Mohammad AskarBack to knowledge library

Glaucoma · Diagnostics

Glaucoma diagnostics

A practical lecture for ophthalmology training: examine the disc, read OCT and fields, measure pressure and plan follow-up.

Clinical lectureAbout 25 minutesMedical and editorial review: 06.10.2026 · 19:28 CEST · Europe/Berlin

Medically reviewed · Mohammad Askar · Published: 6 October 2026

In practice: measure pressure, examine the disc and angle, check OCT images and compare them with the field. One abnormal reading usually is not enough. For a new finding, check test quality and compare earlier examinations first.

Abbreviations and study numbers explained
POAG
Primary open-angle glaucoma.
MRW / cpRNFL
MRW measures a minimum width of the disc rim. cpRNFL is the nerve fibre layer around the disc.
HR
Hazard ratio compares event rates over time. HR 2.6 indicates a higher rate in the studied group, not a personal 260% chance of disease. Association does not prove causation.
95% CI
Confidence interval shows statistical uncertainty around an estimate. It does not capture every source of error in a study.
NNT
Number needed to treat to prevent one additional endpoint over a specified time. OHTS NNT ≈20 concerns POAG onset over five years, not prevention of blindness.
c-statistic
Shows how well a prediction model separates later events from non-events. It is not tonometer accuracy.

01Suspected glaucoma: what do we examine first?

When glaucoma is suspected, assess six areas: history, intraocular pressure (IOP), drainage angle, optic disc, retinal nerve fibre layer (RNFL) and visual field. The disc is the visible optic nerve head; the visual field tests visual function. 1 3

Then ask: Is damage already present? Is glaucoma suspected? Or is only the risk increased? One pressure reading, a large cup or a red OCT mark cannot answer this alone. 1 3

Ask about family glaucoma, previous pressures, operations, myopia and eye injuries. Review all medicines, especially corticosteroids as drops, tablets or other preparations. 2

Check previous results and treatment. Look for pseudoexfoliation material and pigment dispersion at the slit lamp. Symptoms and systemic disease help determine urgency and later treatment. 2

Ocular hypertension (OHT) means raised pressure without demonstrated glaucomatous damage. Pressure, disc appearance, OCT or visual fields may raise suspicion of glaucoma. 1

In preperimetric glaucoma, convincing structural damage is present without a typical visual field defect yet. Before using this diagnosis, check for measurement errors and other optic nerve diseases. 1

Compare OCT with the visual field. OCT may show early damage more clearly; visual fields may be more useful in advanced disease. There is no fixed sequence for every eye. 1 15

Histological estimates of 20–50% cell loss are not OCT cut-offs. They cannot tell us the GCL thickness at which an individual must develop a visual field defect. 1 15

02Optic disc: what do we see and what does it mean?

Assess the disc stereoscopically at the slit lamp. Start with size, shape and orientation. Then check rim width and colour, notches, RNFL defects, haemorrhages and vessels. 1

Inspect the surrounding tissue too. Photograph the baseline for later comparison. Always judge a large cup against the size of the disc. 1

The ISNT rule describes usual rim width: inferior > superior > nasal > temporal. It is most helpful in medium-sized discs. 1

Large, small and tilted discs often differ. A violation calls for further assessment but does not diagnose glaucoma. A small disc can be damaged despite having little cupping. 1

The cup-to-disc ratio (CDR) compares cup size with disc size; here we use the vertical measurement. A CDR of 0.7 can be normal in a large disc. A small disc may be damaged even with a CDR of 0.3. 1 2

With an inter-eye difference of about 0.2, first compare disc sizes. Then check the rim, OCT and visual field. Do not diagnose from a fixed CDR cut-off alone. 1 2

Measure vertical disc diameter with a narrow slit. Record the lens and its correction factor. EGS 6 gives about 1.9 mm as an average, not a boundary between health and disease. 1 12 13

The older EGS 4 table gives factors of 1.36–1.59 for different 90 D lenses. Example: with an appropriate factor of 1.36, 1.2 mm × 1.36 is about 1.63 mm. Check your lens factor; it does not apply to every 90 D lens. 1 12 13

Notching is a focal indentation of the rim. Bayoneting is a sharp bend of a vessel at the cup edge. Exposed vessels along the cup may also suggest rim loss. 1

Visible laminar pores and cup depth add descriptive detail but cannot determine stage alone. Marked pallor or unexpectedly poor acuity should prompt assessment for other optic nerve diseases, especially when disc and field findings do not fit. 1

Myopia can tilt, rotate or stretch the disc and nerve fibre pattern. The tilt direction alone cannot establish a neurological diagnosis. 1 8

A grey crescent is a grey area inside the disc margin. It can make the visible rim appear smaller than it is. Locate the true disc edge and distinguish this from atrophy outside the disc. 1 8

