Abstract
Purpose :
Bioptic telescopic spectacles (BTS) allow people with vision impairment to obtain driving privileges. Hazard perception is the ability to anticipate potential road hazards to avoid a collision and has been shown to predict collision rates. The purpose of this study was to analyze the predictive value of various threshold central visual field calculation methods on hazard perception.
Methods :
ETDRS visual acuity and log contrast sensitivity were measured. A binocular visual field was constructed using the better total deviation value from corresponding monocular points from Humphrey 24-2C SITA Faster tests. Values for superior, inferior, left, and right hemifields were also calculated. Hazard perception testing was conducted using 30 commercially available first-person driving video clips. Hazard detection times for each clip were converted to z-scores, converted back to seconds using the average response time across all videos, and then compared among conditions. Backward selection with liberal criterion for removing potential predictors was used to compose regression models that maximized explanation of variance.
Results :
Twenty-one licensed bioptic drivers and 21 control drivers completed the study. There was no significant difference in age (mean±SD = 50±15 years for bioptic drivers vs 50±12 years for controls; P=0.926). Control drivers had significantly better VA (logMAR -0.05±0.07) than visually impaired drivers (logMAR 0.69±0.20, P<0.001), even when these drivers used their bioptic telescope (logMAR 0.14±0.12, P<0.001). Log CS was poorer among bioptic drivers (1.57±0.21 vs 1.80±0.06, P<0.001). Bioptic drivers had significantly poorer binocular visual field total (-4.11±3.65 dB vs. 0.65±1.88 dB, P<0.001) and superior deviation (-4.28±3.99 vs. 0.86±1.44 dB, P<0.001). Controlling for age, binocular superior visual field deviation (P<0.001) as a better predictor of hazard response time than binocular inferior (P=0.079), left (P=0.002), right (P=0.021), or overall deviation (P=0.005), and better eye mean deviation (P=0.002). The best predictive model (R2=0.35) of driving hazard response time included age, sex, visual acuity, and binocular superior visual field mean deviation.
Conclusions :
A combination of vison measurements is useful for predicting performance on a hazard perception test in bioptic drivers with low vision. Further work to determine how these findings are related to on-road performance and outcomes is warranted.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.