Safe AF episode 2 live! Only 8.5 mins, so check it out.
Are the contributing and causal factors similar between Severe Incidents & Fatalities (SIFs) and non-SIFs?
Will addressing the minor potential, low-energy hazards help prevent the major potential hazards?
Today we explore a paper by Bayona et al. 2024 investigating these questions.
Links below.
If you find these pods useful then please help share on LinkedIn, and subscribe, rate and review on Spotify/Apple etc.
Spotify:
Transcription:
Here is your transcript with the numbers removed and formatted for readability:
We pour resources into investigating minor hand injuries and ensuring proper lifting techniques. We count the stitches and our injury rates look fantastic, but we face a perpetual conundrum. We’re good at measuring and managing the minor stuff, but we still kill people because sometimes the things that hurt us aren’t the same things that kill us.
G’day everyone, I’m Ben Hutchinson and this is Safe As, a podcast dedicated to the thrifty analysis of safety, risk and performance research. Visit safetyinsights.org for more research.
Today’s study is “The Things That Hurt People Are Not The Same As The Things That Kill People: Key Differences in the Proximal Causes of Low and High Severity Construction Injuries” from Bayona and colleagues, published in 2024 in the Journal of Construction Engineering and Management.
This paper dived into the differences between serious injuries and fatalities, or SIFs, and less severe incidents across a range of industries—utilities, oil and gas, construction, and more. For background, it briefly touched on Heinrich’s work, noting that while he didn’t actually suggest that the safety pyramid was causal, his ideas have been used to support the belief that stopping minor injuries naturally prevents the major ones.
The paper does a really commendable job highlighting the empirical landscape of minor events versus SIFs. First, importantly, there’s no empirical suggestion that the ratios in Heinrich’s work are fixed. Nor should we expect them to be, outside of some very specific scenarios and conditions.
Now that we’ve got that out of the way, some research suggests that there’s really no correlation between minor and major injuries over time. This seems to be a popular position within the SIF communities. However, the published evidence isn’t so clear, despite what people may claim.
As noted in this paper, prior research, looking at the contributing factors surrounding SIFs, found some interesting distinctions: So, fatal accidents tend to involve higher magnitudes of energy, somewhere over around 1500 Joules of energy, with some caveats.
Working without supervision was more often linked to fatalities than less serious injuries.
Younger workers were more vulnerable to SIFs than older workers. And one study, from Linda Bellamy, specifically examined the barrier and management system failures that led to 23,000 serious Dutch accidents. She found that all of the identified causes of fatal accidents were also present in the non-fatal accidents or injuries. Suggesting that there can be a connection between minor and major accidents, provided that we’re dealing with the same hazards. I’ll cover that study at a later date.
They argue that SIFs are paradoxical. Quoting the paper: “Although it’s fortunate that SIF events are, relatively speaking, rare, the low rate of occurrence hampers our ability to trend and learn from them.” So, what were the research questions: One, what are the differences, if any, in the immediate causes of SIFs and low severity injuries? What are the differences, if any, in the immediate causes of SIFs and potential SIFs?
Next was their methods, and what made SIFs different? A team of construction safety experts and academics hypothesized seven unique factors that might lead to high severity injuries.
They then collect details on 38 incidents: 38 actual SIFs, 12 potential SIFs, and 13 low severity injuries. The team assessing these incidents were blind to the outcome; they didn’t know if it was minor or major. They simply checked for these seven factors. Finally, the researchers compared these factors across the injury types to see what explained the difference.
So, what did they find in this study? One, there was no differences between SIFs and P-SIFs. This means that events that could have been serious or fatal often share the same causal or contributing factors as those that were fatal.
Two, they found two factors differentiated low severity incidents from the SIFs and potential SIFs. One was absent direct controls, and two, absent or not followed work plans. Specifically, they say, the odds that a SIF involves a missing direct control is about two orders of magnitude higher than a low severity incident. It’s a big difference.
What is a direct control? Well, a direct control is a safeguard that’s targeted to this energy source. It effectively mitigates the energy when installed, verified, and used properly, and is effective even if someone makes an unintentional error during work.
Therefore, they suggest a differentiated approach to prevent SIFs in construction, emphasizing control of the high energy hazards, those involving 1500 joules or more, through proper planning, discipline, and execution. No big surprises there. The study reinforces a growing idea in safety. That is, the things that hurt people aren’t the same as the things that kill people, and this is said to be replicable.
So let’s take a look at some of the findings that SIFs have found around absent or not followed work plans. Issues with work planning, discipline, and execution have consistently contributed to SIFs, or at least been found to, or constructed to, contribute to SIFs, but this study found that they were also contributors to P-SIFs, potential SIFs, so the work implies that proper work planning and execution is needed to specifically address high energy hazards, so this might be a cornerstone of a program that targets SIFs. No real surprises there.
Also, absent direct controls are another differentiator for SIFs and potential SIFs versus low severity. A missing or insufficient direct control almost always precedes a SIF but is almost never involved in a low severity incident.
So low severity incidents often happen when there are no specific controls missing. Absent direct controls and absent or not followed work plans were involved in almost all of the same SIF cases. Practically, this means organizations should focus on ensuring the presence of direct controls against these high energy hazards through expected things like effective work planning and execution. And, again, not surprisingly, safety should be measured by the extent to which life-threatening hazards are observed to be controlled.
Another finding was that human factors that they studied in this paper weren’t necessarily differentiators between low severity and high severity injuries. It was unexpected that things like task familiarity, risk normalization, hazard recognition, weren’t significantly linked to SIFs. They were equally present in SIFs and low severity incidents. Therefore, quoting the paper, generalized safety practices that focus on human behavior might be important for more general injury prevention, but they may not help to target SIFs directly.
And another finding. SIFs happened for the same reasons. No differentiators were found between the SIFs and PSIFs, suggesting similar contributing factors. This supports focusing on potential events to expand learning and improvement.
So the limitations. Well, some notable limitations are the retrospective data method is almost certainly influenced by some degree of outcome and confirmation bias. As is virtually every type of retrospective analysis, essentially, what you look for is what you find. And it’s not easy to clearly separate the investigators’ own mental and accident models from the description of the accident. Also, there was quite a small number of incidents that were analyzed in this study.
So what can we make of these findings? Well, I think some clear learnings are that we need to be clear about the barrier systems in place, their function, elements, performance specifications, allocations, resources, and more.
We should focus on direct controls, however you define them. Full barriers. Controls. Safeguards. Whatever. And ensure their implementation, resourcing, and their performance meets assumptions. We should focus on high-energy tasks and their interactions with people. Finally, we shouldn’t rely on incidents.
The statistical issues with TRIFR, for instance, is even worse with the rarest SIFs.
That’s it on Safe As. I’m Ben Hutchinson and hope you found this useful. We’ll see you next time.
SafetyInsights.org 