Abstract
Background:
The evolution of visual acuity screening has seen a paradigm shift towards electronic charts and mobile apps for smart devices. However, these tools lack proper validation, raising significant concerns about their safety and accuracy. This study aimed at determining the validity of the WHOeyes app against the gold standard Snellen chart in community vision screening. Secondly, to evaluate the app’s inter-rater reliability and its agreement with the Snellen chart in community vision screening.
Methods:
A community-based analytical cross-sectional study was conducted from February to May 2024, involving participants aged 9 years and older. Visual acuity was tested using both the WHOeyes app and the Snellen chart by different examiners. Measurements were converted to the logarithm of the minimum angle of resolution (MAR) units for analysis.
Results:
The study involved 601 participants with a mean (SD) age of 42.4 (±17.9) years. The WHOeyes app showed a good range of sensitivity (66.67% to 100%), specificity (85.7% to 100%), and positive and negative predictive values between 36.3% to 100% and 69.4% to 100% at different cut-offs of 6/12, 6/18, 6/60, and beyond. Kappa value of the WHOeyes app vs. the Snellen chart showed moderate to substantial agreement, given visual impairment as a categorical outcome. Receiver operating characteristic curve (ROC) WHOeyes app demonstrated an area under the curve of 0.96 with optimal cut-offs between 6/12 and 6/18.
Conclusion:
WHOeyes is a valid and reliable screening application that may be used to identify visual impairments among people of different ages in the community.
Keywords: Reliability, validity, vision screening, visual acuity, WHOeyes
INTRODUCTION
Vision impairment (VI) is a critical public health concern that affects millions worldwide.[1] Currently, there are an estimated 4.95 million blind persons and 70 million vision-impaired persons in India.[2] VI not only limits economic and educational opportunities but also significantly reduces quality of life and increases the risk of premature mortality.[3,4,5] Despite the fact that over 80% of vision impairment cases are preventable through early diagnosis and treatment, a large number remain undetected. This is largely due to a lack of patient education and barriers in accessing healthcare services.[6,7] In recent years, modern electronic charts and easily downloadable mobile apps are increasingly available, reflecting the growing demand for portable vision screening tools. These tools are particularly crucial in remote and low-resource areas, where access to eye care is often limited.[8] While these mobile apps offer a promising alternative for vision screening, a significant concern lies in the lack of proper validation. This raises serious questions about their safety and accuracy, which are essential for effective and reliable vision care.[8,9] To address these challenges, the World Health Organization (WHO) recently launched the WHOeyes application on the occasion of World Sight Day (October 12, 2023). WHOeyes is an automated, smartphone-based app designed to revolutionize vision screening. It utilizes a tumbling E chart to test both distant and near visual acuity.[10] Our study aims at determining the validity of the WHOeyes app against the gold standard Snellen chart in community vision screening. Also, to evaluate the app’s inter-rater reliability and its agreement with the Snellen chart in community vision screening.
MATERIALS AND METHODS
This analytical cross-sectional study was conducted as a part of postgraduate research in Thaduthatkondur village of Villupuram district and at Thirubuvanai village in Puducherry (which were the field practice areas of our institution) for a period of 3 months duration (March–May 2024).
