Defense Date


Document Type


Degree Name

Doctor of Philosophy


Health Related Sciences

First Advisor

Christine Reid


In Minnesota’s workers’ compensation system, injured employees at risk for sustaining permanent disability may be eligible for receipt of vocational rehabilitation (VR) services if they are determined to be capable of benefitting from such services. VR services can be a valuable resource to injured employees who need assistance minimizing their work disability and maximizing their residual wage-earning capacity. However, for VR services to be effective at a system level, it is necessary to precisely and accurately identify an injured employee’s rehabilitation potential. Failure to do so is likely to result in the misallocation of a scarce and costly resource. Given recent trends in Minnesota’s workers’ compensation system (e.g., higher VR service costs and lower RTW rates among injured employees with indemnity claims), this study was conducted with the purpose of developing and validating an objective, evidence-based method of predicting the RTW status as of claim closure of injured Minnesota employees who sustained permanent impairment and received VR services. To accomplish this purpose, a closed-claim, retrospective design was implemented. Data for this cross-sectional study was obtained from the Minnesota administrative claims database. There were 15,372 claims that met all eligibility criteria. With guidance from the biopsychosocial disablement models developed by Nagi and the World Health Organization, 15 discrete predictor variables that represented medical, individual, and workplace factors were selected for study inclusion. Descriptive and predictive analyses were used to assess the relationship between this study’s RTW outcome and its set of RTW predictors. Using logistic regression, an optimal RTW model was first developed and then internally validated with a split-dataset approach. The optimal RTW model included four main effects (attorney involvement; severity of permanent impairment; age; job tenure) and three first-order interaction effects (pre-injury average weekly wage X pre-injury industry; attorney involvement X severity of permanent impairment; attorney involvement X job tenure). Though not retained in the optimal RTW model, part of body affected and education also had notable bivariate relationships with the outcome. The optimal RTW model’s performance regarding goodness-of-fit and clinical usefulness suggests it may be of value to those assessing rehabilitation potential within Minnesota’s workers’ compensation system.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission