The limitations of Snellen acuity testing in assessing visual function have inspired the development of alternative modes of visual testing. In particular, glare and contrast sensitivity testing provide important information about visual function that can supplement Snellen acuity readings.1,2 These tests are particularly significant in patients who have relatively good Snellen acuities yet complain of significant visual disability, especially at night.3,4
Accurate glare disability testing has the potential to play a crucial role in public safety. The Impact of Cataracts on Mobility project demonstrated drivers with cataracts are 2.5 times more likely to be involved in motor vehicle accidents than age-matched drivers without cataracts.5 Patients with cataracts do not typically report substantial visual impairment under normal lighting conditions but tend to experience exaggerated visual impairments under conditions of glare,3,6 which can cause temporary night blindness. As a result, one of the most common complaints from patients with cataracts is that glare impairs their night driving.3,6
Accurate documentation of functional visual loss due to glare is important, but existing studies evaluating the relative sensitivity and accuracy of current glare testing devices have been inconclusive,3,7–11 suggesting a more accurate screening method is warranted. Previous studies of disability glare in patients with mild-to-moderate cataracts have failed to detect an association between glare symptoms and scores on disability glare tests.9,12–16 Because cataract surgery may be indicated by glare symptoms without a significant reduction in Snellen acuity, clinicians need a reliable glare testing method that mimics patients’ visual symptoms.
In the United States, published policies recognize reduced vision due to glare disability as medical justification for cataract surgery, and these policies are consistent with the American Academy of Ophthalmology’s Preferred Practice Pattern on cataracts.2 A Corcoran Consulting Group review of random cataract surgeries reimbursed by U.S. government or other U.S. third-party payors from 2009 to 2012 (n = 2,220 procedures/112 practices) determined the documentation of medical necessity for cataract surgery was weak or missing altogether in 21% of those procedures; in 4% of those procedures, the flawed chart documentation was directly related to lack of glare testing (personal communication, January 29, 2013).
The Brightness Acuity Tester (BAT)17 uses diffuse illumination and is often used to assess glare, but results obtained can be variable because the intensity and angle of light to the eye is not consistent. Several reports have found the BAT device to have poor sensitivity and validity.3,7–11 Other methods of glare testing include having technicians ask patients to read an eye chart while shining a small flashlight in the eye, but results are highly variable and subjective; testers and testing approaches should be repeatable and reliable, and should be able to detect eyes with some degree of opacification.11,18 It has been suggested that glare tests include measurements of glare effects under conditions that mimic those found in night driving (ie, when ambient light is low).3
The primary outcome measures of this study were to determine the change in corrected distance visual acuity (CDVA) created by the device and its validity in the measurement of glare-induced visual disability caused by a cataract. A secondary objective was to determine the correlation between the glare-induced changes in CDVA with functional glare disability, as measured by the glare subcategory questions from the Refractive Status Vision Profile (RSVP) questionnaire.19,20 Participants were also asked if the glare source simulated what they actually experience at night while driving.
Patients and Methods
The controlled point source light-emitting diode (LED) glare tester (EpiGlare Tester; Epico, LLC, Columbus, OH) can be attached directly to a phoropter, and provides point source illumination with a constant intensity and light spectrum and defined angle of incidence to the eye. By design, when diffuse illumination is used (as with the BAT), there is a wide range of veiling luminance due to the wide range of incidence angles. By using a point source illumination, a single point on the curve (Figure 1) is established along with its resulting known and consistent veiling luminance. Automobile headlights, sun reflections, and lampposts also act as a point source, and have been specified as major sources of veiling luminance.
Relationship between equivalent veiling luminance and angular separation between line of sight and glare source.
The brightness of the individual LED on the controlled point source LED glare tester is designed to approximate the median glare dosage for high intensity discharge and incandescent tungsten halogen systems at 40 m distance misaligned by 1.5° at average passenger car mounting height, as described by the U.S. Department of Transportation and National Highway Traffic Safety Administration.21
The device places four sources of light evenly distributed around the viewing aperture to maintain the patient’s attention along a specific line of sight axis to minimize participant distraction. The interior of the device cup is black by design to minimize reflections and fully control the angular source of the glare light.
