[Skip to Content]
[Skip to Content Landing]
Figure 1.  Virtual Reality Performance in the Supermarket Shopping Simulation
Virtual Reality Performance in the Supermarket Shopping Simulation

A, Screen captures from the supermarket shopping virtual reality simulation. B, Comparison of the duration to complete the simulation between patients with glaucoma and healthy individuals. C, Comparison of the number of incorrect selections between patients with glaucoma and healthy individuals. Error bars indicate SE.

Figure 2.  Virtual Reality Simulations of Stair Navigation
Virtual Reality Simulations of Stair Navigation

Top view of stair navigation path and screen captures (shown in inserts) from the daytime (A) and nighttime (B) virtual reality simulations. To minimize learning effects, the types and locations of the obstacles in the navigation tasks vary in each round of simulation (shown in inserts). Comparisons of the duration to complete the navigation (C) and the number of collisions (D) between the glaucoma and healthy groups in the daytime and nighttime simulations show that patients with glaucoma fared worse in nighttime than daytime navigation. Error bars indicate SE.

Figure 3.  Virtual Reality Simulations of City Navigation
Virtual Reality Simulations of City Navigation

Top view of city navigation path and screen captures (shown in inserts) from the daytime (A) and the nighttime (B) virtual reality simulations. To minimize learning effects, the types and locations of the obstacles in the navigation tasks vary in each round of simulation (shown in inserts). Comparisons of the duration to complete the navigation (C) and the number of collisions (D) between the glaucoma and healthy group in the daytime and nighttime simulations show that patients with glaucoma fared worse in nighttime than daytime navigation. Error bars indicate SE.

Figure 4.  Identifying Patients With Vision-Related Disability
Identifying Patients With Vision-Related Disability

The example shows the performance locations of all healthy individuals (blue dots) with reference to a multivariate, 3-dimensional space constructed from the variables measured from the supermarket shopping simulation and age. The 95% confidence region is demarcated by an ellipsoid. The performance locations of 2 patients with glaucoma (aged 60 years) (red dots) are outside the age-adjusted boundary (black curvilinear line), suggesting that they had vision-related disability. The severity of vision-related disability can be inferred from the Mahalanobis distance (dotted red lines), which is a unitless, scale-invariant measure taking the correlations of the different parameters in the multivariate space into consideration. The Mahalanobis distance is measured between the performance location of an individual and the centroid axis of the healthy group at a specific age.

Video 1. Virtual Reality Shopping Task to Characterize Glaucoma-Related Visual Disability

Comparison of virtual reality performance between a healthy individual (A) and a patient with glaucoma (B) in a supermarket shopping task. Participants were asked to identify 10 shopping items from a rack. In this example, the patient with glaucoma took 167.5 seconds to complete the shopping and incorrectly identified 1 item, while the healthy individual took 39.7 seconds to correctly identify 10 shopping items without misidentification. MD indicates mean deviation; VFI, visual field index.

Video 2. Virtual Reality Nighttime Stair Navigation Task to Characterize Glaucoma-Related Visual Disability

Comparison of virtual reality performance between a healthy individual (A) and a patient with glaucoma (B) in the nighttime stair navigation task. Participants were asked to walk up and then down 2 flights of stairs. In this example, the healthy individual took 179.5 seconds whereas the patient with glaucoma took 465.1 seconds to complete the navigation. The numbers of collisions during the navigation were 0 and 3, respectively. MD indicates mean deviation; VFI, visual field index.

Video 3. Daytime vs Nighttime Stair Navigation Task to Characterize Glaucoma-Related Visual Disability

Comparison of virtual reality performance of a patient with glaucoma in the daytime (A) and nighttime (B) stair navigation task. In this example, the patient with glaucoma took almost double the time required in daytime navigation (93.4 seconds) to complete the nighttime navigation (181.6 seconds). The number of collisions was 1 in daytime navigation and 6 in nighttime navigation. MD indicates mean deviation; VFI, visual field index.

Video 4. Virtual Reality City Navigation Task to Characterize Glaucoma-Related Visual Disability

Comparison of virtual reality performance between a healthy individual (A) and a patient with glaucoma (B) in the nighttime city navigation task. The participants navigated a virtual distance of approximately 90 m in a city area modeled on Hong Kong. In this example, the healthy individual took 78.7 seconds whereas the patient with glaucoma took 167.8 seconds to complete the navigation. The numbers of collisions during the navigation were 0 and 1, respectively. MD indicates mean deviation; VFI, visual field index.

Video 5. Daytime vs Nighttime City Navigation Task to Characterize Glaucoma-Related Visual Disability

Comparison of virtual reality performance of a patient with glaucoma in the daytime (A) and nighttime (B) city navigation task. In this example, the patient with glaucoma took almost double the time required in the daytime navigation (85.1 seconds) to complete the nighttime navigation (167.8 seconds). The number of collisions was 2 in the daytime navigation and 1 in the nighttime navigation. MD indicates mean deviation; VFI, visual field index.

Supplement.

eMethods 1. Measurements of Visual Acuity, Contrast Sensitivity, Visual Field Sensitivity, and Vision-Related Quality of Life

eMethods 2. Design and Specifications of the VR Simulations

eMethods 3. Real-World Shopping Simulation

eMethods 4. Calculation of VR Disability Scores

eResults. Test-Retest Variabilities of VR Simulation Measurements

eDiscussion. Study Limitations

eReferences.

eFigure 1. Study Flow Chart

eFigure 2. Relationship Between Duration to Complete Supermarket Shopping Simulation and Visual Function Measurements and Age in Glaucoma Patients

eFigure 3. Relationship Between Duration to Complete Stair Navigation in Daytime and Visual Function Measurements and Age in Glaucoma Patients

eFigure 4. Relationship Between Duration to Complete Stair Navigation in Nighttime and Visual Function Measurements and Age in Glaucoma Patients

eFigure 5. Relationship Between Duration to Complete City Navigation in Daytime and Visual Function Measurements and Age in Glaucoma Patients

eFigure 6. Relationship Between Duration to Complete City Navigation in Nighttime and Visual Function Measurements and Age in Glaucoma Patients

eFigure 7. Detection of Glaucoma Patients With Vision-Related Disability

eFigure 8. Virtual Reality Disability Score and Patient-Reported Quality of Life

eFigure 9. Real World Shopping Simulation

eTable 1. Comparisons of Age, Visual Acuity (Better Eye), Binocular Contrast Sensitivity, Binocular Visual Field Sensitivity, and National Eye Institute Visual Function Questionnaire-25 (NEI VFQ-25) Rasch Score Between the Healthy Individuals and Glaucoma Patients

eTable 2. A Multivariable Linear Regression Model Investigating Factors Associated With the Duration Required to Complete the Shopping Simulation

eTable 3. Multivariable Linear Regression Models Investigating Factors Associated With Stair Navigation Duration in the Nighttime and Daytime Simulations

