Lee Ashendorf1,2, Erika L Clark1,3, Michael A Sugarman1. 1. a Edith Nourse Rogers Memorial Veterans Hospital , Bedford , MA , USA. 2. b Department of Psychiatry , Boston University School of Medicine , Boston , MA , USA. 3. c Department of Psychiatry , Massachusetts General Hospital and Home Base , Boston , MA , USA.
Abstract
OBJECTIVE: While recognition memory has been the primary tool for the assessment of performance validity in neuropsychological evaluations, some consideration has also been given to embedded measures from other cognitive domains, including processing speed. The present study evaluated the classification accuracy of several speed-based measures in a Veterans Affairs Medical Center Polytrauma sample. METHOD: The present sample consisted of 114 military veterans (Mean age = 35.5, SD = 9.4) referred for a suspected history of mild traumatic brain injury who were administered a full neuropsychological protocol that included several validity checks. Veterans were assigned to Valid (n = 80) or Invalid (n = 34) groups based on outcomes of performance validity measures (PVMs). RESULTS: Several processing speed measures yielded acceptable or excellent classification accuracy; sensitivity values ranged from 29 to 53% with specificity values above 90%. Efforts to identify an improved algorithm that would collapse across multiple processing speed PVMs were unsuccessful compared to classification based on single measures. CONCLUSIONS: Processing speed measures can serve as efficient performance validity assessment tools.
OBJECTIVE: While recognition memory has been the primary tool for the assessment of performance validity in neuropsychological evaluations, some consideration has also been given to embedded measures from other cognitive domains, including processing speed. The present study evaluated the classification accuracy of several speed-based measures in a Veterans Affairs Medical Center Polytrauma sample. METHOD: The present sample consisted of 114 military veterans (Mean age = 35.5, SD = 9.4) referred for a suspected history of mild traumatic brain injury who were administered a full neuropsychological protocol that included several validity checks. Veterans were assigned to Valid (n = 80) or Invalid (n = 34) groups based on outcomes of performance validity measures (PVMs). RESULTS: Several processing speed measures yielded acceptable or excellent classification accuracy; sensitivity values ranged from 29 to 53% with specificity values above 90%. Efforts to identify an improved algorithm that would collapse across multiple processing speed PVMs were unsuccessful compared to classification based on single measures. CONCLUSIONS: Processing speed measures can serve as efficient performance validity assessment tools.
Authors: Nathaniel W Nelson; Seth G Disner; Carolyn R Anderson; Bridget M Doane; Kathryn McGuire; Gregory J Lamberty; James Hoelzle; Scott R Sponheim Journal: Neuropsychology Date: 2019-09-23 Impact factor: 3.295