Considering 27.1% prevalence of vision impairment in India as per study by Vashist P et al. and expected sensitivity and specificity of 64.7% and 98%, respectively, as per study by Abdul Rahman SNA et al. with 95% confidence interval, 8% absolute precision, and 15% non-response rate, the sample size for the present study was calculated as 601 using open epi software version 3.[11,12]
Individuals aged 9 years and above who were physically fit and were able to communicate with a reliable mental status were included in the study. Individuals aged ≥9 years were included as they could reliably interpret the tumbling E chart and follow test instructions, as per WHOeyes app recommendations.[10] Participants who were ambulatory and did not report acute illness, injury, or disability that would prevent completion of the vision screening tests were considered “physically fit.” “Reliable mental status” was operationally defined as the ability to understand and follow visual acuity test instructions, confirmed through pre-test interaction.[12]
Following ethical clearance from Institutional Ethics Committee (IEC) and permission from relevant authorities (head of the locality) in the community, informed written consents were obtained from participants who were aged more than 18 years. Assent was obtained from participants aged between 12 and 18 years along with written consent from their parents. For participants aged between 9 and 12 years old, verbal consent was obtained along with written consent from their parents. The participants were selected using purposive sampling technique. Subsequently, house-to-house surveys were conducted, during which basic sociodemographic details were collected using a Google Form, and two vision tests, one with WHOeyes app and the other with the Snellen chart were performed on the same day. The examiners included community medicine postgraduate doctors, medical interns, and 3rd-year medical students. The principal investigator first demonstrated the screening test, after which the students and medical interns (CRMI) practiced it with their peers. This was followed by supervised field sessions led by postgraduate doctors. The screening was conducted during the students Family Adoption Program visit as part of their experiential learning. For distant vision testing using WHOeyes app and Snellen chart, each eye was individually tested by covering the non-tested eye with an opaque occluder. Using WHOeyes app, the vision test was conducted by randomly portraying a single letter E one at a time, and the participant was asked to point out the letter’s direction. Then, the examiner would swipe the screen as per participant’s response. Distant vision testing using app was done two times once by the first examiner and second time by examiner 2 for testing inter-rater reliability. A brief break of 10 min was given in between the two tests. Vision test using Snellen chart was done once by the postgraduate doctors [Figure 1]. At the end of the screening, 121 participants with impaired visual acuity were referred to ophthalmologist to our nearest rural health training center located approximately 3 km from the study site.
Figure 1.
Flow chart of the study protocol
Statistical analysis
Data collected using Google Forms was downloaded as an Excel sheet in Microsoft Excel 2016. Following cleaning and coding, analysis was done using Jamovi software version 2.4.11. The definitions, cut-off points, and diagnostic terms used throughout the study were based on the WHO classification of visual impairment.[13] To align with existing literature and exclude dependent data, only vision levels obtained from the right eye were analyzed.[12] Continuous variables were summarized as mean (SD) and categorical variables as frequency and percentage. All visual acuity (VA) measurements were converted to logMAR units. Tests of validity like sensitivity, specificity, positive predictive value, negative predictive value, accuracy, positive likelihood ratio, and negative likelihood ratio with 95% confidence interval were calculated for each cut-off points, namely 6/12, 6/18, 6/60, and >6/60. Inter-rater (user 1 and user 2) and the app’s reliability (WHOeyes app Vs Snellen chart) with visual impairment as outcome were obtained using Cohen’s Kappa. Comparison of logMAR values with 95% confidence interval (CI) for each cut-off point between the two methods was done using a paired t-test. Intraclass correlation (ICC) of logMAR was performed to determine the degree of agreement between the two methods. A Bland–Altman plot was used to demonstrate the consistency of measurements obtained between the Snellen chart and WHOeyes app. P value <0.05 was considered significant.
RESULTS
The table of sociodemographic details of the study participants [Supplementary Table 1] shows that the mean (SD) age of the 601 study participants was 42.4 (±17.9) years. The majority of 427 (71%) were females, and about 163 (27.1%) individuals completed their secondary school education. A total of 18.3% of the individuals were illiterate. Less than half, i.e., 268 (44.6%) of them were unemployed, and of those employed, 123 (20.4%) indulged in agriculture and 68 (11.3%) as skilled workers. Nearly three fourth 475 (79.03%) of them had no usage of spectacles. The recommended frequency of eye checkups among the general population without any symptoms of eye pathology was once every two years. Among the study participants, most of them around 290 (48.3%) participants had never tested their VA previously, and 101 (16.8%) had an eye check-up more than three years ago. Around 148 (24.6%) participants underwent VA testing within a year, and 62 (10.3%) had testing done before two years. Using the WHOeyes app, 480 (79%) of the participants were found to have normal vision, 34 (5.7%) had mild visual impairment, 75 (12.5%) had moderate visual impairment, and 12 (2%) had severe visual impairment.
Supplementary Table 1.
