COVID-19 research update: How many pandemic papers have been published?

A briefing on developments in coronavirus research publishing.

  • Nature Index

Jinqing Liu, a research technician, adjusts his mask in the sealed-off lab at the Ragon Institute of MGH, MIT, and Harvard.
Credit: Boston Globe / Contributor / Getty

COVID-19 research update: How many pandemic papers have been published?

A briefing on developments in coronavirus research publishing.

28 August 2020

Nature Index

Boston Globe / Contributor / Getty

Jinqing Liu, a research technician, adjusts his mask in the sealed-off lab at the Ragon Institute of MGH, MIT, and Harvard.

Number of coronavirus articles reaches over 23,500 on major databases

28 August 2020:

In just six months, major databases have been flooded with research articles, letters, reviews, notes, and editorials related to COVID-19.

A new study estimates that 23,634 unique published articles have been indexed on Web of Science and Scopus between 1 January and 30 June 2020.

Jaime Teixeira da Silva, an independent researcher based on Kagawa, Japan, and researchers at Thompson Rivers University, Canada and the University of Malaya, Malaysia analysed 12,331 coronavirus articles on Web of Science and 12,602 papers tracked by Scopus that had been published in the six-month period.

Research papers accounted for 48% and 37% of all COVID-19-related articles on Scopus and Web of Science, respectively. Letters, reviews, editorials, and notes made up the rest of the share of coronavirus articles on each database.

In both databases, the United States, China, and Italy were the leading countries by publishing volume, while BMJ, Journal of Medical Virology, and The Lancet published the most papers related to the coronavirus.

The findings were published in Scientometrics.

Data reported by countries remains patchy

19 August 2020:

There has been no shortage of coronavirus research since the beginning of the pandemic, but inconsistent data sharing and formatting could hinder global efforts to tackle the outbreak.

A new study has found that while most countries are keeping track of total confirmed cases and deaths, they differ widely in how they report other types of data, such as the number of tests conducted and patient demographics.

Jonathan LoTiempo, a PhD candidate at the George Washington University in Washington D.C., and his colleagues analysed data reported by public health agencies in 15 countries with the highest COVID-19 burden as of 23 April 2020, which included the United States, Spain, Italy, and France.

Data was presented in plain text, HTML, or PDF, with 11 of the 15 offering an interactive web-based data dashboard and seven making data available for download as comma-separated files. These formats are not compatible, and there was also a lack of information about where the source data was archived, making it difficult for researchers to access, the authors said.

The findings were published in Science & Diplomacy.

Depression and anxiety spikes in postgraduate students

18 August 2020:

The rate of major depression among postgraduate students has more than doubled during the pandemic, according to a new survey of more than 45,000 students in the United States.

Led by Igor Chirikov at UC Berkeley’s Center for Studies in Higher Education, the team surveyed 30,725 undergraduates and 15,346 postgraduates across all subjects at nine public research universities in the United States between May and July 2020.

Doctoral researchers had the highest rates of major depressive disorder (43%) and generalized anxiety disorder (36%) among the study participants.

Both disorders were more common among women, caregivers, students of colour, non-binary and LGBTQ students, and those from a low-income background.

The report recommends that universities devote more resources to mental health services and training to help academic staff identify mental health issues in students.

Fast-tracked studies riddled with conflicts of interest

14 August 2020:

The pandemic may have accelerated the adoption of fast peer review, but a new preprint raises concerns that conflicts of interest are hindering fairness and transparency.

Lonni Besançon at Monash University in Melbourne, Australia and his colleagues analysed 12,682 COVID-19 articles indexed on the PubMed biomedical and life sciences literature database. They found that 8% had been reviewed and accepted for publication on the day they were submitted, suggesting that in some cases, the peer-review process had been rushed.

The researchers identified editorial conflicts of interest in 43% of these papers, including authors who were also editors at the journals they submitted to.

The findings highlight concerns about the fairness and transparency of the peer-review process when such short acceptance times are in play, the researchers say.

“Fast-tracking of peer-review should therefore only be done when scientific rigour can be maintained, as its loss might lead to disastrous consequences for patients and for public health as a whole,” they say.

The preprint, which has not been peer-reviewed, was posted on bioRxiv.

How the pandemic is impacting early- and mid-career researchers

12 August 2020:

Almost 90% of early- and mid-career researchers say they have experienced a huge disruption to their work due to the pandemic, which is severely affecting their productivity and mental health.

That’s the finding of a survey of 333 early- and mid-career researchers in Australia conducted by the Australian Academy of Science’s Early- and Mid-Career Researcher (EMCR) Forum.

The survey found that 46% of the respondents had carer responsibilities. Competing priorities, interruptions, extended work hours, and home-schooling tasks have had the greatest impact on their mental health and productivity during lockdown.