The alpha zone has irregular pigmentation beside the disc. The beta zone is closer to the disc edge, with visible sclera or larger choroidal vessels. 1

Beta atrophy also occurs in myopia and in older healthy eyes. Document its size and changes. A larger area alone does not prove glaucoma progression. 1

Look for optic disc haemorrhages (Drance haemorrhages) and photograph a new finding. Compare OCT and visual fields with earlier tests, then reassess pressure and risk factors. 1 9

A haemorrhage increases concern about disease or progression but may have other causes. A rule based on half a disc diameter cannot prove its cause. 1 9

Interpret optic disc findings together
FindingUseful questionLimitation
Disc size / CDRDoes the cup fit the disc size and rim?A large CDR alone does not prove damage; small discs can be diseased despite a small CDR.
Notch / RNFL defectIs there corresponding OCT and visual field topography?Consider artefacts, myopia and other optic neuropathies.
Grey crescent / atrophyWhere is the true outer disc margin?Do not count a pigment variant as absent rim; atrophy alone does not establish glaucoma.
Disc haemorrhageDocument a new haemorrhage and reassess trajectory and risks.Association does not prove current progression or a single cause.

1 8 9

Study appraisal: grey crescent and haemorrhage risk

Reykjavik 2005: A cross-sectional study assessed photos of 1,012 right eyes in people aged 50 or older. 22% had a grey crescent. An association with glaucoma was not demonstrated: OR 1.05; 95% CI 0.49–2.26. OR means odds ratio, a statistical comparison of the presence of disease. 8 9

In practice: A grey crescent can make the rim look too small. The study proves neither that it is congenital nor that it remains harmless throughout life. 8 9

Budenz/OHTS 2017: This was a prospective observational analysis within OHTS. Participants were followed over time. Annual stereo photos covered 3,236 eyes of 1,618 people. The median follow-up was 13 years: the middle of the follow-up distribution. Haemorrhages were detected in 179 eyes. 8 9

Result: Over the studied period, 25.6% of eyes with haemorrhage and 12.9% without developed POAG. The time-dependent HR adjusted for baseline risks was 2.6 (95% CI 1.7–4.0). Nevertheless, 78% of haemorrhage eyes did not reach a POAG endpoint during a median 49 months after haemorrhage. 8 9

Limits: Annual photos can miss short-lived haemorrhages. Eyes with baseline haemorrhage were excluded. Haemorrhage was the only risk factor occurring during follow-up that was included in the model. 8 9

Random assignment to medication reduced haemorrhage rate: HR 0.70 (95% CI 0.52–0.95). This differs from asking whether every new haemorrhage should trigger a particular treatment. That strategy was not tested by randomisation. Document and monitor a new haemorrhage more closely; do not infer current progression or a specific treatment from it alone. 8 9

03Reading OCT: images before colours

OCT images retinal layers and measures their thickness. It does not count living ganglion cells. Circle scans around the disc measure RNFL; radial disc scans assess the rim. Macular scans show inner retinal layers. 1

A 3.4 mm circle is only a protocol example. Scan shape and analysis depend on the device. Spectral-domain and swept-source are different OCT techniques; this does not establish universal superiority of a brand. 1

GCL means ganglion cell layer. GCIPL includes GCL and the inner plexiform layer. Depending on the software, GCC may also include RNFL. 1

Check which layers and area the device report actually measures. Do not directly compare absolute values across devices. Use the same system for follow-up where possible and record software changes. 1

Open the B-scans, the cross-sectional images, first. Are they sharp and centred? Are there motion errors or cloudy media? Do the drawn layer boundaries follow the actual anatomy? This automatic boundary detection is called segmentation. 1 14

Only then interpret the colours. A Spectralis Q value around 15 is a minimum in certain protocols, not a guarantee of a correct scan. Segmentation may be wrong even with a good signal. 1 14

Check alignment with the line between disc and fovea. Follow-up scans must return to the same location; this is called registration. The colour compares the measurement with a reference database. 1

Myopia, torsion, unusual disc size or wrong segmentation can mark healthy areas red ('red disease'). A green overall value may hide a local defect. Check individual areas and original images. 1

With central field defects or a difficult disc, also inspect the macula. Glaucoma damage often follows arcuate nerve fibre bundles and the horizontal raphe, the line between upper and lower bundles. 1

Vertically bounded patterns may fit disease at or behind the chiasm. A 'snail-shaped' pattern alone is not a diagnosis. Macular disease and myopia can also alter GCIPL. 1

04Follow-up OCT: real change or measurement error?

