NEAR MISS

In its inspector guidance notes, the HSE sums up the benefits of near miss recording by saying it may help to prevent a reccurrence, identify any weaknesses in operational procedures, and — when reviewed over time — may reveal patterns from which lessons can be learned.
There is no disagreement among safety professionals on the need to report, record and assess near misses to improve safety. Where there is some confusion is in deciding what counts as a near miss.
But before we try to unpick this question, the classic accident triangle provides a useful reminder of why it is important to report and investigate near misses. The triangle is presented with different numbers in each layer, and with different numbers of layers, depending on the data used, but the principle is illustrated in Figure 1 (opposite).
The chance to learn from the investigation of a death is an opportunity no health and safety professional would seek, and that most would never have had. Nationally, and within particular industries, numbers of deaths may be a (lagging) indicator, but a single large accident or changes in the economy leading to a reduction in the workforce can cause figures to vary enormously.
Though recording and investigating injuries presents a more detailed picture, this is still a lagging indicator — measuring after the event. And in small organisations, recorded injuries may only be in single figures, making small, chance variations appear significant and providing few learning opportunities.
Recording and investigating near misses, on the other hand, not only helps you to assess the strength of your safety management system but also provides an opportunity to fix problems before injuries occur. At its most simple, reporting the frayed carpet that nearly tripped someone can trigger a repair to prevent the accident in which someone actually trips. Likewise, if someone noticed that many train drivers missed the same red signal at the same place on the many occasions where the train went on to arrive safely, something might have been changed before a driver missed the signal with fatal results.

Free lessons

In Inspector Guidance Note (IGN) 1.08, the HSE states: “Recording non-reportable near misses … may help to prevent a re-occurrence.  Recording these near misses can also help identify any weaknesses in operational procedures as deviations from normal good practice may only happen infrequently but could have potentially high consequences. A review of near misses over time may reveal patterns from which lessons can be learned.”
Later in IGN 1.08, the HSE defines a near miss as “any incident, accident or emergency which did not result in an injury”, and goes on to explain that particular types of near miss are defined under the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations (RIDDOR) as “dangerous occurrences” and must therefore be reported to the HSE.
Schedule 2 of the Regulations provides a detailed description of what type of occurrences are considered dangerous enough to report, including some failures of lifting equipment, pressure systems and scaffolding, as well as some incidents involving explosives, biological agents and breathing apparatus.
Reporting only these “dangerous” near misses, however, is not sufficient in a proactive system of safety management. To be useful, the definition of “near miss” must be broader than the dangerous occurrences in RIDDOR.
This is where it becomes tricky. Some people regard a near miss as the opposite of an accident. James Reason (The Human Contribution, 2008), for example, outlines four types of “error outcomes”. The first two types refer to outcomes where the results could have been harmful, but in the event there was no harm. He calls these “exceedances” where the error was at the edge of a known range of equipment or human behaviour (doing something too fast, too high, too much) or “free lessons” for less well defined “inconsequential unsafe acts”.
The third category Reason calls “incidents”, which he describes as “events of sufficient severity to be reported”, admitting there is no general agreement on what counts as “sufficient”.
He goes on to suggest that incidents tend to involve some financial loss or damage to equipment or production, and may even cause “minor harm” to people, but nothing too serious. The fourth category is “accidents”, which are events with significant adverse consequences, including injury and death.

Hit or miss?

This use of the word “accident” to mean only those events where harm is caused is common and the term “accident book” reinforces Reason’s view.
But if you look again at the HSE definition of a near miss in IGN 1.08, it includes the idea that an accident that does not result in an injury is a near miss, suggesting that, for the HSE, accidents can be both harmful and harmless. When a child says “it was an accident” after breaking your favourite ornament, they are not considering the extent of injury to a person; they are explaining that the event was unintended. This presents a completely different definition of an accident — that which is accidental. In the first case, a “near miss” is defined as the opposite of an accident; in the second case, a near miss can be the result of an accident (or a violation).
These different meanings are represented in Figure 2 (opposite). If there is to be a system within an organisation to collect and analyse near miss data there must be a common agreement on how the organisation is using these words.
Unfortunately, other HSE sources either do not define the word accident, or present contradictions. HSG 245 Investigating accidents and incidents uses the word accident 53 times, and the word incident 23 times, without defining either. And INDG 453 Reporting accidents and incidents at work offers: “An accident is a separate event to a death or injury, and is simply more than an event; it is something harmful that happens unexpectedly.” Never has the word “simply” been so out of place.
RIDDOR does not define an accident, but states that it includes acts of non-consensual physical violence done to a person at work, and suicide on a transport system. Definitions incorporating the idea that an accident is “unintended” are therefore inconsistent with RIDDOR, since suicide and violence are clearly intended by someone.
Other definitions of accident include the terms unexpected, undesirable, unforeseeable or unforeseen. The World Health Organisation (WHO) defines an accident as “an unplanned, unexpected, and undesired event, usually with an adverse consequence”. Note the ambiguity on the requirement for harm.
Daniel Shears, of the GMB union, also argues that the word “unexpected” — and its running mate “unforeseen” — is inappropriate within the definitions of accident or near miss. “In many cases, it would have been reasonable to expect that particular behaviour could result in an injury or ill health, even if the person involved did not expect it,” he explains.
As a member of PABIAC (the Paper and Board Industry Advisory Committee) involved in developing health and safety strategy for that sector, Shears has been involved in many discussions over how to define a near miss. The final PABIAC version wisely leaves out any debate about foreseeability, and includes consideration of what stopped the miss being a hit, (something that is missing in the HSE IGN and WHO definitions): “An unplanned event or situation that could have resulted in injury, illness, damage or loss but did not do so due to chance, corrective action or timely intervention.”

