Development of a computerized Digit Vigilance Test and validation in patients with stroke.

Abstract

To develop a computerized Digit Vigilance Test (C-DVT) with lower random measurement error than that of the DVT and to examine the concurrent validity, ecological validity, and test-retest reliability of the C-DVT in patients with stroke. A cross-sectional study. Forty-four patients with stroke. We developed and tested the C-DVT. To examine the psychometric properties, the participants completed both the C-DVT and DVT twice with a 14-day interval. We developed the C-DVT on the basis of expert input and examinee feedback. C-DVT scores were highly correlated with DVT scores (ρ = 0.75), supporting the concurrent validity. The C-DVT scores were moderately correlated with the scores of the Barthel Index and the Activities of Daily Living Computerized Adaptive Testing system (ρ = -0.60~-0.57), supporting the ecological validity. The test-retest agreement of the C-DVT was excellent (intra-class correlation coefficient = 0.92). The random measurement error of the C-DVT (minimal detectable change percent change (MDC%) = 15.4%) was acceptable and lower than that of the DVT (33.0%). The practice effects of the C-DVT were statistically significant, but the effect size d was small (0.15). A C-DVT with a limited amount of random measurement error was developed. These preliminary findings show that the C-DVT demonstrates satisfactory concurrent validity, ecological validity, and test-retest reliability in patients with stroke.