eTable 4. Multivariable Ordered Logistic Regression Models Investigating Factors Associated With the Risk of Collision During the Nighttime and Daytime Stair Navigations

eTable 5. Multivariable Linear Regression Models Investigating Factors Associated With City Navigation Duration in the Nighttime and Daytime Simulations

eTable 6. Multivariable Ordered Logistic Regression Models Investigating Factors Associated With the Risk of Collision During the Nighttime and Daytime City Navigations

eTable 7. Univariable Logistic Regression Model Investigating Factors Associated With Vision-Related Disability

eTable 8. Multivariable Logistic Regression Model Investigating Factors Associated With Vision-Related Disability

eTable 9. Comparisons of Age, Visual Function Measures, and National Eye Institute Visual Function Questionnaire-25 (NEI VFQ-25) Rasch score Between 20 Healthy Individuals and 9 Glaucoma Patients Who Had Completed Five Sets of Virtual Reality Simulations in Two Separate Visits Within Four Months for the Evaluation of Test-Retest Variability of the Virtual Reality Measurements

eTable 10. Intraclass Correlation Coefficients (ICC) of the Duration to Complete the Simulation for the Individual Virtual Reality Simulations and the Overall Visual Performance Score

1.
Flaxman  SR, Bourne  RRA, Resnikoff  S,  et al; Vision Loss Expert Group of the Global Burden of Disease Study.  Global causes of blindness and distance vision impairment 1990-2020: a systematic review and meta-analysis.  Lancet Glob Health. 2017;5(12):e1221-e1234. doi:10.1016/S2214-109X(17)30393-5PubMedGoogle ScholarCrossref
2.
Tham  YC, Li  X, Wong  TY, Quigley  HA, Aung  T, Cheng  CY.  Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis.  Ophthalmology. 2014;121(11):2081-2090. doi:10.1016/j.ophtha.2014.05.013PubMedGoogle ScholarCrossref
3.
Weinreb  RN, Leung  CK, Crowston  JG,  et al.  Primary open-angle glaucoma.  Nat Rev Dis Primers. 2016;2:16067. doi:10.1038/nrdp.2016.67PubMedGoogle ScholarCrossref
4.
Haymes  SA, Leblanc  RP, Nicolela  MT, Chiasson  LA, Chauhan  BC.  Risk of falls and motor vehicle collisions in glaucoma.  Invest Ophthalmol Vis Sci. 2007;48(3):1149-1155. doi:10.1167/iovs.06-0886PubMedGoogle ScholarCrossref
5.
McGwin  G  Jr, Huisingh  C, Jain  SG, Girkin  CA, Owsley  C.  Binocular visual field impairment in glaucoma and at-fault motor vehicle collisions.  J Glaucoma. 2015;24(2):138-143. doi:10.1097/IJG.0b013e3182a0761cPubMedGoogle ScholarCrossref
6.
Baig  S, Diniz-Filho  A, Wu  Z,  et al.  Association of fast visual field loss with risk of falling in patients with glaucoma.  JAMA Ophthalmol. 2016;134(8):880-886. doi:10.1001/jamaophthalmol.2016.1659PubMedGoogle ScholarCrossref
7.
Montana  CL, Bhorade  AM.  Glaucoma and quality of life: fall and driving risk.  Curr Opin Ophthalmol. 2018;29(2):135-140. doi:10.1097/ICU.0000000000000455PubMedGoogle ScholarCrossref
8.
Centers for Disease Control and Prevention. Deaths from unintentional injury among adults aged 65 and over: United States, 2000-2013. Accessed July 7, 2018. https://www.cdc.gov/nchs/data/databriefs/db199.htm
9.
Nelson  P, Aspinall  P, O’Brien  C.  Patients’ perception of visual impairment in glaucoma: a pilot study.  Br J Ophthalmol. 1999;83(5):546-552. doi:10.1136/bjo.83.5.546PubMedGoogle ScholarCrossref
10.
Viswanathan  AC, McNaught  AI, Poinoosawmy  D,  et al.  Severity and stability of glaucoma: patient perception compared with objective measurement.  Arch Ophthalmol. 1999;117(4):450-454. doi:10.1001/archopht.117.4.450PubMedGoogle ScholarCrossref
11.
Chung  DC, McCague  S, Yu  ZF,  et al.  Novel mobility test to assess functional vision in patients with inherited retinal dystrophies.  Clin Exp Ophthalmol. 2018;46(3):247-259. doi:10.1111/ceo.13022PubMedGoogle ScholarCrossref
12.
Medeiros  FA, Weinreb  RNR, R Boer  E, Rosen  PN.  Driving simulation as a performance-based test of visual impairment in glaucoma.  J Glaucoma. 2012;21(4):221-227. doi:10.1097/IJG.0b013e3182071832PubMedGoogle ScholarCrossref
13.
Glen  FC, Smith  ND, Crabb  DP.  Impact of superior and inferior visual field loss on hazard detection in a computer-based driving test.  Br J Ophthalmol. 2015;99(5):613-617. doi:10.1136/bjophthalmol-2014-305932PubMedGoogle ScholarCrossref
14.
Russell  S, Bennett  J, Wellman  JA,  et al.  Efficacy and safety of voretigene neparvovec (AAV2-hRPE65v2) in patients with RPE65-mediated inherited retinal dystrophy: a randomised, controlled, open-label, phase 3 trial.  Lancet. 2017;390(10097):849-860. doi:10.1016/S0140-6736(17)31868-8PubMedGoogle ScholarCrossref
15.
Bainbridge  JW, Smith  AJ, Barker  SS,  et al.  Effect of gene therapy on visual function in Leber’s congenital amaurosis.  N Engl J Med. 2008;358(21):2231-2239. doi:10.1056/NEJMoa0802268PubMedGoogle ScholarCrossref
16.
Maguire  AM, Simonelli  F, Pierce  EA,  et al.  Safety and efficacy of gene transfer for Leber’s congenital amaurosis.  N Engl J Med. 2008;358(21):2240-2248. doi:10.1056/NEJMoa0802315PubMedGoogle ScholarCrossref
17.
Wei  H, Sawchyn  AK, Myers  JS,  et al.  A clinical method to assess the effect of visual loss on the ability to perform activities of daily living.  Br J Ophthalmol. 2012;96(5):735-741. doi:10.1136/bjophthalmol-2011-300093PubMedGoogle ScholarCrossref
18.
Ekici  F, Loh  R, Waisbourd  M,  et al.  Relationships between measures of the ability to perform vision-related activities, vision-related quality of life, and clinical findings in patients with glaucoma.  JAMA Ophthalmol. 2015;133(12):1377-1385. doi:10.1001/jamaophthalmol.2015.3426PubMedGoogle ScholarCrossref
19.
De Maesschalck  R, Jouan-Rimbaud  D, Massart  DL.  The Mahalanobis distance.  Chemom Intell Lab Syst. 2000;50(1):1-18. doi:10.1016/S0169-7439(99)00047-7Google ScholarCrossref
20.
Brereton  RG.  Hotelling’s T squared distribution, its relationship to the F distribution and its use in multivariate space.  J Chemometrics. 2016;30(1):18-21. doi:10.1002/cem.2763Google ScholarCrossref
21.
Mangione  CM, Lee  PP, Gutierrez  PR, Spritzer  K, Berry  S, Hays  RD; National Eye Institute Visual Function Questionnaire Field Test Investigators.  Development of the 25-item National Eye Institute Visual Function Questionnaire.  Arch Ophthalmol. 2001;119(7):1050-1058. doi:10.1001/archopht.119.7.1050PubMedGoogle ScholarCrossref
22.
Mangione  CM, Lee  PP, Pitts  J, Gutierrez  P, Berry  S, Hays  RD; NEI-VFQ Field Test Investigators.  Psychometric properties of the National Eye Institute Visual Function Questionnaire (NEI-VFQ).  Arch Ophthalmol. 1998;116(11):1496-1504. doi:10.1001/archopht.116.11.1496PubMedGoogle ScholarCrossref
23.
Sherwood  MB, Garcia-Siekavizza  A, Meltzer  MI, Hebert  A, Burns  AF, McGorray  S.  Glaucoma’s impact on quality of life and its relation to clinical indicators: a pilot study.  Ophthalmology. 1998;105(3):561-566. doi:10.1016/S0161-6420(98)93043-3PubMedGoogle ScholarCrossref
24.
Janz  NK, Wren  PA, Lichter  PR, Musch  DC, Gillespie  BW, Guire  KE.  Quality of life in newly diagnosed glaucoma patients: the Collaborative Initial Glaucoma Treatment Study.  Ophthalmology. 2001;108(5):887-897. doi:10.1016/S0161-6420(00)00624-2PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 518
    Citations 0
    Original Investigation
    March 19, 2020

