RoB 2 traffic light plot for randomized trials
A study-by-domain view for randomized trials.
Source: Sterne et al., BMJ 2019
Illustrative valuesRisk of Bias Plot Generator
FigranEnter or paste your judgments and get traffic light and summary plots for your review.
- RoB 2, ROBINS-I, QUADAS-2 and more
- Imports robvis CSV and Excel
- Free SVG and 300 DPI PNG
Traffic light
StudyStudy actions
Summary
Risk of bias plot examples
Weighted risk of bias summary plot
Compare judgments using study weights.
Source: McGuinness & Higgins, Res Synth Methods 2021
Illustrative valuesROBINS-I traffic light plot for non-randomized studies
Visualize intervention study assessments.
Source: Sterne et al., BMJ 2016
Illustrative valuesQUADAS-2 risk of bias and applicability concerns
Show risk of bias and applicability together.
Source: Whiting et al., Ann Intern Med 2011
Illustrative valuesCochrane risk of bias tool (RoB 1): risk of bias summary and graph
Display the original Cochrane domains.
Source: Higgins et al., BMJ 2011
Illustrative valuesROBINS-E traffic light plot for exposure studies
Explore judgments in exposure studies.
Source: Higgins et al., Environ Int 2024
Illustrative valuesQUIPS risk of bias for prognostic factor studies
Summarize prognostic factor assessments.
Source: Hayden et al., Ann Intern Med 2013
Illustrative values+12 more
Risk of bias plot in colorblind-friendly colors
A larger review in the colorblind palette.
Source: robvis colourblind palette
Illustrative valuesHow to make a risk of bias traffic light plot
1Choose your tool
Choose a tool to load its domains.
2Enter or paste judgments
Click a circle to set a judgment, or import a table.
3Download both plots
Download either plot, or both combined, as PNG or SVG.
Risk of bias tools and domains
| Tool | Study design | Domains | Judgments | Reference |
|---|---|---|---|---|
| RoB 2 | Randomized trials | Bias arising from the randomization process; Bias due to deviations from intended interventions; Bias due to missing outcome data; Bias in measurement of the outcome; Bias in selection of the reported result | Low, Some concerns, High, No information | Sterne et al., BMJ 2019 |
| ROBINS-I | Non-randomized intervention studies | Bias due to confounding; Bias in selection of participants into the study; Bias in classification of interventions; Bias due to deviations from intended interventions; Bias due to missing data; Bias in measurement of outcomes; Bias in selection of the reported result | Low, Moderate, Serious, Critical, No information | Sterne et al., BMJ 2016 |
| ROBINS-E | Non-randomized exposure studies | Bias due to confounding; Bias arising from measurement of the exposure; Bias in selection of participants into the study (or into the analysis); Bias due to post-exposure interventions; Bias due to missing data; Bias arising from measurement of the outcome; Bias in selection of the reported result | Low, Some concerns, High, Very high, No information | Higgins et al., Environ Int 2024 |
| QUADAS-2 | Diagnostic accuracy studies | Patient selection; Index test; Reference standard; Flow and timing | Low, High, Unclear | Whiting et al., Ann Intern Med 2011 |
| RoB 1 (original Cochrane tool) | Randomized trials | Random sequence generation (selection bias); Allocation concealment (selection bias); Blinding of participants and personnel (performance bias); Blinding of outcome assessment (detection bias); Incomplete outcome data (attrition bias); Selective reporting (reporting bias); Other bias | Low, Unclear, High | Higgins et al., BMJ 2011 |
| QUIPS | Prognostic factor studies | Study participation; Study attrition; Prognostic factor measurement; Outcome measurement; Study confounding; Statistical analysis and reporting | Low, Moderate, High | Hayden et al., Ann Intern Med 2013 |
| Custom | Any | Domain 1; Domain 2; Domain 3 | Low, Some concerns, Moderate, Unclear, High, Serious, Critical, Very high, No information, Not applicable |
A traffic light plot shows every study's judgment in every domain; a summary plot shows, for each domain, the share of studies (or of meta-analysis weight) at each level.
Judgments for RoB 2 and ROBINS-I apply to a specific result, so make one plot per outcome when judgments differ between outcomes.
In RevMan the traffic light plot is called the risk of bias summary and the summary plot is called the risk of bias graph. The plot format and color conventions follow robvis.
Risk of bias plot FAQ
Does this tool assess risk of bias?
No. You make the judgments with the tool your protocol names (RoB 2, ROBINS-I, QUADAS-2 or another), usually two reviewers independently. This page turns those judgments into the traffic light and summary plots that reviews publish.
How is this different from robvis or RevMan?
robvis is the R package and web app that set the standard for these plots, and it is what most reviews cite. This page draws the same two plots with the same symbols and color options, imports the same CSV or Excel layout, and lets you edit judgments directly in the browser. RevMan draws risk of bias figures inside a Cochrane review. Cite the assessment tool you used; if you want to cite the plot format, cite McGuinness and Higgins (2021).
Which risk of bias tools are supported?
RoB 2, ROBINS-I, ROBINS-E, QUADAS-2 (with applicability concerns), the original Cochrane tool (RoB 1), QUIPS, and a Custom option for any other domain-based checklist. ROBINS-I version 2 is still a draft, so it has no preset; rename the domains in the ROBINS-I preset or use Custom.
What is a weighted summary plot?
The summary plot (called the risk of bias graph in RevMan) has one bar per domain. An unweighted summary plot counts studies: each bar shows the percentage of studies at each judgment. A weighted plot uses each study's weight in the meta-analysis, so large studies count for more. Add a Weight column or turn on Weights, then choose Weighted.
Can I make a risk of bias table instead?
Yes. Export ▾ Copy table pastes a formatted table into Word or Excel. Export ▾ CSV saves the judgments in the robvis column layout so you can keep them with your review or import them here again.
Are the colors colorblind-friendly?
The Cochrane colors (green, yellow, red) are the usual convention, and every circle also has a symbol (+, −, ×, !, ?), so the plot can be read without color. The Colorblind option uses an orange scale that stays distinct for red-green color vision deficiency.
What format should my data be in?
One row per study: the study name, one column per domain, then Overall and an optional Weight. Judgments can be words (Low, Some concerns, High) or symbols (+, −, ×). Paste it from Excel or upload a CSV or XLSX file.
Should I make one plot per outcome?
RoB 2 and ROBINS-I judge risk of bias for a specific result, and the same study can be judged differently for different outcomes. If your judgments differ by outcome, make one plot per outcome and say which outcome each one covers in the figure legend.
Is it free, and is my data private?
Yes. No account is needed to make or download plots, and your judgments stay in this browser; a draft is kept only in this browser tab.
Ready to plot your risk of bias assessment?
Choose your tool and enter your judgments.