Systematic Review Screener

Best practice: two independent reviewers screen all titles/abstracts, with third reviewer resolving disagreements. This dual screening process increases reliability and reduces bias. For large reviews (5000+ articles), some screen 100% with two reviewers while others screen a subset (20-30%) in duplicate to assess agreement, then have single reviewers screen remainder if agreement is high (κ > 0.80). Full-text screening typically requires two reviewers for all articles. Document your screening process, inter-rater reliability (kappa statistic), and how disagreements were resolved. PRISMA guidelines recommend dual screening - single screening is not considered rigorous for systematic reviews.

Define criteria using PICOS framework: Population (who was studied?), Intervention/Exposure (what was examined?), Comparison (what was it compared to?), Outcome (what was measured?), Study design (what types of studies?). Also specify: language restrictions, publication date range, publication status (published only vs grey literature), geographic settings. Exclusion criteria mirror inclusions (e.g., exclude non-human studies, exclude conference abstracts). Pilot test criteria on sample of articles to ensure clarity and consistency. Document any criteria changes during review with justification. Pre-register criteria in PROSPERO before screening to prevent outcome-driven decisions.