In the process of evaluating studies to be included in your review, you should include a Validity or Risk of Bias assessment. LATITUDES is a collection of tools, and training, for validity assessment in evidence synthesis projects.
For an evidence synthesis, some sort of software for managing citations is required.
Zotero would be my first choice, Mendeley my second. Endnote has been criticized for de-duplication problems, but here is a suggested protocol for de-duping in Endnote using some advanced features.
A fully featured citation manager is really useful for capturing and organizing the papers to be reviewed. It can also help document the searching process if you make a separate folder for every search. Remember it's an important part of the process to capture all the results (to reduce bias in selection), so using a database that exports results in bulk and/or a citation manager that can import from results lists (such as Zotero or Mendeley) is useful. Adding each set of results to separate folders allows you to easily document how many results were produced by each search. De-duplication features are also useful (you have to do it somehow!) but usually need careful monitoring. Collaboration features make reviewing easier.
Features to look for:
There are also specialty tools that can be used in place of or with citation managers. COVIDENCE is a fee-based tool that is the gold standard of current systematic review tools. Rayyan and CADIMA are free systematic review tools. These are mostly used for multi-person systematic reviews and allow for selection and review by multiple people. Qualtrix and other surveying tools can be used for targeting the review process to the research question. AbsTrackr is a abstract review tool that uses machine learning to serve as a "second reviewer."
The SR Toolbox (down as of Feb 2024) is a review site for systematic review tools from protocol development to document management.
Be sure to carefully review tools to be sure they will fit the needs of your study.
It's always been a danger that you might cite something that had been retracted or criticized without realizing it. But now there are tools to help with that.
