Highlight
In a large US cohort of 83,305 patients with newly diagnosed open-angle glaucoma suspect (OAGS), 20.6% converted to primary open-angle glaucoma (POAG) over follow-up, with an overall annual conversion rate of 6.1%.
Conversion was front-loaded: the annual rate was 9.4% in the first year after OAGS diagnosis and 5.3% in years 2 through 5.
Older age, male sex, Black race, residence outside the Northeast, gonioscopy, and glaucoma treatment during the suspect stage were associated with higher hazard of diagnostic conversion.
Using a pragmatic per-visit conversion threshold of 5%, the estimated monitoring interval ranged from approximately 6.1 years in the lowest-risk subgroup to 0.6 years in the highest-risk subgroup, illustrating the potential value of risk-stratified surveillance.
Background
Primary open-angle glaucoma remains one of the leading causes of irreversible blindness worldwide. Because structural optic nerve injury and functional visual field loss accumulate silently over years, clinicians often monitor patients labeled as glaucoma suspects before definitive disease is established. This creates a familiar tension in practice: too little surveillance risks delayed diagnosis and preventable visual loss, while too much surveillance increases cost, patient burden, and pressure on eye care capacity.
Open-angle glaucoma suspect is a heterogeneous designation. Patients may be classified as suspects because of elevated intraocular pressure, suspicious optic disc appearance, borderline retinal nerve fiber layer changes on optical coherence tomography (OCT), unreliable or equivocal visual field findings, family history, or some combination of these. Not all such patients progress to POAG, and progression risk is known to vary substantially. Yet in day-to-day practice, follow-up intervals are often determined by clinician judgment, local convention, and logistical constraints rather than a formalized estimate of near-term conversion risk.
The study by Yoo and colleagues addresses this practical gap using a very large US administrative claims database. The investigators sought not only to quantify annual rates of conversion from OAGS to POAG, but also to develop a simple framework for estimating per-visit risk and tailoring follow-up frequency accordingly. The clinical appeal of this approach lies in its direct relevance to resource allocation: if some patients have very low short-term conversion risk, less frequent monitoring may be reasonable, while higher-risk groups may justify tighter surveillance.
Proposed Section Structure
This topic is best organized around clinical applicability rather than a purely descriptive summary. A logical structure includes: clinical background and unmet need; study design and cohort definition; key quantitative findings; interpretation of the risk-stratified monitoring model; clinical implications for scheduling surveillance; limitations and generalizability; and a concise conclusion with research priorities.
Study Design
Design and Data Source
This was a retrospective cohort study using Optum’s de-identified Clinformatics Data Mart Database, covering the years 2007 through 2021 in the United States. Administrative claims data were used to identify individuals newly diagnosed with OAGS and to track subsequent diagnostic conversion to POAG.
Population
Eligible patients had newly diagnosed OAGS and were required to have continuous enrollment during a 3-year lookback period before the index diagnosis and a 5-year study period afterward. To improve diagnostic validity, the OAGS diagnosis had to be made by an ophthalmologist or optometrist, and patients needed at least one OCT or visual field test. Patients with any prior glaucoma-specific treatment or prior POAG diagnosis were excluded.
The final cohort included 83,305 patients. Of these, 57.6% were female; 5.4% were Asian, 11.1% Black, 12.7% Hispanic, and 70.8% non-Hispanic White.
Outcome
The primary outcome was diagnostic conversion from OAGS to POAG, identified through diagnosis codes in claims data.
Statistical Approach
The investigators used multivariable Cox proportional hazards modeling to assess factors associated with conversion. In a secondary analysis, they created subgroups based on age and treatment status and estimated predicted probabilities of conversion for each group. These probabilities were then used to derive monitoring intervals anchored to a per-visit conversion risk threshold.
Key Findings
Overall Conversion Rates
Among the 83,305 OAGS patients, 17,134 converted to POAG, representing 20.6% of the cohort. The overall annual conversion rate was 6.1%.
Importantly, the timing of conversion was not uniform. The annual conversion rate was 9.4% during the first year after the index OAGS diagnosis, then fell to 5.3% annually during years 2 through 5. This pattern suggests that the first year after initial designation as a glaucoma suspect carries a distinctively higher diagnostic yield. Several explanations are plausible. Some patients may already have had early POAG that was not fully documented at index presentation, while others may undergo intensified testing shortly after the initial suspect diagnosis, increasing the likelihood of clarification and formal reclassification.
Factors Associated With Higher Hazard of Conversion
On multivariable Cox regression, several factors were associated with greater hazard of conversion to POAG. Older age showed a meaningful association, with hazard ratios of at least 1.38 depending on age stratum. Male sex was associated with a modestly increased hazard (HR 1.12). Black race was also associated with higher hazard (HR 1.14), consistent with a broader body of literature documenting greater glaucoma burden and earlier or more severe disease in Black populations.
