Safety Reporting, Production Rates, and their effects on the Severity of Occupational Injuries

Introduction Despite significant advancements in occupational safety and health, there will likely always be individuals who sustain injuries (Hassan & Khalifa, 2025). Prior work has examined observation and feedback as potential means to reduce injury rates over time (Ludwig & Laske, 2023). However, there is limited research on how to reduce the severity of injuries. If we can understand factors that relate to injury severity, we can reduce injury severity, which would lead to a safer work environment. In this study, we examine how production rates and safety variables influence the severity of injuries in the workplace. Methods This research will utilize three years of archival safety data from a large oil refinery. An observational study will be conducted examining the relationship between production rate, safety variables, and incident severity. Production rate is composed of actual production values, how much oil is being extracted, and flaring rates, how much excess gas is being burnt off. Safety variables are composed of safety audits, corporate-led job site inspections, behavioral observations, employee-led job site inspections, and safety moments, moments where safe employee behavior is recognized. A multinomial regression model will be used to determine the base rate for incident severity, and then use the previous variables to determine if there is a relationship between them. Expected Results It is anticipated that results will show a relationship between operational practices and the severity of workplace incidents. Specifically, increases in safety audits, behavioral observations, and safety moments during the average production periods are expected to predict lower incident severity. Conversely, higher flare rates and production levels are expected to be correlated with higher severity incidents. The effect sizes are expected to be moderate, consistent with prior evidence on the impact of proactive safety interventions, and may reveal diminishing returns at very high levels of auditing and observation (Yang et al., 2023). Practical Implications The findings of this study have the potential to expand and refine the occupational safety literature by shifting the focus from injury frequency to injury severity, a dimension that remains comparatively underexplored. By using a multinomial regression on a large archival data set, the research also contributes a methodological example of how operational and safety metrics can be integrated to predict severity outcomes in high-risk industrial settings. Beyond theory, these findings could help HR and safety professionals target the most effective interventions, such as audits, observations, and safety moments, where they matter most.