Glaucoma screening is a testing program to detect glaucoma early, before symptoms appear, and prevent irreversible vision loss. Glaucoma is a chronic progressive optic neuropathy characterized by damage to the optic nerve and retinal nerve fiber layer5)9), and it often progresses slowly without symptoms in the early stages2).
As of 2020, the global number of glaucoma patients is estimated at approximately 76 million (prevalence 3.54%)2). This number is projected to increase to 111.8 million by 20402). Primary open-angle glaucoma (POAG) accounts for about 69% of all cases2).
At least half of those affected are undiagnosed and untreated, and this proportion is even higher in areas with limited access to healthcare2). In the United States, glaucoma is the second leading cause of irreversible blindness, but it is the leading cause among Black and Hispanic populations2).
The main risk factors for primary open-angle glaucoma are as follows6)7).
The prevalence in individuals aged 75 and older is 23.2% in Black people and 9.4% in White people, showing significant racial disparities2).
Glaucoma is asymptomatic in the early stages and progresses slowly, so at least half of patients remain undiagnosed2). Without treatment, it leads to irreversible blindness, but early detection and appropriate treatment can prevent vision loss. Especially in high-risk groups, targeted screening can efficiently detect undiagnosed cases due to high prevalence1)2).
Screening recommendations from various countries and international organizations differ significantly1).
| Organization | Recommendation for general population | Notes |
|---|---|---|
| AAO | Examination at age 40 | Earlier if risk factors present1) |
| USPSTF | Not recommended | Insufficient evidence1) |
| EGS | Not recommended | Effectiveness and cost-effectiveness unclear1)8) |
WHO considers general population screening to have low cost-effectiveness but recommends periodic examination for high-risk groups (elderly, family history, specific race/ethnicity, women, socially vulnerable)1).
The Pan-American Association of Ophthalmology (PAAO) also does not recommend general population screening, but advises examination for high-risk groups aged 65 and older, with positive family history, or of African descent1).
Exceptionally, in sub-Saharan Africa, joint guidelines with ICO recommend opportunistic screening for the general population aged 35 and older1).
The guidelines of the Japan Glaucoma Society emphasize that evaluation of the optic disc and retinal nerve fiber layer is the foundation of glaucoma diagnosis9). Epidemiological data from the Tajimi Study indicate that normal-tension glaucoma is common and many cases are missed by intraocular pressure measurement alone9).
A comprehensive eye examination including glaucoma screening includes the following 5)6).
OCT is widely used as an adjunctive test for glaucoma diagnosis, but glaucoma diagnosis should not be made based on OCT alone 5)8). There is no interchangeability of measurements between different OCT devices 5)8).
Because the prevalence of glaucoma in the general population is low, the current sensitivity and specificity of screening tests are insufficient, leading to many false positives and risks of overdiagnosis and unnecessary treatment 1). The USPSTF, EGS, and World Glaucoma Association all conclude that there is insufficient evidence that screening improves clinical outcomes 1). On the other hand, targeted screening limited to high-risk groups improves positive predictive value and is cost-effective 1)2).
Targeted screening for high-risk groups has higher clinical utility and cost-effectiveness compared to general population screening 1)2).
In a review by Allison et al., the Baltimore SToP Glaucoma Study conducted community-based screening targeting primarily African Americans aged 50 and older. As a result, 39.5% of screened individuals were referred for further examination, and 51% of those who attended were diagnosed with glaucoma 2).
Ladapo et al. estimated that screening Black individuals aged 80 and older could reduce the prevalence of blindness by 10.9%. Multiple screenings starting at age 50 could reduce undiagnosed glaucoma by 33%, visual impairment by 6.8%, and blindness by 9.9% 2).
In the United States, Medicare covers annual glaucoma examinations for the following high-risk groups 2).
In a decision analysis Markov model in China by Tang et al., the ICER (incremental cost-effectiveness ratio) for screening for PACG and primary open-angle glaucoma in rural areas was calculated to be $1,280, and screening was estimated to prevent 246 years of blindness per 100,000 people in rural areas and 1,325 years in urban areas 2).
AI screening using fundus photographs
Method: Deep learning (DL) is applied to color fundus photographs (CFP) to automatically detect optic disc cupping/disc ratio (CDR) and retinal nerve fiber layer defects 3)
Advantages: Eliminates inter-examiner variability, enabling objective and uniform assessment. Can be combined with telemedicine for remote screening 3)
Challenges: Disease severity bias in training data, lack of ethnic diversity, and unestablished minimum threshold for GON detection 3)
AI analysis of OCT and visual field tests
OCT: DL analysis of RNFL thickness and macular inner layer thickness data to automatically determine glaucomatous changes 3)
Visual field test: Pattern recognition of visual field test results for abnormality detection is being researched 3)
Outlook: Integrated AI analysis of CFP, OCT, and visual field data is expected to achieve diagnostic accuracy beyond single modalities 3)
Single-Gene Testing
Myocilin (MYOC): The first identified glaucoma gene, involved in about 5% of primary open-angle glaucoma4)
CYP1B1: Involved in about 20% of primary congenital glaucoma4)
Cascade Screening: Clinical cascade screening using family history is effective; a Tasmanian study found a number needed to screen (NNS) of 19 to detect one case (compared to 68 in the general population)4)
Polygenic Risk Score (PRS)
Overview: Quantifies the cumulative effect of hundreds of genetic variants to stratify an individual’s risk of developing glaucoma4)
Application: 80% of first-degree relatives are not high-risk and can be discharged, allowing resources to be concentrated on the remaining 20%4)
Challenges: Most GWAS are based on Nordic populations, and applicability to other ethnic groups such as African populations remains unvalidated4)
The Rotterdam Study found that the lifetime risk of developing glaucoma in first-degree relatives of glaucoma patients was 22.0%, approximately 10 times higher than the 2.3% in the control group4). This provides scientific evidence for family history-based screening.
In settings where standard automated perimeters are not available, the following tools are used6)7).
Mackey et al. point out that ensuring ethnic diversity in GWAS is essential to enhance the clinical utility of PRS4). The association of the LOXL1 gene with exfoliation glaucoma shows different results in Asian and African populations, and the universal application of PRS based solely on Nordic data has limitations4).
In the clinical implementation of PRS, ethical issues remain, such as when to inform families of risks identified through neonatal genetic testing, when to start clinical screening, and how to handle health literacy issues4).
In a review by Camara et al., the potential of multimodal AI analysis integrating fundus photography, OCT, and visual field testing is demonstrated3). However, there are many challenges for clinical implementation, including disease severity bias in training data, setting bias, and the lack of established minimum thresholds for GON detection3).
The development of polygenic risk scores (PRS) is progressing, and it is becoming possible to estimate an individual’s risk of developing glaucoma from the cumulative effect of multiple genetic variants 4). However, at present, most research data are based on Nordic populations, and applicability to other ethnic groups has not been verified. Additionally, ethical issues such as when and how to disclose genetic risk information need to be addressed 4). Further research is required for clinical implementation.