Geography appeared relevant as well: residence outside the Northeast was associated with a higher hazard of conversion, with hazard ratios of at least 1.24. A record of gonioscopy was strongly associated with conversion (HR 1.90). OAGS treatment was also associated with higher hazard, with hazard ratios of at least 1.31.
These latter associations deserve careful interpretation. Gonioscopy and treatment are unlikely to be causal drivers of conversion. Rather, they probably function as markers of clinician concern, disease complexity, or higher baseline suspicion. In claims-based analyses, this type of confounding by indication is expected: patients judged to be at higher risk are more likely to undergo more thorough evaluation and receive treatment before definitive POAG coding appears.
Risk-Stratified Subgroups
The study’s most practice-oriented contribution came from its subgroup analysis by age and treatment status. Among the lowest-risk patients, defined as younger than 50 years and untreated, the annual conversion rate was 2.0%, and observed follow-up occurred every 0.9 plus or minus 0.6 years. Among the highest-risk patients, defined as older than 70 years and treated, the annual conversion rate was 16.7%, with follow-up every 0.4 plus or minus 0.4 years after year 1.
When the authors standardized surveillance to a 5.0% per-visit conversion threshold, the implied monitoring interval varied markedly across risk strata. For the lowest-risk group, the interval was 6.1 years; for the highest-risk group, it was 0.6 years. This contrast sharply illustrates how current follow-up patterns may not always align with estimated near-term conversion risk.
Clinical Interpretation
Why the First Year Looks Different
The elevated first-year conversion rate of 9.4% is probably not solely biological progression. In routine glaucoma care, a new suspect diagnosis often triggers confirmatory testing, repeat visual fields, OCT review, pachymetry, gonioscopy, and closer optic nerve assessment. Some eyes initially coded as OAGS are therefore better understood as having prevalent but not yet fully adjudicated POAG. This makes the first post-diagnosis year a hybrid interval containing both true incident conversion and diagnostic refinement.
From a clinical scheduling standpoint, this supports maintaining relatively close follow-up soon after the initial suspect designation, even in lower-risk individuals, until the diagnostic picture stabilizes.
How to Use Per-Visit Risk in Practice
The study’s conceptual advance is the shift from annualized conversion risk alone to per-visit conversion risk. Clinicians do not manage glaucoma suspects in abstract yearly units; they choose when the next visit should occur. A risk-per-visit framework therefore maps more naturally onto clinical workflow.
That said, the proposed 5% per-visit threshold should be viewed as pragmatic, not definitive. Different practices may prefer different thresholds depending on patient preferences, medicolegal environment, access limitations, baseline testing quality, life expectancy, and the consequences of delayed detection. A younger patient with stable disc appearance and normal fields may tolerate a higher threshold than a monocular patient, someone with strong family history, or a patient with poor follow-up reliability.
Interpreting Treatment as a Risk Marker
One potentially counterintuitive finding is that treated OAGS patients had higher conversion hazard. This should not be interpreted to mean treatment worsens outcomes. Rather, clinicians likely treated those considered more likely to convert based on features not fully captured in claims data, such as intraocular pressure level, corneal thickness, optic nerve cupping, disc hemorrhage, retinal nerve fiber layer thinning, or suspicious field change. In observational datasets, treatment often identifies a high-risk phenotype rather than a treatment effect.
Strengths of the Study
The study has several notable strengths. First, the sample size was very large, improving precision and allowing robust subgroup analyses. Second, the inclusion criteria attempted to increase diagnostic specificity by requiring eye care provider coding and at least one OCT or visual field test. Third, the time horizon of up to 5 years is clinically meaningful for surveillance planning. Fourth, the risk-stratified framework is directly translatable to operational decisions in high-volume glaucoma and comprehensive ophthalmology clinics.
The work is also relevant to health systems planning. As populations age and demand for glaucoma care rises, identifying low-risk patients who may safely undergo less frequent surveillance could help preserve specialist capacity for higher-risk individuals and for patients with established glaucoma who require closer management.
Limitations and Cautions
Despite its practical value, this study should not be mistaken for a definitive progression model. The outcome was diagnostic conversion based on claims coding, not biologically verified onset of optic neuropathy. Coding behavior varies across clinicians and practices, and some cases of true POAG may be undercoded or delayed in coding. Conversely, some apparent conversions may reflect administrative relabeling rather than objective worsening.
Claims data also lack key ophthalmic variables central to glaucoma risk assessment, including intraocular pressure, central corneal thickness, cup-to-disc ratio, disc hemorrhage, OCT thickness metrics, baseline and longitudinal visual field indices, laterality, medication adherence, and family history detail. As a result, the model cannot incorporate the granular features that ordinarily drive individualized surveillance decisions.
