EHS OutLoud Blog
Opinion: The Real Top Cause of Incidents

Opinion: The Real Top Cause of Incidents

I often see posts, blogs and articles on the common causes of incidents or the top three, or top ten, causes of incidents. Many of these discussions revolve around unsafe acts, or errors, by workers. The unfortunate truth is far too many believe these are actual causes.

Other blogs, articles or discussions revolve around underlying causes or “root” causes. Root cause is a term that predates modern safety and has no specific meaning within the safety world. Since the safety lexicon lacks any real specificity in terms, these discussions often go off the rails as many demonstrate they do not even understand the principle of multiple causation or underlying causes in company systems or processes.

These last two paragraphs could summarize most of the discussions I see in online forums regarding safety incidents and investigations. All of these diverge into other areas as people argue over theories and their validity. As human beings, we feel the need to categorize things (common causes or top three) but really never use specific language. That is something that continues to get us into trouble.

Now, I know you are reading this looking for subheads to explain the most common cause and could be secretly wondering if it really is unsafe employees. Well, let’s get past that – is just isn’t. It is not “unsafe acts” or “worker error.” You can even leave out “human factors.” It is not people, and it never has been.

To promote understanding, we humans need categories, and we often want to present the clearest picture we can. Incidents are complex because of randomness. There are so many variables that we cannot ever hope to control them all. That includes people. We look at people as the weakest part of a system, as they are responsible for many random factors that reasonably cannot be controlled. Many programs strive to control people’s perception and action. They see the obvious problem as that people’s behavior must be controlled. Strangely, people simultaneously also are what gives a system its resilience and strength.

There is an incident cause that absolutely is in my top three. It is one I never see mentioned. It is easy to understand why. It is not possible for to know something unless you actively seek that knowledge. Ignorance not only is bliss, but it also is instantaneous. We often really only understand only that which we seek to understand. These things that confirm our place in the world and reinforce our view of things. Confirmation bias is very common. We see what we expect to see. Our amazing brain fills in the gaps with assumptions based on our experience.

So here it is – a top cause of incidents: poor investigations – poor investigative processes, techniques and investigators. I know that is three things, but they group together nicely. Now some are thinking they do a good investigation because they use some sort of tool. I would beg to differ. A tool only is as good as the person wielding it. Being given a paintbrush does not mean you can replicate Michelangelo or Vermeer.

A poorly-investigated incident means that an opportunity has been lost – an opportunity to fix the underlying issues and reduce the likelihood of a recurrence or similar incident. Interestingly, it is also an opportunity to demonstrate value. Safety people often conduct, or lead, most investigations. Yet, from my experience, the quality could be better.

Investigators usually fall into three categories, supervisors, safety people, and those assigned. Supervisors investigating themselves are not likely to find that the employee was not properly supervised or improperly trained on the job. Those assigned may include supervisors or may just be a manager or someone else. They usually lack any training (like most supervisors) and focus on filling out the report. Most of the time there is a real demand for a report to be done quickly.

Besides, this stuff is so easy. Once we gather a few facts, we automatically know what probably happened. Our amazing brain fills in the rest and assumptions become facts; the report almost writes itself. That is the focus after all right? Complete a multipage report to submit to someone (the author having no idea what happens to it after that). Job Done.

Then there are the safety people. They may, or may not, have training. They almost certainly have some experience, and this makes them much more able than the other two groups to fill in the gaps and find a “story” (hypothesis) that matches the actual facts available. Could it be that in some cases safety people are worse at investigations than someone untrained?

Well, it could be. If you believe in Heinrich (88 percent of all incidents are caused by unsafe acts), behavior based safety (90 percent of all incidents are caused by unsafe behaviors) or pyramids that tell the future (pyramid ratios, various authors), then you are probably not great at investigating incidents. There may just be some of the previously mentioned confirmation bias at work.

I get to see a lot of incident investigations, and frankly, most of them are not good. Some organizations do not have a lot of incidents, so they get very little practice in investigations. Most still lack a good process for conducting, reviewing, and analyzing incidents. So when one occurs they, having very little practice, they are unable to seize on the golden opportunity to find and fix problems in their system.

The underlying cause of that may lie in a system that is focused on hazard rather than risk. Focused on compliance, rather than risk management. These sorts of systems tend to believe interesting things like “be careful” is a real control. If you check your hazard assessments and “situational awareness” is in the control column, chances are your incident investigations will do a great job of finding out that the employee involved lacked situational awareness. Such incident investigations contain actions like “remind the employee to be aware of surroundings” or “brief/rebrief employee on procedure and sign off.”

More sophisticated investigations may find that retraining is need since there was a “failure to follow procedure.” Retraining is a really popular one. Now a procedure is a real control (administrative) but isn’t there a hierarchy? Is the problem that the employee did not get the training, the training is ineffective, or that the training is not relevant? More ineffective training is unlikely to solve any issues.

It is tough to rank your incidents by risk if you use a 3X3 risk matrix as it lacks the specificity needed to really understand risk. 5X5 is the most common because it is the most intuitive and gives almost three times the possibilities of a 3X3 matrix. It is true that how you manage risk and assess hazards will bias your incident reports in a positive or negative way.

In other cases, the pressure to produce a report makes the investigation cursory since many do not differentiate between a report and an investigation. Reports are there to communicate to the organization and even external stakeholders the basic information about the incident. “We are investigating” ought to be an acceptable response to a query about substantive details. Differentiating between report and investigation is critical to success. Rushed by a deadline and carrying a good deal of confirmation bias, the result is a poor investigation completed on time. It is a missed opportunity to identify real gaps in company systems.

The truth here is that an investigation takes time, diligence and skill. Investigating is not really a natural process. Although we spend time training people, that only is simulation. The real thing is a good deal more complex. Tools and theories are available, but using them takes practice and a questioning attitude to override the natural confirmation bias. Practice is something we often do not get enough of – unless we are investigating near misses. Not every near miss is worth investigating, but it often takes a bit of time to even determine that.

The next time you have to investigate something, get out your report in the allotted time frame and then finish your investigation. In the end, you need to ask yourself if you have solved the system or organization issue(s). Ask yourself whether you have seen past the immediate or proximate causes to something beyond those causes focused on the involved employee(s).

Safety professionals can make a difference in the quality of investigation by helping others see what they would not normally see: to understand why something happened rather than simply what and how. It is very easy to do a quick investigation and file it away. Do not waste the opportunity presented by an incident investigation to peer into company systems and actually address gaps. A more critical approach and deeper understanding of causation can lead us to solve issues that are not obvious and show value to the employer.

TAGS: Safety
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.