Date of Graduation


Document Type


Degree Type



School of Public Health


Occupational & Environmental Health Sciences

Committee Chair

Lesley Cottrell

Committee Co-Chair

Anna Allen

Committee Member

Steven Wheeler

Committee Member

Weimin Gao

Committee Member

Christa Lilly


Introduction: Minimal research has been conducted on healthcare utilization among injured workers residing in rural areas. These workers are often at high risk for work-related injury due to the industries in which they work and commonly experience barriers to accessing healthcare. The first study in this dissertation series explored general characteristics of injured workers receiving care at occupational medicine clinics in West Virginia (WV). The second and third studies explored the moderating effects of rurality, pain medication, and psychiatric medication on sociodemographic and health status factors and number of medical encounters.

Methods: Each study used a cross-sectional design with data from the electronic health records of individuals receiving care at any West Virginia University Hospitals affiliated Occupational Medicine clinic over a two-year term (January 1, 2017-December 31, 2019). The outcome variable for all three studies was number of medical encounters – number of visits to the occupational medicine clinic for issues resulting from a work-related injury. Predictor variables for the first study were gender, race, ethnicity, religion, rurality, payor source, smoking status, pain medication, psychiatric medication, location/type of injury, and median severity of injury. Variables that had an alpha value of 0.25 or less in bivariate analyses in the first study were included as predictor variables in the second and third studies. Generalized linear regression with Poisson distribution was completed in each of the studies to determine covariates (first study) and to determine main effects and moderating effects of specified predictor variables (second and third studies).

Results: A total of 784 participants received care that accounted for 2,433 encounters over the course of two years. General characteristics of injured workers in WV were somewhat similar to those in other areas with the majority being white (n=746, 96%), non-Hispanic (n=759, 99%), and male (n=466, 59%). However, approximately one third of these workers resided in a rural area and rurality was significantly associated with a decreased number of encounters (p=.0122*). Rurality also had a moderating effect on median severity of injury, worker’s compensation, psychiatric medication, and all other injuries. Pain medication and psychiatric medications were also found to moderate some of the predictor variables. The presence of pain medication in the electronic health record had a moderating effect that changed the interaction of psychiatric medication, religious, worker’s compensation, upper extremity wound or traumatic injury, and sprain, strain of head/neck/face that either increased or decreased the likelihood of number of encounters. Religious, rural, current smoking status, worker’s compensation, pain medication, sprain, strain, or tear of trunk, and upper extremity wound or traumatic injury were moderated by psychiatric medication.

Conclusion: The association between number of encounters and the sociodemographic and health status of injured workers in WV is moderated by rurality, and documented pain medication, or psychiatric medication. The impact of rurality and documented medication regimens impacted the predictor variables in slightly different ways demonstrating that the interaction of factors that influence healthcare utilization is complex. It is important to explore these differences to guide programming, support, research, and policy development to optimize healthcare access and utilization for positive return to work outcomes in this population of workers. To do this we need to look at data outside of that provided by the Bureau of Labor Statistics to gain a deeper understanding of the healthcare utilization and return to work processes for injured workers. Utilizing electronic health records data in future studies can guide in the development of programming that addresses whole worker health following reported work-related injury, mitigates barriers to care, and identifies those at risk for increased healthcare costs and long-term disability.