The EMCR Forum made several recommendations, including fellowship and contract extensions, reducing full-time work hours for carers, and extending JobKeeper subsidies to universities and other STEM workplaces. JobKeeper is a temporary subsidy paid by the Australian government to businesses that have been significantly affected by coronavirus.

In a press release, Michael Bowen, chair of the EMCR Forum, said that without rapid and continued support from employers, funders, and the government, there will be a “substantial brain drain and lost future capacity and capability to provide solutions to future challenges, such as the next pandemic”.

Most researchers expect to reuse old data during lockdown

7 August 2020:

The majority of researchers who are locked out of their labs due to the pandemic expect to reuse their own data over the next 12 to 18 months, according to a new survey of more than 3,400 researchers from around the world.

The widespread reuse of old data raises concerns that questionable practices such as salami slicing - writing multiple, even contradictory, papers based on the same dataset.

A team at Digital Science and Springer Nature (Nature Index’s publisher), surveyed 3,436 academic researchers between 24 May and 18 June 2020 on their attitudes towards open data sharing and research data management.

More than 60% of the respondents expected to reuse their own data in the coming months and 51% said they were likely to reuse data collected by other researchers before lockdown measures came into effect.

Roughly one-third of academic researchers in the survey said that their research had been “extremely” or “very” impacted by the pandemic, which is a higher percentage than those working in professional settings.

Researchers working in chemistry, biology, medicine, materials science have been the most severely impacted by the pandemic, according to the survey, while those working in social sciences and the humanities were the least affected.

Majority of early COVID-19 articles provide no new information on the disease

6 August 2020:

Original findings on the COVID-19 pandemic are being crowded out by reviews, opinion pieces, and editorials, slowing down our understanding of the disease.

That's the conclusion of a recent study, which compared the scientific articles published during the first three months of the coronavirus pandemic to articles published during the 2009 H1N1 swine influenza pandemic. The articles were retrieved through the Medline database by searching PubMed for the terms “COVID-19” or “COVID”.

Led by Nicola Di Girolamo from the Center for Veterinary Health Sciences at Oklahoma State University, the study found that roughly half (50.6%) of the 1,165 COVID-19 papers they analyzed were secondary articles, such as reviews and guidance articles. This means that only half of the publications offered original data, for example, from human medical research or in silico and in vitro studies.

The study found that the proportion of secondary articles is higher and the proportion of studies offering original data is lower than during the H1N1 pandemic.

"Compared to the H1N1 pandemic, the majority of early publications on COVID-19 does not provide new information, possibly diluting the original data published on this disease," the authors state.

The study also found that only a few published articles (3.6%) reported limitations in the abstracts, such as the potential for selection bias and the need for more evidence, which could be driving an overemphasis of the article findings and recommendations.

The study was published in Scientometrics.

Just 29% of COVID-19 studies meet 'gold standard'

27 July 2020:

The number of registered clinical trials is growing by the day, but how many of these studies present solid evidence?

According to a new analysis, just 29.1% of COVID-19 studies tracked by the global register have the potential to produce high quality evidence.

Additionally, only 29.3% of registered clinical trials were blinded and placebo-controlled – considered the ‘gold standard’ for epidemiological research.

Led by Krishna Pundi at Stanford University’s School of Medicine in California, the team analyzed 1,551 COVID-19 studies that were registered up to 19 May 2020 on

“Even before results are known, most studies likely will not yield meaningful scientific evidence at a time when rapid generation of high-quality knowledge is critical,” the team concludes in their paper, published in JAMA Internal Medicine.

Meanwhile, enrolment in US-based clinical trials is showing signs of improvement, according to a report released by Medidata, a healthcare data services company.

Lockdown and stay-at-home orders have caused a sharp decline in the number of new patients enrolling in non-COVID-19-related clinical trials across the country.

In April 2020, clinical trial enrolments had dropped by 70%. By June, enrolments had picked up again, down by just 38% from pre-pandemic levels.

US states’ reporting on COVID-19 cases “abysmal”

21 July 2020:

Most US states are reporting incomplete data on the progression of the pandemic, according to a report by the non-profit, Resolve to Save Lives.

Led by Tom Frieden, former director of the US Centers for Disease Control and Prevention, the report outlines 15 essential indicators that should be reported by all states, such as new confirmed and probable cases, and the percentage of hospital beds occupied by COVID-19 patients.

Just 2% of these essential indicators were reported in full by US states, the District of Columbia, and Puerto Rico, while another 38% were only partially reported. Contact tracing and testing accounted for the majority of the missing data.

“State reporting on contact tracing is abysmal,” the authors wrote. “Only eight states report data on the source of exposure for cases, which is determined during case interviews.”

