The state of preprints at PLOS

In the weeks since announcing our expanded partnership with medRxiv, we’ve received lots of thoughtful questions from our authors, editors, and the scientific community. In this post, we aim to answer some of them. 

But first, a quick recap of the role preprints have played at PLOS. While PLOS journals have always welcomed submissions with associated preprints, it wasn’t until 2016 that we formalized our support for preprints in journal policy and began accepting incoming direct transfers from bioRxiv. In 2018, we expanded that partnership and made it a 2-way street, giving authors the option to have their manuscripts forwarded to bioRxiv as part of the submission process at select PLOS journals. In 2019 came direct transfers from medRxiv to PLOS, and just this month, facilitated posting from PLOS to medRxiv.

The rates at which researchers have taken advantage of these integrations demonstrate how highly PLOS authors value preprints as a tool for disseminating their work. Since 2016, we have received close to 3,000* incoming direct transfers from bioRxiv and medRxiv for consideration at PLOS. Since May 2018, more than 6,000 (13%) authors submitting life science research to PLOS reported that they have already posted their manuscript as a preprint, with nearly 9,000 (19%) opting in to have their manuscript forwarded to bioRxiv at the point of submission. 

Both author-posted and opt-in rates for preprints vary by journal, research area, and geographic region, and have shifted over the years. In some subject-specific journals (Pathogens, Genetics) for example, rates spiked in 2020 before declining slightly in 2021, possibly relating to strong preprint support during the initial phases of SARS-CoV-2 pandemic. Other journals with broader scope (ONE, Biology) show more gradual, consistent growth over time. 

Percent of submissions from May 1 2018-Dec 31 2021 with either facilitated preprint opt-in or author-posted preprints by journal

Now, on to your questions.

Preprint questions & answers

Question: Could preprints potentially spread misinformation, or pose a potential risk to the public?

As we all know, peer review and publication take time. Preprints let researchers share new work before and during the journal editorial process. Early sharing creates new opportunities for collaboration, while maximizing availability for vetting, feedback and review. But early sharing also poses potential risks: for example, the risk that readers may treat preprints uncritically, as proven fact, or that preprints may contain information that could be dangerous to the public. That’s why responsible preprint posting is so important.

Responsible preprinting is a top priority for PLOS. Every preprint undergoes a thorough screening by bioRxiv, medRxiv, or PLOS editors. Basic checks include fit and appropriateness—for example: Does the preprint report research? Is it within scope for the server? Is it free from inflammatory or derogatory language? More complex checks by subject area experts look at potential concerns such as risk to public health, dual use research of concern (DURC), or risk of revealing patient identities.

The rigor of this screening process is demonstrable. Across PLOS on average 71%** of the eligible manuscripts whose authors opted-in to have their manuscript forwarded to bioRxiv pass screening and go on to appear on the server.

Question: How does posting a preprint impact citations?

Research suggests that early sharing via preprint is associated with increased attention and citations for the corresponding peer-reviewed publication. In one study, following a cohort of over 74,000 articles published from 2015 to 2018, articles with preprints enjoyed 49% higher Altmetric scores and 36% higher citations than those without (Yu & Hughey, 2019). Another study, looking at bioRxiv preprints specifically, found that the articles with preprints had higher Altmetric and citation rates than articles in the same journals without associated preprints (Frazer, Momeni, Mayer & Peters, 2020).

Importantly, it’s not yet clear whether this association is causal. There may be many factors at work. For example, perhaps authors who share preprints are also more likely to engage in other behaviors that are known to increase attention, like sharing open data or methods, presenting at conferences, or posting on social media. Maybe authors are likely to preprint their most promising work. Or, perhaps preprints aid in driving online traffic to published articles.

While the relationship between preprints and citations is interesting insofar as it demonstrates the quality and interest of the research being posted, the correlation between preprinting and citation rate is in some sense beside the point. The main aim of preprints is to increase the efficiency of research communication, and of the peer review process.

Question: What if preprints are never published in a journal, or are retracted after publication?

There is no evidence that articles with preprints are less likely to be published, or more likely to be retracted from the journals in which they are published, compared with other articles. Indeed, at PLOS, we’ve observed an equal or slightly higher acceptance rate for submitted manuscripts with an associated preprint. We aim to share detailed data in the future, after a more thorough analysis. 

Many journals have embraced preprints, especially in response to the SARS-CoV-2 pandemic. What makes PLOS’ preprint program notable is the workflow integration that we offer at scale. We treat preprint posting as a free author service by integrating it into the submission process with a single question. That ease and accessibility makes preprinting more practical for our authors, and helps to normalize the use of preprints in the communities that we serve.

* All numbers are current through Dec 31 2021 unless otherwise noted
**Pass rate data is current through Jun 30 2021