Welcome Guest! The IOSH forums are a free resource to both members and non-members. Login or register to use them

Postings made by forum users are personal opinions. IOSH is not responsible for the content or accuracy of any of the information contained in forum postings. Please carefully consider any advice you receive.

Notification

Icon
Error

Options
Go to last post Go to first unread
Bill6152  
#1 Posted : 29 November 2017 08:18:16(UTC)
Rank: Forum user
Bill6152

Over the last few months within our business I have seen an increasing number of accidents where route cause is either " not working as trained" or "lack of attention" and for me it’s just too easy to put these down as a cause with nothing else identifying why people are not working as trained or why is the lack of attention a factor in the accident occurring. Was just wondering what other people do to tackle this, how you get more detailed valuable information from accident investigations and reports 

hilary  
#2 Posted : 29 November 2017 08:44:22(UTC)
Rank: Super forum user
hilary

Bill

It's root cause - getting to the root of the problem, rather than route cause which would be following the path of the accident.

You need to adopt a methodology for gettign to the root - 5 whys is a good starter.  You ask why the accident happened and when you get an answer like "he didn't follow the procedure" you then ask "why" and so on until you get to the ultimate reason or reasons for the accident.  Sometimes it is simply operator error, but other times you will find a significant underlying cause but you have to dig deep for it ie getting to the root of the issue.

thanks 1 user thanked hilary for this useful post.
A Kurdziel on 29/11/2017(UTC)
Safety Shadow  
#3 Posted : 29 November 2017 08:45:44(UTC)
Rank: Forum user
Safety Shadow

Hi, In my opinion I feel that this may not be the root cause as you haven’t exhausted all your questions, such as …why people are “not working as trained”. Can this be because it’s a job related factor such as unrealistic time frames for a job to be complete, hence the reason why people might be cutting corners and “not working as trained” to save time?. You may also want to dive deeper in to the training that staff receive for the task in hand, has it been delivered effectively and have the learners (staff) fully understood why they have to carry out a job “as trained”. Has there been an assessment process at the end of the training sessions to ensure all learners have fully grasped the course outline and where necessary has feedback been given to learners who may have struggled, or have they just been sent on their way to carry out the task? As for “lack of attention” you need to ask yourself why? Is it that staff are exhausted? Can frequent breaks be added in to task to refresh staffs attention spam? Etc. I hope this helps a little.
Hazzard41579  
#4 Posted : 29 November 2017 09:08:06(UTC)
Rank: Forum user
Hazzard41579

Bill

Where there is lack of attention, I would also add as a root cause - Lack of risk awareness.

Are they fully aware of all the risks associated with the operation and the consequences of lacking attention?

Agree with previous posts on asking Why then why again until satisfied you have nailed it.

fairlieg  
#5 Posted : 29 November 2017 09:55:29(UTC)
Rank: Forum user
fairlieg

I investigate using a few methods and RCA to me is too linear a process which is why you get to those end points which are really starting points, great for machine/equipment failure but limiting for operational failures.  If you’re getting to these issues you must ask what’s wrong with the training, procedures, risk assessments etc that causes people not to follow them.  If the folk are just bad people how did you manage to hire them.  I am sure however, there are other reasons, it's probably not those things that are the problem it’s usually how the work is arranged in reality and the variability that doesn’t allow people to follow training, procedures etc.

Human error, training etc should be the starting point.  RCA's tend to look at conditions rather than methods, so does 5 why etc.  You need to understand how the work is done v's how it was planned (look up Erik Holnagell and safetyII).  There you will find the gaps, these are going to be where the workers deviated from the plan but there will be a reason why it made sense for them to make that deviation (this is called local rationality, have a look at Sidney Dekkers Field Guide to Human Error and Todd Conklin Better Questions).

Avoid Hindsight Bias (he should have, could have etc)

Avoid Confirmation Bias (finding evidence that confirms yours or others’ ideas of what happened)

Avoid Fundamental Attribution Error (Google the term)

For better investigation outcomes, try to include folk who do the work and find out how they do similar work, how the work is arranged what the priorities are and the variability they face when they are doing the work.  Facilitate discussions don't investigate and in interview.  Don't punish and blame, you can’t do that and learn and improve.  Get them (Spvs, workers, managers involved in the operation) in a room for at least an hr talk about the incident, and find out how they do the work, where are the pain points, what rules they break (where are the deviations), why and when do they break them.  Every reason is valid even if they don't believe the rules make sense or they are tool complex.  Take some time away and let it soak then reconvene to look at solutions....  You will get more out of this than RCA and 5 Whys.  then you need to manage the solutions like a MOC project.  Keep feeding back to the participants so they can see what actions have been taken and the improvements this helps future engagement.