    Use of Virtual Reality Simulation to Identify Vision-Related Disability in Patients With Glaucoma

    Author Affiliations
    • 1Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Hong Kong, People’s Republic of China
    • 2Department of Ophthalmology, University of California, San Diego, La Jolla
    • 3Singapore Eye Research Institute, Singapore National Eye Centre, Singapore
    • 4Division of Biomedical Engineering, The Hong Kong University of Science and Technology, Hong Kong, People’s Republic of China
    JAMA Ophthalmol. Published online March 19, 2020. doi:10.1001/jamaophthalmol.2020.0392
    Key Points

    Question  Can real-world visual performance be estimated with virtual reality simulations to assess vision-related disability in patients with glaucoma?

    Findings  In this cross-sectional study including 98 individuals with glaucoma, vision-related disability was associated with task and lighting condition in patients with glaucoma, with 8.0% to 30.7% having vision-related disability in supermarket shopping, stair navigation, or city navigation. A higher proportion of patients had vision-related disability in nighttime (30.0%-30.7%) than daytime (8.0%-19.8%) navigations.

    Meaning  These results appear to support the hypothesis that virtual reality simulation augments the evaluation of visual disability in clinical care by providing clinicians a new perspective to understand how visual impairment imparts vision-related disability in patients with glaucoma.

    Abstract

    Importance  Clinical assessment of vision-related disability is hampered by the lack of instruments to assess visual performance in real-world situations. Interactive virtual reality (VR) environments displayed in a binocular stereoscopic VR headset have been designed, presumably simulating day-to-day activities to evaluate vision-related disability.

    Objective  To investigate the application of VR to identify vision-related disability in patients with glaucoma.

    Design, Setting, and Participants  In a cross-sectional study, 98 patients with glaucoma and 50 healthy individuals were consecutively recruited from a university eye clinic; all participants were Chinese. The study was conducted between August 30, 2016, and July 31, 2017; data analysis was performed from December 1, 2017, to October 30, 2018.

    Exposures  Measurements of visual acuity, contrast sensitivity, visual field (VF), National Eye Institute 25-item Visual Function Questionnaire Rasch score, and VR disability scores determined from 5 VR simulations: supermarket shopping, stair and city navigations in daytime, and stair and city navigations in nighttime. Duration required to complete the simulation, number of items incorrectly identified, and number of collisions were measured to compute task-specific and overall VR disability scores. Vision-related disability was identified when the VR disability score was outside the normal age-adjusted 95% confidence region.

    Main Outcomes and Measures  Virtual reality disability score.

    Results  In the 98 patients with glaucoma, mean (SD) age was 49.8 (11.6) years and 60 were men (61.2%); in the 50 healthy individuals, mean (SD) age was 48.3 (14.8) years and 16 were men (32.0%). The patients with glaucoma had different degrees of VF loss (122 eyes [62.2%] had moderate or advanced VF defects). The time required to complete the activities by patients with glaucoma vs healthy individuals was longer by 15.2 seconds (95% CI, 5.5-24.9 seconds) or 34.1% (95% CI, 12.4%-55.7%) for the shopping simulation, 72.8 seconds (95% CI, 23.0-122.6 seconds) or 33.8% (95% CI, 10.7%-56.9%) for the nighttime stair navigation, and 38.1 seconds (95% CI, 10.9-65.2 seconds) or 30.8% (95% CI, 8.8%-52.8%) for the nighttime city navigation. The mean (SD) duration was not significantly different between the glaucoma and healthy groups in daytime stair (203.7 [93.7] vs 192.9 [89.1] seconds, P = .52) and city (118.7 [41.5] vs 117.0 [52.3] seconds, P = .85) navigation. For each decibel decrease in binocular VF sensitivity, the risk of collision increased by 15% in nighttime stair (hazard ratio [HR], 1.15; 95% CI, 1.08-1.22) and city (HR, 1.15; 95% CI, 1.08-1.23) navigations. Fifty-eight patients (59.1%) with glaucoma had vision-related disability in at least 1 simulated daily task; a higher proportion of patients had vision-related disability in nighttime city (27 of 88 [30.7%]) and stair (27 of 90 [30.0%]) navigation than in daytime city (7 of 88 [8.0%]) and stair (19 of 96 [19.8%]) navigation. The overall VR disability score was associated with the National Eye Institute 25-item Visual Function Questionnaire Rasch score (R2 = 0.207).

    Conclusions and Relevance  These findings suggest that vision-related disability is associated with lighting condition and task in patients with glaucoma. Virtual reality may allow eye care professionals to understand the patients’ perspectives on how visual impairment imparts disability in daily living and provide a new paradigm to augment the assessment of vision-related disability.