To simulate the effect of straylight in night-driving conditions, we used the Stiles-Holladay disability glare formula to approximate the veiling luminance as a reference for the device.7,21,22 This equation clearly implies that intraocular scatter is the primary source of glare: Lv = 10 Eglare / θ2, where Lv is the equivalent veiling luminance in cd/m, Eglare is the illuminance of the glare source at the eye measured in lux, and θ is the angular distance between the line of sight and the glare source in degrees (Figure 2).7,21,22 The angle and intensity of the light source are critical features to standardize while measuring glare.15,20,22–24 Glare conditions that induce disability are most significant when the light source is at an angle less than 25° of arc separation from the object being viewed.7,16,25 This is based on the veiling illuminance curve; Figure 1 shows that the veiling luminance (y axis on a log scale) is highly dependent on angle of incidence (x axis). The equivalent veiling luminance increases exponentially as the angle becomes small and it falls off exponentially as the angle becomes large. However, the curve is relatively flat between 15° and 25°.16 Glare testing in this region would have the advantage of eliminating the nonlinear effects at low and high angles and provides a minimum of variation of veiling luminance with respect to angle change.
The classic Stiles-Holladay disability glare formula for a point glare source: Lv(θ) = [10 E / θ2] for 1° < θ < 30°; where Lv = veiling luminance, E = illumination (lux) incident on the cornea and θ = the angle in degrees between the line of sight and glare source.
The device consists of a housing unit with two moveable eyepieces, two rechargeable batteries, and a battery charger and is attached to a phoropter for use during an eye examination (Figure 3).
Close up view of the device prototype attached to a phoropter.
The glare source consists of four LEDs that produce a constant illumination level with known spectral properties, including a total bandwidth in the range of 430 to 685 nm. The device has two intensity levels that are electronically controlled to provide uniform test conditions. The device was designed and tested to conformity with IEC 62471:2006 “Photobiological Safety of Lamps and Lamp Systems” and is certified beneath the retinal blue light hazard limit specified in IEC 62471:2006.
Eighty-nine participants were enrolled in this prospective multicenter study: 40 patients (80 eyes) with senile cataract and 49 normal participants (98 eyes) without cataract.
Prospective candidates were identified from the investigator’s usual clinic population. Slit-lamp examinations were performed in undilated eyes before any examination drops were instilled. Participants were classified as having either bilateral cataracts or clear lenses in both eyes. The opacity subtype (cortical, nuclear, or posterior subcapsular) and grade (clear, trace, 1+, 2+, 3+, or 4+) were determined for those eyes with cataract.
Study participants were required to meet all of the following criteria: 18 years or older, presenting with a cataract of lens grade 2+ or greater (cataract group) or having a clear or trace lens grade (non-cataract group), ability to understand and provide written informed consent as required by the site’s Institutional Review Board, and be willing and able to comply with testing according to the Investigator. Eligible participants were assigned to either a non-cataract group (normal participants) or a cataract group, based on the lens finding of the screening slit-lamp examination performed by the investigating ophthalmologist. Normal participants were required to have a clear or trace lens in both eyes (cataract classification = none, clear or trace nuclear or cortical). Cataract group eligibility required that both of the participant’s eyes have a lens grade 2+ or greater (cataract classification = nuclear, cortical, or posterior subcapsular).
Participants were excluded if the study eye showed evidence of or if the patient reported the following: CDVA of logMAR 0.50 or Snellen 20/63 or worse in either eye, ocular pathology, including corneal or macular disease or glaucoma, an inability to comply and cooperate with the testing device, and a presence or history of any other condition or finding or concomitant medication that, in the investigator’s opinion, made the participant unsuitable as a candidate for the device or study participation or may confound the outcome of the study.
The study conformed to the tenets of the Declaration of Helsinki and was conducted in accordance with the U.S. Food and Drug Administration’s Good Clinical Practice regulations as a nonsignificant risk study in accordance with the abbreviated Investigational Device Exemptions requirements of 21 Code of Federal Regulations (CFR)§812.2(b). Under the abbreviated requirements, nonsignificant risk study investigations of a device are considered to have approved applications for an Investigational Device Exemption if the sponsor and investigators comply with requirements of 21 CFR§812.2(b).