Work with leadership to change their reaction to failure too this can impact what people are prepared to share bad reactions will obviously dive people away.  And try not to wait for an accident to happen, successful work will be riddled with deviations from the plans, review your high risk activities and try this process you might be surprised at what you find

Good article here https://www.nv-reliability.com/single-post/2017/07/03/Introduction-to-the-New-View-Safety2-Safety-Differently

Podcast from Todd http://preaccidentpodcast.podbean.com/

Good Information from Ivan here http://www.safetydifferently.com/understanding-and-adding-to-the-investigation-toolbox/

Edited by user 29 November 2017 09:59:40(UTC)  | Reason: Not specified

thanks 1 user thanked fairlieg for this useful post.
A Kurdziel on 29/11/2017(UTC)
A Kurdziel  
#6 Posted : 29 November 2017 12:55:19(UTC)
Rank: Super forum user
A Kurdziel

Essentially, when you ask enough ‘whys’ you get to the final root cause which is that management is failing in some way or other. This is why root cause analysis is always a bit controversial: it is easier to blame employees rather that the bosses.  

Our current accident reporting system has 22 “root causes” which seem to have been selected at random and largely meant to blame employees. Once we have sorted it out, I hope to reduce the system to about 4 root causes. Note the plural; it is rare for there to be a single “root cause”.  To give me some sort of framework I suggest:

  • Failure at the SSOW level- not following whatever controls have been agreed- there are of course reasons why that has happened.
  • Failure due to the employees lack of competency - managers should make sure that an employee’s training and experience aligns with the requirements of the job
  • Inappropriate supervision- unrealistic deadlines,  wrong training, equipment not fit for purpose
  • Cultural issues- basically no one cares how a job is done as long as it is sort of done
jontyjohnston  
#7 Posted : 29 November 2017 13:28:38(UTC)
Rank: Super forum user
jontyjohnston

Interesting perspective fairleig!

The problems with many investigation techniques is that they are riddled with bias to begin with so they only take you so far. There are a couple of very useful and simple models for taking the investigation beyone the lack of attention type of finding. Look at James Reasons GEMS model or HSG48 for some examples.

Taking the "lack of attention" example this could be further distilled to a skill based lapse or perhaps rule based mistake, the former normally caused by internal distractions (football, divorce, etc.) or the the latter external distraction (someone interupting the employee).

Very interesting subject once you get into it.

RayRapp  
#8 Posted : 29 November 2017 14:04:52(UTC)
Rank: Super forum user
RayRapp

I agree with most of the comments in that the concept of a 'root cause' is essentially flawed because most noteworthy accidents/incidents have multiple causes - deciding which one is the 'root' is subjective and unnecessary in my opinon.

I used to work for a company who had software for accident causation and it was a nightmare. There are so many potential causes for each and every failure the list is almost endless.

fairlieg  
#9 Posted : 29 November 2017 15:27:52(UTC)
Rank: Forum user
fairlieg

Jonty, the problem with Reason is that he focuses on the congnitive factors in human error.  The OP was asking for methods to get away from the worker being the problem.

Focus on the cognitive factors and you end up going to fix behaviours (BBS) while not really looking at environmental or other context-related factors (variation, organisation of the work, ETTO etc.).  Although he does say in his book Human Error, 2001 “you can't change the human condition but you can change the condition under which humans work”, the problem is with GEMS you start looking at skill-based, rule based and knowledge based errors so like you say there is aleady a bias there because those will be the target aspects of the investigation.

Johan Bergtrom explains the difference here https://www.youtube.com/watch?v=rHeukoWWtQ8

jontyjohnston  
#10 Posted : 29 November 2017 16:06:17(UTC)
Rank: Super forum user
jontyjohnston

Dont disagree with you fairlieg, was just giving a wider perspective to illustrate that there are a range of causal factors to be considered in a robust investigation, which often gets curtailed with failed too...lack off....etc.

Mentioned Reason only because he would be most widely recognised, I did a presentation once on his accident trajectory model entitiled "The Swiss Cheese Model of Accident Trajectory...a Model with more hols than swiss cheese"......looks very like dominos with holes instead of dots...so I have researched his work but not such a big fan myself.

As a professional accident investigator the key is to avoid all bias so I start with a blank A3 sheet of paper to map the causal factor, and if I cant get it on one sheet I have not properly understood the sequence.

Again just tossing perspectives around in the interests of a balance debate.

Jonty

Users browsing this topic
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.