    Introduction

    Glaucoma, a chronic, progressive optic neuropathy and the leading cause of irreversible global blindness, is estimated to affect 76.0 million of the global population by 2020.1,2 Although functional assessment of the optic nerve in glaucoma currently relies on measurement of visual field (VF) sensitivity,3 VF testing does not adequately inform disability experienced by patients in real-world situations. Many patients with glaucoma are at risk of falls and motor vehicle crashes,4-7 which are the 2 leading causes of unintentional injury-related deaths in adults aged 65 years or older.8 Difficulties with shopping, staircases, and crossing the road are frequent vision-related disabilities reported by patients with glaucoma.9,10 However, to our knowledge, practical measures to assess these individuals’ real-world activity are not available and used in clinical practice. Whereas a number of custom-designed platforms, such as tests of mobility11 and driving simulation,12,13 have been developed to evaluate vision-related disability, the tests require special equipment that can be difficult to set up and costly to implement in clinical practice. For example, the multiluminance mobility test has been applied in clinical trials14-16 to investigate the treatment effect of ocular gene therapy on visual performance in patients with Leber congenital amaurosis; however, the test requires special setup and design of an obstacle course in a designated research facility, which can be challenging to standardize and resource intensive to replicate. Evaluating the patients’ ability to find objects in a room, read street signs, recognize facial expression, and detect computerized motion in the Assessment of Ability Related to Vision test is another performance-based instrument and is easier to administer.17,18 Yet, examining vision-related disability that connects to the complexity of real-world experiences demands a more adaptable design because of the varieties of real-world situations in daytime and nighttime settings. Devising a versatile and affordable clinical instrument for measurement of vision-related disability is an unmet need.

    This study investigated the application of virtual reality (VR) technology for evaluation and measurement of vision-related disability via simulating activities of daily living relevant to patients’ real-world experience. We designed 5 interactive VR environments displayed in a commercially available VR headset, simulating common activities including (1) supermarket shopping, (2) stair navigation in daytime, (3) stair navigation in nighttime, (4) city navigation in daytime, and (5) city navigation in nighttime to evaluate vision-related disability in patients with glaucoma. We compared the performance between healthy individuals and patients with glaucoma in simulated daytime and nighttime environments and investigated the association of measurements obtained from the VR simulations with measurements gathered from standard visions tests and patient-reported quality-of-life questionnaire (National Eye Institute 25-item Visual Function Questionnaire [NEI VFQ-25] Rasch score). In addition, we derived disability scores from the individual VR simulations with a goal to differentiate patients with visual disability from those without visual disability.

    Methods

    A total of 98 glaucoma patients and 50 healthy individuals were consecutively recruited from the University Eye Center, the Chinese University of Hong Kong, between August 30, 2016, and July 31, 2017. The inclusion/exclusion criteria, clinical examination, and investigations are described in eMethods 1 in the Supplement. The study was approved by the Kowloon Central Research Ethics Committee with written informed consent obtained. Participants did not receive financial compensation. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.

    The VR environments and objects were designed and programmed in Unity software, version 5.3.3p1 (Unity Technologies). The codes developed for the VR simulations can be adapted to all commercially available VR headsets. In this study, we selected the HTC Vive (HTC Corp), which had a screen resolution of 2160 × 1200 pixels enabled by the PenTile OLED technology with a refresh rate of 90 Hz. With the orientation and positional tracking in the HTC Vive programmed using the SteamVR SDK plugin (Valve Corp), study participants immersed into and interacted with the virtual 3-dimensional environment in a binocular stereoscopic view. Five VR simulations were developed and each participant was asked to complete the 5 simulations in the following sequence: (1) identifying 10 shopping items from a supermarket rack (Figure 1A), (2) walking up 2 flights of stairs and then walking down 2 flights of stairs in simulated daytime (Figure 2A) and (3) nighttime (Figure 2B) environments, and (4) navigating a city area over a virtual distance of approximately 90 m in simulated daytime (Figure 3A) and (5) nighttime (Figure 3B) environments. The VR simulations were administered in a quiet room with a technician monitoring the safety of the participants during the time they were wearing the VR headset. All participants received a training session to test-run each of the 5 simulations before the testing session. The training and testing sequence is shown in eFigure 1 in the Supplement. Participants’ responses in the simulations were recorded in real time in a computer and the following parameters were measured: duration to complete the simulation, number of shopping items incorrectly identified in the shopping simulation, and number of objects collided with in the stair and city navigations. Details of the designs and specifications of the VR simulations are described in eMethods 2 in the Supplement. The test-retest variabilities of the VR simulations were relatively low (eResults in the Supplement). To compare VR simulation with real-world performance, we also designed a real-world shopping simulation and recruited 50 patients with glaucoma to undertake both VR shopping simulation and real-world shopping simulation. Further details on the specifications and administration of the VR and real-world testing are described in eMethods 3 in the Supplement.

    Statistical Analysis

    Comparisons of VR measurements, visual function measurements, and biometric variables between the glaucoma and healthy groups were performed with an independent t test. Within-group comparisons of VR measurements between daytime and nighttime simulations were performed with a paired t test. The associations between the duration needed to complete the simulation and visual function measurements were evaluated with univariable and multivariable linear regression analysis. Factors associated with vision-related disability were determined from logistic regression analysis. Significant variables (P < .05) identified in the univariable analysis were included in the multivariable analysis. The associations between the risk of collision during navigation and binocular VF sensitivity were determined by ordered logistic regression analysis after adjustment of covariates.

    The coefficients of correlation between visual performance score measured by the Assessment of Ability Related to Vision test and visual function measurements, including visual acuity, contrast sensitivity, and visual field sensitivity measurements, in patients with glaucoma have been reported to vary between 0.326 and 0.366.18 The number of patients with glaucoma required to detect a correlation of at least 0.3 between the duration required to complete a VR simulation and a visual function measure with 85% power using a 1-sided, 5%-level test was 79. The number of healthy individuals required to provide a power of 80% using a 1-sided, 5%-level test to detect a mean (SD) difference in duration of 15 (30) seconds (estimates were derived from the supermarket simulation) to complete a VR simulation between the glaucoma group and the healthy group was 37. We hypothesized that patients with glaucoma would take longer to complete a task compared with healthy individuals. We included 98 patients with glaucoma and 50 healthy individuals considering that 20% to 25% of the participants might not be able to complete all 5 simulations because of VR-related motion sickness. Statistical analysis was performed using Stata software, version 14.0 (StataCorp), and R, version 3.2.5 (R Foundation for Statistical Computing), from December 1, 2017, to October 30, 2018.