Patients who elected to participate in this study gave their written informed consent prior to the use of the investigational device. The study followed the procedures written below and the work was HIPPA compliant. This protocol and the informed consent form were approved by Chesapeake and Ohio Health Institutional Review Boards.
Functional Vision Questionnaire. Functional visual ability was evaluated before glare testing and was performed using the driving and glare subscales of the RSVP questionnaire19,20 (Figure A, available in the online version of this article). Each subscale consisted of three questions for driving and three questions for glare. Each question was answered for each type of specified optical correction.
Refractive Status Vision Profile (RSVP) Questionnaire: Driving and Glare Subscales.
Glare Testing. The glare testing measurements for the study were obtained in a dimly lit room prior to the instillation of any eye drops administered for examination purposes. CDVA was obtained using an Early Treatment Diabetic Retinopathy Study (ETDRS) chart. Manifest refraction and CDVA were measured through the phoropter in each eye without glare. Those who were unable to read all letters correctly on the 20/63 line (logMAR 0.50) were excluded from the study for failure to meet inclusion criteria. Visual acuity was measured as the total number of correct letters identified and expressed in logMAR. CDVA with glare was then measured with the device as the glare source using the following procedures.
The device eyepiece was placed over the lens of the phoropter. The patient was aligned with the center of the phoropter lens. The device power switch was turned “On” and the “High” glare setting was used per protocol. Using the participant’s manifest refraction in the phoropter, the CDVA was measured with the glare source. The test was performed monocularly by occluding the contralateral eye. The requirement that participants have either clear lenses or cataracts in both eyes was instituted to prevent biased results. Starting at the 20/63 (logMAR 0.50) line of the ETDRS visual acuity chart, the participant was asked to read each letter left to right. If the participant read all of the letters on the 20/63 line (logMAR 0.50), he or she was asked to begin reading each letter on the next lower line. The participant was encouraged to continue reading each line until three or more letters on a line were missed. Disability glare was scored as the decline in the number of letters correctly identified in the presence of glare.
The following question was asked when CDVA with glare testing was completed in both eyes: “With respect to driving at night, in thinking about how your vision is affected when you see headlights from an oncoming car, in your opinion, how much does this device accurately represent the difficulty you experience while driving at night?”
The following response options were provided to each participant: 0 = unable to determine if device simulates headlight glare; and/or vision is not affected by headlight glare; 1 = very little, if any; 2 = somewhat similar; 3 = very similar; and 4 = almost exactly.
Finally, the device was rated by the examiners using the qualitative questionnaire (Figure B, available in the online version of this article).
User assessment of the device.
Tabulation of summary statistics, graphical presentations, and statistical analyses were performed using Stata 10 software (Stata Corp., College Station, TX). All two-sided testing and confidence intervals used a significance level of 5%.
The Wilcoxon signed ranks test was used to evaluate whether the change in visual acuity was different from zero for each group individually (ie, patients with cataracts and participants with clear lenses) after testing with the device. If the distribution of change in visual acuity showed good conformance to normal distribution, then the t test for individual samples was used. Visual acuity values were converted to the logMAR equivalent and to total numbers of letters observed. All statistical calculations were formed from the logMAR equivalent and from the number of letters observed.26
The study’s secondary outcome was whether patients in the non-cataract and cataract groups perceived the device as an accurate simulation of their nighttime driving experience. We applied the binomial test for percent positive response. Null hypothesis – positive response = 50% (random chance).
Baseline data did not exist to facilitate sample size calculations. The sample size was based on the Wilcoxon rank sum test, and assumed that the probability is 0.70 that visual acuity would decline by a greater amount with glare for patients in the cataract group than for those in the non-cataract group. The type 1 error rate was controlled at 5% and the power was selected at 90%.