New app pinpoints COVID-19 hotspots, symptoms

21 July 2020:

A literature review has identified seven key artificial intelligence applications for the pandemic. These included early detection and diagnosis of infection, treatment monitoring, contact tracing, and development of treatments and vaccines. The review was published in Diabetes & Metabolic Syndrome: Clinical Research & Reviews.

Among a wave of new tools using artificial intelligence approaches to analyze massive COVID-19 datasets is the COVID Symptom Study app, which allows almost four million users to report the symptoms they experience each day, reports Emma Yasinski for The Scientist.

Led by Andrew Chan, physician and epidemiologist at physician and epidemiologist at Massachusetts General Hospital in Boston, researchers are using the data gathered by the smartphone app to pinpoint COVID-19 hotspots, new symptoms, and to help inform plans for future outbreaks.

A study published in Nature Medicine in May found that a loss of taste and smell was a reliable predictor COVID-19 infection. The study used data gathered by the COVID Symptom Study app.

How a computer scientist detects hundreds of duplicated images in coronavirus preprints

21 July 2020:

As the deluge of coronavirus preprints elevates concerns about poor quality research, computer scientist Daniel Acuna at Syracuse University in New York has applied a software tool he developed to extract and compare some 21,000 images in 3,500 preprints posted on medRxiv and bioRxiv, Richard Van Noorden reports in Nature.

Applying algorithms that can identify matching images in tens of thousands of papers at a time, Acuna’s tool in June 2020 identified around 400 images that were potentially duplicated within four hours, and flagged 24 papers on a website he created and flagged them on PubPeer, a review and discussion platform.

Acuna’s findings led to a flurry of discussion, with some researchers questioning the accuracy of the tool and others arguing that the images still needed to be checked by human eyes to ensure they are not being falsely labelled as duplicated.

“I still think it is useful because it is picking up things that would be hard for humans to catch,” Acuna told Nature.

Survey reveals gaps in coronavirus research

15 July 2020:

Thousands of studies and clinical trials are underway to find vaccines and treatments, but research on detection, prevention, and care is still lacking, according to a survey of over 4,000 researchers in 130 countries.

Led by Trudy Lang, who studies global health at the University of Oxford in the UK, the respondents reported that there needs to be a stronger focus on the effectiveness public health measures - such as handwashing and social distancing - and access to maternal care.

The survey also found that research on public trust, mental health interventions, and social issues, such as domestic violence, was being neglected.

“The gaps that came up were familiar from my experience fighting Ebola and Zika: they concerned community health care, case detection and public communications,” Lang wrote in Nature. The report was shared on The Global Health Network.

Travel bans lead to less accurate weather forecasts

15 July 2020:

The inadequacy of appropriate data from which to draw firm research conclusions has been a recurring theme since the start of the COVID-19 pandemic. The problem now extends to weather research.

In a twist, it turns out that travel bans to slow the spread of the virus have meant that up to 75% of the meteorological observations normally collected by commercial flights are no longer being made, resulting in less accurate weather forecasts.

Ying Chen at Lancaster University in the United Kingdom compared the accuracy of global weather forecasts between March and May 2020 with the same months in 2017-19. Chen also analyzed forecasts made in February this year before the pandemic intensified.

While weather forecasts were more accurate in February 2020 than previous years, temperature, relative humidity, windspeed, and air pressure measurements collected by commercial aircraft were less accurate between March and May than previous years.

Weather forecasts were hit hardest in busy regions, including North America and southeast China. Western Europe was an exception, with grounded flights only having a small impact on weather forecasting accuracy. This is likely due to the high number of meteorological stations spread throughout the region, the author writes.

“It’s a good lesson which tells us we should introduce more observation sites, especially in the regions with sparse data observations,” Chen said in a press release. “This will help us to buffer the impacts of this kind of global emergency in the future.”

The findings were published in Geophysical Research Letters.

Using informatics as a tool to guide COVID-19 decision-making

13 July 2020:

Raw counts of cases and deaths are essential for tracking the progression of COVID-19, but how useful are they for public health decision-making?

A new study applies biomedical informatics to produce data visualizations aimed at providing a clear picture comparing the effectiveness of different pandemic mitigation across different states in the US.

Jay Ronquillo, founder of bioinformatics company Grinformatics in Haymarket, Virginia, and his team analyzed all COVID-19 cases and deaths in the US by state up to 24 April 2020. They compared cases with each state’s population, resource, and statewide intervention data.

As of 24 April 2020, the US reported 903,696 COVID-19 cases and 51,859 deaths, with 42 states and Washington, DC implementing statewide stay-at-home orders for all residents between 15 and 27 days after their first reported case.

States without stay-at-home orders had the highest cases and deaths per million, indicating the importance of implementing “responsible mitigation strategies focused on protecting their most vulnerable populations", the authors write.

COVID-19 cases (a) and deaths (b) per million residents, and cases per hundred doctors (c) and ventilators (d) by state and quartile, as of 24 April 2020. Jay G Ronquillo et. al

The findings were published in the Journal of Public Health.

