Driving cessation and driving limitation in glaucoma: the Salisbury Eye Evaluation Project

Abstract

Objective: To determine if glaucoma is associated with driving limitation or cessation.

Design: Cross-sectional analysis within a longitudinal, population-based cohort study.

Participants and controls: One thousand one hundred thirty-five ever-drivers between 73 and 93, including 70 subjects with unilateral and 68 subjects with bilateral glaucoma.

Methods: All subjects reported their driving habits during each of 4 study rounds. During the fourth and final study round, subjects were assessed systematically for the presence of glaucoma.

Main outcome measures: Self-reported driving cessation or driving limitation, including cessation of night driving, driving fewer than 3000 miles annually, or cessation of driving in unfamiliar areas.

Results: Fifteen percent of subjects without glaucoma no longer were driving at the end of the cohort study, compared with 21% of unilateral glaucoma subjects (P = 0.2) and 41% of bilateral glaucoma subjects (P<0.001). Multivariate regression analysis showed that bilateral (odds ratio [OR], 2.6; P = 0.002), but not unilateral (OR, 1.5; P = 0.3), glaucoma subjects were more likely no longer to be driving when compared with subjects without glaucoma. The odds that bilateral glaucoma subjects no longer were driving doubled for every 5 dB of visual field (VF) worsening in the better eye (P<0.001). Driving cessation within the previous 2 years was analyzed using separate multiple regression models, and both bilateral (OR, 3.6; P = 0.004) and unilateral (OR, 2.4; P = 0.06) glaucoma subjects were more likely to stop driving over this period than subjects without glaucoma. Driving cessation associated with bilateral glaucoma was present in 0.82% of the population, or 1 in every 122 individuals. Greater numbers of driving limitations were not more likely among subjects with glaucoma than subjects without glaucoma. However, bilateral glaucoma subjects did attribute more driving limitations to difficulties with their vision than subjects without glaucoma (OR, 2.2; P = 0.02).

Conclusions: Bilateral, and possibly unilateral, glaucoma is associated with significantly higher rates of driving cessation among the elderly. The substantial difference in driving patterns seen with different degrees of better-eye VF damage suggests that minimizing VF loss in the better-seeing eye is associated with better functional outcomes.

Figure 1

Timeline of Salisbury Eye Evaluation and Testing Performed Visual asses = Visual Assessment, including binocular acuity with habitual correction and contrast sensitivity in each eye; Driving Qnr = Driving Questionnaire; ST VFs = suprathreshold visual fields; 24−2 VFs = Visual fields using Humphrey SITA fast 24−2 algorithm. Covariates include time-invariant variables (age, race, sex, education) and time-varying covariates (medical comorbidities and cognitive status). Data for time-varying covariates was taken from study round 4.

Figure 2

Modeled probability for driving cessation amongst subjects with bilateral glaucoma. Both the probability of not driving, and mean deviation of the better-seeing eye are adjusted for confounding variables. Outcomes for study subjects with bilateral glaucoma are displayed as individual points, with subjects who are still driving having a probability of 0, and subjects who have stopped driving having a probability of 1.

Figure 3

Duration of discontinued driving by glaucoma status. Unilateral and bilateral glaucoma subjects were compared to subjects without glaucoma for each time interval using multinomial logistic regression. * p=0.06, **p<0.05 in multinomial regression model when compared to subjects without glaucoma. In all other time periods, no significant difference in driving cessation was observed when compared to subjects without glaucoma.