Sociodemographic characteristics of eligible participants (n=601)
| Characteristics | n (%) | |||||
|---|---|---|---|---|---|---|
| Age in years, mean (SD) | 42.4 (17.9) | |||||
| Gender | ||||||
| Male Female |
174 (29) 427 (71) |
|||||
| Educational status | ||||||
| Illiterate Primary Secondary Higher Secondary Graduate and above |
110 (18.3) 134 (22.3) 163 (27.1) 80 (13.3) 114 (19) |
|||||
| Occupation | ||||||
| Unemployed Agriculture Factory workers Skilled workers Students Professionals Computer users |
268 (44.6) 123 (20.4) 34 (5.7) 68 (11.3) 73 (12.1) 20 (3.3) 15 (2.5) |
|||||
| Previously visual acuity tested | ||||||
| Not tested <1 yr ago 2 yrs ago >3 yrs ago |
290 (48.3) 148 (24.6) 62 (10.3) 101 (16.8) |
|||||
| Use of spectacles | ||||||
| Yes No |
126 (20.96) 475 (79.03) |
|||||
SD: Standard deviation
Table 1 shows the validity measures of the VA tested using the WHOeyes app among the study participants at various cut-offs. At cut-offs beyond 6/60, sensitivity, specificity, PPV, and accuracy were found to be 100%. Also, the sensitivity (92.6% with 95%CI of 90 to 94.8) and PPV (97.1% with 95%CI of 95.2 to 98.4) were maximum at the 6/12 cut-off point. At 6/18 cut-off, NPV (98.7% with 95% CI of 97.2 to 99.4) was higher. The specificity (96.7% with 95% CI of 94.7 to 97.9) and accuracy (95.53% with 95% CI of 93.4 to 96.9) were maximum at 6/60 cut-off point. Also, at 6/60 cut-off, the VA test using the WHOeyes app was 25.58 times more likely to be positive. Whereas at the 6/12 cut-off and 6/18 cut-off, the test was likely to be negative in only 0.34% and 0.16% of patients.
Table 1.
Validity measures summary of the WHOeyes app against the gold standard, the Snellen chart (n=601)
| Cut-off point | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) | Accuracy (95% CI) | LR+ (95% CI) | LR− (95% CI) |
|---|---|---|---|---|---|---|---|
| 6/12 | 92.6 (90,94.8) | 85.7 (77.2,92) | 97.1 (95.2,98.4) | 69.4 (60.4,77.5) | 91.5 (89,93.6) | 6.48 (3.99,10.53) | 0.08 (0.06,0.12) |
| 6/18 | 66.67 (43.75,83.72) | 95.69 (93.5,97.16) | 36.3 (22.19,53.38) | 98.7 (97.25,99.42) | 94.65 (92.33,96.3) | 15.46 (12.98,18.42) | 0.34 (0.25,0.48) |
| 6/60 | 85.45 (73.84,92.44) | 96.68 (94.68,97.95) | 74.6 (62.66,83.72) | 98.31 (96.76,99.14) | 95.53 (93.44,96.98) | 25.58 (22.45,29.14) | 0.16 (0.12,0.19) |
| <6/60 | 100 (34.24,100) | 100 (99.18,100) | 100 (34.24,100) | 100 (99.18,100) | 100 (99.19,100) | * | * |
PPV: Positive predictive value, NPV: Negative predictive value, LR+:Positive likelihood ratio, LR−: Negative likelihood ratio, CI: confidence interval, α was set at 0.05, *by default it is undefined
The overall validity measures of the WHOeyes app for vision screening in children aged 9 to 17 years showed high sensitivity (98.3%, 95% CI: 90.9–100.0%) and PPV (98.3%, 95% CI: 90.9%–100.0%) but extremely low specificity (0.0%, 95% CI: 0.0%–97.5%) and NPV (0.0%, 95% CI: 0.0%–97.5%). Despite low specificity, the diagnostic accuracy was 96.7% (95% CI: 88.5–99.6%). The likelihood ratio of a positive test (LR+) was 0.98 (95% CI: 0.95–1.02), and the likelihood ratio of a negative test (LR−) was infinite, indicating poor performance in ruling out disease. Similiarly the validity measures of the whoeyes app among the illiterate participants is mentioned in the Supplementary Table 2.
Supplementary Table 2.