Tracking COVID-19 around the world

6 July 2020:

Wealthy countries hit the hardest by the pandemic produce more COVID-19 research than countries with low infection rates, according to a new analysis.

Led by Xunzhi Zhu at the Chinese Academy of Sciences’ Institute of Botany, the team gathered all COVID-19-related papers on PubMed published between 21 January 2020 and June 15 2020. They compared the number of papers with the number of cases in each country.

Between 21 January 2020 and June 15 2020, 11,859 papers from 157 countries had been posted on PubMed. The United States, China, Italy, and the United Kingdom had the highest number of cases and also accounted for the majority of COVID-19 papers during the study’s timeframe.

Belgium and Spain were the quickest to respond, publishing their first papers before their first COVID-19 cases were confirmed.

The findings, which have yet to be peer-reviewed, were posted on the preprint server, medRxiv.

A revised paper changes its conclusions

7 July 2020:

In May, a preprint concluded that a mutated version of the SARS-CoV-2 virus (which causes COVID-19) is more infectious than the original virus that emerged in Wuhan, China in late 2019.

Led by Bette Korber at the Los Alamos National Laboratory in New Mexico, the study found that a mutation in the virus that causes an amino acid change in its spike protein was “more transmissible” and was of “urgent concern”.

The bioRxiv preprint drew a lot of attention, and has been mentioned by almost 7,000 tweets and 417 news outlets to date.

After criticism from the research community, citing a lack of experimental evidence, among other factors, the journal Cell published a peer-reviewed version of the study this month.

The peer-reviewed paper includes additional experimental and clinical data on the D614G variant of the virus. The authors say that while the results indicate that the mutated variant of the virus may be able to replicate at higher levels in human cells, it is not clear whether it is more transmissible or results in more severe cases of COVID-19.

The study is one of several suggesting that the mutation is more infectious in cultured cells, though it remains unclear whether this holds for human infections, Nature reports.

Analysis identifies countries with the most successful responses

2 July 2020:

A variation of a statistical technique known as cluster analysis has identified countries with pandemic response outcomes that are better – or worse - than expected.

Mathematicians Nick James at the University of Sydney in Australia and Max Menzies at Tsinghua University in China analyzed the cumulative daily cases of COVID-19 in 208 countries, and the related deaths, between 31 December 2019 and 30 April 2020. They sourced the data from the European Centre for Disease Prevention and Control.

Their technique allowed them to cluster countries according to how similar their outbreaks were, and track how the pandemic has evolved over time in each location.

The results show that, in the early stages of COVID-19, Iran and Italy had higher death rates than expected – “severely impacted clusters” – while deaths in Singapore, South Korea, Qatar, and Australia were “anomalously low”.

The authors write that the technique “may assist in identifying the characteristics of the most and least successful government strategies for managing COVID-19”.

The study has been published in the journal Chaos.

New journal aims to speed up COVID-19 peer review

30 June 2020:

A new open-access journal has been launched to review COVID-19 preprints on a fast-tracked timeline, in an effort to thwart the spread of misinformation.

Launched by MIT Press, the new journal, Rapid Reviews: COVID-19, will use artificial intelligence to identify the preprints that are of most relevance to health officials, clinicians, and the public. After being screened by a group of volunteers, manuscripts will be evaluated by up to three reviewers in a matter of days, The Scientist reports.

The journal’s editorial team will be led by Stefano Bertozzi, dean emeritus of the University of California, Berkeley's School of Public Health in California. The first round of reviews will be published in July 2020.

“This project signals a breakthrough in academic publishing, bringing together urgency and scientific rigor so the world’s researchers can rapidly disseminate new discoveries that we can trust,” Vilas Dhar, trustee of the Patrick J. McGovern Foundation in Boston, Massachusetts, said in a press release.

In April, the Royal Society Open Science journal announced it was expediting its Registered Report review process for COVID-19 research, aiming to return its initial review to authors within one week of submission.

How fast are journals publishing COVID-19 research?

26 June 2020:

Journals are publishing COVID-19 papers up to eight times faster than articles on other topics, according to an analysis of over 850 papers.

Led by Amr Barakat at the University of Pittsburgh Medical Center, Pennsylvania, the team identified 294 COVID-19-related articles published between 1 January 2020 and 16 May 2020 in 16 journals.

The control group included 588 articles published in the same journals.

From submission to online publication, journals took a median of 20 days to publish COVID-19 papers compared to 119 days for other articles.

Journals also accepted COVID-19 articles faster than other papers, taking a median of just 13 days to accept pandemic-related articles versus 102 days for the control group.

The findings, which have not been peer-reviewed, were posted on medRxiv.

