Rasch Analysis of the Behavioral Assessment Screening Tool for Adolescents (BAST-A) in Mild Traumatic Brain Injury

The Behavioral Assessment Screening Tool for Adolescents (BAST-A) assesses behavioral and emotional symptoms and experiences in adolescents with mild traumatic brain injury (mTBI). We used modern test theory to validate the BAST-A by analyzing person-level and item-level properties of its subscales (Negative Affect & Fatigue, Executive & Social Function, Risk Behaviors) in a sample of adolescents with concussion/mTBI. Following Rasch Guidelines in Rehabilitation Research (RULER), we assessed unidimensionality, item hierarchies, targeting (including items at different severity levels), and strata. The Executive & Social Function and the Negative Affect & Fatigue subscale included items at different severity levels and displayed adequate targeting. In contrast, the Risk Behaviors subscale displayed mistargeting, indicating that different item severity levels may not have been adequately captured by the BAST-A or represented in the sample. The Executive & Social Function subscale distinguished three strata, or distinct levels of symptom frequency, while the Negative Affect & Fatigue subscale distinguished five strata. Risk Behaviors did not distinguish any strata in this sample. Overall, the BAST-A displayed adequate psychometric properties for the Negative Affect & Fatigue subscale and Executive & Social Function subscale, but not for the Risk Behaviors subscale. While the other subscales capture common symptoms and experiences in their respective domains, the Risk Behaviors subscale may only include symptoms for which any endorsement would indicate a severe or concerning problems (e.g., substance misuse). While continuous scores are appropriate for the Negative Affect & Fatigue and Executive & Social Function subscales, dichotomizing the Risk Behaviors score is recommended.