    Results

    A total of 148 participants, including 50 healthy individuals and 98 patients with glaucoma (94 with primary open-angle glaucoma and 4 with primary angle-closure glaucoma), were recruited for VR simulations and measurements of visual acuity, VF sensitivity, contrast sensitivity, and vision-related quality of life. In the glaucoma group, mean (SD) age was 49.8 (11.6) years, and 60 patients (61.2%) were men. Ninety-four patients (95.9%) with glaucoma had VF defects in at least 1 eye (ie, 4 [4.1%] had preperimetric glaucoma) and 122 of 196 eyes (62.2%) had moderate to advanced VF defects with VF mean deviation less than −6 decibel (dB). Mean (SD) age (49.8 [11.6] vs 48.3 [14.8] years, P = .49) and visual acuity (better eye) (0.03 [0.10] vs 0.01 [0.07] logMAR, P = .16; Snellen equivalent, 20/22 [20/25] vs 20/21 [20/24]) were comparable between the glaucoma and healthy participants. However, findings in the glaucoma vs healthy group were lower for binocular VF sensitivity (24.35 [6.87] vs 31.63 [1.12] dB, P < .001), binocular contrast sensitivity (1.67 [0.25] vs 1.85 [0.12] log units, P < .001), and NEI VFQ-25 Rasch score (44.86 [16.78] vs 61.55 [16.68], P < .001) (eTable 1 in the Supplement).

    Patients with glaucoma required a mean of 34.1% (95% CI, 12.4%-55.7%) more time or 15.2 seconds (95% CI, 5.5-24.9 seconds) longer to identify 10 items in the shopping simulation compared with healthy individuals (Figure 1B) (Video 1). The mean (SD) number of shopping items incorrectly identified was 1.0 (2.0) in the glaucoma group and 0.7 (1.2) in the healthy group (Figure 1C). The duration required to complete the simulation was associated with binocular VF sensitivity (R2 = 0.281) (eFigure 2A in the Supplement). In the multivariable analysis, the duration required to complete the simulation increased by 1.3 seconds (95% CI, 0.4-2.1 seconds) for each decibel decrease in binocular VF sensitivity, 7.5 seconds (95% CI, 3.4-11.5 seconds) for each line (0.15 log unit) decrease in binocular contrast sensitivity, and 1.1 seconds (95% CI, 0.6-1.6 seconds) for each year increase in age (adjusted R2 of the model = 0.451) (eTable 2 in the Supplement). Univariable analyses are shown in eFigure 2 in the Supplement.

    In the nighttime stair navigation, patients with glaucoma required a mean of 33.8% (95% CI, 10.7%-56.9%) more time or 72.8 seconds (95% CI, 23.0-122.6 seconds) longer, to complete the navigation than healthy individuals (Video 2); however, in the daytime simulation, the time required to complete the navigation did not differ significantly between the groups (203.7 [93.7] vs 192.9 [89.1] seconds, P = .52) (Figure 2C). The difference in the duration between daytime and nighttime simulations was greater in patients with glaucoma (92.8 seconds, 95% CI, 71.9-113.7 seconds) than healthy individuals (29.7 seconds, 95% CI, 18.7-40.7 seconds) (Video 3). In the multivariable analysis, variables associated with navigation duration were age in the daytime simulation and binocular VF sensitivity and age in the nighttime simulation (eTable 3 in the Supplement) (univariable analyses are shown in eFigure 3 and eFigure 4 in the Supplement). For each decibel decrease in binocular VF sensitivity, the duration to complete the nighttime stair navigation increased by 8.4 seconds (95% CI, 3.4-13.5 seconds) after adjustment of covariates (adjusted R2 = 0.319). Patients with glaucoma experienced more collisions than healthy individuals in the nighttime (4.5 [6.7] vs 1.8 [2.1], P = .01) and daytime (1.7 [1.9] vs 0.6 [1.2], P < .001) simulations, and the difference in the numbers of collisions between daytime and nighttime simulations was greater in patients with glaucoma (2.9; 95% CI, 1.6-4.3) compared with healthy individuals (1.2; 95% CI, 0.5-1.8) (Figure 2D). For each decibel decrease in binocular VF sensitivity, the risk of collision increased by 1.06-fold (95% CI, 1.00-1.12) in the daytime stair navigation and 1.15-fold (95% CI, 1.08-1.22) in the nighttime stair navigation (eTable 4 in the Supplement), after controlling for covariates.

    In the city navigation, patients with glaucoma required 30.8% more time (95% CI, 8.8%-52.8%), or 38.1 seconds (95% CI, 10.9-65.2 seconds) longer, than healthy individuals to complete the nighttime simulation (Video 4), whereas the duration to complete the daytime simulation was similar between the groups (118.7 [41.5] vs 117.0 [52.3] seconds, P = .85) (Figure 3C). Patients with glaucoma required 36.1% (95% CI, 27.0%-45.1%) more time to complete the nighttime than daytime simulation (Video 5); in healthy participants, the difference was not significant. In the multivariable analysis, age was associated with navigation duration in the daytime simulation, whereas age and binocular VF sensitivity were associated with navigation duration in the nighttime simulation (eTable 5 in the Supplement) (univariable analyses are shown in eFigure 5 and eFigure 6 in the Supplement). For each decibel loss in binocular VF sensitivity, the duration required to complete the nighttime navigation increased by 3.5 seconds (95% CI, 1.2-5.9 seconds) after adjustment of covariates (adjusted R2 = 0.409). Patients with glaucoma experienced more collisions than healthy individuals in the nighttime navigation (1.60 [2.3] and 0.67 [0.86], respectively, P = .02) (Figure 3D) and they experienced more collisions in the nighttime than the daytime navigation (difference, 0.83; 95% CI, 0.40-1.26). For each decibel decrease in binocular VF sensitivity, the risk of collision increased by 1.07-fold (95% CI, 1.00-1.15) in the daytime city navigation, and 1.15-fold (95% CI, 1.08-1.23) in the nighttime city navigation, after controlling for covariates (eTable 6 in the Supplement).

    Because the performance indicators measured from the VR simulation are interconnected (eg, patients who experience more collisions may take a longer time to complete a navigation task), we visualized and evaluated informational complementarity in a 3-dimensional multivariate space and derived a task-specific VR disability score for each participant with reference to the Mahalanobis distance scaled between 0 and 100 (eMethods 4 in the Supplement).19 The Mahalanobis distance, which is a unitless and scale-invariant measure taking the correlations of the parameters of interest into consideration, was measured between the performance location of a participant and the age-specific location on the centroid axis of the healthy controls in the multivariate space formed by the VR performance parameters and age (Figure 4). We mapped out the 95% confidence region of the VR disability scores of the healthy individuals using the Hotelling T2 statistic20 and identified the age-adjusted boundaries of the multivariate space to differentiate patients with (ie, outside the 95% confidence region of the healthy controls) or without (within the 95% confidence region of the healthy controls) vision-related disability for a specific VR simulation task (the calculations of the VR disability scores and the 95% confidence region are described in the eMethods 4 in the Supplement). Twenty-seven patients with glaucoma (30.7%) showed vision-related disability in nighttime city navigation (eFigure 7E in the Supplement) and 27 patients (30.0%) showed vision-related disability in nighttime stair navigation (eFigure 7C in the Supplement). Twenty-four patients (24.5%) had vision-related disability with shopping (eFigure 7A in the Supplement), 19 patients (19.8%) showed disability with daytime stair navigation (eFigure 7B in the Supplement), and 7 patients (8.0%) showed disability with daytime city navigation (eFigure 7D in the Supplement). Fifty-eight patients (59.2%) with glaucoma had vision-related disability in at least 1 task.

