A comparison of NES2 and traditional neuropsychological tests in an AIDS patient sample and a control sample

Selected subtests of the Neurobehavioral Evaluation System - 2 (NES2) were compared to similar traditional neuropsychological tests in an effort to establish their validity and determine their usefulness in a clinical setting. The paired tests included measures of attention, motor speed, divided attention, verbal learning and memory and mood assessment. The performance of an "affected" HIV positive group and an "unaffected" control group matched for age, education and gender were analyzed. The variables, clustered conceptually into traditional and computerized. tests, were first analyzed with, MANOVA to confirm or disconfirm that there were broad differences between the groups. The result in both analyses were statistically significant. In general, the MANOVAs indicated that the HIV positive group performed less well than the control group. To determine the contribution of each of the tests to the global finding an analysis of variance (ANOVA) was performed on each test. There were significant differences between the HIV positive group and the control group on all measures except the Continuous Performance Test (a test of sustained attention) which was eliminated from the sensitivity/specificity analysis that followed. The sensitivity/specificity analysis proved the NES2 subtests to be sensitive to cognitive impairment but to a lesser degree than traditional tests, with the exception of the memory measures. Mode of administration and response as well as some differences in the tests themselves were thought to account for the sensitivity differences. The paired tests were highly correlated with each other, emphasizing that they tap into similar neurobehavioral domains and offer a notable degree of validity for the NES2. Their use may be interchangeable once better norming is achieved. Benefits that derive from comparing traditional and computerized neurobehavioral measures provide a clearer understanding of some neuropsychological functions and their behavioral correlates. Additionally, validated computerized measures offer the promise of greater efficiency in time expended for assessment and, in some cases, greater accuracy of measurement.