Cameron Neylon, a researcher on scholarly communications at Curtin University in Perth, Australia. “If we think openness of communication is valuable in a crisis, it should surely be valuable in normal times as well,” he says.
Although the experience might prod individual scientists into sharing work faster and more openly, this might not in itself lead to a publishing revolution, Neylon says. “I don’t see this as a tipping point,” he says. “The big changes that are going to come are going to be structural.”
The current system would have to shift wholesale to rewarding open, early sharing of findings to give scientists incentives to communicate their work in this way. A few funders and research institutions were already advocating that approach, and the pandemic could nudge them further along this path. But the crisis could unleash other forces that might reshape science communication: not least an economic downturn that could disrupt research budgets, job markets and the scientific-publishing industry.
Some changes in publishing are probably here to stay. Scientific communities that embrace preprints tend never to look back, says John Inglis, co-founder of the medRxiv and bioRxiv preprint sites and the executive director of Cold Spring Harbor Laboratory Press near New York City. And May was the busiest-ever month for both sites, says Inglis. (The arXiv preprint server, which hosts physics and mathematics manuscripts, still receives more papers per week, however.)
Submissions to bioRxiv were increasing even before the pandemic; an influx of coronavirus papers only partly contributed to its growth this year. But the growth in medRxiv, which is co-run with Yale University in New Haven, Connecticut, and BMJ Publishing Group in London, was due almost entirely to the more than 3,700 COVID-19 papers it hosts (see ‘Torrent of preprints’). Inglis thinks the pandemic has raised the site’s profile and that it could soon see growth in other areas of medical research. “We have a long, long way to go, but I think there is more awareness,” he says.
With the outbreak emerging first in China, it’s no surprise that preprint servers saw many more posts than before from Chinese authors, too. That change could stick — as might a growing tendency for scientists to publish in journals from Chinese publishers, as efforts to bolster the country’s science-publishing industry gather pace, says Jie Xu, who studies scholarly communication at Wuhan University in China.
Some of the results that researchers have been posting online in the crisis are more like dispatches from the front line than carefully crafted papers intended to stand the test of time, Inglis suggests. That trend might not last after the pandemic is over.
Yet scientists still have a desire to produce polished peer-reviewed work quickly. Accordingly, journals are racing to publish peer-reviewed COVID-19 papers (see ‘COVID preprints take priority’). A study posted on bioRxiv last month surveyed 14 medical journals (S. P. J. M. Horbach. Preprint at bioRxiv http://doi.org/dt3r; 2020), and found that they published papers on the coronavirus nearly twice as quickly as they did other papers at the time, largely due to quicker peer review (see ‘Rapid review’). It’s unlikely that journal-based peer review could regularly work at this pace, says Stefano Bertuzzi, chief executive of the American Society for Microbiology in Washington DC. “I don’t see room for efficiency improvement in that respect on a regular basis. I think this is just the emergency situation that we’re dealing with,” he says.
But the crisis has inspired experiments in review that might persist. Some journals and publishers, including PLOS, eLife, the UK Royal Society and Hindawi, have launched an initiative to create a pool of scientists who are willing to rapidly review papers on COVID-19, as well as to share reviews between journals. An effort called Review Commons, launched in early December 2019, allows scientists to have their manuscripts reviewed even before being posted as a preprint. The manuscript and its reviews appear together on bioRxiv and are then submitted to a participating journal.
Some researchers have taken the initiative to curate preprints themselves. Scientists at the Precision Immunology Institute at Icahn School of Medicine at Mount Sinai in New York City, for instance, have reviewed dozens of bioRxiv and medRxiv preprints, posting their assessments alongside the manuscripts. Efforts such as these show that it is possible to do thorough peer review outside journal-organized mechanisms, says Inglis — which might be a theme of science publishing’s future.
Many experts say some of the biggest impacts on the scientific-publishing industry are likely to be financial, and will play out over years. If economies continue to nosedive, the budgets that support the scholarly publishing enterprise will come under pressure.
“It’s really hard to see the big publishers not facing a very substantial revenue hole,” says Neylon. Institutions might try to cancel or renegotiate contracts with publishers, he says. (In March, the non-profit organization Jisc in Bristol, UK, which negotiates contracts on behalf of British university libraries, asked publishers to delay or minimize subscription increases, and to take other measures to cushion the blow to university budgets.)
Neylon thinks that university presses, learned societies and other small publishers will be under the most pressure. And Joseph Esposito, a senior partner at publishing consultancy Clarke & Esposito in Washington DC, wonders whether an economic downturn could even slow growth in open-access publishing. “We might find that some of the heat is going to go out of the open-access movement, and that you’re going to find fewer journals, fewer articles being written, and greater attempts in a tight job market to publish in highly regarded publications,” he says.
But Robert-Jan Smits, president of Eindhoven University of Technology in the Netherlands and architect of the European-led ‘Plan S’ for open-access publishing, thinks that the coronavirus crisis will be looked back on as the event that tipped science in general towards fast, open publishing. “It’s the final push that is necessary,” he says.
Nature 582, 167-168 (2020)