Use of the preprint server MedRxiv for industry-sponsored research: COVID-19 versus non-COVID-19 researchk

ABSTRACT

Objective: Use of preprint servers by academics has accelerated during the SARS-CoV-2 pandemic, but whether their use by industry has increased over this period is unclear. We compared the use of medRxiv for the dissemination of COVID versus non- COVID research, focussing on industry-sponsored studies. Research design and methods: Using the search function in medRxiv (https//www.medrxiv.org/), we determined the total numbers of preprints mentioning 'COVID-19 or SARS-CoV-2' (COVID) and all others (non-COVID) deposited between 1 January and 30 September 2020. We then manually screened all COVID and non-COVID preprints deposited during a peak-activity week (24-31 May) to determine subject area, study type, author affiliations and funding source. Results: As of 30 September, 7449 COVID and 3335 non-COVID preprints had been deposited on medRxiv, including 404 and 129, respectively, over 24-31 May. The top two subject areas represented by COVID preprints were 'Infectious disease' (35.1%) and 'Epidemiology' (27.7%). Non-COVID preprints were deposited under a range of subject areas, most frequently 'Epidemiology' (10.9%) and 'Neurology' (8.5%). The highest proportion of preprints (COVID, 32.7%;non-COVID, 38.0%) described observational research (32.4%) and modelling studies (12.4%). Ten (2.5%) COVID preprints acknowledged commercial funding, versus 6 (4.7%) non-COVID. Industry authors were listed on 9 (90%) and 5 (83%) of these, respectively. While half of industry-sponsored non-COVID preprints described observational research, those on COVID covered a range of study types, including artificial intelligence, diagnostics, modelling, and observational research. Conclusions: Although the use of medRxiv has accelerated during the pandemic, utilization by industry has remained low. Further research is needed to identify potential barriers to the industry embracing preprint servers.