Primary Efficacy Criteria
CDVA was evaluated with and without glare using the device as the glare source. For each eye, the difference in CDVA with and without glare was estimated using 95% confidence intervals. The mean changes in CDVA without and with glare, as determined by a paired analysis, were estimated using 95% confidence intervals.
Functional Vision Questionnaire
Functional visual ability was estimated using the subscale scores for driving and glare from the RSVP questionnaire. The composite score for driving and glare subscales and the individual subscale scores were correlated with the glare-induced changes in CDVA. The RSVP subscale questions administered before glare testing established the participant’s baseline functional visual ability/disability for glare and driving difficulty.19,20
Other Outcome Measures
Device performance, as estimated from the participants’ subjective assessment of the degree of similarity between glare produced by the device and their nighttime driving experience, was tabulated and summarized.
Eighty-nine participants were enrolled at five sites. The mean age of the cataract and non-cataract groups was 69.7 and 35.1 years, respectively. Females represented 80% of the cataract group and 71.4% of the non-cataract group. The cataract group comprised 85% whites and 15% African Americans. The non-cataract group comprised 79.6% whites, 12.2% Hispanics, 6.1% African Americans, and 2.0% Pacific Islanders.
In the non-cataract group, the mean ± standard deviation manifest refractive spherical error was −1.43 ± 2.4, median: −0.5, range: −10.5 to 1.5. In the cataract group, the mean manifest refractive spherical error was −0.15 ± 2.6, median: 0.69, range: −10 to 4.1. The non-cataract group had a mean age of 35.1 ± 8.9 years, median: 33 years, range: 18 to 55 years. The cataract group had a mean ± SD age of 67.9 ± 8.2 years, median: 70 years, range: 49 to 82 years.
Primary Efficacy Criteria
The CDVA reduction with glare from the device was greater for the cataract group than the non-cataract group (P < .001), determined using the Mann–Whitney test to compare the mean change in both groups. The cataract group had a mean change in logMAR of −0.49 ± 0.3 (median: −0.46, range: −1.08 to 0.0). In the non-cataract group, the mean change in CDVA (logMAR) using the device was −0.13 ± 0.2 (median: 0.0, range: −0.7 to +0.3). This equates to a 5-line Snellen reduction (0.49 logMAR) in the cataract group and a 1-line reduction (−0.13 logMAR) in the non-cataract group. The Wilcoxon rank sum test showed a statistically significant difference in change in CDVA between the normal and cataract groups. Mixed models were also used, taking into account the potential correlation between eyes from the same participant, and results were the same: a larger reduction in CDVA for those in the cataract group compared to those in the non-cataract group. We controlled for age in the mixed effect model by comparing the change in CDVA between patients with cataract and normal participants. We found the patients with cataract were significantly more likely to experience a decrease in their CDVA compared to those without cataract.
Functional Vision Questionnaire
We used the RSVP questionnaire (Driving and Glare Subscales) to quantify visual performance and disability,19,20 and assist in validating the glare tester (Figure A). The controlled point source glare test consistently caused a larger change in logMAR value in cataract compared with normal participants across all RSVP subscale levels.
The majority of participants in both groups responded the device truthfully represented the participant’s nighttime driving experience. In the cataract group, 83% of the patients stated the device accurately represented the difficulty they experience while driving at night. In the non-cataract group, the device accurately represented the difficulty with night driving (or ease of night driving, in this group) in 71%. The differences were statistically significant (P < .001 and = .003, respectively).
The main goal of this study was to evaluate objectively and subjectively a new technology in detecting glare-induced decline in visual acuity. Our results indicate that the device accurately detects glare-induced visual disability caused by cataract. Patients with cataract were significantly more likely to experience a decrease in their CDVA compared to those without cataract (P < .001) after being exposed to glare conditions simulated by the device. In addition, the majority of patients with cataract (83%) reported the device accurately reproduced the difficulty they experience with night driving. Furthermore, study investigators reported the device was easy to use and useful in daily practice.