Validity measures of the WHOeyes app compared to the Snellen chart (Gold Standard) among illiterate participants (n=110)
| Cut-off point | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) | Accuracy (95% CI) | LR+ (95% CI) | LR− (95% CI) |
|---|---|---|---|---|---|---|---|
| 6/12 | 88.6 (75.4–96.2) | 80.3 (68.7–89.1) | 75 (61.1–86) | 91.4 (81–97.1) | 83.6 (75.4–90) | 4.5 (2.73–7.4) | 0.142 (0.06–0.326) |
| 6/18 | 57.1 (18.4–90.1) | 92.2 (85.3–96.6) | 33.3 (9.9–65.1) | 96.9 (91.3–99.4) | 90 (82.8–94.9) | 7.3 (2.9–18.5) | 0.465 (0.197–1.095) |
| 6/60 | 75 (57.8–87.9) | 89.2 (79.8–95.2) | 77.1 (59.9–89.6) | 88 (78.4–94.4) | 84.5 (76.4–90.7) | 6.93 (3.51–13.7) | 0.280 (0.158–0.496) |
| <6/60 | 100 (34.24,100) | 100 (99.18,100) | 100 (34.24,100) | 100 (99.18,100) | 100 (99.19,100) | * | * |
PPV: Positive predictive value, NPV: Negative predictive value, LR+:Positive likelihood ratio, LR−: negative likelihood ratio, CI: confidence interval, α was set at 0.05, *by default it is undefined
Figure 2 shows the receiver operating curve characteristics for the overall cut-off points of the WHOeyes app. The area under the curve for the WHOeyes app was 0.96, showing a better sensitivity and specificity at a cut-off between 6/12 and 6/18.
Figure 2.
Receiver operating curve of all the cut-off points using the WHOeyes app (N = 601)
Table 2 summarizes the results of reliability testing using the kappa statistic. Kappa values of the inter-user reliability of the WHOeyes app test and against the gold standard Snellen chart were 0.708 and 0.603, respectively, showing moderate to substantial agreement.
Table 2.
Results of reliability testing using the kappa statistic (n=601)
| Category | Kappa value | Z | P |
|---|---|---|---|
| Inter-user | 0.708 | 22.8 | <0.001 |
| WHOeyes vs. Snellen chart | 0.603 | 19.4 | <0.001 |
The overall intraclass correlation coefficient (ICC) of logMAR units based on a single rater, absolute-agreement, one-way random effect model was 0.89 (with 95% CI of 0.883–0.913), showing an excellent reliability between the raters.
The Bland–Altman plot of logMAR units by two methods [Supplementary Figure 1 (116.3KB, tif) ] to find the difference between logMAR values of the WHOeyes and Snellen chart was plotted on the vertical (Y) axis against the mean logMAR acuity of the two methods on the horizontal (X) axis. The blue line on the figure represents the mean logMAR difference between two methods, while the green and pink lines show 95% limits of agreement (+/−1.96 SD) of the mean logMAR difference. Overall, the plot depicts an acceptable level of agreement between logMAR values of the WHOeyes and the Snellen Chart. Table 3 shows the mean difference between logMAR values measured using the WHOeyes app and the Snellen chart at different cut-off points. Overall, there was a slight underestimation of VA by the WHOeyes when compared to Snellen at the 6/6 cut-off point, with a mean difference of −0.011 (+/−0.49), which was statistically significant.
Table 3.
