Cochrane Risk of Bias Tool 2.0 for Randomized Controlled Trials
Also known as: RoB 2.0, RoB 2
RoB 2 is the Cochrane Collaboration's updated methodology for assessing the risk of bias in randomized controlled trials (RCTs). Published in 2019, it replaced the original Cochrane RoB tool with a more structured, transparent approach using signalling questions and domain-based judgments to evaluate five critical sources of bias.
Key highlights
- Domain-based structure aligns with mechanisms of bias, making judgment more transparent and defensible
- Signalling questions provide consistent guidance across reviewers and studies, improving inter-rater reliability
- Updated judgment incorporates modern trial design practices (e.g., blinding of outcome assessors, adjustment for confounding)
- Flexible approach accommodates diverse trial designs and contexts; guidance includes publication-specific algorithms
- Explicitly addresses intention-to-treat analysis and missing data mechanisms, reflecting current evidence standards
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
RoB 2 is mandatory in Cochrane systematic reviews of randomized trials. Use when appraising RCT quality for clinical guidelines, meta-analyses, or evidence synthesis. Essential when trials will inform clinical decision-making or policy recommendations.
Strengths & limitations
- Domain-based structure aligns with mechanisms of bias, making judgment more transparent and defensible
- Signalling questions provide consistent guidance across reviewers and studies, improving inter-rater reliability
- Updated judgment incorporates modern trial design practices (e.g., blinding of outcome assessors, adjustment for confounding)
- Flexible approach accommodates diverse trial designs and contexts; guidance includes publication-specific algorithms
- Explicitly addresses intention-to-treat analysis and missing data mechanisms, reflecting current evidence standards
- Requires detailed extraction of trial methods and analysis; assessments may be hampered by inadequate reporting in published papers
- Signalling questions sometimes have ambiguous answers, necessitating reviewer judgment and potentially introducing subjectivity
- Scoring algorithm does not generate a total 'bias score'; integration of five domain judgments is categorical, not quantitative, limiting ranking or weighting of trials
- User training and familiarity with RoB 2 guidance is essential; complex algorithm requires consultation of domain-specific algorithms
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
Does RoB 2 produce a numerical score?
No. RoB 2 uses categorical judgments (Low Risk, Some Concerns, High Risk) for each domain and overall. Some systematic review software tools display a visual summary (e.g., traffic-light chart) but no total score is calculated.
Can I use RoB 2 for observational studies or non-randomized trials?
No. RoB 2 is designed exclusively for randomized trials (RCTs and cluster-randomized trials). For observational studies, use Newcastle-Ottawa Scale or other tools; for non-randomized interventional studies, consult ROBINS-I.
How long does RoB 2 assessment take per trial?
Typically 30–90 minutes per trial, depending on report quality, complexity, and reviewer familiarity. Detailed trial reports with protocols available reduce time; sparse reports increase time due to contacting authors or marking as unclear.
What is the difference between 'Some Concerns' and 'High Risk' in a domain?
'Some Concerns' indicates that bias cannot be ruled out but there is insufficient evidence of likely bias; 'High Risk' means there is substantial evidence (e.g., lack of blinding without objective outcomes) that bias probably did occur and may have affected results.
Should I exclude High Risk trials from my meta-analysis?
Not automatically. The effect of bias on a particular outcome must be evaluated. Instead, conduct a main analysis and a sensitivity analysis excluding High Risk trials; report both. Direction and magnitude of potential bias inform interpretation.
Sources
- 1.Sterne, J. A., Savović, J., Page, M. J., Elbers, R. G., Blencowe, N. S., Boutron, I., ... & Higgins, J. P. (2019). RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ, 366, l4898.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 3). Cochrane RoB 2.0. ScholarGate. https://scholargate.app/research-methodology/cochrane-risk-of-bias