Carrie Levinson, MSLIS 
Manager of Reference & Instruction Services 

Gloria Willson, MLIS, MPH 
Director of Education & Research Services 

Gustave L. and Janet W. Levy Library 
Icahn School of Medicine at Mount Sinai 

Introduction 

Founded in 2014, Veritas Health Innovation is an Australian not-for-profit organization that created Covidence, an online software platform intended to streamline the evidence synthesis process. Covidence facilitates title and abstract screening, full text review, and data extraction sections for these types of syntheses. Users can upload citation files from different literature databases and generate documentation that can be added to the supplemental material of manuscripts. It includes functionality for adding and managing reviewers, uploading PDFs, and creating templates. 

Content 

Evidence syntheses like systematic reviews, scoping reviews, and meta-analyses require a great deal of organization and project management and involve rigorous methodology that must be followed(1). Covidence allows users from many different disciplines, ranging from medicine and health sciences to social sciences and education, to efficiently organize their projects by creating reviews within the tool from which they can extract relevant data. It is intended to help users automate several parts of the evidence synthesis process, including progress tracking and documentation, and ensures rigor for each step in the study selection process.  

The resource, available online at www.covidence.org, includes templates for data extraction forms and quality assessment tools. Its audience is researchers ranging from students to faculty and research staff – basically, anyone who might find themselves working on an evidence synthesis project. 

Features/Functionality 

Users must create an account to use Covidence. Once logged into the site, they have the option to start a new review and may choose to create a general review or a Cochrane review. Additional information requested includes the type of review being conducted, the question type being answered, and the disciplinary area. Depending on whether the user is utilizing an institutional license or an individual one, the number of reviews available for creation may be limited. 

Once the review is created, there is a lot of room for customization in settings. The creator of the review can add other reviewers (even if they are outside the creator’s institution), choose which data extraction template to use, set up reviewer settings for each section to keep the project consistent, and even add eligibility criteria and study tags for easy classification. The search strategy and the date it was last run can also be added, which is helpful for recordkeeping and reporting. A PRISMA flow diagram updates as the team goes through the screening process, and generates required information for reporting, such as total number of imported references, the number of full-text references assessed for eligibility, and the number of studies included in the review (2). This can be accessed at any time, edited as needed, and exported to be included as supplemental material for the manuscript. 

The import section generally starts off with the title and abstract screening section, but imported files can also be added directly to other sections instead if needed. This is directed by the user. The tool automatically de-duplicates citations from each imported file, which updates the PRISMA flow diagram. Once citations are in the title and abstract screening section, reviewers can simply read the titles and abstract presented to them and click “Yes,” “Maybe,” or “No” to vote on whether they believe the citation should move to the full text review section. “Maybe” in this case is functionally equivalent to “Yes”, which might be the most confusing part of this section, which is otherwise straightforward to use. Reviewers do not know how the other reviewer(s) voted until they themselves have voted on a record, which is intended to reduce selection bias. 

The default setting for votes is two reviewers, and ones that have been voted “Yes” or “Maybe” then move to the full text review section, where the same type of voting takes place. In this section, however, votes are classified as “Include” or “Exclude” with the additional requirement to provide reasons for exclusion, which Covidence prompts automatically. Full text, if available as open access, will be uploaded automatically into this section. PDFs behind a paywall must be added manually, either through a bulk upload or individually. At times, the bulk upload helps with making this part of the review more efficient, but it is not the easiest process, even with the tutorials provided by the platform. The chances that the team will have to individually upload article PDFs is quite high, which can be frustrating and tedious for large reviews. Covidence does offer an integration with the LibKey Nomad extension (which provides links and access to PDFs from library subscriptions) that aids in populating the articles in this section. 

As mentioned previously, data extraction is highly customizable, with two different templates available. Teams can use the default settings or choose which fields they want to include. There is also an editable quality assessment template available in this section. Users are able to export CSV files for different parts of this section if they so choose. 

For help and guidance, Covidence offers an extensive knowledge base that can be customized for specific institutions. It also offers very responsive technical support via email and regularly gives webinars on different sections of the platform. 

Business Model 

Covidence offers free trials for one review with the ability to screen up to 500 citations, a paid subscription for either one review or a package of three reviews (with unlimited collaborators), and institutional licenses. Reviewers can also be added to one review by a team with a subscription, but they are limited to working on that review. Single and package subscriptions are valid for 12 months. 

Breakthrough 

Overall, Covidence is an easy-to-use tool that makes a significant portion of the evidence synthesis process more efficient. Any user embarking upon such a project, from beginners to seasoned researchers, will appreciate the way Covidence is structured for their use, especially with the large amount of assistance available. Covidence also helps librarians execute comprehensive searches and upload search results in an efficient manner. Health science libraries that offer systematic review services should definitely consider subscribing to this platform, especially if there are many teams that come to the library asking for this type of assistance.  

​​References 

​1. Munn Z, Peters MDJ, Stern C, Tufanaru C, McArthur A, Aromataris E. Systematic review or scoping review? Guidance for authors when choosing between a systematic or scoping review approach. BMC Med Res Methodol. 2018;18(1):143. Available from: https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-018-0611-x  

​2. Covidence. FAQ: How can I use my Covidence PRISMA figure in my report/write-up/publication? [Internet]. Covidence.org. 2023 [cited 2025 Mar 18]. Available from: https://support.covidence.org/help/faq-how-can-i-use-my-covidence-prisma-figure-in-my-report-write-up-publication ​