Assess quality systematically
Eight validated tools, one consistent workflow. mapped picks the right risk-of-bias instrument for your study type, generates AI-assisted domain judgments with source quotes, and outputs traffic-light visualizations ready for your manuscript.
Updated April 2026
Risk of Bias
Methodological Quality Assessment
| Study | D1 | D2 | D3 | D4 | D5 | Overall |
|---|---|---|---|---|---|---|
| Gao et al. 2021 | + | − | + | + | − | − |
| Hong et al. 2015 | + | + | − | + | − | − |
| IVUS-ACS 2024 | + | + | + | + | + | + |
| Tan et al. 2015 | − | + | + | − | + | − |
| Kim et al. 2013 | + | − | + | + | − | − |
| Bendary et al. 2024 | × | − | − | + | − | × |
| Jakabčin et al. 2009 | − | + | + | − | + | − |
| Chieffo et al. 2013 | + | − | − | + | + | − |
| Testa et al. 2026 | + | + | − | − | + | − |
| Kim et al. 2015 | − | + | + | + | − | − |
| Diletti et al. 2026 | + | + | + | + | + | + |
| Liu et al. 2018 | − | + | + | + | − | − |
Key Capabilities
Eight-Tool Catalogue
RoB 2 (RCTs, recommended default), RoB 2 Cluster (cluster RCTs), ROBINS-I (non-randomized interventions), ROBINS-E (non-randomized exposures), QUADAS-2 (diagnostic accuracy), QUIPS (prognostic studies), Newcastle-Ottawa Scale (observational), AMSTAR 2 (systematic reviews / umbrella reviews). mapped suggests the right tool for your study type automatically.
AI-Assisted Domain Judgments
The Advanced Reasoning Engine reads each included study in full and drafts domain-level judgments — Low risk / Some concerns / High risk for RoB 2; tool-specific scales for the others — with supporting quotes from the paper. Every AI judgment is presented alongside the source text so reviewers can accept, modify, or override with full transparency.
Dual-Reviewer with Conflict Resolution
Independent dual review is enforced for risk of bias just as it is for screening. Each reviewer's domain judgments are blinded until both are complete, then disagreements are routed to a designated resolver with both rationales visible. Cohen's kappa is calculated automatically per tool and per project.
Adaptive Tool Selection
When you select a study type at project setup (intervention, prognostic, DTA, NMA, scoping), mapped recommends the right risk-of-bias tool — RoB 2 for RCTs, QUIPS for prognostic, QUADAS-2 for DTA, and so on. You can override the suggestion per study; the workflow adapts to mixed-design reviews.
Traffic-Light + Lightbar Visualizations
Two complementary plots out of the box. Traffic-light plots show domain-level judgments per study (green / yellow / red). Lightbar summaries show the proportion of studies at each risk level per domain. Both export to PNG, SVG, PDF, or EPS at publication-ready 600 DPI for journal submission.
Per-Domain / Per-Study Export
Risk-of-bias data exports as Word tables (with domain rationales), Excel matrices (for cross-checking), or PDF reports (for ethics committees and supplementary materials). Domain-level rationales are preserved so reviewer 2 can audit any judgment back to the source paper.
Frequently asked questions
- What risk-of-bias tools does mapped support?
- Eight: RoB 2 (RCTs, default), RoB 2 Cluster, ROBINS-I, ROBINS-E, QUADAS-2 (diagnostic accuracy), QUIPS (prognostic), Newcastle-Ottawa Scale (observational), and AMSTAR 2 (systematic reviews / umbrella reviews). mapped recommends the right tool for your study type and supports mixed-design reviews per study.
- How does mapped's risk-of-bias compare to RevMan or EPPI-Reviewer?
- RevMan supports RoB 2 well but is RCT-focused. EPPI-Reviewer covers more tools but the AI assistance is limited. mapped offers the broadest tool catalogue (8 instruments) plus AI-drafted domain judgments with quoted source text — a meaningful time saving for reviews with 30+ studies.
- Are the AI judgments trustworthy?
- Treat them as a strong starting point, not a final answer. Every judgment is presented alongside the quoted source text, and you accept, modify, or override per domain. Dual-reviewer protocols still apply. Inter-rater reliability between AI-drafted and human-confirmed judgments is published in our methodology docs.
- Does mapped do RoB 2 for cluster RCTs?
- Yes. RoB 2 Cluster is included as a separate tool in the catalogue with the cluster-specific signalling questions and additional bias domains (e.g. identification or recruitment of participants into clusters). It's selected automatically when you mark a study as cluster-randomized.
- How much does the Risk of Bias module cost?
- Included in every Mapped Project (list $119/project, currently $79 launch pricing) — all 8 tools, AI-assisted judgments, dual-reviewer, conflict resolution, and visualizations. Custom Enterprise plans extend it to unlimited projects with SSO. Pricing: mappedresearch.com/pricing.
- Can mapped handle mixed-design reviews?
- Yes. A single review can mix RCTs, observational studies, and diagnostic-accuracy studies — each one gets the appropriate tool (RoB 2, Newcastle-Ottawa, QUADAS-2 respectively). Visualizations group studies by tool, and the manuscript module narrates the methodology correctly per design.
Comparing tools? See how mapped stacks up against EPPI-Reviewer on the workflow you actually run.
Mapped vs EPPI-Reviewer