Methodology
A step-by-step guide to conducting a rigorous systematic literature review using EvidenceGrid AI, aligned with PRISMA 2020 and Cochrane standards.
Frame your review using the PICO framework — Population, Intervention, Comparison, Outcome. A precise research question determines which studies are eligible and drives every subsequent decision.
Use the EvidenceGrid search panel to query PubMed, OpenAlex, arXiv, Europe PMC, and CORE simultaneously. Combine Medical Subject Headings (MeSH) with free-text keywords and Boolean operators (AND, OR, NOT). Apply year filters to scope the literature.
In the Research Plan tab, define criteria before screening begins. Common filters include study design (RCT, cohort, case-control), language, publication type, and population characteristics. Document your rationale — PRISMA requires it.
EvidenceGrid's AI pre-screens each paper against your criteria and flags borderline cases for human review. Two independent reviewers should resolve conflicts to maintain inter-rater reliability. The platform calculates kappa scores automatically.
Papers that pass title/abstract screening advance to full-text review. Upload PDFs manually or let EvidenceGrid fetch open-access versions. Use the built-in PDF reader with annotations to highlight key passages.
Define custom extraction fields per research question — sample size, intervention details, outcome measures, risk-of-bias indicators. AI extracts values from full text; reviewers approve or edit each field. All decisions are timestamped and attributed.
Apply a validated tool (Cochrane RoB 2, ROBINS-I, or NOS depending on study design). EvidenceGrid lets you create extraction fields that mirror the tool's domains, so risk-of-bias assessment stays inside the same workflow.
Export your evidence grid to Excel, CSV, or structured JSON for meta-analysis software (RevMan, R metafor, Stata). The export includes PRISMA flow counts, extraction data, and approval audit trails.
Preferred Reporting Items for Systematic Reviews and Meta-Analyses. EvidenceGrid tracks the four-stage flow (Identification, Screening, Eligibility, Included) and generates counts automatically.
The authoritative guide on systematic review methodology. Chapters 4-6 cover search strategy, screening, and data extraction — the three stages EvidenceGrid automates.
Register your review protocol at PROSPERO before starting screening. Include your registration number in the published review. Pre-registration reduces reporting bias.
The Grading of Recommendations, Assessment, Development and Evaluations framework rates the certainty of evidence across outcomes. Use your extracted data to populate GRADE summary-of-findings tables.