Mean difference in visual acuity measurements of WHOeyes and Snellen chart using paired samples test (n=601)
| Paired samples test | ||||||||
|---|---|---|---|---|---|---|---|---|
| Paired differences |
t | df | Sig. (2-tailed) | |||||
| Mean | Std. deviation | Std. error mean | 95% confidence interval of the difference |
|||||
| Lower | Upper | |||||||
| LogMAR WHOeyes – Snellen 6/60 | 0.051667 | 0.126367 | 0.036479 | −0.028623 | 0.131956 | 1.416 | 11 | 0.184 |
| LogMAR WHOeyes – Snellen 6/24 | 0.113316 | 0.211206 | 0.048454 | 0.011518 | 0.215114 | 2.339 | 18 | 0.031 |
| LogMAR WHOeyes – Snellen 6/18 | 0.125500 | 0.177076 | 0.055996 | −0.001173 | 0.252173 | 2.241 | 9 | 0.052 |
| LogMAR WHOeyes – Snellen 6/12 | −0.007857 | 0.196478 | 0.028068 | −0.064292 | 0.048578 | −0.280 | 48 | 0.781 |
| LogMAR WHOeyes – Snellen 6/9 | 0.004250 | 0.064937 | 0.018746 | −0.037009 | 0.045509 | 0.227 | 11 | 0.825 |
| LogMAR WHOeyes – Snellen 6/6 | −0.011313 | 0.049466 | 0.003056 | −0.017331 | −0.005295 | −3.702 | 261 | 0.000 |
DISCUSSION
The effectiveness of a new screening test hinges primarily on its accuracy and validity. Despite an increase in digital eye testing applications, validated apps remain scarce, presenting a significant challenge.[9] Most studies on validated digital VA apps have focused primarily on reliability, with less emphasis on critical validity parameters such as sensitivity, specificity, predictive values, and likelihood ratios. Hence, this study aimed to determine the validity of the WHOeyes app besides evaluating its inter-rater reliability and its agreement with the Snellen chart in community vision screening.
Our study indicates that at the 6/60 cut-off point, all validity parameters for the WHOeyes app were 100%, except for the likelihood ratio, which was undefined. For other cut-off points, the WHOeyes app demonstrated a sensitivity of 92.6% at the 6/12 cut-off point, which was notably high. This sensitivity was superior to other apps like VI screen, which reported sensitivity ranging from 55.6% to 88.4%, and peek acuity, which showed sensitivities between 83 and 86% in general studies and 48% to 78% in studies involving school children.[12,14] In contrast to sensitivity, the WHOeyes app consistently achieved high specificity across all cut-off points, especially at the 6/60 cut-off point, where the specificity was 96.68%. This high specificity aligned with values reported by the VI screen (94.7% to 99.3%) and even exceeded the 91% specificity observed in studies of peek acuity among school children.[12,14]
The WHOeyes app showed the highest positive predictive value of 97.1% at the 6/12 cut-off point. This high PPV indicates that the app was effective in identifying participants with visual impairments, particularly in populations with higher disease prevalence. The negative predictive values for the WHOeyes app were excellent across various cut-off points, peaking at 98.7% at the 6/18 cut-off point. However, direct comparisons with other apps were challenging due to differences in disease prevalence and participant age. The app achieved excellent area under the curve (AUC) scores at all cut-off points, demonstrating its robust performance in differentiating between normal and visually impaired participants. A similar AUC score was demonstrated by the Vis-screen application, with values ranging from 0.93 to 0.97, which was comparable with the WHOeyes app.[12]
Our study found that inter-rater reliability, measured by the kappa coefficient, showed a slight decline due to changes in examiners across different screening programs. Despite this, the agreement levels were still comparable to those of validated apps. A similar study on WHOeyes also has demonstrated a kappa value of 0.75, which was consistent with the findings of our study.[15] Kappa coefficient of the Vis-screen app was 0.7 and 0.6 for inter-rater reliability and its agreement against the Snellen chart, respectively.[12] The WHOeyes app also demonstrated moderate agreement with the Snellen chart. Kappa values of the inter-user reliability of the WHOeyes app test against the gold standard Snellen chart were found to be 0.708 and 0.603, respectively, showing moderate to substantial agreement. The ICC of 0.89 (95% CI of 0.883–0.913) showed excellent internal consistency for the WHOeyes app. This indicates a high level of agreement between raters, as per Landis and Koch’s 1977 cut-offs, reinforcing the app’s reliability in consistent measurement.[16] This was comparable with ICC values of smartphone-based VA testing among school students, peek acuity, and the Langone eye test application by New York University, which ranged from 0.85 to 0.98.[17,18,19,20,21,22] In fact, ICC value of WHOeyes was comparatively higher than Langone eye test.[21] Bland–Altman analysis showed an acceptable level of agreement between the visual acuities of the WHOeyes app and Snellen chart.
