Why Safety Professionals Should Be Losing Sleep Over Key Safety Statistics
How we gather, record, and analyze safety information can give us a misleading picture of our safety performance.
Do numbers keep you at night?
If they're safety stats, they should.
Workers are dying, workers are injured, workers are sick, and we do not have accurate reporting. That last one keeps me up at night because reporting and information management is absolutely essential for accurate data and for the rest of our safety math. (Learn more about Safety Data: The Game Changer You Might Be Ignoring.)
What about the numbers we do have? How do those looks?
In a word, bad. Both in terms of how many incidents we have and why we continue to see them, and also how we get and use the data.
Some Stats Worth Losing Sleep Over
Here's one example:
In 2017, there were 5,147 (reported) fatalities at work in the United States.
We can compare that to the statistics from 2003: 5,575 reported fatalities, along with 1.3 million injuries resulting in days lost.
That seems like a definite improvement, but here's the problem: this 2003 data on injuries was not collected in the same manner as the data from 2017. That makes it difficult to know whether we're comparing apples to apples.
What about Canada?
In 2016, there were 904 worker fatalities in Canada, coupled with 240,682 injuries resulting in days lost from work.
Now, consider this. in the United Sates, reporting is for private industry and does not consider medical aids. Injuries are reportable as lost time incidents on the second day missed.
In Canada, on the other hand, some jurisdictions don't require lost time incidents to be reported until the third day missed. In some jurisdictions, the employer can pay the first two days lost, and then, on the third, report it as a days lost incident to their insurance carrier.
So, we have lost time incident figures, but incidents are only reported as such on the third day that is missed.
Confusing? You bet.
Worth losing sleep over? Very much so.
Variability in Reporting
What this example shows is that reporting varies across borders and throughout the world.
In researching this article, I noted that the United States Bureau of Labor Statistics uses whole list of resources to come up with their final numbers, including coroners' reports and hospital records. Despite this, they did not track the total number of work-related fatalities.
Sources of data on fatal work injuries, 2017
Number of documents
State workers' compensation
In Canada, most of our information comes from Workers Compensation Boards, but again, we're not always reporting the same information, in the same categories.
For example, most jurisdictions are only starting to record data from agriculture where the farm work is done on what is considered a family farm. In other words, you need a WCB account. That makes a difference: in Alberta alone, there were 23 fatalities recorded on these types of farms in 2008 but not included in provincial OH&S stats. Legislation changed that in 2016 but again, only for farms with a WCB account.
Why This Matters
This matters because so much of what we do depends on statistical information. Our performance is judged by incidents and injury rates, fatalities, and days lost. Safety professionals spend hours preparing reports, the information in which is subjective to say the least. Then, we start multiplying and dividing the totals to produce rates, frequencies, severities, indexes, and ratios, all in an attempt to demonstrate that we are getting better.
We would like to think that what we do as safety professionals has a positive impact on the safety performance of our workplaces.
And that’s what should worry us about the safety stats we rely on. Because the bottom line here is that we simply do not know if what we're doing is changing anything because we don't have all the data we need. What we can demonstrate with one set of data and calculations can be presented differently with a different set of data and new calculations.
Workplace Safety Data Needs to Be Standardized
In researching this article, I came across a common recommendation. It was made by several agencies and reports across North America.
That recommendation, in short: standardize workplace data reporting.
Can't argue with that.
(For related reading, see How to Use Standard Work Instructions to Improve Workplace Safety.)
There's no denying that we have a lot of data. If our job as safety professionals was just to collect a pile of raw numbers, we could pat ourselves on the back.
But those numbers don't always give us a clear picture and aren't always easy to use.
To see what I mean, let's take a look at another set of stats. These are the U.S. Department of Labor's list of fatal work injuries by industry.
Number and rate of fatal work injuries, by industry sector, 2017
Number of fatal work injuries
Fatal work injury rate
(per 100,000 full-time equivalent workers)
Transportation and warehousing
Agriculture, forestry, fishing, and hunting
Professional and business services
Leisure and hospitality
Other services (exc. Public admin.)
Educational and health services
At a glance, it looks like agriculture and forestry are by far the worst performers.
But are they? Hard to tell without more information to paint a complete picture.
Now, consider this other chart compiling data from the Department of Labor. It compares the number of fatalities between self-employed workers and waged or salaried employees. The numbers on the chart show that 20% of workplace fatalities in 2017 were of self-employed individuals and 80% for salary and wage workers
Number of fatal work injuries by employee status
Wage and salary
Based on this data, you might assume that salary and wage workers are more likely to have access to safety professionals, which makes them safer. But how could we know? To be meaningful, this data would have to specify the number of workers in each category, as well as the hours worked.
Both charts have useful information. But without more comprehensive data, we can't compare them against each other. And this is true of plenty of safety data, even those collected and analyzed by the same organizations.
We recognize that construction trades, oil and gas, transportation, and healthcare providers have high numbers of fatalities, days lost, and general incidents. They are undeniably high-hazard occupations.
However, we have to stay within an industry – and sometimes within a sector of an industry – to get any kind of value added comparison and get real benchmark data for analysis. Now, if we could compare construction companies that build highways to other construction companies that build highways, that would give us a value-add and a better picture of where each company's safety system stands.
Likewise, when we look at the stats for transportation, heavy truck operators, and semi-truck operators, we have to do an in-depth analysis to remove all fatality statistics where the fault is assigned to a second party, not the driver. That's the only way to get accurate numbers that tell us exactly what we want to know.
Use Stats Wisely
The big lesson here is that we have to use our stats wisely.
People think in pictures. That's why charts, graphs, and tables are such effective ways of conveying information. Add some color and some animation to it, and you've go yourself a real Captain Billy Whiz Bang show.
But remember that, as a safety professional, you're not in the entertainment business; you're in the loss prevention and life-saving business. You need to use numbers carefully. Make sure that the charts and graphs represent what is actually happening in your business.
One Final Statistic
I want to end with one final statistic that I think needs more attention.
Adding up the Canadian and the American workforce (full-time FTEs) shows that there were 142 million full-time workers in North America in 2017.
We have roughly 44,000 safety professionals across the two countries. That means each of us is responsible for an average 3,228 workers.
Now, that's a statistic that should definitely keep you up at night!