Researchers at work at a mobile soil testing lab in Mabanga, Kenya.

Soil testing being carried out at a mobile lab in Mabanga, Kenya.Credit: Joerg Boethling/Alamy

Data suggest that a lack of resources is making it difficult for researchers in low-income countries to turn preprints into peer-reviewed papers, according to one of the authors of a study examining the issue.

Information scientist Anita Bandrowski and computer scientist Peter Eckmann, both at the University of California, San Diego, created a tool called PreprintMatch, which looked for the peer-reviewed final-versions of preprints.

Analysing the almost 140,000 papers posted on the preprint servers bioRxiv and medRxiv since their inception (in 2013 and 2019, respectively) until 2021, the researchers found that 58% of preprints were later published in a journal, with 61% of bioRxiv preprints having an associated final paper, compared with 37% of medRxiv preprints. The team’s results were published in PLoS ONE earlier this year1.

The study found that researchers in low- and lower-middle-income countries (LLMICs) and upper-middle-income countries had turned their preprints into peer-reviewed papers around 40% and 48% of the time, respectively. Preprints by authors in high-income nations, by comparison, were published as papers roughly 61% of the time. “We don’t know why this is,” says Bandrowski. “This seems to be quite a significant effect.”

When they are published, the study also found, preprints from LLMICS made the transition to papers more quickly — in an average of 178 days, compared with 203 days for preprints from richer countries. And LLMIC preprints were altered more between preprint and final version than were preprints by authors in high-income countries (HICs), often changing their titles, abstracts and author lists.

Bandrowski suspects that authors from LLMICs are converting fewer preprints to published papers owing to a lack of financial resources. The study found that when an author from a HIC was added to the paper’s author list, the conversion rate from preprint to paper increased.

Jessica Polka, executive director of ASAPbio — an organization in San Fransisco, California, that promotes innovation in the life sciences — agrees with this assessment. “The availability of resources is a strong contributor to the eventual fate of a manuscript,” she says.

Monopoly of science literacy

The study complements a 2022 analysis2 in Lancet Global Health, which found no indication that preprints that went on to be published in peer-reviewed journals were of a higher quality than those that didn’t. “This is a concern that many people bring up,” Polka says.

Najat Saliba, an atmospheric chemist at the American University of Beirut in Lebanon, notes that preprints don’t tend to be assessed when managers at LLMIC institutions consider whether to promote researchers, which might contribute to lower preprint-posting practices in these settings, compared with authors in middle- and higher-income nations.

Abdulrahman Bamerni, a geologist at the University of Duhok in Iraq, says that the ethics around preprints are not yet established in many LLMICs. One issue is that it’s more common for ideas expressed in preprints to be copied by other scientists, something he says has happened to him.

It’s important to acknowledge that there are notably fewer papers from the global south being published in esteemed international journals, says Saliba. “There is a monopoly of science literacy in the north, when it is compared to the south,” she says. “This is not a secret to anybody. There are many publications that show that journals have been favouring scientific publications from the north versus the south.”

A 2021 analysis3 of nearly 25,000 papers in 20 development journals revealed that only 16% of these articles were authored by researchers in the global south, whereas 73% were authored by researchers in the global north. A further 11% were collaborations between northern and southern authors.

Saliba finds it difficult to publish papers from her location in the Middle East. “Usually, when I publish on my own from this part of the world, the attitude of the journal is one way,” she says. “When I publish with colleagues from the United States or from Europe, the attitude of the journal is different.”

Limitations to spotting trends

There are limitations to what PreprintMatch can do. For example, Bandrowski and Eckmann’s study tracked whether final-version papers were published in journals indexed by the citations database PubMed, which doesn’t cover all scholarly journals. It also analysed English-language papers only.

Despite this, the study uses a “very interesting” technique to examine publishing inequity through the lens of preprints, says John Inglis, co-founder of bioRxiv and medRxiv and executive director of Cold Spring Harbor Laboratory Press in New York.

According to Inglis, authors from LLMICs sometimes find it difficult to explain what preprints are to local journal editors, who might not be familiar with the format. This could contribute to the lower rates of final-version papers, says Inglis, because editors might reject a manuscript thinking that the authors are attempting to publish the same paper twice to gain more credit.

Bandrowski hopes that other researchers will build on her freely available data set, for example by surveying authors on the reasons for their publication patterns. In the meantime, she plans to investigate patterns in the quality of studies from different locations and demographics.