Guides
Guides
How to report AI-assisted screening in your PRISMA 2020 methods
Checklist and example wording for reporting AI or machine-learning screening tools in the PRISMA 2020 methods section and flow diagram.
How to audit AI exclusions in systematic review screening
A step-by-step method to check AI-excluded records with a blind random sample and a Wilson confidence interval, and how to report it.
How to check for retracted studies in a systematic review
A practical routine for finding retracted studies and expressions of concern at import, before analysis and before submission, with example methods wording.
Dual screening and Cohen's kappa: a practical guide
How to run blinded dual screening, calculate and interpret Cohen's kappa (with a worked example), handle the prevalence paradox, and report agreement.
How to validate AI screening before using it in your review protocol
A practical checklist for validating an AI screening tool before you commit to it in a protocol: reading vendor evidence, piloting on a finished review, pre-specifying checks and reporting.
SyRev vs Covidence vs Rayyan for AI title/abstract screening
A dated, sourced comparison of how the three tools use AI at the title/abstract stage: what each automates, published validation, per-review checks and entry prices.