COVID-19 paper downloads surpass best-selling book sales

24 June 2020:

COVID-19 papers have been downloaded more than 150 million times, collectively, since publishers brought down paywalls on research related to the pandemic, according to the International Association of Scientific, Technical and Medical Publishers.

To put that into perspective, it took 17 years for Dan Brown’s popular novel, The Da Vinci Code, to reach 130 million sales. The 2013 video game, Grand Theft Auto V, hit 120 million sales earlier this year.

Cost plays a major role in the accessibility of COVID-19 papers.

Since January, several major publishers, including Elsevier and Springer Nature, which publishes the Nature Index, have made around 50,000 COVID-19-related papers freely available.

The Da Vinci Code and Grand Theft Auto V, cost USD$14.95 and $60, respectively, at the time of release.

With that said, the number of downloads within weeks of publication indicate the widespread interest in COVID-19-related research.

How has the pandemic impacted life scientists?

22 June 2020:

Life scientists may have had their lab hours cut, but more than 40% say they are using the extra time to focus on data analysis and manuscript writing.

Jan Korbel and Oliver Stegle at the European Molecular Biology Laboratory in Heidelberg, Germany surveyed 881 life scientists across eight countries to find out how lockdown measures have impacted their work.

According to the survey, more than 70% of researchers working in wet labs reported that they had lost up to six months of work due to lockdown measures. By comparison, 31% of dry lab scientists reported losing a similar amount of work.

Women made up more than two-thirds of the wet lab scientists surveyed, and they reported having fewer productive hours during the pandemic than their male colleagues.

Despite the challenges, Korbel and Stegle concluded that life scientists had adapted well to the changes. They suggest that the ability to work from home and collaborate over videoconferencing “might ultimately even result in benefits for scientific communities and society as a whole”.

The findings were published in Genome Biology.

Scientists push for retraction of face mask study

18 June 2020:

Over 40 scientists have penned a letter calling for the retraction of a paper claiming that face masks trump social distancing measures to prevent the spread of COVID-19, Apoorva Mandavilli reports for The New York Times.

Led by Mario Molina, who was a co-recipient of the Nobel Prize in Chemistry in 1995, the study claimed that airborne transmission is the “dominant for the spread of COVID-19” and that wearing a face mask is the most effective strategy for reducing the number of infections.

The authors of the paper, published in Proceedings of the National Academy of Sciences, were permitted to select their own reviewers. The open letter calls on PNAS to reconsider the policy whereby this privilege is accorded to members of the National Academy, saying the mechanism “effectively bypasses editorial decisions and undermines peer review”.

Molina defended the conclusions of the study but conceded that the language he and his co-authors used was “perhaps too strong.”

According to a PNAS spokeswoman, “the journal is aware of concerns raised about this article and is looking into the matter.”

Machine learning reveals gaps in coronavirus research

12 June 2020:

COVID-19 papers are more focussed on public health and clinical care than basic microbiology, according to an analysis of more than 100,000 papers.

A team led by Anhvinh Doanvo, a data science consultant at Deloitte in Washington, DC, used machine learning techniques to analyse the abstracts of 137,326 papers in the CORD-19 dataset, a freely available resource of COVID-19 and coronavirus-related articles.

Almost 60% of the COVID-19 abstracts fell into just 5 of 30 research categories, such as healthcare services, public health issues, and testing for coronaviruses.

By contrast, research published about other coronaviruses covered a broader range of topics, including viral structure, pathogenesis, and host cell interactions.

The findings, which have yet to be peer-reviewed, were posted as a preprint on bioRxiv.

How to safely reopen universities during a pandemic

12 June 2020:

Switching to online classes has been a major challenge for universities, but the next hurdle is figuring out how to reopen campuses safely.

A new preprint study suggests that keeping large classes online while allowing smaller ones to resume in the classroom could be an effective strategy.

A team led by Sing Chen Yeo, a data analyst at the Duke-NUS Medical School in Singapore, analysed more than 24 million student Wi-Fi connections at the National University of Singapore to gauge how they clustered together on campus before and after lockdown measures were put in place.

When classes of more than 50 students were moved online, the number of students gathering on campus fell by 70%, indicating that a partial transition to e-learning could be an ideal approach as the pandemic eases, the authors write.

The findings were posted to bioRxiv.

Science's response to past public health emergencies compared

9 June 2020:

Whether it’s COVID-19 or Ebola, researchers have always been quick to respond to public health emergencies, according to an analysis comparing publication patterns for six disease outbreaks.

Led by Lin Zhang at Wuhan University in China, the analysis focussed on articles published on SARS, H1N1, Zika, COVID-19, and two Ebola outbreaks, and found that there was a spike in the number of papers after each disease was declared a public health emergency by the World Health Organization. The findings were published in Scientometrics.

