I recall a fatality investigation I did a few years ago was no exception. A 27 year old male, father of a 4 year old daughter, fell 50’ through a skylight to his death while replacing metal deck. He was a qualified rigger and was properly trained in fall protection. It was especially difficult to learn that this individual had his personal fall arrest harness on, with means to tie off available, but never secured his lanyard to the anchor point. The traditional theory that frequency breeds severity should suggest the employer should have seen signs of minor incidents (i.e. previous safety inspections that identified this same violation) that would have lead to this severe accident. In an article I wrote some time ago, I described the benefit of using the traditional frequency/severity relationship approach to preventing serious accidents. If we applied that approach to this case, previous inspections of this employees crew should have indicated some infractions as it relates to fall protection. There had to be some incidents at the company that related to fall protection violations. After all, safety professionals are taught that a high frequency in minor incidents will eventually lead to a major event…………right?
Another interesting fact about this case was the activity being performed at the time of the accident was something the employee very seldom did. As a matter of fact, the company he worked for very rarely performed this activity, as it is not a part of their every day activities. So if there is little to no exposure to this activity on a regular basis, then how could we have foreseen that the company was increasing its chances for a severe incident? Was this accident an anomaly, or do we need to rethink our approach to predicting potentially disastrous accidents?
In a recent Professional Safety article, the concept of frequency breeding severity is challenged. Recent statistical data from the Bureau of Labor Statistics (BLS) shows that although less severe accidents resulting in days away from work (10 days or less) have decreased significantly from 1995 to 2006, accidents involving 10 or more days away from work have increased significantly during the same time period. Manuele (2008). According to the article, in order to reduce serious injuries and fatalities, safety professionals must address the long-held and still applied belief that reducing incident frequency will equivalently reduce incidents that result in severity. The author of this article is not alone. At the a recent Behavioral Safety Now Conference, James Johnson, a managing director at Liberty Mutual Insurance Co., said “I’m sure that many of us have said at one time or another that frequency reduction will result in severity reduction. This popularly held belief is not necessarily true. If we do nothing different than we are doing today, these types of trends will continue. ” (Manuele, 2008). In a study performed by the author involving over 1,200 incident investigations, a large proportion of incidents resulting in serious injuries and fatalities occur:
It's also important to look at some recent stats released by the Bureau of Labor Statistics (BLS) that might support the idea that old approach may not be as effective as we might have thought. According to the BLS, 4,693 workers died on the job in 2011 – a number virtually even with the 4,690 fatalities in 2010 (BLS, 2013). It’s important to keep in mind that these statistics represent all employees in the country covering multiple industries. I make that note because there is some good news to report for the construction industry. Local efforts, specifically those by members of the Midwest Roofing Contractors Association, appear to be paying off. The 2011 final figure for the private construction sector represent a 5 percent decline for the final 2010 total, making the fifth consecutive year of decline in the industry (BLS, 2013).
The fatality I investigated included most of these characteristics. Using the traditional methods of predicting severe accidents, the accident may never have been predicted. The point is that safety professionals in the roofing industry need to broaden our efforts in identifying the causes of severe accidents. There are several ways to improve our ability to predict these accidents before they happen. The most crucial being effective pre-planning. If a roofing contractor can plan ahead for predictable exposures that can be expected (as well as those that are not expected), these accidents can be eliminated.
I still believe in the traditional frequency/severity approach to reducing accidents. But the statistics are clear; we are not effectively reducing the number of severe accidents. It will take a change in approach to affect some change in result. It’s like Yogi Berra said “If you keep doing what you did, you keep getting what you got ”.