← allmeta

Workflow · PRISMA 2020

Systematic review

A start-to-finish path through the allmeta tools that mirror the PRISMA 2020 process — from question to synthesis. Each step is a tool you can open in a new tab; the sequence mirrors the methodological order a rigorous review follows. Cochrane-level conformance still depends on dual screening, peer-reviewed search, and PROSPERO registration done outside these tools.

  1. Frame the question

    Translate a clinical question into PICO components. Crisp PICO is what makes the search tractable and the inclusion criteria reproducible.

    Why first: the search string, eligibility, and outcome list all derive from PICO. Get this right or rerun everything.

    PICO Formulator → Learn the theory: PICO module

  2. Search every database

    Translate one database's syntax (PubMed) into Embase, Cochrane CENTRAL, Web of Science. Then forward/backward citation-chase the seminal papers.

    PRISMA item 7. Reviewers will check whether you searched ≥3 databases plus citation chasing.

    Search Translator → Citation Chaser → Learn the theory: search strategy

  3. Deduplicate & screen

    Run an automated dedupe pass (DOI / title / author overlap), then screen titles & abstracts. Two reviewers independently in a real review; one is acceptable for rapid reviews.

    PRISMA items 8–9. Record screening decisions for the flow diagram.

    Citation Dedup → PRISMA Screening → Learn the theory: rapid screening

  4. Assess risk of bias

    Use RoB 2 for randomised trials, ROBINS-I for non-randomised, ROBINS-E for exposures. Render a traffic-light summary for the manuscript.

    PRISMA item 11. The single most-scrutinised step at peer review.

    RoB 2 → ROBINS-I → ROBINS-E → Traffic-light plot → Learn the theory: RoB Mastery course (45 min)

  5. Synthesise quantitatively

    Pairwise meta-analysis with Hartung-Knapp + prediction interval. Sensitivity analyses (cumulative, leave-one-out, subgroup, bias). Use Bayesian or multilevel where the data warrant it.

    PRISMA items 13–15. The point estimate is rarely the headline; the prediction interval and sensitivity analyses are what reviewers cite.

    Forest Plot → Heterogeneity → Meta-regression → Pub-bias tests → Learn the theory: MA Methods course

  6. Grade the certainty & report

    Build a GRADE Summary of Findings table (5 domains, with upgrade/downgrade reasoning). Render the PRISMA 2020 flow diagram. Run the PRISMA-2020 checklist before submission.

    PRISMA items 22–25, GRADE Working Group guidance. The SoF table is what the reader actually remembers.

    GRADE SoF → PRISMA Flow → PRISMA Checklist → Learn the theory: GRADE Certainty course

Want a deeper foundation before starting? Take the Evidence Synthesis Foundations course or browse the full 26-course collection. Doing a network or DTA review instead? See network meta-analysis or diagnostic test accuracy.