Shifts in research culture, incentives, and technology would be needed for wide adoption

For decades, peer reviewing technical manuscripts before they were published in a journal was a regular duty for senior scientists. But James Fraser, a structural biologist at the University of California, San Francisco (UCSF), says he hasn’t reviewed a paper for a journal in years. Instead, Fraser and members of his lab focus on reviewing preprint studies that are posted online whenever authors like, bypassing a journal’s peer-review process.

Critiquing preprints offers big advantages over traditional, journal-based peer review, Fraser argues. Authors can get expert feedback more quickly—sometimes in days instead of months—and from a bigger and more diverse pool of reviewers. The process also can be more efficient: Reviewers can focus solely on the quality of the science without worrying about whether a paper is a good fit for a particular journal. And the reviews can be shared with everyone, not just authors and editors. “This is valuable stuff that’s often locked up” in the traditional review process, which at most journals is confidential, he notes.

Fraser isn’t alone in seeing promise in preprint reviews, which represent a radical shift in scholarly communication. Some regard preprint reviews as valuable input for journals that might ultimately publish a paper. Others hope reviewed preprints could ultimately take the place of journal publications. But the widespread adoption of preprint reviewing faces some thorny obstacles. This month, Fraser joined more than 200 scientists, journal editors, and research funders at a workshop that explored ways to overcome them.

Although preprints have long been common in some fields, especially physics, the COVID-19 pandemic prompted many life scientists to embrace them to quickly communicate potentially critical research findings. But many researchers worried these preprints were too often rushed and they needed at least some rapid vetting. Although the pandemic’s surge of preprints has now leveled off, the overall number of life sciences preprints has grown 100-fold since 2014, to 150,000 a year. Preprints now represent 7% of all articles in the U.S. National Institutes of Health’s PubMed database, up from 0.2% in 2015.