| Publication Abstract Display | | Type: Published Abstract | | Title: Trail Making Test and the relative efficacy of discrepancy analysis in an HIV positive sample. | | Authors: Dawes S, Cysique L, Lazzaretto D, Moran LM, Woods SP, Heaton RK, Grant I, and the HNRC Group | | Year: 2006 | | Publication: Journal of the International Neuropsychological Society | | Volume: 12 Issue: S2 Pages: 92 | | Abstract:Discrepancies between Trail Making Test (TMT) Parts A and B are commonly derived for use in research and clinical practice; however, this interpretive approach fails to consider an individual's basic processing speed ability level based on TMT-A demographically-adjusted T-scores. If the TMT-A score level is not taken into consideration when deriving discrepancy scores, then under-detection of abnormal differences in low scoring individuals (T<40) and over-detecting problems of high scoring individuals (T>55) may occur systematically. The aim of this study was to derive normative standards for ability-based TMT discrepancy scores and to examine their classification accuracy relative to conventional TMT measures. A sample of healthy adults (n=201), mean age 39.6 years (sd=11.5) and education 13.6 years (sd=2.6) were examined to derive the normal distributions of discrepancy scores for the simple and ability-level based scores. The resultant normative discrepancy cut-scores which were then applied to a demographically comparable sample of 239 HIV+ individuals, selected to approximate current prevalence estimates of HIV-associated neurocognitive disorders (36% impaired). Results showed that the conventional TMT-B measure provides better sensitivity (54% versus 25-26%, p<.01), but lower specificity (79% versus 83-89%, p<.05) than both derived discrepancy scores in classifying HIV-associated neurocognitive disorders. Findings from this study do not support the incremental predictive validity of the TMT discrepancy analyses relative to traditional TMT scores in classifying persons with HIV-associated neurocognitive disorders. Nevertheless further analyses are needed to examine the construct and incremental validity of the ability-based TMT discrepancy score as a marker of divided attention. |
return to publications listing
|