The WHOeyes app demonstrates impressive accuracy and reliability as a digital VA screening tool. With high sensitivity and specificity, coupled with excellent predictive values and strong ROC analysis results, it outperforms many existing apps. The peek acuity test performed among old-aged Kenyan adults utilized multiple handsets of the same phone model only throughout the study, raising the concern of the generalisability of the study results.[18] Similarly, Langone eye test exhibited potential accuracy concerns as it employed two fixed smartphone devices.[21] The app’s reliability, while slightly impacted by changes in examiners, remains competitive and consistent.
Strengths and limitations
This study benefits from reduced observer bias, use of real-world settings such as a diverse age range, various mobile phone types (Android and iOS), offline app use, and the Tumble E charts. However, it faces few limitations including challenges in maintaining consistent testing distance leading to slight alteration in VA measurements. Additionally, results are influenced by screen size, pixel density, and brightness. While consistent raters could improve inter-rater reliability, intra-rater reliability was not assessed due to time constraints and lack of participant support. The study sample had a higher proportion of female participants, likely due to their availability during daytime household visits. This may have introduced selection bias, potentially affecting the generalizability of the findings.
CONCLUSION
We conclude from our study that the WHOeyes app is a valid and reliable, dependable, screening tool compared to Snellen’s chart in identifying people with visual impairment. It can be used as a screening test in remote places and especially among those who need regular VA checks.
Ethical clearance
This study was approved by the Institutional Ethics Committee with reference number IEC:EC/127/2024.
Key message
WHOeyes app demonstrates impressive accuracy and reliability as a digital visual acuity sscreening tool in our study as compared to other existing apps.
Conflicts of interest
There are no conflicts of interest.
Bland–Altman plot showing distribution of logMAR values tested with the WHOeyes and the Snellen chart. X axis – Average logMAR acuity of the two methods being compared. Y axis – The difference in visual acuity between two methods. Blue line – the mean visual acuity difference between two methods with confidence interval (+/−1.96 SD). Green and pink lines – 95% limits of agreement (+/−1.96 SD) of mean visual acuity difference
Funding Statement
Nil.
REFERENCES
- 1.World Health Organization . Geneva: World Health Organization; 2023. Blindness and visual impairment. Available from: https://www.who.int/news-room/fact-sheets/detail/blindness-and-visual-impairment . [Last accessed on 2024 Sep 08] [Google Scholar]
- 2.Mannava S, Borah RR, Shamanna BR. Current estimates of the economic burden of blindness and visual impairment in India: A cost of illness study. Indian J Ophthalmol. 2022;70:2141–5. doi: 10.4103/ijo.IJO_2804_21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Eckert KA, Carter MJ, Lansingh VC, Wilson DA, Furtado JM, Frick KD, et al. A simple method for estimating the economic cost of productivity loss due to blindness and moderate to severe visual impairment. Ophthalmic Epidemiol. 2015;22:349–55. doi: 10.3109/09286586.2015.1066394. [DOI] [PubMed] [Google Scholar]
- 4.Crews JE, Chou CF, Zhang X, Zack MM, Saaddine JB. Health-related quality of life among people aged ≥65 years with self-reported visual impairment: Findings from the 2006–2010 behavioral risk factor surveillance system. Ophthalmic Epidemiol. 2014;21:287–96. doi: 10.3109/09286586.2014.926556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zheng DD, Christ SL, Lam BL, Arheart KL, Galor A, Lee DJ. Increased mortality risk among the visually impaired: The roles of mental well-being and preventive care practices. Invest Ophthalmol Vis Sci. 2012;53:2685–92. doi: 10.1167/iovs.11-8794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Evans BJW, Rowlands G. Correctable visual impairment in older people: A major unmet need. Ophthalmic Physiol Opt. 2004;24:161–80. doi: 10.1111/j.1475-1313.2004.00197.x. [DOI] [PubMed] [Google Scholar]
