Free — No Login Required

Risk of Bias Visualization Generator

Draw risk of bias traffic light and summary plots for RoB 2 and ROBINS-I from judgments you have already made. Import a CSV, check the plot and download the figure.

Your data stays in your browser. Nothing you enter is uploaded, stored or used for analytics, so export your work before closing the page.

1. Choose the assessment tool

You enter judgments you have already made. This tool never assesses or suggests risk of bias.

RoB 2 (randomised trials)5 domains + overall · Low, Some concerns, HighROBINS-I (non-randomised studies, 2016 version)7 domains + overall · Low, Moderate, Serious, Critical, No information

2. Enter your judgments

Leave a judgment as “Not entered” if you do not have one; it is shown as missing and never counted as low risk. Weights are optional and only used for the weighted summary.

Study judgments
Study	Year	D1	D2	D3	D4	D5	Overall	Weight	Remove

Domain names

D1: Bias arising from the randomization process
D2: Bias due to deviations from intended interventions
D3: Bias due to missing outcome data
D4: Bias in measurement of the outcome
D5: Bias in selection of the reported result

3. Plot and export

Figure title

The weighted summary needs a valid weight for every study. Missing: Study 1.

6 judgments are not entered. They appear as “Not entered” in the figure and are never counted as low risk.

Fix these before exporting (1)

Row 1 has no study name.

Export is disabled until the issues above are resolved.

What this tool does

This tool turns your risk-of-bias judgments into the two figures most systematic reviews report: a traffic light plot, with one row per study and one column per domain, and a summary plot showing the share of judgments in each category for every domain.

It visualises; it does not assess. You enter or import the judgments you reached using the RoB 2 or ROBINS-I assessment, and the tool never proposes, infers or changes a judgment. A missing judgment is shown as “Not entered” and is never counted as low risk.

Each assessment uses its own domains and judgment categories. RoB 2 judgments are Low risk, Some concerns and High risk. ROBINS-I (2016 version) judgments are Low, Moderate, Serious, Critical and No information. The two are never mixed. Every judgment has a symbol as well as a colour, so the figures stay readable in greyscale and for readers with colour-vision deficiency.

If you supply a weight for every study, for example the study weights from your meta-analysis software, the tool can also draw a weighted summary plot. The weights are used exactly as supplied and are not calculated here.

Who should use it

Authors of systematic reviews who assessed randomised trials with RoB 2 or non-randomised studies with ROBINS-I.
Students and residents who need a clear figure for a thesis or manuscript without writing code.
Teams that already use the robvis R package or Excel and want a quick browser alternative.

How to use it

Select the assessment tool, RoB 2 or ROBINS-I. Changing it clears judgments that do not belong to the new tool, after asking you to confirm.
Add studies one at a time or import a CSV. Download the example CSV to see the expected columns.
Choose a judgment for each domain and for the overall judgment. Add a weight only if you have one for every study.
Switch between the traffic light plot and the summary plots and check the legend and the data table.
Download PNG, SVG or PDF. Export your data as CSV to keep a copy, because nothing is saved.

Worked example (hypothetical data)

Suppose a review of four randomised trials assessed with RoB 2 judges domain 1 as Low for three trials and Some concerns for one. The summary plot then shows the first domain as 75% Low risk and 25% Some concerns, because 3 of 4 studies (75%) and 1 of 4 (25%) fall in each category.

If those trials carry meta-analysis weights of 40, 30, 20 and 10 and the single Some concerns trial is the one weighted 10, the weighted summary shows 90% Low risk and 10% Some concerns for that domain. Weighting changes the picture when large studies differ from small ones, which is why the weighting method is printed on the figure.

Use “Load hypothetical example” in the tool to see sample data in both plots.

Common mistakes

Mixing RoB 2 and ROBINS-I categories

The tools have different domains and different judgment scales. Plot each tool separately and do not relabel “Moderate” as “Some concerns”.

Treating missing judgments as low risk

A blank is not a judgment. This tool shows it as its own category so that it is visible rather than flattering.

Reporting a weighted plot with partial weights

A weighted summary needs a weight for every study. Otherwise the proportions describe only some of the evidence.

Plotting one overall judgment per study when outcomes differ

Risk of bias is assessed per result. If you assess several outcomes, plot each outcome separately and say which one the figure shows.

Relying on colour alone

Journals may print in greyscale. Keep the symbols in the legend visible and describe the categories in the caption.

Frequently asked questions

Does this tool assess risk of bias for me?

No. You make the judgments using the official RoB 2 or ROBINS-I guidance. The tool only draws the figure from your judgments.

Which assessment tools are supported?

RoB 2 for randomised trials and the original 2016 ROBINS-I for non-randomised studies of interventions. Newer revisions are not supported yet.

What CSV format should I use?

Columns Study, Year, one column per domain (D1 to D5 for RoB 2, D1 to D7 for ROBINS-I), Overall and optionally Weight. Download the example file from the tool.

Is my data uploaded?

No. Everything runs in your browser. CSV files are read locally and nothing is stored or sent anywhere.

How is the weighted summary calculated?

For each domain, the weight of the studies in a category divided by the total weight of all studies. Studies with no judgment contribute to Not entered. The weights are the ones you supply.

Is this the same as robvis?

No. It is an independent browser tool that draws similar figures. robvis is a separate R package and web app by McGuinness and Higgins. No robvis code or assets are used here.

References

Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898. doi:10.1136/bmj.l4898
Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ 2016;355:i4919. doi:10.1136/bmj.i4919
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