    The NEI VFQ-25 Rasch score represents a summary measure quantifying vision-related difficulties from a variety of daily tasks, which includes finding something on a crowded shelf, going down stairs in dim light or at night, and noticing objects off to the side while walking.21,22 To determine the association between the NEI VFQ-25 Rasch score and the overall performance in the VR simulations, we derived an overall disability score. Different from the task-specific disability score, the multivariate space for the overall disability score was derived from 11 dimensions, including the duration to complete the simulation (ie, 5 measurements from the 5 VR simulations), the number of incorrect selections in shopping simulation and the number of collisions in the 4 navigation simulations, and age. The Mahalanobis distance (scaled between 0 and 100) was then measured between the performance location of a participant and the age-specific location on the centroid axis of the healthy group in that multivariate space. Defining vision-related disability as having an overall VR disability score outside the age-adjusted 95% confidence region of the healthy controls, 53.5% of patients with glaucoma showed vision-related disability (eFigure 8A in the Supplement). The NEI VFQ-25 Rasch score was significantly lower in patients with vision-related disability compared with those without vision-related disability (eFigure 8B in the Supplement). Moreover, the overall VR disability score was significantly associated with the NEI VFQ-25 Rasch score (R2 = 0.207, P < .001) (eFigure 8C in the Supplement).

    A worse binocular VF sensitivity, worse VA (in the better eye), worse contrast sensitivity, and lower NEI VFQ-25 Rasch score were all associated with a higher odds ratio of vision-related disability (ie, an overall VR disability score outside the age-adjusted 95% confidence region of the healthy controls) in the univariable models (eTable 7 in the Supplement). Binocular VF sensitivity was the only variable associated with visual disability (hazard ratio, 1.19; 95% CI, 1.05-1.35 for each decibel decrease) in the multivariable model (eTable 8 in the Supplement).

    We compared VR shopping simulation with real-world shopping simulation in 50 consecutively recruited patients with glaucoma (eFigure 9 in the Supplement). Their mean age was 58.6 (12.7) years; 42 eyes (42.0%) had moderate to advanced VF defects (ie, VF mean deviation<−6 dB). Patients with glaucoma performed worse in VR simulation vs real-world simulation; they needed more time to select 10 shopping items (92.0; 95% CI, 74.6-109.4 vs 61.0; 95% CI, 51.9 -70.1 seconds) and they showed a greater number of incorrect selections (1.2; 95% CI, 0.8-1.7 vs 0.2; 95% CI, 0.1-0.4). Nevertheless, there was an association in the duration to complete the selection of 10 shopping items between VR simulation and real-world simulation (R2 = 0.467; P < .001). The association between duration to complete the selection and binocular VF sensitivity was stronger in the VR simulation (R2 = 0.407; P < .001) than in the real-world simulation (R2 = 0.253; P < .001). We did not perform real-world stair and city navigation simulations because of potential risks of fall-related injuries.

    Discussion

    Virtual reality simulation of activities of daily living has provided an intuitive approach to visualize and measure disability experienced by patients with visual impairment in the real world; the process integrates different aspects of visual function to assess task performance and provides the opportunity for patients and health care professionals to understand how visual impairments affect daily activities. The fact that the duration required to complete the simulations and the number of collisions in the navigation tasks were significantly associated with binocular VF sensitivity measured by the standard automated perimetry implies that the VR simulations are relevant to evaluate vision-related disability in patients with glaucoma with different degrees of VF loss. By integrating various VR performance parameters to calculate the VR disability scores, we suggest that the process is feasible to identify patients with glaucoma who have vision-related disability. Empowering clinicians and eye care professionals to better understand a patient’s perspective on how visual impairment imparts vision-related disability in daily living, as well as translating standard VF test results to performance measures familiar to patients (eg, risk of collision during navigation), our findings indicate that VR simulation has the potential to improve the evaluation of vision-related disability in clinical care for patients with glaucoma.

    Understanding the effects and consequences of VF abnormalities is challenging because vision-related disability is difficult to measure. Although the NEI VFQ-25 is a current standard to gauge vision-related disability and quality of life,21,22 subjective interpretation of no difficulty, little difficulty, moderate difficulty, and extreme difficulty in the questionnaire can limit the standardization and generalizability of the assessment. Not only does the understanding and interpretation of the levels of difficulty for a particular task vary among individuals, glaucoma often compromises visual performance before subjective recognition. One potential advantage of the VR simulations is the ability to recapitulate the complexities and varieties of daily tasks in a standardized environment for review and analysis. We provide evidence suggesting that vision-related disability is associated with task. For example, whereas patients with glaucoma needed a mean of 34.1% more time to complete the shopping simulation than healthy individuals, the performance appears to be comparable between the groups in the daytime city navigation, both in terms of the duration to complete the simulation (Figure 3C) and the number of collisions experienced during the navigation (Figure 3D). Recognizing and locating objects of interest are likely to be more visually demanding than daytime city navigation because shopping items are generally smaller than objects encountered in city navigation (eg, pedestrians and vehicles). The task-dependent disparity in the evaluation of vision-related disability is also evident from the different proportions of patients identified to have vision-related disability (from 8.0% to 30.7%) (eFigure 7 in the Supplement). Another observation is that vision-related disability depended on lighting conditions. Patients with glaucoma appeared to fare worse at nighttime than during the daytime. The differences in the numbers of collisions between the daytime and nighttime navigations were greater in patients with glaucoma than healthy individuals (Figure 2D and Figure 3D). This finding is consistent with the observation that 30.0% to 30.7% of patients with glaucoma showed vision-related disability in nighttime navigations, whereas 8.0% to 19.8% showed vision-related disability in daytime navigations (eFigure 7 in the Supplement). Self-reported poor night vision is common among patients with glaucoma,9,23,24 but clinicians may underestimate the degree of disability because many of these individuals have normal visual acuity even at the late stages. Virtual reality simulations can bridge the disconnection between clinicians and patients, enabling both parties to visualize the disparities of vision-related disability in daytime and nighttime conditions. Our study appears to support the hypothesis that patients with glaucoma are more visually disabled at night, although such differences have not been quantified and validated by available clinical tools. Understanding such differences would allow eye care professionals to devise appropriate treatment, visual aids, and counseling to improve the quality of vision and quality of life in patients with glaucoma.