Virology, infectious diseases, and immunology were the most active disciplines in all six outbreaks, and while research in Europe and North America had a focus on public health, China and Japan were found to place more emphasis on biomedical research and clinical pharmacy findings, respectively.

New tool to measure COVID-19 research impact

6 June 2020:

The number of COVID-19 papers is growing fast, making it increasingly difficult for researchers to spot the most relevant findings. A new open dataset calculates impact scores for coronavirus-related articles to highlight stand-out research.

Developed by Thanasis Vergoulis at the Athena Research Centre in Greece and his colleagues, BIP4COVID19 ranks articles based on citations, social media attention, and number of Mendeley readers.

The tool, which is available as a web interface, “can be used to rank articles based on their impact to help researchers prioritise their reading”, the team explains.

Scientists have jumped into studying coronavirus

5 June 2020:

Thousands of scientists have switched their research interests to COVID-19, which could have a lasting effect on the scientific landscape, Elizabeth Gibney reports for Nature.

Of the 1,000 COVID-19-related papers posted on the preprint server arXiv, 45% are authored by researchers who usually publish in high-energy physics and condensed-matter physics.

The pandemic could also influence student choices, with several major research universities reporting that interest in postgraduate programmes on infectious-disease modelling was unusually high.

What will science look like after the pandemic?

3 June 2020:

High-speed publishing and more effective online classes are among the changes that could remain after the pandemic, according to a series of Nature reports on how COVID-19 will shape the research sector.

While researchers are embracing the push to share results openly and rapidly, it will take deep structural changes – such as incentivizing the early, open sharing of results – to ensure that they are lasting, reports Ewen Callaway.

The shift to online classes during lockdown has revealed the benefits of the virtual classroom, and some educators expect that the crisis will make tertiary education more accessible for students in developing countries, Alexandra Witze reports.

How COVID-19 publishing compares to previous coronavirus outbreaks

1 June 2020:

In less than five months, COVID-19 studies accounted for almost 70% of all coronavirus studies published in the last 50 years, according an analysis on preprint server, bioRxiv.

Milad Haghani and Michiel Bliemer at the University of Sydney in Australia analysed 19,518 papers on SARS, MERS, and COVID-19. They found that papers on SARS and MERS were largely published in specialist journals, such as the Journal of Virology, while COVID-19 articles appeared in broader publications, including The Lancet and The BMJ.

While papers covering public health concerns and epidemic control were the first studies to emerge during the SARS, MERS, and COVID-19 outbreaks, they received the least number of citations compared to research on the characterization of the viruses and vaccine development in all three cases.

Journals should rethink their publishing processes

26 May 2020:

The pandemic has laid bare the need to update academic publication practices argues an opinion piece in The BMJ.

They authors including The BMJ’s head of research suggest improvements such as sharing peer review comments with authors in real-time, prioritizing papers with openly available data, and the rapid assessment of research methodology before papers are submitted for formal peer review.

The authors contend that making improvements to the current peer review system is a better strategy in the long-run than getting rid of it entirely.

COVID-19 is transforming global teams

26 May 2020:

The United States and China are leading partners in COVID-19 research, according to an analysis of publishing and collaboration patterns before and after the pandemic.

Led by Caroline Fry at the Massachusetts Institute of Technology in Cambridge, US, the authors found that the two countries are collaborating more with each other since the outbreak, but they are partnering with fewer other countries and working in smaller teams than previously.

The findings, which have yet to be peer-reviewed, were posted as a preprint on SSRN.

The authors note that researchers are opting for efficiency over international teamwork in the race to tackle the crisis, which could shift the direction of global collaborations in the long term.

The role of preprints during the pandemic

23 May 2020:

Roughly 40% of articles on COVID-19 are preprints, according to an analysis of more than 16,000 papers related to the pandemic. On the preprint servers bioRxiv and medRxiv, COVID-19 articles are accessed and downloaded 15 times more often than other articles, the analysis finds.

These preprints are also being shared more widely on Twitter, with the top COVID-19 preprints being tweeted more than 10,000 times, compared to 1,323 tweets for the most popular non-pandemic preprint.

The findings were posted as an unreviewed paper on bioRxiv.

On average, COVID-19 preprints are being published in journals 26 days faster than other preprints. When comparing these preprints with their published versions, roughly 73% had no alterations to the wording or numbers in their abstracts, while graphics and tables had not been changed in just over 60%.

Virtual conferences could be the new normal

18 May 2020:

Giving a conference presentation through a screen may feel awkward, but there’s a lot to love about virtual conferences, reports Chris Woolston for Nature.

For some researchers, the convenience and low expense outweigh the drawbacks of virtual conferences, such as the lack of face-to-face networking. The format also makes international meetings more accessible for junior researchers and those from developing countries.

In an online Nature poll of 485 researchers who have attended virtual conferences, 82% respondents said that they would attend an online conference in future.

