Predicting Driving Fitness in a Low Vision Clinic: Correlating Two Tasks

Purpose: The Useful Field of View (UFOV) computer task is a visual processing assessment tool used to aid in predicting driving fitness in low vision patients. The Dynavision 2000 board tests visuomotor reaction time, and also has been investigated for training skills pertinent to driving. However, very few studies to date have investigated Dynavision scores with relation to criteria for determining driving fitness, and standard normative scores have not yet been established. This study investigated the correlation between performance on the UFOV and Dynavision tasks, as the tasks could be used interchangeably if the correlation were strong. Performance data on both tasks were gathered for normal-vision adults and low vision patients of driving age. The effect of age and gender on subject performance on each task also was analyzed.

Methods: Fifty-one normal-vision adults and 17 low vision patients participated on the UFOV and the Dynavision Mode A tasks. Each subject completed one UFOV computer trial, which measures visual processing speed (ms) as flashed images must be identified and located on the screen, and three 60-s trials on Dynavision Mode A, which determines visuomotor reaction times as a random sequence of illuminated buttons are struck on a large wall-mounted board. The UFOV Selective Attention task (Task 3) was analyzed in isolation because most normal subjects achieved the optimal threshold for Task 1, Processing speed (100% normal vision vs. 59% low vision) and Task 2, Divided Attention (84% normal vision vs. 6% low vision).

Results: UFOV and Dynavision scores correlated positively and significantly (p<0.05) in normal adult (r2 = 0.392) and low vision subjects (r2 = 0.479). Increasing age was associated with poorer scores on both the UFOV and Dynavision tasks in the normal adult and low vision cohorts. Gender effect on performance was significant only for Dynavision performance in normal vision subjects, as males performed with shorter reaction times. Gender did not significantly affect performance on UFOV or on Dynavision for low vision patients.

Conclusion: Performance on the UFOV and Dynavision tasks correlate moderately in both low vision and normal vision adult subjects. The amount of variability on one task accounted for by the other (39% for normal-vision adults and 48% for low vision patients) is not enough to support the interchangeability of the two tasks clinically. Dynavision should be further and more specifically investigated for its individual cutoffs, sensitivity, and specificity for identifying driving candidates and correlating to crash risk. If found to be related to driving capability, the results of both UFOV and Dynavision tasks could be combined to improve the recommendations concerning driving by low vision patients.