Run a PRISMA systematic review with AI

A systematic review lives or dies on process: explicit criteria, recorded screening decisions, and an extraction table you can defend. The Synthesis Lab is systematic-review software that runs that whole workflow, with the Fact Checker verifying every extracted cell against the paper it came from.

1. Define the review and its criteria

Create a review in Synthesis Lab and write your inclusion and exclusion criteria explicitly. These criteria drive screening. Every include/exclude decision is recorded against them, which is what makes the review systematic rather than a pile of papers you liked.

2. Screen your candidates

Papers flow in from your Literature Finder searches and saved-search alerts. Screen each candidate against your criteria; each decision (and its reason) is stored. New papers matching your saved searches keep queueing for screening. Your review stays alive as the field moves.

3. Extract into the evidence matrix

Define your extraction columns (population, design, outcome, effect, whatever your questions need) and run extraction over the included papers. Every filled cell is pulled from the paper's full text, carries a link to the exact source span, and is verified before you see it. A cell the AI can't ground stays empty rather than being guessed.

4. Appraise study quality

Apply ROB2, GRADE, or CASP presets to score each study, with quoted justifications from the paper. Claims that rest only on low-quality evidence are flagged separately in synthesis, so a weak study can't quietly carry a strong conclusion.

5. Generate the PRISMA flow diagram and synthesis

The flow diagram (identified → screened → included, with exclusion reasons) is generated from your actual recorded decisions, not filled in by hand at the end. Theme synthesis then drafts a narrative in which every claim links back to matrix cells, ready to hand to Thesis Editor as the skeleton of your literature review chapter.