medRxiv is a not-for-profit health sciences preprint server for complete but unpublished manuscripts that launched June 1, 2019. It is managed in partnership with Yale and the British Medical Journal (BMJ) and is owned and operated by Cold Spring Harbor Laboratory. medRxiv is billed as a “service, not a product,” and is freely available to use.
Covering all subjects in the health sciences, including dentistry, emergency medicine, health informatics, medical education, medical ethics, nursing, and sports medicine, medRxiv lists over 50 specific research areas. Accepted manuscripts must fall into the categories of research articles, systematic reviews and meta-analyses, clinical research design protocols, and data articles. There is also an exhaustive list of what will not be accepted. medRxiv is intended to serve as a location for sharing research and propelling the scientific process. One of my favorite aspects of medRxiv (and preprint servers in general) is that it is an opportunity for complete result reporting, with places to link to protocols, analyses, and supplementary materials, and is an outlet to publish studies with negative or inconclusive results. Traditionally, publishing value has been limited to positive results or conclusive science, and preprint servers fill this void in the scientific literature. Globally, more than 3,000 institutions are represented in medRxiv’s preprint publications.
medRxiv’s use increased exponentially in 2020 due to the Covid-19 pandemic. From its inception to the end of 2019, medRxiv was averaging 50 submissions per week; by September 2020, medRxiv was averaging around 62 submissions per day. Submission requirements follow International Committee of Medical Journal Editors (ICMJE) guidance and the EQUATOR Network (Enhancing the QUAlity and Transparency Of health Research) reporting guidelines checklist, and require statements on funding, competing interests, IRB approval, study registration location, any protocols, and a data sharing/availability statement. Once a manuscript is submitted, it is screened for content, article type, and inappropriate or unsafe content that could harm patients and the public. Manuscripts are screened for plagiarism by Crossref Similarity Check. It is important to note that manuscripts do not undergo formal peer review or any editing before being published to the server. Further, manuscripts cannot be removed once added to medRxiv, and any corrections require the submission of a revised manuscript. medRxiv prefers the term “withdrawal” in lieu of retraction, and in the rare instance that a preprint is withdrawn, it will be marked as such and remain on the sever.
Features & Functionality
medRxiv has its own built-in search. The basic search is one line, similar to the format of many other databases. Using the advanced search, researchers can limit by date posted, preprint server, subject area, author, or DOI. The advanced search also offers the ability to create search strings by title, abstract/title, and full text/abstract/title, using the “any,” “all,” and “phrase” specifications for keywords. Results can be formatted by selecting the number of results on each page, and sorted by best, oldest, or newest. There is an option to select standard versus condensed format, but I was unable to determine the difference. Another useful feature is the ability to edit a search from the results page. Boolean operators and some modifiers are supported; a list of search tips is provided from a link on the advanced search page. Because medRxiv is considered a sister site to bioRxiv, users can perform a federated search across both servers using the basic search option. Alternatively, results can be limited to one preprint server by using the advanced search. Compared tox other gray literature databases, medRxiv has more robust search features and one of the more user-friendly formats.
A link on the homepage leads to Covid-19 preprints from medRxiv and bioRxiv. This citation list can be limited further to only one of the preprint servers. The medRxiv archive can be searched by year and month to retrieve any preprints published during a specific time period. Preprints on medRxiv are archived in Portico. Researchers can also sign up for email alerts or RSS feeds by search keywords and subject categories, respectively. Preprint full text is accessible through the server itself, but can also be found in Google, Google Scholar, Microsoft Academic, Semantic Scholar, Dimensions, CZI Meta, and PubMed using the preprint filter. Users can export individual preprints to several different types of citation management software and can add all results from a page to an export list. The export option is incredibly helpful since this is not a universal database feature. As someone who works on systematic reviews, I would like to see an export option where all resulting citations could be added without going through each page; however, the page-by-page method is relatively common with gray literature databases.
medRxiv’s open format allows for immediate informal peer review, feedback, and collaboration thanks to a comment box under each posted preprint. Comments posted to preprints are moderated for inappropriate content. Altmetrics and usage statistics are collected for each preprint and can be found using the metrics tab on the preprint page. Author information and information about the article’s history and copyright also are found on the preprint page using the info/history tab. Authors of preprints on medRxiv are the copyright holders and may choose to add a Creative Commons license to their work. medRxiv offers an opportunity called “M2J” for authors to directly transfer their preprint publication to a robust list of journals, including BMJ, Frontiers, JAMA Open, and PLOS. Once a preprint is published in a journal, a link to the journal article is added to the preprint on medRxiv.
The server is supported in part by the Chan Zuckerberg Initiative. In keeping with the open science initiative of preprint servers, medRxiv does not offer paid subscription models and is free to use by accessing medRxiv.org or using any of the other free access options for full text. medRxiv is suitable for any health sciences institution or organization interested in the most current research on more than 50 subjects. The “News and Notes” section hasn’t been updated since March 2021, but Twitter is updated frequently.
medRxiv is a valuable resource for those interested in open science, rapid research, and nontraditional publishing avenues. medRxiv is clear in its caution that research published on the server is not peer reviewed and should not be used to guide clinical practice. This caveat is an especially important consideration for those involved with supporting systematic reviews.