With a large change over a few weeks, open both original scans side by side. Were the same location and region measured? Check signal, segmentation, media and software. 1

Then decide whether the change is real or a measurement error. The teaching example '31 µm in 20 days' is retained. Without the original data, we cannot attribute it to vascular pulsation. 1

Healthy RNFL also thins with age. A healthy study group's average does not tell us how much loss is safe for an individual. 1 7

A trend line shows how quickly measurements change. A statistically significant decline must also fit image quality, defect location, stage and visual field. One test or non-comparable devices cannot establish a reliable trend. 1 7

The 'floor' is the range where OCT can barely measure further thinning. Residual tissue and the analysis method set this limit. There is no universal 40 µm point at which OCT should be stopped. 1 6

Give visual fields more weight in advanced glaucoma. Also check whether the macula or local OCT areas can still show change. GCC and RNFL do not necessarily reach their floors together. 1 6

Study appraisal: ageing – Leung 2012

Leung 2012: An observational study assessed 100 healthy people once. A random subset of 35 also underwent repeated tests over a mean 30 months. 7

Result: Global RNFL thinned by an average 0.52 µm/year, reported as −0.52 µm/year (95% CI −0.86 to −0.17). Individual sectors changed at different rates. 7

Limits: Only 35 people in follow-up, limited duration and a particular device context. The abstract and results report were read, not the full text; complete individual longitudinal data were unavailable. 7

In practice: This explains the commonly cited figure of about 0.5 µm/year. Do not subtract that value from every patient's trajectory. 7

Study appraisal: measurement floor – Bowd 2017

Bowd 2017: This study assessed OCT measurement limits. In 41 eyes with moderate-to-advanced damage, five measurements over five weeks estimated variability. A second, independent group had 87 advanced eyes and only two measurements over a mean 1.8 years. The full text was read. 6

Estimated floors: MRW 105 ±15.9 µm; GCIPL 38 ±3.4 µm; cpRNFL 38 ±4.2 µm. Abbreviations are explained in the terms box. The respective measured areas still above floor were 19%, 36% and 14%. 6

Important: These percentages describe areas within the study group. They do not describe how many diseased eyes the test detects; they are not diagnostic sensitivities. 6

Limits: With only two follow-up measurements, the rate of decline is imprecise. It is difficult to check whether change follows a straight line. The study used Spectralis and its own SALSA layer analysis in selected stages. 6

In practice: Macular GCIPL may still have measurable areas in advanced glaucoma. The study did not compare treatment decisions or demonstrate better preservation of vision. 6

05Measuring and interpreting eye pressure

Goldmann applanation tonometry (GAT) is the clinical reference method. It measures the force needed to flatten a defined corneal area. The Imbert–Fick principle explains this in a simplified model; a real cornea does not fully fit that model. 1

Check calibration, fluorescein amount and alignment of the semicircles. Keep the patient relaxed. Fluorescein rings ('mires') that are too broad or narrow can alter the reading. 1

Hold the lids open without pressing on the globe. Ask the patient to breathe normally and avoid straining. Posture and time of day also affect pressure; a fixed 6 mmHg addition is inappropriate. 1

Repeat an unexpected reading under good conditions. High pressure with pain or sudden visual loss requires urgent assessment. 1

CCT means central corneal thickness. Thickness, curvature and mechanical properties affect pressure measurement. Interpret readings especially carefully after refractive surgery or with corneal oedema. 1 5

Measure CCT and record it alongside the measured IOP. A blanket formula cannot give a reliably 'true' pressure. A value within the statistical normal range does not exclude glaucoma. 1 5

Other tonometers can help in children, difficult positioning or unusual corneas. Record device, model and measurement conditions during follow-up. 1

Two devices may track high and low values similarly yet differ substantially in the same eye. Good correlation is not the same as good agreement. Examine the differences and their variability too. 1

A diurnal profile or supervised home measurement can reveal fluctuations missed at one visit. Triggerfish is a contact lens sensor. It records shape changes at the corneoscleral region for up to 24 hours and may help identify a useful time for tonometry. 1 10

Its readings are not directly mmHg values. Implantable sensors are also being studied. Assess added benefit, cost and procedural risk; they do not simply replace the baseline examination. 1 10

The water drinking test measures the pressure response to a defined amount of water under controlled conditions. Studies use, for example, 800 mL, 1000 mL or 10 mL/kg. It cannot diagnose alone and is not a self-test. 11

The earlier 'four litres' statement is not a verified instruction. Assess medical suitability and fluid restrictions first. Simply waiting 1–2 hours cannot replace assessment of a concerning finding. 11

Pressure measurement: principle and clinical limitations
MethodMeasurement / useInterpretation
Goldmann / PerkinsApplanation; Perkins is portable and can be used supine.Consider calibration, fluorescein, cornea and examination technique.
Tono-PenSmall contact area; useful when access is restricted.Not automatically interchangeable with Goldmann; consider variability and examination context.
iCare / reboundBrief probe contact; selected devices permit supervised home measurements.Check corneal effects, position, training and device-specific differences.
Non-contact / ORAAir pulse; ORA also measures biomechanical parameters.Do not equate different outputs; clinically confirm concerning values.
Diurnal profileRepeated measurements at documented times and under documented conditions.An office profile does not capture every nocturnal peak; fluctuation alone does not establish damage.
TriggerfishContact lens sensor: patterns of pressure-related dimensional change over 24 hours.Electrical output, not directly mmHg; supplementary information rather than a replacement for tonometry.