    Virtual reality simulation may permit the translation of clinical test results into performance measures familiar and meaningful to a patient. Visual field sensitivity is the current standard to stage and monitor the status of patients with glaucoma. However, how 1-dB loss in VF sensitivity affects the execution of daily tasks in real-world situations remains poorly understood. We noted that, for each decibel loss in binocular VF sensitivity, the risk of collision increased by 6% (95% CI, 0.2%-12%) in daytime stair navigation, 7% (95% CI, 0.3%-15%) in daytime city navigation, 15% (95% CI, 8%-22%) in nighttime stair navigation, and 15% (95% CI, 8%-23%) in nighttime city navigation, after controlling for covariates (eTable 4 and eTable 6 in the Supplement). By connecting visual function measurements to functional consequences, VR simulations may provide a new perspective for patients and clinicians to gain insights into the implication of the clinical test results for risk assessment, patient education, and improvement of treatment adherence.

    Age was associated with the VR performance measured across all the simulations (eTables 2-6 in the Supplement). Adjusting for age-related decline in visual performance therefore appears to be relevant to identify patients with vision-related disability. We defined vision-related disability when the VR disability score was outside the age-adjusted 95% confidence region of the healthy individuals (Figure 4). Some patients with glaucoma and VF defects did not show vision-related disability in the 5 VR simulations examined in this study (eFigure 7 in the Supplement). In other words, these results suggest that an abnormal visual function test result may not indicate disability, especially when the patient is young. Virtual reality simulations can better inform vision-related disability than the current standard visual function measures.

    Limitations and Strengths

    This study has limitations. As with all visual function tests and performance-based tests, learning effect and mastering of VR devices can confound the interpretation of test results. Multiple measures had been taken to minimize learning effect in the VR simulations. For example, all participants needed to complete a training session to familiarize themselves with the 5 VR simulation environments before the tests were conducted. In addition, the types and locations of VR objects in a VR environment were different between the training and testing sessions and between the daytime and nighttime simulations, as shown in the inserts in Figure 2A and B and Figure 3A and B. The finding that the test-retest variabilities of the VR simulation measurements were relatively low (eResults, eTable 9, and eTable 10 in the Supplement) suggests that the influence of learning effect on the VR simulations, if any, would be small.

    Although the association between VF sensitivity and performance measures obtained from the Assessment of Ability Related to Vision in glaucoma is weak,18 binocular VF sensitivity demonstrated relatively strong associations with the VR performance parameters. For example, 28.1% of the variation in the duration to complete the supermarket shopping simulation was explained by binocular VF sensitivity (eFigure 2A in the Supplement) and the proportion increased to 45.1% when age and other visual function parameters were included in the model (eTable 2 in the Supplement). There are other visual (eg, peripheral VF sensitivity beyond 24°, color vision, and stereopsis) and nonvisual function measures (eg, cognitive function) that can be associated with vision-related disability, but they were not measured in the study. Other limitations of VR simulation are described in the eDiscussion in the Supplement.

    Conclusions

    The innovation of VR simulation of activities that are relevant to patients’ real-world experience may provide a new paradigm to integrate various domains of visual function and cognitive function to assess and monitor vision-related disability for enhancement of clinical care and advancement of vision research.

    Back to top
    Article Information

    Accepted for Publication: January 15, 2020.

    Corresponding Author: Christopher K. S. Leung, MD, MB, ChB, Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Hong Kong, PRC (cksleung@cuhk.edu.hk).

    Published Online: March 19, 2020. doi:10.1001/jamaophthalmol.2020.0392

    Author Contributions: Drs Lam and Leung had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Lam, To, Yu, Cheng, Leung.

    Acquisition, analysis, or interpretation of data: Lam, To, Weinreb, Yu, Mak, Lai, Chiu, Wu, Zhang, Guo, Leung.

    Drafting of the manuscript: Lam, Wu, Cheng, Leung.

    Critical revision of the manuscript for important intellectual content: Lam, To, Weinreb, Yu, Mak, Lai, Chiu, Wu, Zhang, Guo, Leung.

    Statistical analysis: Lam, To, Yu, Mak, Wu, Guo, Leung.

    Obtained funding: Lam, To, Leung.

    Administrative, technical, or material support: Lam, To, Weinreb, Yu, Mak, Lai, Chiu, Wu, Zhang, Cheng, Leung.

    Supervision: Leung.

    Conflict of Interest Disclosures: Dr Lam reported receiving grants from the Innovation and Technology Commission, the Government of Hong Kong Special Administration Region (HKSAR), and The Chinese University of Hong Kong during the conduct of the study; in addition, Dr Lam reported having had patent application US Non-Provisional Application No. 15/466,348 pending, related to the virtual reality technology used in this study. Dr To reported receiving grants from the Innovation and Technology Commission Innovation and Technology Fund, Technology Start-up Support Scheme for Universities, The Chinese University of Hong Kong Knowledge Transfer Project Fund, and The Chinese University of Hong Kong Technology and Business Fund during the conduct of the study; in addition, Dr To reported having had patent application US Non-Provisional Application No. 15/466,348 pending, related to the virtual reality technology used in this study. Dr Weinreb reported receiving personal fees from Aerie Pharmaceuticals, Allergan, Eyenovia, and Implantdata; and nonfinancial support from Heidelberg Engineering, Carl Zeiss Meditec, Genentech, Konan, Optovue, Topcon, Optos, Centervue, and Bausch & Lomb outside the submitted work. Dr Leung reported receiving grants from HKSAR Innovation and Technology Commission and The Chinese University of Hong Kong during the conduct of the study; in addition, Dr Leung reported having had patent application US Non-Provisional Application No. 15/466,348 pending, related to the virtual reality technology used in this study. No other disclosures were reported.

    Funding/Support: The study was supported by Hong Kong Innovation and Technology Fund (ITS/338/15), Hong Kong Innovation and Technology Commission Technology Start-up Support Scheme for Universities (2015-2018), The Chinese University of Hong Kong Technology and Business Funds (2015/16).

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Meeting Presentation: The findings reported herein were partially presented as a poster at the World Glaucoma Congress; June 30, 2017; Helsinki, Finland, and an oral presentation at the Association of Research in Vision and Ophthalmology Annual Meeting; May 10, 2017; Baltimore, Maryland.

