Two Versions of the Near-Miss Story
You are auditing a 24/7 distribution centre with 210 workers and roughly 180 forklift movements per shift. In a pedestrian walkway an operative tells you, unprompted, that nobody reports near misses any more because shift supervisors say it costs time and the numbers get used against them. Later, in a meeting room with the operations manager present, a shift supervisor states that all incidents and near misses are reported within 24 hours. The near-miss log you sample contains four entries for the last 12 months. The OH&S manager then asks you to route all further worker interviews through her so that people are not disrupted, and offers an anonymised summary instead. Your audit plan has 90 minutes left for Clause 10.
1. How do you treat the two statements about near-miss reporting, and what does an evidence-based approach require here?
Read the worked answer
Two contradictory statements are not evidence of a fact; they are evidence that something needs testing. The audit principle is that audit evidence should be verifiable and drawn from more than one source, and that conclusions should not be based on a single interview in either direction.
Test both claims against records and observation:
- The near-miss log itself: four entries in 12 months against roughly 47,000 forklift movements a year. Low reporting is a signal, not proof - reporting rates are not proportional to movements, but four is low enough to ask what the expected rate is and whether the organization has ever measured it.
- Clause 10.2 records: incident reports, investigations, actions and their closure. Look for near misses recorded in other channels (maintenance logs, damage reports, safety committee minutes) that never reached the log.
- Clause 5.4 evidence: safety committee or worker representative minutes. Did workers raise near misses there? What was the response?
- Clause 9.1.1: is reporting data used as a performance indicator, and is it presented at management review (9.3)? If reporting is a KPI and low numbers are praised, that supports the operative's account.
- Observation: stand in the walkway for 20 minutes. Near misses observable and unreported are stronger evidence than any statement.
Interview two or three more operatives, individually and away from supervisors, before you form a view. Record what you did and what you found, including the negative results.
2. How do you respond to the request to route all worker interviews through the OH&S manager?
Read the worked answer
Interviews are only useful if people can speak freely, and the request as framed would compromise the reliability of the evidence. But the manager's stated reason - disruption - is legitimate, so negotiate rather than confront.
- Explain the purpose: you need to sample worker experience of reporting, and a manager-mediated summary is not audit evidence of what workers think; it is the auditee's interpretation.
- Offer a workable alternative: short interviews (10-15 minutes) in a quiet room off the floor, scheduled around handovers, no supervisor or manager present, names not recorded, and a summary of themes shared with management afterwards.
- Be clear that access to workers and to records is part of the audit arrangement. If access is ultimately refused or only granted with a manager present, you have an audit limitation: record it, raise it with the auditee's top management, and inform the certification body's programme manager. If evidence essential to the audit objective cannot be obtained, the audit may need to be suspended rather than concluded with an unqualified opinion.
Do not let the negotiation consume the 90 minutes you have for Clause 10. Escalate early and continue with records while the access question is resolved.
3. If the suppression of reporting is confirmed, what finding or findings would you raise, and against which requirements?
Read the worked answer
Discipline yourself to anchor to a requirement; 'poor safety culture' is an observation, not a nonconformity.
- 10.2.1 is the primary anchor. The organization must establish, implement and maintain a process for reporting, investigating and taking action on incidents and nonconformities. A process that exists in a procedure but is actively discouraged at supervisory level is not implemented. Word it around the evidence: shift supervisors telling workers that near misses cost time, and the reporting rate falling.
- 7.4.1/7.4.2 (internal communication) if the message being communicated contradicts the reporting process.
- 5.4 (consultation and participation) if workers have no effective means of participating in hazard reporting and the safety committee is not functioning as a channel.
- 5.1 (leadership) only if you can show that top management knew or should have known - for example, low reporting appeared at management review and was not questioned. Leadership findings must be evidence-based, not inferred from the shop floor.
Grading depends on extent: if the behaviour is confined to one shift supervisor and the site otherwise detects and acts on under-reporting, that is minor; if it is accepted practice across shifts and undetected by the site's own monitoring and management review, it is major. Note the point in 9.1.1 and 9.3 - the organization had the data and drew no conclusion from it.