Funding worries weigh down early-career researchers

18 May 2020:

Only 10% of early career researchers on contracts that end this year have been offered additional funding due to the COVID-19 pandemic, according to a survey of 4,800 doctoral students and early-career researchers working at UK universities during lockdown.

As reported by Simon Baker at Times Higher Education, roughly 70% of the survey respondents said that they were worried about their finances, while another two-thirds reported that they were concerned about their future plans.

“These researchers are at critical stages in their careers,” Janet Metcalfe, who heads Vitae, a non-profit professional development organization for researchers, told Times Higher Education. “The restrictions due to COVID-19 are not only having a significant impact on their current research activities, but are likely to have long-term implications for their future careers.”

COVID-19 widens the gender gap in medical research

14 May 2020:

Journal articles that include female authors have fallen by 16% since the coronavirus outbreak, according to an analysis of more than 1,000 medical papers published in 2020.

Led by Jens Peter Andersen, who studies bibliometrics at Aarhus University in Denmark, the study found that female first authors have published 23% fewer COVID-19 papers compared to all medical articles published in the same journals in 2019. Articles with female last authors have also dropped by 16%.

The analysis, posted as an unreviewed preprint on arXiv, is one of a growing number of studies revealing how the pandemic is negatively impacting women’s productivity.

Wikipedia favours peer-reviewed research

2 May 2020:

Editors of Wikipedia’s COVID-19 content are largely relying on research that is peer-reviewed, highly cited, and widely shared online, according to a new analysis.

Giovanni Colavizza, a bibliometrics researcher at the University of Amsterdam, analyzed the citations and Altmetrics data of 64,040 papers related to COVID-19 and other coronaviruses.

Colavizza found that highly cited papers and those frequently mentioned on social media were more likely to be among the 3.4% of those that were referenced in Wikipedia’s 4,500 pages of COVID-19 content. The findings, which have yet to be peer-reviewed, were posted on bioRxiv.

COVID-19-related pages on Wikipedia have attracted almost 250 million pageviews since the beginning of April 2020. - Gemma Conroy

Australian researchers face job cuts due to coronavirus shut-down

12 May 2020:

Some 7,000 researchers working at Australian universities could lose their jobs within the next six months as a result of the coronavirus pandemic, according to a new report.

The report predicts that Australian universities could lose at least AU$3 billion (US$1.9 billion) in revenue this year due to travel bans and visa restrictions faced by international students. International student fees make up around one-quarter of universities’ revenue, covering staff salaries, research facilities, and other research expenses, Nature reports.

“We’re seeing a significant impact on our capacity to support high-quality research teams,” Susan Dodds, the deputy vice-chancellor of research and industry engagement at La Trobe University in Melbourne, told Nature.

It is estimated that over 9,000 international PhD students will have to postpone or abandon their projects this year due to travel bans. - Gemma Conroy

How could coronavirus change academic careers?

12 May 2020:

Online teaching could become the new essential skill for academics in a post-pandemic job market, reports Times Higher Education.

“Being seen as an online enthusiast, innovator or expert will be both more important and more demanding for interviewees,” said Robin Grimes, a material physicist at Imperial College London, UK.

The COVID-19 crisis could also result in a greater push towards virtual conferences and meetings, which are more inclusive, “especially to those who lack the financial support to go to increasingly expensive meetings”, Grimes told Times Higher Education.

Grimes also pointed out that online conferences could urge universities to consider job candidates from other countries, a move that could make the hiring process “less unfair by levelling the playing field”.

While such changes are a welcome silver lining of the pandemic, Rebecca Jarrett, head of resources at Cranfield University, said it could take time to see results in a slow academic job market.

“We know the impact on student recruitment, so people will be reluctant to move roles, particularly if they have a permanent post,” Jarret told Times Higher Education. - Gemma Conroy

Preprint servers enhance screening to tackle false coronavirus claims

7 May 2020:

Preprint repositories are tightening their screening procedures to slow the spread of questionable COVID-19 findings, reports Nature.

BioRxiv and medRxiv have enlisted outbreak specialists to scrutinise papers more closely for claims that could cause harm, contradict public health advice, or fuel conspiracy theories. BioRxiv is also no longer accepting papers predicting treatments using computational modelling alone.

“We can’t check the side effects of all the drugs and we’re not going to peer review to work out whether the modelling they’re using has any basis,” Richard Sever, co-founder of bioRxiv and medRxiv, told Nature. “There are some things that should go through peer review, rather than being immediately disseminated as preprints.”

The stricter screening measures were put in place after bioRxiv posted a controversial paper reporting similarities between HIV and SARS-CoV-2. The study was withdrawn after scientists widely criticised it as poorly conducted.

