OBJECTIVE: As part of routine care at Veterans Affairs facilities, veterans with a service-related traumatic brain injury (TBI) are administered a self-report post-concussive symptom measure, the Neurobehavioral Symptom Inventory (NSI). Interpreting the NSI can be problematic given that over-reporting on self-report measures is often found in both civilian and military patient populations. This study investigates embedded scales on the NSI that identify possible and probable symptom exaggeration.
METHOD: 183 veterans with a history of mild TBI were administered the Minnesota Multiphasic Personality Inventory, 2nd edition, Restructured Form (MMPI-2-RF) and the NSI. The participants were divided into symptom validity testing pass and fail groups based on their performance on the MMPI-2-RF symptom validity scales. Cut scores on the NSI Total and Validity-10 scores were then established and applied to two additional veteran populations.
RESULTS: Sensitivity and specificity values were derived for all NSI Total and Validity-10 values. Optimal cut scores were determined based on specificity levels of ≥95%. The NSI Total cut score was ≥57 for possible and ≥67 for probable symptom exaggeration and the Validity-10 cut score was ≥22 for possible and ≥27 for probable symptom exaggeration, with sensitivity ranging from 27 to 43%. Applying these cut scores to a broader clinical and research sample resulted in lower rates of suspected exaggeration.
CONCLUSIONS: Both the NSI Total and Validity-10 cut scores consistently identified potential symptom exaggeration across three mild TBI samples. Clinicians and researchers who use the NSI are encouraged to utilize either embedded validity measure in their practice.