Dear All,

Writing this due to some ongoing discussion with my team who is overtly focussing on human behaviors as root causes of accidents, and want to undermine the significance of the technical and procedures factors contributing to it.

To a great extent, I understand and accept that there is an human element involved in all the facilures, be it technical or procedural. But, while dissecting an accident should we ignore the weakness in engineering and procedural issues and focus more on 'why' the individual failed to pick up those weakness and modify his behavior.

I would like to hear more from experts here as to why it wouldnt be appropriate to attribute root causes of accidents to human behaviors? And to focus on engagement of people alone can fix the larger technical and system issues? How reliable is this approach, while I completely agree that engagement and focussing on human behavior is a very critical aspect to build an overall culture. But, does engagement alone solve the problem?

Looking forward to hear more on this.

Regards.

P

I agree with above post, a big subject.

Historically human error was blamed on the majority of accidents. Management was content to allow the operator to take the blame. However, we now have a better understanding of human factors and human behaviour. So whilst the majority of accidents are still caused by human error the more serious incidents normally involve a combination of causal factors i.e. planning, supervision, training, equipment, etc.

Taking an example of railways accidents, drivers passing a signal at danger (SPAD) the driver was blamed and often prosecuted following a fatal accident (assuming the driver survived). The Cullen enquiry in 2000 into the Ladbroke Grove and Southall train disaster completly blew the lid off the notion that train drivers were the main cause of train crashes. Since the enquiry, rail infrastructure has been fitted with TPWS, which prevents trains from passing a red signal at a junction. Hence train disasters have been consigned to history.   

We shouldn't shy away from identifying human error as a direct cause of accidents but it is never a root cause.

For a simple resource to encourage understanding of why human errors occur this prompt list from HSE of Performance Influencing Factors is useful https://www.hse.gov.uk/h...nfactors/topics/pifs.pdf

Violations are another issue, where people break rules. But, in the vast majority of cases people violate for very good reasons, usually to get the job done because they get no person benefit. People can be squeamish about identifying violations because they can lead to blame, but I would prefer to know about them provided people act like grown-ups when considering why they happen.

Originally Posted by: andybz

... in the vast majority of cases people violate for very good reasons, usually to get the job done...

Not sure I concur that to get the job done is a very good reason to violate albeit the excuse most often proferred

To me a very good reason to violate is when following the rules, training and procedures would directly result in death or major injury and by violation you prevent the outcome.

Normally such circumstances prove that the rules, training and procedures were wanting.

Human error/behaviour can be seen in cases as the immediate cause but certainly not the Root Cause. the only real time that human behaviour could be a root cause is in a violation scenario which is extremely rare cases. there are complex antecedents as to why a person chose to act in a certain way that led to the incident. it is only through investigating those antecedents shall you find the Root Cause. look at the work completed by Rasmussen for a more in-depth idea.

Some years back I completed an Accident/Loss Investigation and Evidence Gathering course sanctioned by IOSH, the trainer stated that the term 'Root Cause' was no longer appropriate. Accidents and incidents have multiple causes. Okay, you could argue some causal factors have more gravitas than others.

Since completing the course I have never used the term root cause. Unfortunately some accident/incident forms still use the term, which should be banished for good as far as I'm concerned. 

Whilst sympathising with Ray's view that perhaps we shouldn't use the term "root cause", I think that human failure could be such a cause - in the scenario where senior executives have deliberately chosen not to the right thing.

Plenty of examples of this happening!

Most of the recorded ones involve large corporations, but there was also the owner of the car wash prosecuted for manslaughter after a victim of modern slavery was electrocuted.

But in such a case, nobody wastes time on a formalised root cause analysis.

“Indeed, in my experience the vast majority of accidents have been either someone just not thinking or violating procedures and rules to "get the job done" which I would never ever class that as a good reason.”

The question is why people not are thinking or violating rules.

If as suggested you look at HSG 48- “Reducing error and influencing behaviour” This explains why people   do these things and what we can do to either stop them doing this or mitigate them ie prevent them doing something stupid even if they want  to.

That is down to management, which as someone said is something we can deal us with. Simply putting it down to human error is fairly pointless since we cannot reprogramme people to think or follow rules. The only way to eliminate the human factor is to totally automate the process.

Agree with the above two posts

Workers violate rules and take short cut all the time, it’s normal behaviour. Maybe not desirable but normal, as such there are more questions to ask when we find violations rather that pointing at them as a cause.  What was the cause of the violation and more to the point why did they not get away with it this time. (ref Challenger Disaster, Normalisation of Deviation)

I think it was the late Jens Rasmussen that said workers experiment all the time.  Sometimes the experiment goes well and sometimes it doesn’t

Most of time things go well for the worker regardless of if they are following the rules (most of the time they don’t).  There is a tendency to tie the violation to the incident in isolation but we only find out about violations when things go wrong.

I said before what’s happening when nothing bad is happening?

Also we cant push blame up the hierarchy either don’t investigate functions in silos understand how they interact as this can cause leaders to behave in a way that can influence the context the worker operates in (organisations culture & psychology)

If you’re getting to the point where you think the worker is just an idiot you have to ask what’s wrong with the recruitment process and supervision that allows the “idiot” to get there in the first place….

and the systems were not robust enough to accomodate, trap or correct that mistake

There is your underlying/root cause - the human error is the immediate cause. For me the question nowwould be is it reasonably practicable to correect that root cause - and sometimes the answer will be no. Yes we all make mistakes, the predictable ones we should try to control, the unpradictable ones we have no chance with.