Rayyan includes a built-in Risk of Bias tool for assessing the methodological quality of your included studies. This article covers how to add articles, complete an assessment, and work with your team.
Note Risk of Bias currently supports the RoB 2 framework for randomized controlled trials (RCTs). If your included studies are not RCTs, this tool does not apply to your review. Support for additional frameworks — including non-randomized studies and diagnostic accuracy studies — is planned for future releases.
Add articles to your assessment
- Open your review and click the Risk of Bias tab.
- Click Add articles and select the included studies you want to assess.
- The selected articles appear in the assessment table. The Full text column shows whether a PDF is attached to each article.
Note A full-text PDF is needed to review each study against the RoB 2 signaling questions. If an article shows no attachment in the Full text column, you can attach its PDF directly from this table before starting the assessment.
Complete an assessment
- Click Start Assessment or double-click an article in the table to open it — the RoB 2 assessment panel appears on the right alongside the article.
- The panel shows five domains (D1–D5), each with a set of signaling questions. Expand a domain to begin.
- Answer each signaling question: Yes, Probably yes, Probably no, No, No information, or N/A.
- Optionally, add a short justification in the text field below each answer — this is not required but helps transparency and makes your assessment reproducible.
- Once all questions in a domain are answered, the domain judgment appears automatically: Low risk, Some concerns, or High risk.
- Complete all five domains. When the last domain is done, Rayyan calculates and displays the overall risk-of-bias judgment for the article.
- Click Save changes, then Mark as Completed.
- Repeat for each article in your assessment.
Best practice Fill in justifications even though they are optional. They make your assessment defensible during peer review and easy to re-check months later when the reasoning behind individual answers is no longer fresh.
The five RoB 2 domains
RoB 2 evaluates each study across five domains, each covering a different type of bias:
| Domain | What it assesses |
|---|---|
| D1 — Randomization process | Was the allocation sequence random and adequately concealed? |
| D2 — Deviations from intended interventions | Did knowledge of the assigned intervention affect what participants or personnel did? |
| D3 — Missing outcome data | Were outcome data available for nearly all participants? Could missing data depend on the true outcome? |
| D4 — Measurement of the outcome | Was the outcome measured appropriately, and were assessors blinded to the intervention? |
| D5 — Selection of the reported result | Was the result selected from multiple analyses in a way that depends on the findings? |
Note The RoB 2 form is fixed — you cannot add, remove, or reword questions. This keeps your assessment aligned with the published methodology and comparable to other RoB 2 reviews. For guidance on answering each signaling question, refer to the official Cochrane RoB 2 crib sheet.
Working with a team
Risk of Bias supports the same collaborative workflow as Data Extraction.
- When Blind Mode is on, each collaborator assesses articles independently without seeing the other's answers.
- When Blind Mode is turned off, Rayyan flags any differences between collaborators at the domain or overall judgment level as unmatched answers.
- The team reviews the flagged articles together, discusses, and updates their answers to reach alignment — there is no automatic resolution.
Important Resolve all unmatched answers before reporting. An export with unresolved assessments will contain inconsistent results across collaborators, which complicates synthesis.
Stage complete
| Category 5 — Data Extraction & Risk of Bias | Next: Export & Report → |
Comments
0 comments
Article is closed for comments.