Neuropsychological Screening for Cognitive Impairment Using Computer-Assisted Tasks

The aim of this study was to validate a computer-assisted screening battery for classifying patients into two groups, those with and without cognitive impairment. Participants were all patients referred to the neuropsychology clinics at four VA medical centers during a 1-year period. Patients meeting the study inclusionary criteria (N = 252) were administered the Neurobehavioral Evaluation System-3 (NES3) computer-assisted battery. A detailed neuropsychological examination was carried out by an experienced neuropsychologist, who diagnosed the patient as cognitively impaired or not impaired. The neuropsychologist's diagnosis was the gold standard. Recursive partitioning analyses yielded several classification procedures using the NES3 data to predict the gold standard. These procedures produced a set of six NES3 tasks that provide good sensitivity and specificity in predicting diagnosis. Sensitivity and specificity for the least random classification procedure were 0.87 and 0.67, respectively. The results suggest that computer-assisted screening methods are a promising means of triaging patients.