When Systems Alone Are Not Effective...

Due to much welcomed advancements in equipment, technology, automation, processes, application of safety management system systems and principles, robust risk controls etc. most safety critical industries look at their incident statistics and are happy to see a declining rate of significant incidents such as; major plant damage, disruptions to operations, serious injuries and fatalities. However, we know that sometimes it still goes tragically wrong. From my experience, there are a few primary reasons why the advancements are not preventing significant incidents and some simple things organisations can do to further enhance their safety efforts.

Issues of Concern

Statistically speaking, we can definitely see safety enhancements have resulted in lower rates of significant incidents…. However, what concerns me as an Incident Investigator are two common issues I tend to find when I am investigating a significant incident:

When I interrogate the organisation’s incident database searching for previous similar events involving that particular operation or plant and find a host of previous low level incidents have occurred. However, due to the actual consequence of these prior events being so low level they were considered as “nothing events” and in most situations simply filed away. Unfortunately another repeat occurred and this time, the consequence escalated up.
The fact that personnel were aware of the hazard and nobody was greatly surprised that the significant incident occurred, evidenced by the common first words witnesses utter at interview, such as: “I knew that was going to happen one day…”.

There’s nothing worse than investigating a fatality and very quickly seeing the reality was it could have been so easily prevented IF the incident data was used to conduct trends and identify and address the deficiencies and vulnerabilities shown by the data. Or IF the known hazard was officially reported and addressed.

Established vs. Effective

So what can organisation’s do to address these two issues? They say the first step in resolving anything is to identify the problem, the reasons for it and then come up with a solution. I believe some simple tips can help address these two issues if we broaden our view to focus not only if a system is established but spend the time to focus on why it may not be effective.

Time, Trending and Systemic Analysis of Data

Just finding the time to conduct an investigation can be a challenge. Rarely, do organisations have dedicated Incident Investigators who are just waiting for adverse incidents to occur so they can investigate them. In the majority of cases, organisations have personnel within a section called something like the Health, Safety and Environment Section (or, as I saw recently - the Health, Safety, Environment, Quality and Compliance Section). Most personnel who work in such sections of an organisation already have very busy workloads promoting safety or conducting compliance activities.

The fact these people also have to investigate incidents as and when they occur is problematic as they’re struggling to just do their usual activities, let alone have the time to conduct a quality incident investigation. When trying to conduct a low level investigation, rarely, if ever, are the Investigating personnel told “Don’t worry about your day job. We’ll take care of that while you focus on the investigation...”. More so, they are left to juggle their normal workload as well as the investigation (and don’t even get me started on the question if they’ve received adequate training to be able to conduct the investigation competently, have a well defined organisational investigation process or an Investigation Quality Standard - perhaps some issues I can write about next time…).

Consequently, for many personnel there is just not enough time to conduct quality investigations - particularly of the low level events. Many of these are simply entered into a database with minimal data collection with the focus on the actual consequence, not the potential consequence.

The sheer nature of a significant incident means that the organisation knows it will have to conduct a comprehensive investigation and adequate resources are usually allocated. However, there is not the same appreciation that the more frequent, lower level events can actually result in a huge pay-off for an organisation in terms of revealing the areas of vulnerability / deficiency that need to be addressed before a significant actual consequence.

From a risk based point of view, organisations can’t investigate every incident to the same depth. There are competing challenges such as operational pressures, resource challenges, a transient workforce and skills shortages. However, there should be recognition that the Incident Management Database is not just a repository of “stuff”, but should contain some very valuable data that can be used by an organisation to conduct trending. The challenge is that sometimes the validity of the data in the Incident Management Database is questionable.

While low level incidents should not be investigated to the same depth as a significant incident it is important to keep in mind the value of them. One by one, very minor incidents may be perceived as “nothings”, however, in actively trending the reports from the database it can often be seen that there is a trend of these low level “nothings” occurring. This should be a trigger to signify that there should be a systemic investigation undertaken to identify the common contributing factors of all these low level similar incidents.... because the next repeat incident could lead to an escalation resulting in a significant incident. We never want to find this during post-incident data collection. Analysis of the reports and trends in the incident management database should result in formulation of proactive programs addressing systemic issues identified.

Reporting of Hazards

In relation to the second issue where it’s found that personnel knew about the hazard and were not surprised by the incident, I believe we need to focus more on why that hazard was not recognised, reported or adequately addressed. It’s not enough to have a system or form for hazard reporting – all that shows is that something is established. We also need to assess the effectiveness of the system or process and this is where we find not enough consideration has been given as to how the system will work in an applied setting.

Listed below are some common views expressed from workers in the field why they might not adhere to reporting protocols:

“It’s too complicated”.
“It takes too long”.
“I’m too busy”.
“I don’t know how to use computers”.
“Why bother? Nothing ever happens anyway”.
“I thought someone else had already reported it”.
“I fixed it when I saw it so didn’t think I had put a report in…it wasn’t a hazard anymore”.
“It’s not up to me. That’s the Safety Department’s job”.

For a system to work, the users need to know how to use it and more importantly - they need to really appreciate why they should report. If we view the situation from the perspective of the field workers and engage them to resolve potential hurdles then we’re more likely to get the result we want.

I believe we also need to make training and awareness more personal. By using case studies highlighting how hazards were known about before the incident and reinforcing that safety is real, not an abstract concept perhaps we’ll get more buy in. If we can show workers the important role they have in maintaining safety, help them realise the personal responsibility they have in this and that it’s not “just another bit of paper” they have to fill in - perhaps we can work together to overcome the reporting constraints and we won’t hear those words “I knew that was going to happen one day” again.

Interested in Knowing More?

Further information on Safety Wise’s Incident Cause Analysis (ICAM) Training is available from our website: http://www.safetywise.com/

Additional ICAM Related Services

Safety Wise also offers the following additional services for sites that adopt the ICAM investigation analysis method: