Authors :
Imani Evans, BA – Vanderbilt University, Vanderbilt University Medical Center
Terra Lee, BS, BA – Vanderbilt University, Vanderbilt University Medical Center
Presenting Author: Shawniqua Williams Roberson, MEng, MD – Vanderbilt University Medical Center
Rationale:
Many people with epilepsy suffer deficits in executive function, e.g. problems with processing speed, working memory, and cognitive flexibility, for which the Trail Making Test (TMT) is a well-validated measure. Despite its common use in neuropsychological evaluations, the TMT requires pen and paper, making it impractical to administer remotely. Yet remote administration of cognitive assessments would greatly improve generalizability of research studies on cognition in epilepsy patients, many of whom are unable to drive to appointments. We designed an analogous test, the Trails-V, that relies only on verbal responses, thus can be administered in person or via videoconference. We report the preliminary feasibility and comparative validity of the Trails-V in a small cohort of epilepsy patients.
Methods:
In this prospective observational study, we included patients with presumed epilepsy diagnosis, upon admission to the Epilepsy Monitoring Unit. We excluded patients who could not self-consent due to cognitive impairment, and those with severe impairments in vision or motor control. We administered the TMT Parts A and B, followed by Trails-V Parts A and B. We assessed feasibility based on completion rates and error rates. We compared completion times of Parts A and B, and of the B/A ratios, using Spearman correlations. Type I error threshold was set at 0.05.Results:
We recruited 16 participants, of whom 3 were administratively withdrawn. Of the 13 remaining participants (10 women, 2 men, 1 nonbinary), 12 identified as White, 2 as African-American, and 1 as Native American or Pacific Islander. Ten (77%) had attained a high-school diploma or greater. Ten (77%) participants completed both assessments. One participant had 1 error on Trails-V, part A. Four participants had 3 or fewer errors on Trails-V, part B. Median (interquartile range – IQR) completion times for the TMT parts A and B were 30 (22, 58) seconds and 60 (37, 252) seconds, respectively. Median (IQR) times for the Trails-V parts A and B were 12.5 (10, 22) and 28 (17, 47), respectively. Trails V part A did not correlate with TMT-A (rho= –0.16, p=0.67). Trails-V part B did correlate with TMT-B (rho = 0.87, p=0.003). Trails-V B/A ratio did not correlate with the TMT-B/TMT-A ratio (rho = 0.35, p = 0.33).Conclusions:
Our novel measure of executive function, using verbal responses, is feasible and shows a wide range of scores in a small group of epilepsy patients. Completion times for part A, but not part B nor the B/A ratio, may correlate with those measured using the traditional Trail Making Test. Our preliminary observations suggest that with optimization of part A, Trails-V may be an alternative to the TMT, to assess executive function when in-person assessment is not feasible. Further study with remote (via video) formats and larger cohorts will be needed to confirm external validity and establish norms after optimization.Funding: This project was funded in part by the Vanderbilt Undergraduate Summer Research Program, and by the National Institutes of Health (K23 AG072030)