A psychometric study of the neurobehavioral symptom inventory
King, Paul R., Jr.
MetadataShow full item record
Traumatic brain injury (TBI) has been identified as the signature injury of modern combat. Considering the large number of troops who have survived brain injuries during military service, the Department of Veterans Affairs needs an effective screening tool to help identify TBI and symptoms of post-concussion syndrome (PCS). One tool intended to assess the nature and severity of PCS symptoms is the Neurobehavioral Symptom Inventory (NSI; Cicerone & Kalmar, 1995). This study explored and detailed the psychometric properties of the NSI to determine whether it is a sensitive and specific screening tool. Data for this study were obtained from an existing test database from a federally funded comprehensive study of the experiences of combat veterans with and without TBI ( Cognitive Assessment of Veterans after Traumatic Brain Injury ; VA HSR&D SDR 06-162). The final data set is comprised of 494 veterans of Operations Iraqi and Enduring Freedom residing in the upstate New York Veterans Integrated Services Network (VISN 2), of whom 219 were determined to have experienced at least one TBI. Each completed a comprehensive neuropsychological interview, several self-reports, and cognitive testing. Based on affective measures, a number of these veterans met diagnostic criteria for generalized anxiety disorder, major depressive disorder, and PTSD. Estimates of the NSI’s internal consistency ranged from acceptable to high (NSI total α = .95). Signal detection theory was used to evaluate the accuracy of the NSI in identifying true symptoms of PCS. Results of multiple receiver operating characteristic (ROC) curve analyses yielded area under the curve (AUC) estimates which ranged from .67 to .75. The modified and original somatic subscales yielded the greatest AUC estimates, followed by the summed total of all NSI responses. As a result of this inquiry, the NSI is viewed as a marginally valuable metric of PCS symptoms, although caveats for usage are offered.