Low back strains, carpal tunnel syndrome, and other soft-tissue musculoskeletal injuries are the most frequent causes of missed workdays in the United States, and most result from ergonomic, slip, trip, or fall hazards, according to the Bureau of Labor Statistics. Fortunately, most of these hazards can be minimized through workplace interventions. One effective approach is ergonomics, which is the study of preventing musculoskeletal disorders through workplace design and policies.

The first step to designing ergonomic interventions is to identify workplaces that need them. In a paper published in the Journal of Occupational and Environmental Medicine, National Institute for Occupational Safety and Health (NIOSH) researchers describe how they used computer programming—specifically, machine-learning—to identify industries at high risk for these prevalent hazards. Study lead author Alysha Meyers, Ph.D., NIOSH epidemiologist, explains the study and its results.

Q: What is unique about the study?

A: We applied machine learning to identify work-related injuries by cause, using workers’ compensation records for about two thirds of Ohio workers. Machine learning uses algorithms to “teach” a computer to perform a certain task. For example, in our study we applied a mathematical machine-learning technique to quickly and accurately code workers’ compensation claims into one of three groups: 1) ergonomic; 2) slips, trips, and falls; and 3) other.

Ultimately, the machine-learning technique enabled us to identify those industries and groups that face a greater risk of injury from ergonomic-related, slip, trip, or fall hazards. Identifying these high-risk industries tells us where we need to focus ergonomic or safety interventions.

Q: What did you find?

A: Workers in Ohio skilled nursing facilities were at the highest risk for severe (more than 7 missed workdays) ergonomic-related claims, and workers in the general freight trucking industry are at the highest risk for severe slip, trip, or fall claims.

Using data from the Ohio Bureau of Workers’ Compensation, we analyzed more than 1.2 million claims from 2001 to 2011 representing more than 200 industries. We then ranked these claims for musculoskeletal injuries that could have been prevented with workplace interventions to prevent ergonomic-related injuries, or slips, trips, and falls.

Q: What are the next steps?

A: Already, our findings are helping occupational safety and health specialists to focus their injury prevention efforts on high-risk occupations and industries. This activity is especially apparent in Ohio, where our study took place, but researchers in other states could use our study to develop similar approaches.

More information is available:

• Applying Machine Learning to Workers’ Compensation Data to Identify Industry-Specific Ergonomic and Safety Prevention Priorities: 2001 to 2011
• NIOSH: Ergonomics and Musculoskeletal Disorders
• NIOSH: Slip, Trip, and Fall Prevention for Healthcare Workers
• NIOSH: Center for Workers’ Compensation Studies