Patterns of underlying causes of work-related traumatic fatalities – Comparison between small and larger companies in British Columbia

This study mined the Workers’ Compensation Board of British Columbia (WCB) to compare discrepancies in injury and fatalities between small and large businesses.

I’ve skipped a lot.

For background:

·         Data from 2003-07 found the traumatic fatality rate among workers in small businesses (<20 person-years) was 9.7 per 100k person-years vs 2.7 per 100k in large businesses (4.6 per 100k was the average for all BC businesses)

·         When compared between businesses in similar industries, the “fatality rate was up to 5 times higher, even though small businesses had lower overall injury rates than their larger counterparts (2.5 vs. 3.2 per 100 person-years) and their serious injury rate (ICD-9) was only 2.5 times higher”

·         They propose some reasons why these discrepancies may exist:

o   Employees in smaller businesses may “discount minor or soft tissue injuries”

o   Smaller businesses may underreport their payroll for a variety of reasons, including that they tend to employ more seasonal or part-time workers

o   Smaller businesses may be subcontracted to larger businesses in similar classification units, and perform tasks with a higher risk of injury

o   “small businesses may not have the necessary time, budget, knowledge, or expertise to look into health and safety issues compared to larger businesses where there are economies of scale that extend to safety programs”

o   Smaller businesses hire fewer workers overall and more part-time workers, increasing the challenges of effective training, instruction, supervision and more

Results

Overall they found:

·         Fatality rates were 9.7 in small businesses and 2.7 for larger businesses

·         This results in an odds ratio of 3.62 – hence, workers in small businesses were 3.6 times more likely to experience a work-related traumatic fatality than workers in large businesses

·         Statistically significant differences by employer size were: fatalities within one month of employment, higher fatality rates in primary industries and transport, vehicular incident fatality rates, seat belt use

·         “Small businesses in British Columbia. have a significantly higher traumatic fatality rate than do larger businesses”

I’m sick of typing ‘businesses’, so from here on in I’m abbreviating SB (small businesses) and LB (large businesses).

435 traumatic fatalities were reviewed over the 5-year period – 251 in SB and 184 in LB. 243 were left for review in SB after removing several that were not traumatic, and likewise 179 in LB. 93% of fatalities were male.

No significant differences were found for mean age. While the total number of fatalities varied by year, “the percentage of fatalities attributed to small and larger businesses was stable”.

In SB, the mean length of employment was shorter (6 years) versus 8 years for LB. The number of fatalities that occurred with employment durations of <1 month was higher in SB than LB. Nearly 9% of fatalities in SB occurred within the first week of employment, which was closer to 4.5%ish in LB.

Further, another 4% of fatalities in SB occurred between 1 week and 1 month of employment, but this was only 1% in LB.

By Industry & Mechanism

No surprises, but primary industries (fishing and forestry), transportation and warehousing and general construction were the most hazardous industries.

I’ve skipped heaps in the next section.

Primary industries: No significant differences were found between SB and LB. The most common accident type was struck by at 26% (mostly loggers).

Transportation/Warehouse: Fatalities were similar, but there were some differences. For SB, vehicle incidents accounted for 55%, followed by struck by at 12%, and aircraft crashes at 9%. In LB, 45% were vehicular incidents, 29% aircraft crashes and 3% struck by.

Vehicular incidents: Vehicle incidents/crashes were the most common mechanism of traumatic fatality. 36% occurred in SB whereas in LB it was 25%.

Most of these incidents occurred on highways, then next most common was logging road. Logging roads was a common location for SB.

Most of the fatalities were among professional drivers – 61% in SB and 43% in LB, almost 1/3 was among workers whom driving wasn’t their core profession.

The paper covers a lot on drugs and alcohol. Just covering one data point being that drugs or alcohol was implicated in 33% of fatalities in SB and 31% in LB.

Seat belt use: Where it could be evaluated, it was found that 67% of vehicle fatalities in SB weren’t wearing a seat belt, and 53% didn’t in LB.

Falls from elevation: of 20 fatalities of this mechanism in SB, and where data was available (13 cases), 12 of the incidents weren’t using personal fall protection and 1 was.

Of the 20, 8 occurred where there wasn’t any fall protection available. Of the 4 fatal falls in LB, info was available for 3 of them: all 3 were noted not to be using fall protection. 2 of the 4 cases didn’t provide adequate fall protection such as guardrails, fall protection or restraint.

By Formal Procedures, Hazard Analysis, Equipment, Training etc

They tried to evaluate whether procedures, hazard analyses, equipment, training and more were implicated.

Because of patchy data, they only evaluated cases where it was indicated in the report that these factors should have been implemented or used, but were not.

So, caveat emptor – difficult to read into this due to the counterfactuals in investigations.

Written work procedures: in SB, 50 fatalities (20% of this sample) didn’t have the expected documentation or work processes in place, and 7 fatalities (3%) where existing work processes were not followed.

In LB, 43 fatalities (23%) didn’t have expected documented processes or inadequate work procedures, and 18 (10%) of fatalities where existing work procedures weren’t followed.

Hazard analysis or equipment poorly maintained: These items were observed in 18 of 422 of cases.

In SB, 6 cases were “clearly documented “ where inadequate hazard analysis or poor equipment maintenance were listed as causative factors. This was 12 in LB.

Training and education: in SB, 6 incidents were cited as involving inadequate training and education and 8 within LB.

I’ve skipped more on physical mechanisms like guarding and LOTO (wasn’t much there).

Conclusion

They conclude that:

·         “the data does clearly indicate that workers in small businesses were at twofold risk of a fatal injury within the first week of employment, and fourfold risk of fatal injury within the first month of employment compared with new workers employed by larger businesses”

·         “there were several patterns that were associated with fatalities, and would therefore warrant prevention efforts. Loggers tend to use recreational drugs, while fishers and barge hands tend to use alcohol to the point of intoxication. Seat belt use is not prevalent among truckers and may be virtually absent among log hauling truckers. Many general construction workers do not wear the necessary fall protection. As well, operators of equipment who leave their vehicles to assess, repair, or investigate their vehicle can be run over by their own equipment”

[** I wouldn’t read into their overall conclusions too much, as the second dot-point reads a bit like blaming people for not wearing PPE; but the original fatality reports, or this study, don’t comment on whether it was properly provided to the workers, was operatable, whether they were instructed and trained, whether high production pressures existed, local norms, leadership etc.]

Ref: Holizki, T., McDonald, R., & Gagnon, F. (2015). Patterns of underlying causes of work-related traumatic fatalities–Comparison between small and larger companies in British Columbia. Safety science, 71, 197-204.