    References
    1.
    Flaxman  SR, Bourne  RRA, Resnikoff  S,  et al; Vision Loss Expert Group of the Global Burden of Disease Study.  Global causes of blindness and distance vision impairment 1990-2020: a systematic review and meta-analysis.  Lancet Glob Health. 2017;5(12):e1221-e1234. doi:10.1016/S2214-109X(17)30393-5PubMedGoogle ScholarCrossref
    2.
    Tham  YC, Li  X, Wong  TY, Quigley  HA, Aung  T, Cheng  CY.  Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis.  Ophthalmology. 2014;121(11):2081-2090. doi:10.1016/j.ophtha.2014.05.013PubMedGoogle ScholarCrossref
    3.
    Weinreb  RN, Leung  CK, Crowston  JG,  et al.  Primary open-angle glaucoma.  Nat Rev Dis Primers. 2016;2:16067. doi:10.1038/nrdp.2016.67PubMedGoogle ScholarCrossref
    4.
    Haymes  SA, Leblanc  RP, Nicolela  MT, Chiasson  LA, Chauhan  BC.  Risk of falls and motor vehicle collisions in glaucoma.  Invest Ophthalmol Vis Sci. 2007;48(3):1149-1155. doi:10.1167/iovs.06-0886PubMedGoogle ScholarCrossref
    5.
    McGwin  G  Jr, Huisingh  C, Jain  SG, Girkin  CA, Owsley  C.  Binocular visual field impairment in glaucoma and at-fault motor vehicle collisions.  J Glaucoma. 2015;24(2):138-143. doi:10.1097/IJG.0b013e3182a0761cPubMedGoogle ScholarCrossref
    6.
    Baig  S, Diniz-Filho  A, Wu  Z,  et al.  Association of fast visual field loss with risk of falling in patients with glaucoma.  JAMA Ophthalmol. 2016;134(8):880-886. doi:10.1001/jamaophthalmol.2016.1659PubMedGoogle ScholarCrossref
    7.
    Montana  CL, Bhorade  AM.  Glaucoma and quality of life: fall and driving risk.  Curr Opin Ophthalmol. 2018;29(2):135-140. doi:10.1097/ICU.0000000000000455PubMedGoogle ScholarCrossref
    8.
    Centers for Disease Control and Prevention. Deaths from unintentional injury among adults aged 65 and over: United States, 2000-2013. Accessed July 7, 2018. https://www.cdc.gov/nchs/data/databriefs/db199.htm
    9.
    Nelson  P, Aspinall  P, O’Brien  C.  Patients’ perception of visual impairment in glaucoma: a pilot study.  Br J Ophthalmol. 1999;83(5):546-552. doi:10.1136/bjo.83.5.546PubMedGoogle ScholarCrossref
    10.
    Viswanathan  AC, McNaught  AI, Poinoosawmy  D,  et al.  Severity and stability of glaucoma: patient perception compared with objective measurement.  Arch Ophthalmol. 1999;117(4):450-454. doi:10.1001/archopht.117.4.450PubMedGoogle ScholarCrossref
    11.
    Chung  DC, McCague  S, Yu  ZF,  et al.  Novel mobility test to assess functional vision in patients with inherited retinal dystrophies.  Clin Exp Ophthalmol. 2018;46(3):247-259. doi:10.1111/ceo.13022PubMedGoogle ScholarCrossref
    12.
    Medeiros  FA, Weinreb  RNR, R Boer  E, Rosen  PN.  Driving simulation as a performance-based test of visual impairment in glaucoma.  J Glaucoma. 2012;21(4):221-227. doi:10.1097/IJG.0b013e3182071832PubMedGoogle ScholarCrossref
    13.
    Glen  FC, Smith  ND, Crabb  DP.  Impact of superior and inferior visual field loss on hazard detection in a computer-based driving test.  Br J Ophthalmol. 2015;99(5):613-617. doi:10.1136/bjophthalmol-2014-305932PubMedGoogle ScholarCrossref
    14.
    Russell  S, Bennett  J, Wellman  JA,  et al.  Efficacy and safety of voretigene neparvovec (AAV2-hRPE65v2) in patients with RPE65-mediated inherited retinal dystrophy: a randomised, controlled, open-label, phase 3 trial.  Lancet. 2017;390(10097):849-860. doi:10.1016/S0140-6736(17)31868-8PubMedGoogle ScholarCrossref
    15.
    Bainbridge  JW, Smith  AJ, Barker  SS,  et al.  Effect of gene therapy on visual function in Leber’s congenital amaurosis.  N Engl J Med. 2008;358(21):2231-2239. doi:10.1056/NEJMoa0802268PubMedGoogle ScholarCrossref
    16.
    Maguire  AM, Simonelli  F, Pierce  EA,  et al.  Safety and efficacy of gene transfer for Leber’s congenital amaurosis.  N Engl J Med. 2008;358(21):2240-2248. doi:10.1056/NEJMoa0802315PubMedGoogle ScholarCrossref
    17.
    Wei  H, Sawchyn  AK, Myers  JS,  et al.  A clinical method to assess the effect of visual loss on the ability to perform activities of daily living.  Br J Ophthalmol. 2012;96(5):735-741. doi:10.1136/bjophthalmol-2011-300093PubMedGoogle ScholarCrossref
    18.
    Ekici  F, Loh  R, Waisbourd  M,  et al.  Relationships between measures of the ability to perform vision-related activities, vision-related quality of life, and clinical findings in patients with glaucoma.  JAMA Ophthalmol. 2015;133(12):1377-1385. doi:10.1001/jamaophthalmol.2015.3426PubMedGoogle ScholarCrossref
    19.
    De Maesschalck  R, Jouan-Rimbaud  D, Massart  DL.  The Mahalanobis distance.  Chemom Intell Lab Syst. 2000;50(1):1-18. doi:10.1016/S0169-7439(99)00047-7Google ScholarCrossref
    20.
    Brereton  RG.  Hotelling’s T squared distribution, its relationship to the F distribution and its use in multivariate space.  J Chemometrics. 2016;30(1):18-21. doi:10.1002/cem.2763Google ScholarCrossref
    21.
    Mangione  CM, Lee  PP, Gutierrez  PR, Spritzer  K, Berry  S, Hays  RD; National Eye Institute Visual Function Questionnaire Field Test Investigators.  Development of the 25-item National Eye Institute Visual Function Questionnaire.  Arch Ophthalmol. 2001;119(7):1050-1058. doi:10.1001/archopht.119.7.1050PubMedGoogle ScholarCrossref
    22.
    Mangione  CM, Lee  PP, Pitts  J, Gutierrez  P, Berry  S, Hays  RD; NEI-VFQ Field Test Investigators.  Psychometric properties of the National Eye Institute Visual Function Questionnaire (NEI-VFQ).  Arch Ophthalmol. 1998;116(11):1496-1504. doi:10.1001/archopht.116.11.1496PubMedGoogle ScholarCrossref
    23.
    Sherwood  MB, Garcia-Siekavizza  A, Meltzer  MI, Hebert  A, Burns  AF, McGorray  S.  Glaucoma’s impact on quality of life and its relation to clinical indicators: a pilot study.  Ophthalmology. 1998;105(3):561-566. doi:10.1016/S0161-6420(98)93043-3PubMedGoogle ScholarCrossref
    24.
    Janz  NK, Wren  PA, Lichter  PR, Musch  DC, Gillespie  BW, Guire  KE.  Quality of life in newly diagnosed glaucoma patients: the Collaborative Initial Glaucoma Treatment Study.  Ophthalmology. 2001;108(5):887-897. doi:10.1016/S0161-6420(00)00624-2PubMedGoogle ScholarCrossref
    ×