Building Abstractus Around the Researcher Workflow

At Abstractus, we’ve learned that improving systematic literature reviews isn’t just about automating one step of the process. While screening is the largest portion of the process we are focused on streamlining, the more we work with researchers, the more we see how many small parts of the workflow add friction to a review process.

This is why our most recent product updates have focused on making that process a little more frictionless, from the initial search to deduplication, and soon, collaboration and auditability.

Making literature searches easier with Search Assist

One of our newest additions is Search Assist, a free tool designed to help researchers with compiling search terms for each database, before beginning the search process. Finding the right studies is integral to the success of a review, and to do that, it is paramount to have the right keywords in order to find all the articles that could possibly be relevant before beginning the search.

Search Assist is designed to do just that for researchers; after entering the research question and focus points, it will help generate keyword terms, as well as an estimate of the number of sources with those keywords. This helps one quickly triangulate their database search.

Adding deduplication: you asked, we answered

When originally developing the model, the researchers we worked with used Covidence, and as such simply exported the already deduplicated list of abstracts into our model, ran it, and exported it back into Covidence.

As we expanded with who we worked with, we ran into researchers who did not have access to Covidence, and as such, still needed to deduplicate their list of abstracts before moving on to screening. So, we decided to add this feature directly into Abstractus, so people would not have to have access to a different platform to deduplicate their abstracts.

Deduplication is now a part of the workflow Abstractus covers, allowing teams to clean their dataset before screening without having to switch between platforms.

Building toward better collaboration and auditability

Our next set of updates will be focused on collaboration.

As we’ve worked with research teams, one of the things that they appreciate the most about other platforms are their ability to work as a team, as well as clear reasoning for exclusion: i.e. on Covidence, when excluding an abstract, one can select a reason from a dropdown menu, such as number of patients, population, etc. While we already have a reasoning log, talks with researchers have shown the importance of tagging, where instead of a lengthy reason, there is a simple tag attached to exclusion decisions.

As such, we have made the addition of these features a priority, allowing for more ease in the creation of PRISMA tables and collaboration between researchers working on the same project. These features are being developed directly from conversations with researchers using Abstractus today. As we continue to develop Abstractus, we are doing our best to build not for researchers, but with researchers, listening to feedback every step of the way.