Disability glare tests have been unable to detect an association between glare symptoms and scores in patients with mild-to-moderate cataract in other studies.7,9,13–16 The BAT has been shown to be inaccurate, at times overestimating the patient’s glare disability, and even underestimating the disability.7,9–11 In addition, the BAT is not, nor does it claim to be, specifically designed to simulate glare during night driving conditions.17
A study by Prager et al.11 examined two popular commercially available devices, the BAT and Miller-Nadler glare tester, in predicting outdoor acuity in patients with minimal cataract. In outdoor vision, the BAT on its highest intensity was found to be inaccurate, overpredicting glare disability in 76% of the population. Even with the reduction in glare luminance (medium setting), more than 37% of the patients with cataract had better acuity outdoors in the sunlight than predicted. Furthermore, the Miller-Nadler glare tester consistently underpredicted glare disability in 62% of patients. To the best of the authors’ knowledge, there have not been any studies powered sufficiently to evaluate the accuracy of the BAT.
Because the BAT cannot be measured at the phoropter and is variable (eg, intensity relies on battery charge and results rely on the angle at which it is held), a true head-to-head comparison with the investigational device would be difficult to consistently replicate from site to site.
The study authors acknowledge the small sample size as a limitation and suggest future study comparing the device to the BAT and/or Miller-Nadler glare tester. Further, although not a parameter for this study, we likewise would suggest future studies compare this device to other objective measures (eg, ocular light scatter) to determine whether there is a correlation between various instruments. Initial reports with the C-quant (Oculus Optikgeräte, Wetzlar, Germany), another device used to test the influence of stray light produced by a glare source, are encouraging; however, clinical experience with this device is limited and the relationship between straylight results from this test and driving performance has not been established.7
Our results suggest the investigational device will provide an improvement in patient care, will deliver a means to document visual impairment for medical reimbursement, and has the potential to reduce the risk of traffic accidents attributed to drivers with visual disability caused by glare.
- American Academy of Ophthalmology. Contrast sensitivity and glare testing in the evaluation of anterior segment disease. Ophthalmology. 1990;97:1233–1237. doi:10.1016/S0161-6420(90)32431-4 [CrossRef]
- American Academy of Ophthalmology Cataract and Anterior Segment Panel. Preferred Practice Guidelines: Cataract in the Adult Eye. San Francisco: American Academy of Ophthalmology, 2011. Available at: www.aao.org/ppp. Accessed January 31, 2013.
- Bichao IC, Yager D, Meng J. Disability glare: effects of temporal characteristics of the glare source and of the visual-field location of the test stimulus. J Opt Soc Am A Opt Image Sci Vis. 1995;12:2252–2258. doi:10.1364/JOSAA.12.002252 [CrossRef]
- Owsley C, Stalvey BT, Wells J, Slone ME, McGwin G Jr, . Visual risk factors for crash involvement in older drivers with cataract. Arch Ophthalmol. 2001;119:881–887. doi:10.1001/archopht.119.6.881 [CrossRef]
- Owsley C, McGwim G Jr, Sloane M, Wells J, Stalvey BT, Gauthreaux S. Impact of cataract surgery on motor vehicle crash involvement by older adults. JAMA. 2002;288:841–849. doi:10.1001/jama.288.7.841 [CrossRef]
- Shieber F. Recent Developments in Vision, Aging, and Driving: 1988–1994. Ann Arbor: University of Michigan Transportation Research Institute. 1994;Oct. Report No. UMTRI-94-26. Available at: http://deepblue.lib.umich.edu/handle/2027.42/1075. Accessed June 18, 2013.
- Aslam TM, Haider D, Murray IJ. Principles of disability glare measurement: an ophthalmological perspective. Acta Ophthalmol Scand. 2007;85:354–360. doi:10.1111/j.1600-0420.2006.00860.x [CrossRef]
- Rubin GS, Bandeen-Roche K, Huang GH, et al. The association of multiple visual impairments with self reported visual disability: SEE project. Invest Ophthalmol Vis Sci. 2001;42:64–72.