- 7.Armstrong KL, Jovic M, Vo-Phuoc JL, Thorpe JG, Doolan BL. The global cost of eliminating avoidable blindness. Indian J Ophthalmol. 2012;60:475–80. doi: 10.4103/0301-4738.100554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Aruljyothi L, Janakiraman A, Malligarjun B, Babu BM. Smartphone applications in ophthalmology: A quantitative analysis. Indian J Ophthalmol. 2021;69:548–53. doi: 10.4103/ijo.IJO_1480_20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mena-Guevara KJ, Piñero DP, de Fez D. Validation of digital applications for evaluation of visual parameters: A narrative review. Vision (Basel) 2021;5:58. doi: 10.3390/vision5040058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.World Health Organization . Geneva: WHO; WHOeyes app. Available from: https://www.who.int/teams/noncommunicable-diseases/sensory-functions-disability-and-rehabilitation/whoeyes. [Last accessed on 2024 Feb 03] [Google Scholar]
- 11.Vashist P, Senjam SS, Gupta V, Gupta N, Shamanna BR, Wadhwani M, et al. blindness and visual impairment and their causes in india: results of a nationally representative survey. plos one. 2022;17:e0271736. doi: 10.1371/journal.pone.0271736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Abdul Rahman SNA, Naing NN, Othman AM, Mohamad Z, Ibrahim TMM, Hashim SE, et al. Validity and reliability of Vis-Screen application: A smartphone-based distance vision testing for visual impairment and blindness vision screening. Medicina (Kaunas) 2023;59:912. doi: 10.3390/medicina59050912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.World Health Organization . Washington, D.C.: Pan American Health Organization; Visual health 2019. Available from: https://www.paho.org/en/topics/visual-health. [Last accessed on 2024 Jul 21] [Google Scholar]
- 14.Rono HK, Bastawrous A, MacLeod D, Wanjala E, Di Tanna GL, Weiss HA, et al. Smartphone-based screening for visual impairment in Kenyan school children: A cluster randomised controlled trial. Lancet Glob Health. 2018;6:e924–32. doi: 10.1016/S2214-109X(18)30244-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wu Y, Keel S, Carneiro VLA, Zhang S Wang W, Liu C, et al. real-world application of a smartphone-based visual acuity test (whoeyes) with automatic distance calibration. BR J ophthalmol. 2024;108:1613–20. doi: 10.1136/bjo-2023-324913. [DOI] [PubMed] [Google Scholar]
- 16.Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–74. [PubMed] [Google Scholar]
- 17.Bhaskaran A, Babu M, Abhilash B, Sudhakar NA, Dixitha V. Comparison of smartphone application-based visual acuity with traditional visual acuity chart for use in tele-ophthalmology. Taiwan J Ophthalmol. 2022;12:155–63. doi: 10.4103/tjo.tjo_7_22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bastawrous A, Rono HK, Livingstone IA, Weiss HA, Jordan S, Kuper H, et al. Development and validation of a smartphone-based visual acuity test (Peek Acuity) for clinical practice and community-based fieldwork. JAMA Ophthalmol. 2015;133:930–7. doi: 10.1001/jamaophthalmol.2015.1468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Han X, Scheetz J, Keel S, Liao C, Liu C, Jiang Y, et al. Development and validation of a smartphone-based visual acuity test (Vision at Home) Transl Vis Sci Technol. 2019;8:27. doi: 10.1167/tvst.8.4.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Perera C, Chakrabarti R, Islam FMA, Crowston J. The Eye Phone Study: Reliability and accuracy of assessing Snellen visual acuity using smartphone technology. Eye. 2015;29:888–94. doi: 10.1038/eye.2015.60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Iskander M, Hu G, Sood S, Heilenbach N, Sanchez V, Ogunsola T, et al. Validation of the New York University Langone Eye Test application, a smartphone-based visual acuity test. Ophthalmol Sci. 2022;2:100182. doi: 10.1016/j.xops.2022.100182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zhao L, Stinnett SS, Prakalapakorn SG. Visual acuity assessment and vision screening using a novel smartphone application. J Pediatr. 2019;213:203–10.e1. doi: 10.1016/j.jpeds.2019.06.021. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Bland–Altman plot showing distribution of logMAR values tested with the WHOeyes and the Snellen chart. X axis – Average logMAR acuity of the two methods being compared. Y axis – The difference in visual acuity between two methods. Blue line – the mean visual acuity difference between two methods with confidence interval (+/−1.96 SD). Green and pink lines – 95% limits of agreement (+/−1.96 SD) of mean visual acuity difference