Summary
In short
- Report volume is not a measure of programme health. A rising count can indicate improving reporting culture or deteriorating conditions, and the number alone cannot distinguish them.
- Classify potential severity at the point of reporting. Without it, the reports with fatal potential are lost among the ones about spills and trailing cables.
- Reporting has a cost to the reporter: time, attention, and sometimes the implication that they were involved in something. Reducing that cost matters more than promoting the programme.
- Feedback is what sustains reporting. A report that produces no visible response teaches the reporter that the next one is not worth making.
- Research on serious injury prevention indicates that events with fatal potential have different causes from routine ones, which is the case for separating them rather than analysing the whole set together.
- The most useful question on the form is what would have had to be different for someone to be hurt. It produces the control gap directly.
What it is
What it is
What is a near miss?
An unplanned event that did not result in injury, illness or damage but had the potential to do so. The definition is broad by design, which is why the classification of potential severity matters more than the definition itself.
How does it differ from a serious potential incident?
A serious potential incident is a near miss whose credible worst outcome was death or life-altering injury. The distinction matters because research indicates those events have different root causes from routine near misses, so grouping them together means the analysis is dominated by a different failure profile.
When to use it
When to use it, and when not to
This captures events with no outcome but potential for one. Related records handle events that did have outcomes.
Use it for
- Any unplanned event with potential for injury, illness or damage that did not produce one
- Unsafe conditions found and corrected, where recording them supports pattern analysis
- Events reported by contractors, agency staff and visitors on site
- Observations from inspections and tours that identify an absent or defeated control
- Events that would be trivial in isolation but recur in the same place or task
Not for
- The incident report, for events that did produce injury, illness or damage
- The serious potential incident report, where the credible worst outcome was death or life-altering injury
- Hazard reporting for standing conditions rather than events, where a separate route exists
- Maintenance defect reporting, which should route to work orders rather than to safety records
- Quality and food safety deviations, which have their own processes
Standards
What it is built against
Near miss reporting sits within incident investigation and hazard identification requirements rather than a standard of its own.
| Clause | Requirement | Where it lands |
|---|---|---|
| ISO 45001 cl.10.2 | Incident investigation covering incidents, which includes events that did not result in injury | What happened |
| ISO 45001 cl.6.1.2 | Hazard identification drawing on incidents and near misses as a reactive source | Follow up |
| ISO 45001 cl.5.4 | Worker participation, including protection from reprisal for reporting | Event |
| ISO 45001 cl.7.4 | Communication, including feedback to those who report | Follow up |
| 29 CFR 1904.35(b)(1)(iv) | Prohibition on discharging or discriminating against an employee for reporting | Event |
| OSHA recordkeeping guidance | Reporting procedures must not deter or discourage reporting | Event |
| ISO 45001 cl.9.1 | Monitoring with indicators that measure what is intended rather than what is easy | How close was it |
| ISO 45001 cl.8.1.4 | Coordination with contractors, whose near misses belong in the same system | Where |
What it does not cover
- The incident report, for events producing injury, illness or damage.
- The serious potential incident report, where fatal or life-altering potential existed.
- Hazard reporting for standing conditions rather than discrete events.
- Maintenance defect reporting, which should route to work orders.
- Root cause analysis, applied selectively where the potential warrants it.
Filling it in
Filling it in well
Make it cheap to report, classify potential immediately, and close the loop visibly.
A form requiring a login, ten fields and a narrative will be completed by people with time and an office. Reporting from a phone in under a minute, with a photograph and a location, reaches the people who see most of what happens. The barrier is nearly always effort rather than willingness.
The single most productive field. What would have needed to change for this to hurt someone: a step to the left, a second earlier, a heavier load. It produces both the potential severity and the control gap in one answer, and it is easier to answer honestly than a severity rating.
Grade what could have happened, immediately. Reports carrying fatal or life-altering potential should route separately and receive investigation proportionate to that potential, because research indicates they have different causes from routine events and analysing them together buries them.
Tell the reporter what happened, and make the outcome visible locally. Reporting is sustained by evidence that it produces change, and a report absorbed into a system with no response teaches the reporter and everyone they tell that it was not worth the minute it took.
