All Accident Are Incident
All Accident Are Incident
All Accident Are Incident
outline (implied) legal requirements and HSE guidance, particularly HSG 245
Investigating Accidents and Incidents - A Workbook for Employers, Unions, Safety
Representatives and Safety Professionals
outline use of failure tracing methods - such as fault tree analysis and event tree
analysis (ETA) - as investigative tools.
Note that in the second point "causes" is in the plural - there is more than one cause for every
accident.
What is an accident?
An early definition of accidents involving injury at work was proposed by Lord
MacNaughton in the case of Fenton v Thorley & Co in 1903, as follows:
"Some concrete happening which intervenes or obtrudes itself upon the normal cause of
employment. It has the ordinary everyday meaning of an unlooked-for mishap or an untoward
event which is not expected or designed by the victim."for students of the NEBOSH National
Diploma
This definition refers to a worker suffering a mishap which had a degree of unexpectedness,
but it's too narrow as it's only concerned with accidents that result in injury, and not all
accidents do.
A trawl of some 40 accident definitions found in general, legal, medical, scientific and health
and safety literature suggests the ideal accident definition should include reference to causes
and effects. Causes should include unexpectedness, unplanned events, multi-causality and
sequence/chain of events. Effects should include injury, disease, damage, near miss or loss.
Taking the best of the definitions, one that covers all the bases might run as follows: "An
accident is an unexpected, unplanned event, in a sequence of events that occurs through a
combination of causes; it results in physical harm - injury or disease - to an individual,
damage to property, a near miss, a loss or any combination of these effects."
All accidents should be investigated; not just those that result in injury. Any accident
investigation should focus on the multi-causal accident and not uni-causal injury (where there
is one).
environmental factors.
A thorough accident investigation process should therefore highlight all accident causes usually between 10 and 20 for each accident - and then provide the basis to develop control
measures designed to eliminate both immediate and underlying causes, resulting in a
continual improvement in the OSH management system. Remember, every negative needs a
positive and every cause needs a control.
Organisations should ask themselves the following questions about their accident
investigation processes:
investigate an accident on their patch, will come up with "lack of supervision" as one of the
contributory causes. Hence the need for a team of three investigators.
Once you have found all the causal factors and explored all the branches of the event tree to
their ends, then, and only then, can the investigative team get into control mode.
Never discount facts or possible causes that don't fit easily into the picture or the event tree.
These red herrings may be signs of another branch of the event tree that the investigation has
yet to follow. It's usually best to use the term "event tree" rather than "fault tree" as the latter
has definite negative connotations, which have no place in positive accident investigations.
Control measures
In control mode (Step 3) your team should effectively identify all risk control measures that
were missing, inadequate or misused. They should compare activities, conditions and
practices as they actually were in the run-up to the accident with what should have been in
place according to current best practice, agreed systems of work, legal requirements, codes of
practice, guidance and standards.
The team should identify those extra measures that are required to eliminate all immediate
and underlying causes by providing meaningful recommendations which can be properly
implemented to prevent a recurrence, and hence continually improve the OSH management
system.
Particular questions which may help here include:
What risk control systems (RCSs) and workplace precautions (WPs) are needed?
Step 4 - the action plan and its implementation - is the final step in the accident investigation
process. This step should provide a clear action plan with SMARTT objectives (ones that are
specific, measurable, agreed, realistic, time-bound and trackable) to deal effectively with all
the immediate and underlying causes of the accident. It should include lessons that have been
learnt which may be applied to prevent other accidents of a similar type/nature.
It should also provide feedback to people involved at all levels in the organisation to ensure
the findings and action plan recommendations are correct, realistic and fully address all the
issues. This plan should include feeding the findings back into a prompt review of the
existing risk assessment, as any accident is an indicator that a review may well be overdue.
The team should also ensure that the results of the investigation are shown to all concerned,
with the emphasis firmly on the resulting action plan, timescales, responsibilities and
accountabilities, and how the plan will be implemented and its progress chased and
monitored.
The risk control action plan should establish which RCSs and WPs should be implemented in
the immediate, short or medium term. The team should also note which risk assessments and
systems of work need to be reviewed, updated and publicised, and whether the accident
details and the resultant findings and recommendations have been recorded and analysed
from both a numerical and causal viewpoint (that is, reactive monitoring).
They should also flag up whether there are any common causes or trends which suggest the
need for further, deeper and detailed investigation. Finally, they should put a figure on the
overall cost of the accident - both insured and uninsured - and also cost the associated control
measures.
Using the four-step process to investigate all accidents from a causal viewpoint will certainly
improve overall OSH performance in the workplace. Cursory investigations where the only
control measure cited is "employee told to take more care" must be despatched to the health
and safety history books in favour of the much more scientific approach outlined above. They
certainly have no place in the NEBOSH Diploma syllabus.
Need to know
Questions to ask in an accident investigation include the following:
Were there adequate safe systems of work and did people stick to them?
What were the outcomes of the accident - injury, disease, damage, death, near miss,
loss?
Was the risk known? If yes, why was it not controlled? If no, why not?
Did the work organisation (or lack of it) impact on the accident?
Did the nature, shape or form of the materials influence the accident?
Did the work equipment influence the accident? Was it difficult/awkward to use?
Had the people involved received adequate information, instruction and training?
This is an article in the Know-How Series prepared on behalf of the National Examination
Board in Occupational Safety and Health (NEBOSH) by Lawrence Bamber, BSc, DIS,
CFIOSH, FIRM, MASSE