Post a Job

Repeat Accidents – Unlikely but possible

Accident investigators may have been influenced over the years to believe that “repeat accidents” exists. But the chances of an accident repeating itself, a month, year or 10 years from now, is highly unlikely. The only possible explanation for the “repeat accident theory” is the Root Cause Analysis Matrix employed to identify its “point of origin”.

Claims that incident and accident investigations have failed, and therefore the “repeat” of an accident is unfounded when one looks deeper into the circumstances leading up to the event.

Accident causation is pretty much like a sunset. Not one is the same. One can categorise the root causes into various groups; fault of management; fault in design etc etc, but the biggest influencer is human behaviour. It is the single most important variable that no investigator, no matter how experienced, will be able to predict to prevent a repeat or “recurrence” of the incident.

What makes a person put his/her finger into a moving machine? Yes, the machine should have been guarded, but why intentionally injure yourself? Should one not be more careful when working with “unsafe” machinery? Is it really possible to create a scenario to create the exact circumstances for a repeat accident?

So why then the “repeat accident theory”?

The main purpose of safety programs is to prevent incidents and accidents. Should one occur, an investigation is needed to answer three questions.

  • What happened?
  • Why did it happen?
  • What should be done to prevent it from happening again?

Astute investigators know that the first question is the easy part. This can be answered from witness testimony, video footage, computer simulation etc. The list of techniques is as long as a piece of string.

The second part of the answer is then to determine the likelihood that the accident actually happened the way described by the evidence.

The second and third questions are somewhat more difficult to answer, as the one depends on the other. And the “WHY” is often the reason investigators miss the “WHAT”.

Common preventive measures are often limited to “supervision and training or awareness” and occasionally it may call for a new risk assessment and safety work procedures. But it all boils down to shifting the blame to lower levels within the organisation.

Human-machine interaction

Machines do not injure people. (Well, not yet anyway). Machines cannot think and create intent to harm. But when a human operator is interacting with the machine, the risk of injury immediately multiplies exponentially. Regardless of the safety devices fitted to the machine, or not, the human remains in control, and the level of control is influenced by many factors.

Most safety programs are designed to ensure machine operators are competent and machines are “made safe”. But very few takes the human brain into account. What about fatigue, distractions, work pressure, etc, commonly referred to as “job stress”?

Does the RCAT we deploy take this into consideration?

And how is the stress appetite of two employees the same? It is not. It never will be.

And this principle makes each and every incident unique, never again to be repeated.

The worst-case approach

When designing machine safety systems, one cannot cater for every possible scenario. But one can with reasonable certainty design a safety system that will protect 95% of the time under “normal operating conditions”. Additional procedures and systems can also be developed for “abnormal operations” or “failure mode” operations to make it reasonably safe.

A few industries have mastered the art of failure mode safety, the leaders still being NASA.

The paper-cut investigator

Incident investigators often gets two days of training in “investigation and recording of incidents and accidents.” And they are doing a great job in investigating the obvious. And while a paper-cut is not fatal per se, it one of those incidents that has no cure; no prevention, and no safe work procedure. It will happen more than once in a person’s life. And it will hurt every time.

Certain “incidents” cannot be prevented. Yet the common belief is that all incidents can be prevented. False; not true, hogwash. Accept it and live with it.

Back to our paper-cut. Yes, even this can be prevented by issuing suction gloves to the office assistant or typist whenever he or she needs to handle a ream of paper. Or perhaps a charge of static electricity would make the paper “jump up and stick” to your hands? So, when we really want to become ridiculous, all incidents can probably be prevented. In a paper-less society yes, definitely.

That leaves the rock and scissors to fight it out.

The law is not ridiculous

The only people who expect an employer to prevent every possible incident, are those who depend on “a zero incidents” record for their annual bonus. And some will even hide the truth from an employer not to affect that perfect record. But when it comes to compliance, the law expects us to be reasonable and take action if we foresee a problem.

The perfect incident

If one accepts the “repeat accident theory” one also has to accept that an accident can be perfect. Perfect in its effect, causing the same injury over and over to as many workers as possible until management does something about it.

In a company with 10 amputations per year, all 10 must have been caused the same way, under the same circumstances by the same instrumental cause.

The theory of repeat accidents is nothing but fiction if one understands the underlying factors that may influence the cause thereof. So, don’t be too hard on yourself, thinking you have failed. Zero Harm and Reality still needs to reach agreement on the matter.