A Retrospective Study Evaluating the Role of Personal Factors in Underground Miner Injury

INTRODUCTION: The US is a major contributor to the global mining industry, providing over three million working positions. Mining is also considered one of the most dangerous occupations. Both personal characteristics and work-environmental characteristics contribute to the likelihood of work-place injury, however, it is unclear what specific individual factors are associated with injury in US underground mines. Our aim was to determine if a person’s history of being injured is associated with their behavior and perception of their environment. 

METHODS: We conducted a census of 251 employees of a silver mine in the Inland Northwest, matching the results of the questionnaire to their injury records. The questionnaire consisted of four categories: safety climate, behavior, prevention, and work ability. We reduced the features of the data set using Cronbach’s alpha to 26 candidate predictors of injury, and used logistic regression model to identify those significantly associated with injury. 

RESULTS: Of the 222 questionnaires with matched employee records, 48 employees had records of injuries, totaling 248 injuries. From the 26 variables that were used, five were identified as strongly associated with the probability of an employee having been injured. The factors which were associated with injury were whether miner’s felt their supervisor cared about their safety (OR= 0.39, p-value <.004), whether they felt they had no control over their safety (4.2, <.0001), if they took pride in their job (0.42, 0.009), and whether they followed mandated safety regulations such as (not) using ear plugs (3.5, 0.0004) or standing on open platforms (2.0, 0.03). 

CONCLUSIONS: While it may not be possible to change miner’s perspectives, these results suggest it may be possible to identify those more likely to be injured. These results only illustrate an association with injury and prospective studies are needed to confirm these factors are predictive.