Audit findings
Common audit findings
Near miss findings are mostly about the programme rather than the events.
| Finding | Clause | What fixes it |
|---|---|---|
| Report volume used as the primary programme measure. | ISO 45001 cl.9.1 | Measure serious potential proportion, actions arising and time to feedback instead. |
| No classification of potential severity. | ISO 45001 cl.10.2 | Grade at report; without it the significant events are indistinguishable. |
| Reporting requires effort that deters the people who see most. | 1904.35(b)(1)(iv) | Reduce to under a minute from a phone; effort is the usual barrier. |
| No feedback to reporters. | ISO 45001 cl.7.4 | Close the loop visibly; unanswered reports end the reporting. |
| Reports investigated uniformly regardless of potential. | ISO 45001 cl.10.2 | Scale investigation to potential; uniform treatment wastes effort and misses the serious ones. |
| Contractor and agency near misses outside the system. | ISO 45001 cl.8.1.4 | Include them; they perform much of the higher-risk work. |
| Reporting associated with blame or performance consequences. | 1904.35 | Separate entirely; any consequence attached to reporting ends it quickly. |
| Recurring reports from one location with no cumulative response. | ISO 45001 cl.6.1.2 | Aggregate by location and task; recurrence is the finding. |
| Actions closed without verification that the condition changed. | ISO 45001 cl.10.2 | Verify at the point of work; closure on paper is not correction. |
| Programme measured by target per person, producing housekeeping reports. | ISO 45001 cl.9.1 | Targets shape what gets reported; expect what you incentivise. |
Worked case
Case in point: four thousand reports and one that mattered
A site introduced a near miss target of one report per employee per quarter. Volume rose from around two hundred a year to over four thousand. The programme was presented as a success and the reporting rate was benchmarked favourably against sector peers.
A review of a year's reports found that roughly ninety percent concerned housekeeping: spills, trailing cables, obstructed walkways, poor lighting. All genuine, all worth fixing, and all low potential. The reports were triaged by a coordinator who read them in batches.
Among them was a report describing a load that had shifted on a truck as it passed a pedestrian route. It had been categorised alongside the spills, closed with a note to the driver, and never seen by anyone who would have recognised it as an event with fatal potential.
Definitions
Definitions and key terms
- Near miss
- An unplanned event with potential for injury, illness or damage that did not produce one.
- Potential severity
- The credible worst outcome had circumstances differed slightly, classified at the point of reporting.
- Serious potential
- Potential for death or life-altering injury, warranting separate routing and investigation depth.
- Reporting cost
- The time, effort and social risk borne by the reporter, and the main determinant of reporting rate.
- Closing the loop
- Visible feedback to the reporter and the area about what the report produced.
- Triage
- Sorting reports by potential so investigation effort is directed proportionately.
- Recurrence
- Repeated reports from one location or task, which carry more information than any single report.
- Leading indicator
- A measure of activity or condition preceding harm, of which near miss data is one where it is classified.
FAQ
Frequently asked questions
Is a higher near miss count good?+
It is ambiguous, which is why it is a poor primary measure. A rising count can mean improving reporting culture or deteriorating conditions, and the number cannot distinguish them. More informative measures are the proportion of reports carrying serious potential, the proportion producing a change, and the time from report to visible feedback.
What is the most useful question on the form?+
What would have had to be different for someone to be hurt. It produces the potential severity and the control gap in a single answer, it is easier to answer honestly than a numeric severity rating, and it directs the reporter's attention to the mechanism rather than to the outcome that did not occur.
How do we stop the significant reports being buried?+
Classify potential at the point of reporting, by the person who saw the event, and route serious potential separately. Where triage happens later, in batches, by someone reading at volume, the small number of significant reports are sorted alongside housekeeping and receive the same response.
Why do reporting rates fall?+
Almost always because reporting costs something and produces nothing. The cost may be effort, or the implication of involvement, or a conversation with a supervisor. The absence of visible response is the more common cause: a report that disappears teaches the reporter, and everyone they mention it to, that the next one is not worth making.
Should we set reporting targets?+
Cautiously, because targets shape what is reported rather than what is noticed. A per-person quota reliably produces housekeeping reports, which are quick to write and carry no implication about anyone. If a target is used, pair it with measures of potential and of action, or the programme optimises for volume.
The agents
What the agents do with it
The report is a minute of someone's time. What fails is the classification that never happened and the feedback that never came.
Captures reports in under a minute with a photograph and location, classifies potential at the point of report, and routes serious potential separately.
Aggregates by location and task so recurrence surfaces, and tracks time from report to visible feedback as a programme measure.
Brings contractor and agency reports into the same system, where much of the higher-risk work sits.
Routes defect reports to work orders rather than into the safety record, keeping the near miss data about events.
This template lives in KnowSafe — safety and compliance. Incidents, hazards, permits, inspections and the critical controls behind them.
Meet KnowSafe→Sources
Sources
- ISO 45001:2018 clauses 10.2, 6.1.2, 5.4 and 9.1
- 29 CFR 1904.35, employee involvement and prohibition on discouraging reporting, OSHA
- Campbell Institute research on serious injury and fatality precursors
- HSE guidance on incident reporting and investigation (GB)
- ISO 45001:2018 clause 8.1.4, procurement and contractors