Purpose
A specific simulator was used to assess the driving visual performance in patients with dry eye disease (DED) and to determine clinical predictors of visual impairments while driving.
Design
Prospective case-control study.
Methods
The study was conducted in the Center for Clinical Investigation of Quinze-Vingts National Ophthalmology Hospital, Paris, France. Twenty dry eye patients and 20 age- and sex-matched control subjects were included. Vision-related driving ability was assessed using a specific driving simulator displaying randomly located targets with a progressive increase in contrast to be identified. Other examinations included clinical examinations, serial measurements of corneal higher-order aberrations (HOAs), and vision-related quality-of-life questionnaire (Ocular Surface Disease Index [OSDI]). Data collected during driving test (ie, the number of targets seen, their position, and the response time) were compared between groups and analyzed according to clinical data, aberration dynamics, and quality-of-life index.
Results
The percentage of targets missed as well as average response time were significantly increased in DED patients as compared with controls ( P < .01). More specifically, the visual function of DED patients was more impaired in specific situations, such as crossroad or roundabout approaches. In DED patients, the response time was found to positively correlate with the progression index for HOAs ( P < .01) and with the OSDI “symptoms” subscale ( P < .05).
Conclusions
Degradation of ocular optical qualities related to DED is associated with visual impairments during driving. This study objectively has demonstrated the impact of tear film–related aberration changes on activities of daily living in DED.
Dry eye disease (DED) is recognized as a growing public health problem and one of the most frequent reasons for seeking eye care. The DED definition has evolved with recent epidemiologic studies as well as a better understanding of the pathophysiology of the disease. It is estimated to affect from 5% to over 30% of the population, depending on the diagnostic criteria. This common health problem is likely to be overlooked because it tends not to be a common cause of visual morbidity as standardly measured. Nevertheless, there is increasing evidence that DED is a major cause of visual disturbance, which degrades the quality of everyday life and can impact health status.
According to a recent overview arising from the 2007 International Dry Eye Workshop, DED causes damage to the ocular surface and symptoms of ocular discomfort associated with impaired visual quality. Indeed, patients with DED often report vision-related difficulties in doing daily activities. In clinical practice, the main difficulty in managing DED stems from the variability of the symptoms, the lack of a single reliable diagnostic test, and weak correlations between clinical tests, optical and biological examinations, and patient-reported deterioration in quality of life. The precorneal tear film plays an important role in ocular optical quality since it is the most anterior refractive surface of the eye. In the majority of patients with DED, the visual acuity is still 20/20 as standardly measured, but instability of the tear film introduces wavefront higher-order aberration (HOA) changes that always contribute to a decrease in the quality of vision. Our team recently demonstrated that a specific analysis of the time course of HOAs provides objective and quantitative data that are correlated with both clinical signs and patient-reported outcomes, raising the possibility of using this instrument as a new surrogate marker for the disease.
Beyond conventional clinical examination and visual acuity measurement, a specific evaluation of the visual function in daily living tasks is now required to better define the impact of the disease on this population’s health status but also to better assess eligibility or changes over time in clinical trials. Although DED patients commonly complain of difficulties in doing vision-related daily activities, as previously reported using quality-of-life questionnaires, no study has been conducted to determine whether or not DED could be responsible for an objective decrease in visual performance while driving. The present study addresses the impact of DED on a crucial daily activity of modern living. A driving simulator dedicated to visual function evaluation was used in patients with DED and in age- and sex-matched healthy controls in order to better specify the relationship between driving difficulties, objective ocular signs and optical degradation, and patient-reported vision-related quality of life.
Methods
Patients
The study was conducted in the Clinical Center for Investigation of Ocular Surface Pathology (Quinze-Vingts National Ophthalmology Hospital, National Institute for Health and Medical Research 503, Paris, France) in accordance with the Declaration of Helsinki, Scotland amendment, 2000. Previous approval was obtained from the National Ethical Research Committee (Comité de Protection des Personnes Ile de France V, agreement number 10793). All patients gave informed consent to participate in this clinical research study. Twenty white patients with DED and 20 white age- and sex-matched control subjects were prospectively and consecutively included. DED was diagnosed by the association of ocular symptoms and tear film abnormalities (Schirmer I test <5 mm/5 min and/or tear break-up test <10 s), with or without ocular surface damage (corneal and conjunctival staining), according to the DEWS criteria from the modified Delphi Panel Report. Only the subjects with a best-corrected visual acuity of at least 0 logMAR were included, since this study focused on a decrease in visual function related to tear film degradation and ocular symptoms but not to extensive corneal damage. At inclusion time, all patients were treated with tear substitutes only, without any anti-inflammatory or cyclosporin medication, and without changes within the last 3 months. Healthy age- and sex-matched subjects with no ocular pathology, with no treatment, and without any symptoms or signs of DED (Schirmer I test >10 mm/5 min and Oxford score = 0) were included as controls. All participants were in good general health and were licensed drivers with at least weekly driving practice. Exclusion criteria were any ocular pathology but DED, eyelid malposition or dynamic disorders, previous ocular/eyelid surgery, contact lens wear, systemic disorder, pregnancy, and treatment changes within the last 3 months.