Residual confounding is highly likely. Gonioscopy and treatment are almost certainly proxies for clinician concern. Geographic associations may reflect regional variation in coding, access, practice style, or case mix rather than underlying biological differences. In addition, continuous enrollment requirements may have selected for patients with stable insurance coverage, which may limit generalizability to uninsured, intermittently insured, or socioeconomically vulnerable populations.
The extrapolation of a 6.1-year interval for the lowest-risk subgroup should be interpreted with particular caution. While mathematically consistent with the chosen threshold, such a long interval may be unacceptable in real practice because testing quality, adherence, competing ocular disease, and changes in risk profile over time are not static. Most clinicians would still want periodic reassessment at shorter intervals, especially after an initial baseline period.
Relation to Existing Evidence and Guidelines
The findings align directionally with prior evidence that glaucoma risk is heterogeneous and influenced by age, race, and baseline clinical suspicion. The Ocular Hypertension Treatment Study established that risk-based prediction can meaningfully inform surveillance and treatment decisions in ocular hypertension, although the present study addresses a broader suspect population rather than strictly ocular hypertensive eyes. Current guideline-based care similarly emphasizes individualized follow-up based on risk factors, optic nerve appearance, pressure, and structural and functional testing.
The American Academy of Ophthalmology Preferred Practice Pattern for primary open-angle glaucoma suspect recommends that the frequency of optic nerve, retinal nerve fiber layer, and visual field assessment be individualized according to a patient’s level of risk. The present study does not replace that guidance, but it offers a population-based empirical scaffold for operationalizing visit frequency in the absence of complete clinical data.
Practical Implications for Clinicians and Health Systems
For clinicians, the main takeaway is not that low-risk glaucoma suspects should disappear from care for years at a time. Rather, it is that surveillance intensity can be more explicitly calibrated to estimated conversion risk. The first year after suspect diagnosis appears to merit relatively close attention because a substantial fraction of “conversions” likely represent clarification of baseline status. After that, many patients may be safely monitored less frequently than current habits dictate, provided structural and functional testing is of good quality and the overall risk profile remains low.
For health systems, a tiered scheduling model may be especially useful. A practical framework might include an early confirmatory phase after initial OAGS diagnosis, followed by differentiated surveillance based on age, treatment status, race, ocular risk markers, and reliability of follow-up. Such an approach could help reduce unnecessary visit volume while preserving rapid access for higher-risk patients and those with established glaucoma progression.
For researchers, the next step is integration of claims-based pragmatism with clinical precision. Future models should combine administrative data with electronic health record and imaging data, enabling prediction based on intraocular pressure, OCT-derived retinal nerve fiber layer thinning, visual field trend metrics, and optic nerve features. Prospective validation will be essential before any interval algorithm is adopted broadly.
Conclusion
This large retrospective US cohort study suggests that most open-angle glaucoma suspects convert to POAG at a rate below 6% per year after the first year following diagnosis, with substantially lower risk in younger untreated patients and much higher risk in older treated patients. The proposed per-visit risk framework is clinically appealing because it translates epidemiologic risk into scheduling decisions. Its greatest value is likely as a population-level tool to support more rational surveillance intensity, not as a substitute for individualized glaucoma assessment.
In practice, the study supports three messages. First, the first year after OAGS diagnosis is diagnostically high yield and likely deserves closer attention. Second, risk heterogeneity is large enough that uniform follow-up schedules are inefficient. Third, any move toward less frequent monitoring in low-risk patients should be tempered by clinical judgment, baseline test quality, and the recognition that claims-based conversion is an imperfect proxy for true disease onset.
Funding and ClinicalTrials.gov
No ClinicalTrials.gov registration applies because this was a retrospective observational database study. Funding information was not provided in the source summary and should be verified in the full published article.
References
Yoo K, Wu L, James A, Lung K, Camp A, Bolo K, Xu BY. Risk-Stratified Monitoring of Open Angle Glaucoma Suspects Based on Diagnostic Conversion Risk. American Journal of Ophthalmology. 2026-06-03. PMID: 42242412.
American Academy of Ophthalmology Preferred Practice Pattern Glaucoma Panel. Primary Open-Angle Glaucoma Suspect Preferred Practice Pattern. San Francisco, CA: American Academy of Ophthalmology. Latest available edition should be consulted for current recommendations.
Kass MA, Heuer DK, Higginbotham EJ, et al. The Ocular Hypertension Treatment Study: a randomized trial determines that topical ocular hypotensive medication delays or prevents the onset of primary open-angle glaucoma. Archives of Ophthalmology. 2002;120(6):701-713. PMID: 12049574.
Gordon MO, Beiser JA, Brandt JD, et al. The Ocular Hypertension Treatment Study: baseline factors that predict the onset of primary open-angle glaucoma. Archives of Ophthalmology. 2002;120(6):714-720. PMID: 12049575.
