Risk of Bias Plot Generator
Paste your RoB 2 or ROBINS-I judgements, one study per line, and get a traffic-light plot or a summary bar plot in the style of robvis — in the Cochrane colours or a colour-blind-friendly set. Download it as SVG or PNG for your review. It draws figures from judgements you have already made; it does not assess bias. Free, no sign-up.
Study, D1, D2, D3, D4, D5, then an optional Overall judgement and an optional weight. Separate cells with commas, tabs or semicolons; pasting from a spreadsheet works. Judgements: Low, Some concerns, High, No information (or L, S, H, NI, or + − × ?).
Your figure
6 studies read
Show the judgements as a table
| Study | D1 | D2 | D3 | D4 | D5 | Overall |
|---|---|---|---|---|---|---|
| Trial 1 | Low risk | Low risk | Low risk | Low risk | Some concerns | Some concerns |
| Trial 2 | Low risk | Some concerns | Low risk | Low risk | Low risk | Some concerns |
| Trial 3 | Some concerns | Some concerns | High risk | Low risk | Some concerns | High risk |
| Trial 4 | Low risk | Low risk | Low risk | Some concerns | Low risk | Some concerns |
| Trial 5 | High risk | Some concerns | Low risk | Some concerns | Some concerns | High risk |
| Trial 6 | Low risk | Low risk | Low risk | Low risk | Low risk | Low risk |
How to Use the Risk of Bias Plot Generator
- 1. Choose the tool. RoB 2 is for randomized trials (five domains); ROBINS-I is for non-randomized studies of interventions (seven domains). The text box changes to match.
- 2. Paste your judgements. One study per line: the study name, then one judgement per domain, then an optional overall judgement and an optional weight. Copy straight from a spreadsheet, or separate cells with commas or semicolons.
- 3. Pick the plot and the colours. The traffic-light plot shows every judgement for every study. The summary plot shows, for each domain, the share of studies at each level. Choose the Cochrane colours or the colour-blind-friendly set.
- 4. Download the figure. Save an SVG to edit in a vector program, or a PNG at twice the on-screen size. Copy data puts the judgements on your clipboard as a table.
The Two Tools and Their Judgements
| Tool | Domains | Judgements | Colours (Cochrane set) |
|---|---|---|---|
| RoB 2 | D1 randomization process, D2 deviations from intended interventions, D3 missing outcome data, D4 measurement of the outcome, D5 selection of the reported result | Low risk, Some concerns, High risk | Green, yellow, red |
| ROBINS-I | D1 confounding, D2 selection of participants, D3 classification of interventions, D4 deviations from intended interventions, D5 missing data, D6 measurement of outcomes, D7 selection of the reported result | Low, Moderate, Serious, Critical, No information | Green, yellow, red, dark red, blue |
Judgements can be typed in full, as a letter (L, S or M, H, C, NI) or as the traffic-light mark (+, −, ×, !, ?). In ROBINS-I, S means serious; in RoB 2, S means some concerns.
RoB 2 is described by Sterne and colleagues in the BMJ (2019), and ROBINS-I in the BMJ (2016). The figures follow robvis, the risk-of-bias visualization package by McGuinness and Higgins (Research Synthesis Methods, 2021): the same layouts, the same Cochrane and colour-blind palettes, and the same plus, minus and cross marks. This tool is independent of robvis.
Traffic-Light Plot or Summary Plot?
A traffic-light plot lists the studies down the side and the domains across the top, with one coloured circle per judgement. A reader can check any single study, so it suits a review with a manageable number of studies. A summary plot has one horizontal bar per domain, split into the share of studies at each level, plus a bar for the overall judgement. It shows at a glance which domains are the weak points of the evidence.
The summary plot counts each study once by default. If you add a weight as the last column — the meta-analysis weight, for example — you can switch on weighting, and each bar then shows the share of the weight instead. The two can differ. In the built-in RoB 2 example, 50% of the studies are at low risk for deviations from intended interventions, but they carry 61% of the weight; overall, 1 of 6 studies (17%) is at low risk, and those studies carry 24% of the weight. State which version your figure shows.
The Overall Judgement
If a line has no overall judgement, the tool uses the worst of its domains: critical or high if any domain is, otherwise no information if any domain lacks it, otherwise some concerns or moderate, otherwise low. That matches how ROBINS-I is usually read. RoB 2 is more nuanced: its guidance also allows a high-risk overall judgement when several domains raise some concerns, which depends on your assessment. When yours differs from the worst-domain rule, enter it in the Overall column and the tool draws what you entered.
Reporting the Figure
- Name the tool and cite it. “Risk of bias was assessed with RoB 2” or “with ROBINS-I” belongs in your methods, with the reference.
- Say who assessed and how disagreements were settled. The figure shows an outcome; the reader needs the process too.
- Credit robvis when you follow its layout, and say whether the summary bars are weighted.
- Show every study. PRISMA 2020 item 18 asks reviews to present the risk-of-bias assessment for each included study, which the traffic-light plot does; the PRISMA 2020 Checklist shows where each item belongs.
What this tool does not do
This tool draws a figure; it does not judge anything. The judgements come from answering each instrument's signalling questions, ideally by two reviewers working independently. ROBINS-I version 2 reorganised the domains: this tool follows the seven-domain layout of the 2016 tool, so check that your version's domains match before you plot.
How It Works
The page reads your lines, checks every judgement against the scale for the chosen tool, and draws the figure as SVG. A line with an unrecognised judgement or too few columns is left out and reported by line number rather than drawn wrongly. Everything runs in your browser and nothing is uploaded. A table of the judgements sits under the figure for readers who cannot see the colours.
Frequently Asked Questions
- What is a risk of bias traffic-light plot?
- It is a grid with one row per study and one column per bias domain, where each cell is a coloured circle for the judgement: green for low, yellow for some concerns or moderate, red for high or serious. An overall column usually sits at the end. Reviewers use it to show every study-level judgement in one figure.
- What colours does robvis use for risk of bias?
- The Cochrane set is green (#02C100) for low, yellow (#E2DF07) for some concerns or moderate, red (#BF0000) for high or serious, dark red (#820000) for critical and blue (#4EA1F7) for no information. The colour-blind-friendly set runs from pale yellow through oranges to dark brown. This tool uses both sets.
- How is the overall risk of bias worked out?
- In RoB 2, low overall requires low risk in every domain, and high overall follows from a high-risk domain or from several domains with some concerns. In ROBINS-I, the overall judgement is generally the most severe domain judgement. Enter your own overall column if you need to, or let the tool apply the worst-domain rule.
- Should the summary plot be weighted?
- Unweighted bars show the share of studies; weighted bars show the share of the evidence, for instance by meta-analysis weight. robvis draws unweighted bars by default. Use weighting only if you have a weight for every study, and say in the figure legend which you used.
- Can I plot QUADAS-2, ROBINS-E or the original Cochrane risk of bias tool?
- Not here. This tool covers RoB 2 and ROBINS-I. The robvis package itself also supports QUADAS-2, ROBINS-E, QUIPS and a generic template, so use it if you need those.
- How do I cite robvis?
- Cite McGuinness LA, Higgins JPT. Risk-of-bias VISualization (robvis): an R package and Shiny web app for visualizing risk-of-bias assessments. Research Synthesis Methods 2021;12(1):55–61, doi 10.1002/jrsm.1411. Also cite the instrument itself, RoB 2 or ROBINS-I, in your methods.