Clinical Examination and Questionnaire
Slit-lamp evaluations were conducted in a defined sequence and included tear break-up time measurement (s, mean of 3 consecutive tests), ocular surface fluorescein staining (grade 0-5, according to the Oxford score), lissamine green staining (grade 0-9, according to the van Bijsterveld score), and Schirmer I test (mm/5 min, without anesthesia). Before clinical examination, a trained interviewer (G.R.) administered the French version of the Ocular Surface Disease Index (OSDI) questionnaire, which was developed to quantify the specific impact of DED on vision-targeted health-related quality of life. This disease-specific questionnaire includes 3 subscales: ocular symptoms (OSDI-symptoms), vision-related activities of daily living (OSDI-function), and environmental triggers. Each subscale (0-100) was computed, as well as an overall averaged score (0-100).
Dynamic Aberrometry
Serial measurements of corneal and ocular wavefront aberrations were simultaneously performed every second for 10 s after blinking using the dynamic aberrometer KR-1 (Topcon, Clichy, France). The entire procedure has been previously described. Briefly, HOAs were recorded in mesopic conditions without any pharmacologic mydriasis, analyzed by expanding the set of Zernike polynomials up to the sixth order, and expressed for the central 4-mm diameter. The progression index of total (third- to sixth-order) HOAs was defined as the slope of the linear regression line of HOAs throughout the recording period, as previously defined.
Driving Test
We used a driving simulator purchased from Develter Innovation (Ile de France, France). This simulator has an automatic shift. Driving tests were performed with the best spectacle correction in scotopic conditions on a standardized 5-km circuit. Each test had a series of 7 lighted targets, increasing in intensity for 15 s and then disappearing. Lighted targets randomly appeared during the test at various positions and various driving conditions: straight forward, straight backward, at a crossroad entrance, and on the right-hand or left-hand side of a crossroad. For each target seen, the patient had to press a remote button on the wheel. Data included the number of targets seen/missed, their respective location, and the average response time. The results were determined as the mean of 3 consecutive tests.
Statistical Analysis
All data are given as the mean ± SD. For ocular examinations—clinical evaluation, tear osmolarity measurement, and wavefront aberrometry—1 eye per patient was selected using a random number table in order not to bias the statistical relevance of the results. Data were controlled for normality, homogeneity of variances, and sphericity in order to perform the adequate tests. The 2 groups were compared using parametric t tests. In the DED group, scatterplots and Spearman correlation coefficients were used to assess the association between pairs of variables. The probability level of significance was adjusted according to the post hoc Bonferroni procedure in order to maintain an overall type I error equal to 0.05.
Results
The profile, clinical features, and OSDI scores of each group are detailed in the Table . Six patients presented mild-severity DED and 14 patients presented moderate-severity DED, according to the Delphi approach. Significant differences in all the clinical characteristics and OSDI scores were found between DED patients and controls (paired t test, P < .01 for each).
| Dry Eye Patients (n = 20), Mean ± SD (min/max [95% CI]) | Controls (n = 20), Mean ± SD (min, max [95% CI]) | |
|---|---|---|
| Age (y) | 53.4 ± 16.2 (22/84 [46.3-60.5]) | 53.1 ± 16.4 (22/84 [45.9-60.3]) |
| Sex ratio (m/f) | 0.25 | 0.25 |
| Clinical data | ||
| Tear break-up time (s) | 5.9 ± 2.2 (2/10 [5.0-6.9]) | 11.4 ± 3.7 (4/15 [9.9-13.1]) |
| Schirmer (mm) | 9.5 ± 5.4 (1/20 [7.2-11.9]) | 19.6 ± 0.6 (15/20 [19.4-19.9] |
| Oxford (0-5) | 1.1-0.8 (0-4 [0.7-1.4]) | 0 |
| Van Bijsterveld (0-9) | 2.7 ± 1.6 (0-6 [1.9-3.3]) | 0.1 ± 0.1 (0/1 [0-0.1]) |
| Ocular Surface Disease Index | ||
| Overall score | 48.1 ± 18.4 (10.4/89.6 [40.6-56.6]) | 2.2 ± 2.9 (0/10.4 [0.9-3.3]) |
| OSDI symptoms | 43.3 ± 15.6 (15/80 [36.4-50.1]) | 2.1 ± 3.1 (0/15 [0.8-3.5]) |
| OSDI functions | 41.3 ± 27.8 (0/93.8 [29.1-53.4]) | 1.8 ± 2.9 (0/12.5 [0.5-3.1]) |
| OSDI triggers | 58.3 ± 29.2 (8.3/100 [45.6-71.1]) | 2.4 ± 3.9 (0/16.7 [0.7-4.1]) |
Comparative Analysis of Aberration Dynamics Between Groups
Significant variation with time in corneal total HOAs (repeated-measures ANOVA, P < .01), third-order coma ( P < .01), and third-order trefoil ( P < .01) was found in DED patients, whereas no significant change occurred in the control group throughout the recording period. As detailed in Figure 1 , the progression index of corneal total HOAs and of corneal third-order trefoil was significantly higher in DED patients than in healthy controls ( P < .01 and P < .05, respectively).
Driving Visual Performance
The average response time to identify targets was significantly higher in DED patients than in controls ( P < .01) ( Figure 2 , Left). Moreover, a significant difference in the average number of targets seen was found between groups ( P < .01), further depending on target location ( Figure 2 , Right): interestingly, targets appearing at a crossroad entrance and at the right-hand side of a crossroad were more often missed by DED patients than by healthy subjects ( P < .01 and P < .05, respectively). On the contrary, targets appearing straight on (forward or backward) were equally detected in the 2 groups.
