The “safety triangle:” A useful, yet complicated, theory

Occupational safety and health specialists study past work-related illnesses and injuries to understand how to prevent future ones. More than 80 years ago, the occupational safety and health pioneer Herbert Heinrich used this approach to devise the so-called “safety triangle.” After questions arose about the safety triangle’s validity, a National Institute for Occupational Safety and Health (NIOSH) study found that the safety triangle is a useful, yet complicated, theory when considering how to prevent severe work-related injuries. In the sections below, lead author Patrick Yorio, Ph.D., NIOSH statistician, explains the study, which was published in the journal Risk Analysis.

A NIOSH study of the Heinrich safety triangle, pictured above, found that lower-severity events within a mine can help predict a future fatal event within the same mine. Image from NIOSH.

Q: Why did you do this study?
A: In the 1930s, Herbert Heinrich first theorized that severe injuries often occur after many less severe injuries and near misses, and that these events occur in a fixed ratio of 300 near misses to 29 less severe injuries to 1 major injury—the famous safety triangle. In this image, near misses form the base, and smaller, less severe injuries form the middle of the triangle. A severe, even deadly, injury forms the pinnacle. Since then, however, critics have asked whether this ratio is valid and if controlling the risks related to near misses and less severe injuries can prevent a severe injury from occurring.

Our study aimed to find out whether or not the ideas found within the safety triangle are valid. Specifically, we first wanted to understand if near misses and less severe work-related injuries affect the number of work-related fatalities that occur over time in mining establishments. Secondly, we sought to examine if the probability of future work-related fatalities decreases as the severity of near misses and minor injuries decreases—thereby taking on the safety triangle form.

Q: How did you measure these effects?
A: We used records from two databases that the Mining Safety and Health Administration (MSHA) makes publicly available to create a third database addressing our research questions. We included 27,446 establishments over the 13 years from 2000 through 2012. Most of these mines were sand and gravel mines, followed by stone, coal, and nonmetal and metal mines. With this information, we were able to calculate the probability of a work-related fatality due to the previous number of near misses and less severe injuries.

Q: What did you learn?
A: Our results showed that lower-severity events within a mine may be used to predict a future fatal event within the same mine. Further, there is validity to the safety triangle in mining, although it depends upon the definition of injury severity. Specifically, we found that the safety triangle exists, but not for all approaches used to define severity levels. Severity metrics linked to the number of days lost due to work-related injuries produced a systematic decline in the effect as severity decreased. But, that was not the case when defining severity based on MSHA’s predefined categories of severity.

Q: What are the next steps?
A: An important area for future research is to explore potential common causes that contribute to both lower-severity injuries and deaths in the mining industry. Also, research that helps us understand the patterns of causes underlying lower-severity injuries and near misses is important to help improve mining safety.