Unknown User2
Near-Miss Reporting in Korea: Linking Anonymous Reports to Risk Assessment Updates
2026.09.14 10:57
Risk Assessment · Safety System
Near-Miss Reporting in Korea: Linking Anonymous Reports to Risk Assessment Updates
Anonymous near-miss reports can reveal hazards that ordinary inspections miss. The practical challenge is deciding whether several reports describe one event, how to verify an unclear report without identifying the reporter, and when the risk-assessment record needs a new version. This guide focuses on validation, de-duplication and traceable updates rather than repeating a general participation process.
Keep the reporter anonymous, but make the hazard, decision and assessment version traceable.
📌 The operational answer
Do not treat every anonymous report as a separate risk. First compare the event date range, location, task, hazard mechanism and existing control. If two reports describe the same mechanism, link them to one case record while preserving each report’s arrival time and wording. If the hazard or control is materially different, open a separate case even when the location is the same.
Anonymous reporting protects participation; it does not remove the need for proportionate fact checking. Verification should test the condition in the workplace, not attempt to identify who submitted the report.
🔎 1. Define the minimum report fields
Ask for only the information needed to understand the hazard: task or process, approximate time, location or equipment, what could have caused harm, existing control and suggested improvement. Avoid free-text prompts that invite names, medical details or blame.
| Field | Why it matters | Privacy note |
|---|---|---|
| Task and hazard | Connects the report to an existing risk or a new one | Do not ask for a person’s name |
| Time and place | Supports a proportionate site check | Use an approximate window where exact time could identify a shift |
| Existing control | Shows whether the control failed or was absent | Describe the control, not the worker |
🧩 2. De-duplicate without losing the audit trail
Use a case key made from site, task, hazard mechanism and event window. A matching key suggests one case; it is not proof. Record the comparison result as “linked”, “separate” or “insufficient information”, with a short reason. Keep the original report IDs and submission timestamps so the organisation can show how the decision was made without exposing the reporter.
Example: NM-014 and NM-019 both describe a wet floor at the same loading-bay entrance during the same shift. Link them to one hazard case, retain both report IDs and record the additional observation from NM-019.
🧪 3. Verify the condition, not the person
A safety or operations representative can inspect the task, equipment and control described in the report. Use a neutral observation sheet and avoid asking “who submitted this?” If the report is too vague, publish a general follow-up question through the same anonymous channel or check the work area during the relevant task window.
When the condition cannot be reproduced, record the limitation rather than closing the case as false. A report may reflect a temporary setup, an unusual workload or a control that was restored before inspection.
🗂️ 4. Decide whether a new assessment version is needed
Update the risk-assessment version when the report reveals a new hazard mechanism, a changed task or a control that is not working as assumed. A minor wording correction can be recorded in the change log without rewriting the whole assessment. The decision should state the affected task, the control reviewed, the person responsible for the review and the next verification date.
| Decision | Use when | Record |
|---|---|---|
| Link to existing risk | Same hazard mechanism and controls | Case key and reason |
| Amend current version | Control or task description changed | Change log, owner, date |
| Create a new risk | New mechanism or materially different exposure | New identifier and follow-up plan |
📣 5. Share outcomes without exposing reporters
Close the feedback loop with a hazard-focused summary: what was checked, what control will change, who owns the action and when it will be rechecked. Do not publish wording that allows a small team or shift to infer the reporter. Where a report cannot be verified, explain the next check instead of dismissing it.
✅ Practical checklist
| Intake | The form requests task, hazard, location and control—not identity. |
| Comparison | Possible duplicates are linked by hazard mechanism and event window. |
| Verification | The workplace condition is checked without seeking the reporter. |
| Versioning | The assessment ID, change reason, owner and review date are linked. |
| Feedback | Outcome information is shared at a level that protects anonymity. |
🔗 Related guides
Employee Participation in Risk Assessment: Recording Anonymous Feedback
When a Special Risk Assessment Is Needed in Korea
Risk Assessment Documentation: What Inspectors Actually Look For
❓ Frequently asked questions
Q1. Should every anonymous report create a new risk?
No. Compare the hazard mechanism and control first. Several reports may be linked to one case while each original report remains traceable.
Q2. What if the condition cannot be reproduced?
Record the limitation and schedule a proportionate follow-up. An unsuccessful inspection is not proof that the report was false.
Q3. When should a risk-assessment version change?
Change the version when a new mechanism, task or ineffective control affects the assessment. Minor wording corrections can remain in the change log.
Turn anonymous insight into a visible safety decision.
A clear validation and version trail helps global teams explain what changed without exposing who reported it.
Contact Nudge EAP →📚 Sources and note
•Korean Law Information Center
This article provides general operational information. Duties and assessment methods depend on the workplace, task, workforce and current Korean law. Confirm case-specific requirements with the relevant authority or qualified adviser. Information checked on 14 September 2026.
Comments2