How to create risk of bias graph and summary

Creating a risk of bias graph and summary for a systematic review involves assessing the methodological quality of the included studies and presenting the findings in a clear and standardised manner.

The seven domains for assessing risk of bias recommended by Cochrane are: sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and ‘other issues’.

When we read a research report or studies we need to identify the possible bias. Different sections of studies will give clues in searching for the biases. We look for missing information by reviewing study protocol or can even contact the authors.

Below are the steps to guide you:

1. Select a risk of bias assessment tool

Choose an appropriate tool for assessing the risk of bias based on the type of studies in your systematic review.

For example:

• Randomised controlled trials (RCTs): Cochrane RoB 2 tool.

• Observational studies: ROBINS-I (Risk of Bias In Non-randomized Studies – of Interventions)

• Diagnostic accuracy studies: QUADAS-2

2. Assess the risk of bias

Evaluate each study using the chosen tool. Most tools categorise the risk of bias for each domain into three levels: Low risk, Some concerns, and High risk.

3. Record the results

Document your judgments for each domain and overall bias for each study. Use software or tools to facilitate this process:

• RevMan (Review Manager): A free tool from Cochrane to assess and present risk of bias.

• Excel templates or custom tables can also be used.

4. Generate the risk of bias graph and summary

Use software to create graphical summaries of your risk of bias assessments.

Graph Types:

• Risk of bias summary table: A table showing the judgments (e.g., Low, High, Some Concerns) for each domain in each study.

• Risk of Bias Graph: A bar graph summarizing the proportion of studies at each risk level for each domain.

Detailed instructions for entering data and creating graphs can be found on the Cochrane website.