| Publication Abstract Display | | Type: Published Abstract | | Title: Validation of the PASAT 50-item and 100-itdem short forms in an HIV infected sample. | | Authors: Cherner, M, Diehr M, Marcotte T, Heaton R, Miller W, Grant I, and the HNRC Group | | Year: 1999 | | Publication: Archives of Clinical Neuropsychology | | Volume: 14 Issue: 8 Pages: 703 | | Abstract:"The Paced Auditory Serial Addition Task (PASAT) is commonly used to measure speed of information processing, attention/concentration, and working memory. However, its length can prove to be stressful to some patients, making short forms an attractive alternative to the standard 200-item version of the test. Demographically corrected norms have been developed for 50- (lst series) and 100-item (lst + 2nd series) versions (Diehr, Cherner, Marcotte, Heaton, Miller, & Grant, manuscript in preparation). The present study provides clinical validation of the short forms by showing their sensitivity and specificity in a sample of 1064 HIV-infected research participants. Eighty-eight percent of subjects were male, 66% were White, 20% African American, 10% Latino, and 4% other. They ranged in age from 20 to 61 years and had a mean of 13.5 years of education. They received the full version of the PASAT as part of a larger neuropsychological assessment. The 50- and 100-item forms had respective correlations of 0.81 and 0.89 with the full 200-item PASAT. Demographically corrected T scores were assigned to individual raw scores on the full version and the short forms. Performances corresponding to T scores below 40 (1 SD below the mean) were considered to be in the impaired range. Eighty-seven percent of participants who were impaired on the full PASAT were also found to be impaired on the 50-item version, with 79% being correctly classified as unimpaired. Overall accuracy was 81%. Similarly, sensitivity and specificity for the 100-item version were 86% and 82% respectively, with an overall classification accuracy of 83%. Thirty-seven percent of participants showed less than a 5 T-score difference between the 50-item and the full version, and 63% fell within a 10 T-score difference. For the 100-item version, 42% showed less than a 5 T-score difference with the 200-item version, and 57% were within a 10 T-score difference. When participants who obtained above average scores (T score > 55) on the full version were excluded, correspondence between the short forms and the full version improved, implying that the short forms are more effective for detecting level of function in average to below average PASAT performers than in the highest functioning individuals. Results suggest that the 50-item short form shows sufficient sensitivity and specificity to obtain a reasonably accurate classification of impairment without subjecting patients or clinicians to the length and discomfort of the standard PASAT. Adding the second series does not seem to improve diagnostic accuracy significantly.
" |
return to publications listing
|