Get agreement

If the distinction between a near miss and an accident is whether harm is suffered, there needs to be a clear agreement on what is meant by harm.
When a fire destroys an entire factory, but everyone escapes promptly with not so much as a bruised knee or case of smoke inhalation, this is a near miss for those concerned only with the safety and health of people. For others, any fire or explosion, any damage to equipment, property or environment and any interruption to business, services or production are accidents. Some organisations consider an even wider range of outcomes, such as reputational damage or the potential for legal action.
For each category of harm — health, safety, equipment, buildings and so on — a near miss reporting scheme must include clear guidelines on how much harm is no harm, and is therefore a near miss, and how much is harmful, and should be reported as an incident. A key question is what level of harm is insignificant? Is a paper cut or a bumped knee a near miss if no work time is lost? For psychosocial hazards, this decision can be very subjective. One shop worker may wish to report a verbally aggressive customer whose behaviour left them shaken and fearful, while another worker may laugh off the same incident.
Some schemes are very specific in defining what people must report as near misses. Those working at height, for example, may have a specific requirement to report any dropped object, however small, from whatever height. The offshore industry set an excellent example with “DORIS” — the Dropped Object Register of Incidents and Statistics.

A near miss by any other name
A further debate is over whether “near miss” is the correct term. Synonyms to look out for include: near hit, near loss, undesired circumstance, close shave, close call, narrow escape and potential (error, event or accident). Industry specific examples of near misses have their own names — SPADS (signals passed at danger) on the railways or AirProx for the airline industry.

At the other end of the spectrum, some schemes are much more general, perhaps having “incident books” used to collect information as varied as what time someone arrived on shift and what messages have been passed on, to records of equipment alarms and first aid incidents.
The more you specify, the more you may limit what people report, but no limits may result in an avalanche of reports that are difficult to manage. So decide carefully what you are prepared to act on.

Culture is key

A reporting system tacked on to an immature safety culture is unlikely to be successful. When Asda distribution depots decided to develop a near miss system, it was on the back of an existing culture amongst the workforce summed up in its “Clean as you go” and “Don’t walk by” mottos.
“Asda distribution already had a successful accident reporting system, which included the reporting of ‘significant’ near misses,” explains trading law manager Chris Moore. “Our-fiveyear plan is to be the leading safety performer among distribution organisations, so including other near misses was a clear next step.” Significant near misses in the accident reporting system were defined and included vehicles departing on a red light and pallets dropping, but to encourage wide reporting the firm has not restricted what can be included as a near miss.
Shears also sees culture as a key to the success of near miss reporting. “Near miss reporting allows a more holistic approach to safety management — considering equipment, procedural and managerial near misses, not just the behaviour of individuals,” he says.
To be self-sustaining, a near miss system must include feedback. Asda provides feedback via a weekly safety committee at each site, with union safety reps in attendance.
Shears’ ideal is individual feedback to the near miss reporter. “It’s important to manage expectations. If something is going to change as a result of a near miss report, the reporting person should be given information about that to encourage future reporting. If nothing is going to happen as a result they should still be contacted and a reason given.”

Keeping a record

There are many different approaches to recording. Asda, for example, makes use of existing driver debriefing sessions. Drivers are prompted to report near misses to line managers, who complete the paperwork.
Other organisations develop their own systems — spreadsheets, email protocols and paper forms, with a management plan explaining how it all fits together. These systems can do the job, but can be inefficient in terms of time to create, modify and run.
By contrast, London City Airport (LCY) decided to use the ready-made and tested incident management module from SHE Software, which allows any member of staff to report a near miss from any computer, anonymously if preferred.
“SHE has made it easy for the safety team to manage the high levels of reporting,” says Annabelle Thorpe, airside safety coordinator at LCY. “Not only can we set and track actions for each report, but automated emails and escalation processes ensure that those responsible for each action are helped to prioritise their activities.”

"A key question is what level of harm is insignificant? Is a paper cut or a bumped knee a near miss if no work time is lost?"

Whether people complete a form online or on paper — by the person who noted the near miss or a line manager — you need to consider the content of the form. Standard fields for near miss reporting forms include the date, time and location of the event, and a description of the event. Some organisations request the name of the person involved in, or reporting, the incident, but as LCY found, allowing anonymous submissions encourages a higher level of reporting.
Sometimes, forms ask for additional information about the person (role, date of birth, department) but if you want to encourage reporting ask for as little information as you need to identify the person; you can extract other relevant information from the HR system.
Near miss descriptions may be requested as free text, while other forms provide prompts for things like the specific equipment or materials involved, such as personal protective equipment (PPE), hazardous substances or electrical machinery.
Other forms prompt for more detailed information. I have seen a form that asks the reporter to identify immediate and secondary causes and possible remedial actions required, and another that asks the reporter to categorise the incident’s potential severity. For a consistent approach, such analysis is best done by a competent person. 
Shears at the GMB agrees, recommending that to provide consistency, assigning potential severity should be left to the health and safety manager reviewing the reports.
“However,” he notes, “we have found it useful in PABIAC to include some hazard topics to encourage near miss reporting in areas where we are focusing our hazard reduction efforts, such as manual handling and workplace transport.”
Asda prefers not to provide categories. “We don’t want to restrict the near misses that people report,” Moore explains. “A list of near miss types might make people think that if what they had spotted didn’t fit into a category, it wasn’t worth reporting. We’d rather they reported the trivia than left out something important.”
A final but crucial point is that any debate about what constitutes a near miss will do us no good if the effort and resource is not available to put things right.
The HSE advises simply: “Where a review of near miss information reveals that changes to ways of operating, risk assessments or safety management arrangements are needed, these changes should be put into effect.”