Repair a misapplied RoB 2 assessment by remapping every judgement to the correct bias domain and signalling question, rewriting each justification against that question, then rechecking whether the overall judgement changes. RoB 2 assesses a specific randomized-trial result across bias domains using signalling questions and written justifications, producing low risk, some concerns, or high risk judgements (Cochrane Handbook, Chapter 8).
Repair sequence
- Fix the unit of assessment. Confirm the assessment is tied to one specific trial result, since RoB 2 applies to a result rather than to a trial as a whole. If judgements were pooled across outcomes or time points, separate them before remapping.
- Remap each judgement to its domain and signalling question. For every existing judgement, identify which domain it belongs to and which signalling question it answers. Where a judgement was placed under the wrong domain, move it and record the correct signalling question it addresses.
- Rewrite the justification against the signalling question. Replace any justification that argues a different question than the one it sits under. State the trial feature the answer rests on so a second reviewer can follow the reasoning.
- Re-answer the misapplied items. Answer each corrected signalling question afresh, then derive the domain judgement from those answers. Check that the domain-level conclusion follows from the signalling-question answers and from the corrected reasoning.
- Recheck the overall judgement. Recompute the overall risk-of-bias judgement from the corrected domain judgements and record whether it changed. If it did, update the review text and any summary presentation that reported the earlier result.
What to document
Keep a short audit trail: the original domain and signalling question, the corrected mapping, the rewritten justification, and the before-and-after overall judgement. This lets co-reviewers and editors see what changed and why.
Where interpretation connects
A corrected risk-of-bias judgement feeds into how much confidence the review places in the affected result. Cochrane Handbook Chapter 15 covers interpreting results through confidence intervals, exact P values, absolute and relative effects, applicability, indirectness, and certainty, so revisit those discussions when the overall judgement changes.