Objective: To retrospectively examine the clinical utility of neuropsychological tests (NPTs) for predicting employment outcomes in persons with cognitive impairment after moderate to severe traumatic brain injury (TBI).
Methods: 132 individuals of working age with cognitive impairment after moderate to severe TBI were classified into three groups by employment status: competitive employment (CE); supported employment (SE); and unemployed (UE). NPT scores were compared among groups. Using multinomial logistic regression with group allocation as the dependent variable, significant variables were identified, and receiver operating characteristic (ROC) curves were calculated.
Results: Comparison of NPT results among the three groups showed significant differences for all NPTs (all items, p < 0.01). Using multinomial logistic regression analysis, Rivermead Behavioral Memory Test (RBMT) and Behavioral Assessment of the Dysexecutive Syndrome from CE versus SE and Trail Making Test-B and RBMT from SE versus UE were identified. ROC curve analysis indicated small to moderate accuracy (area under the curve, 0.63-0.84).
Conclusion: NPT scores can predict future employment status in patients with cognitive impairment after TBI. These findings may lead to improved clinical assessments when providing work support. Future research should consider occupational categories, managerial categories, and types of re-employment.
Keywords: Cognitive impairments; ROC curve; clinical research; logistic regression; neuropsychological test; return to work; traumatic brain injury.