- Tan JC, Spalton DJ, Arden GB. Comparison of methods to assess visual impairment from glare and light scattering with posterior capsule opacification. J Cataract Refract Surg. 1998;24:1626–1631. doi:10.1016/S0886-3350(98)80354-4 [CrossRef]
- Wachler BSB, Durrie DS, Assil KK, Krueger RR. Improvement of visual function with glare testing after photorefractive keratectomy and radial keratotomy. Am J Ophthalmol. 1999;128:582–587. doi:10.1016/S0002-9394(99)00219-6 [CrossRef]
- Prager TC, Urso RG, Holladay JT. Glare testing in cataract patients: Instrument evaluation and identification of sources of methodologic error. J Cataract Refract Surg. 1989;15:149–157. doi:10.1016/S0886-3350(89)80003-3 [CrossRef]
- Adamsons IA, Vitale S, Stark WJ, Rubin GS. The association of post-operative subjective visual function with acuity, glare and contrast sensitivity in patients with early cataract. Arch Ophthalmol. 1996;114:529–536. doi:10.1001/archopht.1996.01100130521004 [CrossRef]
- Masket S. Reversal of glare disability after cataract surgery. J Cataract Refract Surg. 1989;15:165–168. doi:10.1016/S0886-3350(89)80005-7 [CrossRef]
- Masket S, Belani S. Preoperative evaluation of the patient with visually significant cataract. In: Steinert RF, ed. Cataract Surgery: Technique, Complications & Management. Philadelphia: W. B. Saunders Co.; 1995:21–32.
- Elliott DB, Hurst MA, Weatherill J. Comparing clinical tests of visual function in cataract with the patient’s perceived visual disability. Eye (Lond). 1990;4(Pt.5):712–717. doi:10.1038/eye.1990.100 [CrossRef]
- Elliott D, Bullimore MA, Patla AE, Whitaker D. Effect of a cataract simulation on clinical and real world vision. Br J Ophthalmol. 1996;80:799–804. doi:10.1136/bjo.80.9.799 [CrossRef]
- BAT Brightness Acuity Tester (Product No. BAT-1000) instruction manual. Available at: https://marco.zendesk.com/hc/enus/articles/201585446-Brightness-Acuity-Tester-BAT-1000-Instruction-Manual. Accessed January 31, 2013.
- Elliot DB, Bullimore MA. Assessing the reliability, discriminative ability, and validity of disability glare tests. Invest Ophthalmol Vis Sci. 1993;34:108–119.
- Schein OD, Vitale S, Cassard SD, Steinberg EP. Patient outcomes of refractive surgery: the refractive status and vision profile. J Cataract Refract Surg. 2001;27:665–673. doi:10.1016/S0886-3350(01)00844-6 [CrossRef]
- Pesudovs K, Burr JM, Harley C, Elliott DB. The development, assessment, and selection of questionnaires. Optom Vis Sci. 2007;84:663–674. doi:10.1097/OPX.0b013e318141fe75 [CrossRef]
- United States Department of Transportation and National Highway Traffic Safety Administration. Sensitivity Analysis of Headlamp Parameters Affecting Visibility and Glare. Section 4. Document DOT HS 811 055. December 2008.
- Kalich ME, Lewis LJ, Mora A, et al. for the US Army Aeromedical Research Laboratory Fort Rucker AL (USAARL). Report no 2009–18: Review of Efforts to Develop a Low Luminance-Level Disability Glare Tester. Sensory Research Division. Sept 2009. Available at: http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2&doc=GetTRDoc.pdf&AD=ADA511514. Accessed June 18, 2013.
- Aslam T. Patton N. Measurement validity in ophthalmology. Ophthalmic Epidemiol. 2005;12:155–161. doi:10.1080/09286580590932752 [CrossRef]
- Whitaker D, Elliott DB, Steen R. Confirmation of the validity of the psychophysical light scattering factor. Invest Ophthalmol Vis Sci. 1994;35:317–321.
- Fullerton M, Peli E. Development of a system to study the impact of headlight glare in a driving simulator. Proceedings of the 5th International Driving Symposium on Human Factors in Driver Assessment, Training and Vehicle Design. Iowa City: University of Iowa; 2009:412–418.
- Holladay JT. Proper method for calculating average visual acuity. J Refract Surg. 1997;13:388–391.