As of May 7, bioRxiv and medRxiv have posted almost 3,000 coronavirus-related papers. - Gemma Conroy

Journals change their practices

7th May 2020:

Fast-tracked peer review, acceleration of the drive to open access, AI discovery engines, preprints as a priority, video calls with authors - journals and researchers are rapidly adapting their publishing practices in response to the COVID-19 pandemic.

Turn-around times have shortened. A recent analysis of 669 articles published in 14 medical journals found that COVID-19 articles published since January 2020 are taking 57 fewer days on average from submission to publication compared to non-pandemic articles published before October 2019.

A separate analysis of 11,686 articles found that roughly 68% of COVID-19 papers are now freely available online. For comparison, it’s estimated that approximately 28% of scientific papers across all disciplines are open access.

It is not the first time a disease outbreaks has been the catalyst for innovations in publishing.

Living systematic reviews, for example, whereby the systematic review article format is continually updated with new evidence as it becomes available, were first adopted by researchers studying the 2015-16 Zika virus epidemic to keep pace with the deluge of published findings.

They have since been used to consolidate research on topics such as delayed antibiotic prescriptions for respiratory infections and the use of anticoagulants in cancer patients.

Three changes that journals have made in response to the COVID-19 pandemic with potentially lasting effects on the research publishing industry are listed here:

1. Open access and sharing of data, protocols, and standards

In January 2020, the Wellcome Trust in London released a statement calling researchers, journals, and funders to share their COVID-19 findings and data “rapidly and openly”.

To date, 145 journals, publishers, and institutions have agreed to make their coronavirus research and data freely available, and researchers and journals have been sharing both interim and final research data related to COVID-19 with the World Health Organization (WHO).

Several prominent journals, including Nature, Science, and PNAS, have brought down paywalls on their COVID-19 articles, at least for the duration of the outbreak.

Researchers have also been sharing the data-collection protocols and standards with the WHO.

Magdalena Skipper, editor-in-chief of Nature (published by Springer Nature, which also publishes the Nature Index) says the pandemic could facilitate a further push towards open science.

“We certainly hope that the willingness to collaborate and share information and data pre- as well as post-publication will continue beyond the current pandemic,” says Skipper.

In April, the European Commission launched the COVID-19 Data Portal, where researchers can upload, access, and analyze genetic sequences, protein structures, and data on how the SARS-CoV-2 virus impacts gene expression.

Bernd Pulverer, chief editor of The EMBO Journal, says the outbreak has highlighted a greater need for data-searching tools, such as AI discovery engines and data-directed searches.

“Such technology exists, but needs to be embraced by the community,” he says.

2. Uploading papers to preprints

Many journals, including Cell and the Journal of Experimental Medicine, are encouraging authors to upload their COVID-19 papers to preprint repositories medRxiv and bioRxiv before publication.

This practice has become the default for all submissions at the open access life sciences journal, eLife. To prevent authors from “feeling pressure to post their papers prematurely”, eLife and the EMBO journals have extended their ‘scoop protection’ policy to preprints, meaning they will not decline a preprint (or paper) if similar findings are published elsewhere while it is under review.

3. Speeding up peer review

In March,Nature issued a call-out to researchers who can review COVID-19 manuscripts over short timeframes, and eLife has encouraged early-career researchers to take part in the reviewing process.

Royal Society Open Science is also accelerating its Registered Report review process, with authors receiving their initial reviews within one week of submission.

For a speedier revision process, EMBO editors are consulting with authors through video calls and have allowed referees to comment on each other’s work to help editors reach a decision, says Pulverer.

“COVID-19 may help make these ideas standard,” says Pulverer. “I hope journals will have more focus on essential experiments rather than far-flung and endless revision cycles that damage the current system.” - Gemma Conroy

John Ioannidis hits back at critics

5 May 2020:

In April, Stanford epidemiologist John Ioannidis posted a preprint on medRxiv suggesting that the fatality rate for COVID-19 is likely on par with the seasonal flu, a claim that seemed to support his argument that global lockdown measures are a “once-in-a-century evidence fiasco”.

The controversial paper caused an uproar in the scientific community with some researchers pointing to statistical errors, questionable sampling techniques, and potential problems with the COVID-19 testing kit the team used. In The draft paper has since been revised, and as Undark reports, “acknowledges more uncertainty about the true number of infections”.

“The revised version has tried to address all the major concerns. I think the results are still very robust,” Ioannidis told Undark. “But it’s a single study. You can never say that a single study is the end of the story.”

Ioannidis also addressed the backlash he received after discussing the findings of the original study on Fox News, saying that it “was not possible to just hide it under the carpet.”

“It was a major finding, and I worried that it would be misinterpreted in different ways,” said Ioannidis. “I’m just a scientist. I have no political party affiliation and absolutely no interest to turn this into a political debate, or to have a political agenda supported.”- Gemma Conroy

For updates on coronavirus publication metrics, go here.