Psychometric study of the Neurobehavioral Symptom Inventory

Abstract

The Department of Veterans Affairs (VA) uses the Neurobehavioral Symptom Inventory (NSI) to measure postconcussive symptoms in its comprehensive traumatic brain injury (TBI) evaluation. This study examined the NSI's item properties, internal consistency, and external validity. Data were obtained from a federally funded study of the experiences of combat veterans. Participants included 500 Operations Iraqi and Enduring Freedom veterans, 219 of whom sustained at least one TBI. Data were collected at five VA medical centers and one VA outpatient clinic across upstate New York. Measures included neuropsychological interview, NSI, Beck Anxiety Inventory, Beck Depression Inventory-II, and Posttraumatic Stress Disorder Checklist-Military Version. The NSI demonstrated high internal consistency (total alpha = 0.95; subscale alpha = 0.88 to 0.92). Subscale totals based on Caplan et al.'s factor analysis correlated highly with the NSI total score (r = 0.88 to 0.93). NSI scores differentiated veterans with TBI history from those without but were strongly influenced by variance associated with probable posttraumatic stress disorder, depression, and generalized anxiety. Results suggest that the NSI is a reliable and valid measure of postconcussive symptoms. Scale validity is evident in the differentiation of TBI and non-TBI classifications. The scale domain is not limited to TBI, however, and extends to detection of probable effects of additional affective disorders prevalent in the veteran population.