Repair a missing GRADE domain judgement by locating the outcome and domain with no recorded rating, re-examining the evidence for that domain, recording the judgement, and then checking whether the overall certainty rating changes. GRADE rates certainty at the outcome level using explicit criteria, so the repair belongs in the specific outcome row that lacks a judgement.
- Scan the evidence profile for the outcome row and domain cell that has no recorded judgement. Confirm which domain is blank, such as risk of bias, inconsistency, indirectness, imprecision, or publication bias, and note the outcome it applies to. The expected result is a single identified gap rather than a general sense that the profile is incomplete.
- Re-examine the underlying evidence for that outcome and domain against the GRADE criteria. Record the judgement and the reason for it in the profile so the rating is explicit and traceable. The expected result is a documented domain judgement that matches the evidence reviewed.
- Recalculate the overall certainty for that outcome using the recorded domain judgements. Check whether the added judgement changes the overall rating. The expected result is either a confirmed rating or a revised one, with the change documented.
GRADE provides a transparent approach to rating certainty of evidence, with explicit criteria and outcome-level judgements, so each domain judgement should be recorded against a specific outcome. Cochrane GRADE resources describe GRADE as a systematic approach for rating certainty in systematic reviews and evidence syntheses, including Summary of Findings table guidance and GRADEpro workflows. Use those resources to confirm the domain definitions and the rating you record.
If the evidence for the missing domain is unclear, mark the judgement as unresolved and note what evidence would settle it. That keeps the gap visible for the next reviewer instead of filling it with an unsupported rating.