Trail-Making Test Part B: Evaluation of the Efficiency Score for Assessing Floor-Level Change in Veterans.

Abstract

The Trail Making Test - Part B (TMT-B) is a commonly used executive control measure with a known floor effect, limiting the ability to distinguish impairment among individuals unable to complete this task in the standard time limit. Our group previously proposed the TMT-B Efficiency Score (TMT-Be), which captures performance variability among examinees who fail to complete the task. The present study assesses the TMT-Be in a longitudinal clinical sample. Data were collected via record review of veterans who underwent two clinical neuropsychological evaluations. We identified 30 veterans (mean age Visit 1:69 ± 8.7 years) who were unable to complete TMT-B during at least one evaluation (mean days between visits = 615). Two scoring systems were utilized to examine performance variability: TMT-Be and TMT-B Prorated Score (TMT-Bpr). TMT-Be distribution was less skewed, but more platykuric, compared to TMT-Bpr. TMT-Be and TMT-Bpr were highly correlated. Both metrics correlated with psychomotor speed and another executive task, but not confrontation naming, providing both convergent and discriminant evidence of validity. TMT-Be, but not TMT-Bpr, detected significant decline in performance longitudinally. Age and education were significant predictors of the TMT-Be, but not TMT-Bpr, difference scores. Both metrics captured performance variability in a clinical sample and provided sufficient variance for examining floor-level performance on the TMT-B. TMT-Be appeared to be less prone to creating outliers and more likely to detect change. The results support the utility of the TMT-Be metric in research and clinical settings.