1 10

Study appraisal: CCT correction – Brandt 2012

Brandt 2012: The authors recalculated the OHTS prediction model using five CCT correction formulae. The question was prediction of later POAG, not tonometer accuracy. The full text was read. 5

Participants: 1,433 of 1,636 people had all required data. Median follow-up was seven years. Correcting IOP did not improve prediction: c-statistics 0.754–0.775 versus 0.774 in the original model (Table 3). 5

The c-statistic describes how well the model separates later disease from non-disease. The result supports considering measured IOP and CCT separately. It does not mean that corneal mechanical properties are unimportant. 5

Limits: Only complete datasets from a selected OHT group were analysed. CCT was measured after random treatment assignment. The compared linear models were mathematically similar. 5

In practice: This was not a randomised device comparison or a test of every postoperative cornea. It guarantees neither a personal risk estimate nor a general formula for 'true' IOP. 5

06Gonioscopy: examining the angle

Gonioscopy asks: Is the angle open, narrow or closed? Is there a secondary cause of raised pressure? A normal central anterior chamber, the disc or IOP cannot answer this. 1 3

Use dim light and document all four quadrants. Excessive light or lens pressure can artificially open a narrow angle. 1 3

Identify Schwalbe's line, trabecular meshwork, scleral spur and ciliary body band. Record angle width, pigment, abnormal vessels and signs of trauma or pigment dispersion. PAS are peripheral anterior synechiae: adhesions between iris and angle. 1

During indentation, apply gentle pressure with a suitable gonioscopy lens. Appositional closure may open; an adhesion remains. Examine the angle without indentation too. 1

Van Herick estimates peripheral anterior chamber depth at the slit lamp. AS-OCT images the anterior segment. UBM means ultrasound biomicroscopy and also shows structures behind the iris. 1

These methods complement gonioscopy. They do not assess pigment, synechiae and secondary causes in the same way. A narrow-angle AS-OCT image alone should not determine treatment. 1

07Comparing OCT and fields, recording the baseline

Perform standard automated perimetry as a baseline test. Check strategy, refractive correction, fixation, false-positive responses, fatigue and media opacity. Repeat a new or implausible defect; the AAO benchmark also recommends repeat testing for newly detected glaucomatous damage in a suspect. 1 17

Then compare locations: superior RNFL damage fits an inferior field defect. Exact agreement is not required. Measurement variability and disease stage may explain differences. 1 17

With central symptoms, macular GCIPL change or suspected damage near fixation, test the central field more densely. 24-2c and 10-2 do not sample it identically. 1

PSD measures local irregularity in the field. GHT is the Glaucoma Hemifield Test, comparing upper and lower regions. Neither value can decide alone. Detailed interpretation and progression analysis belong in the visual field chapter. 1

Record at baseline: history and risks, acuity and pupils, anterior segment and fundus, IOP with device and time, CCT, gonioscopy, disc photo or imaging, original OCT images and reliable visual fields. AAO benchmarks list these core areas and also corneal hysteresis, a measure of the cornea's response to loading. 1 17

If tests conflict, check quality first. Consider retinal or neurological causes where appropriate. Record the unresolved question rather than presenting an uncertain diagnosis as definite. 1 17

08Planning follow-up and reviewing target pressure

Assess progression using good, comparable tests. Event analysis asks: Has something changed from baseline? Trend analysis asks: How quickly is it changing across visits? 1

Check whether the change repeats and fits defect location and visual function. Age and remaining life expectancy help assess future visual risk. Do not delay an urgent clinical decision just to obtain a perfect test series. 1

Target pressure is the range we aim to reach with treatment. It is not a diagnostic boundary. Set it according to stage, previous pressure, rate of worsening, life expectancy and treatment burden. 1 17

For POAG requiring treatment, the AAO benchmark starts with at least a 25% reduction from untreated baseline. Arithmetic example: a 25% reduction from 24 mmHg gives 18 mmHg. This is not a ready-made target for everyone. More severe or rapidly worsening damage may need a lower range. 1 17

Below 12 mmHg or single-digit targets may be justified in selected progressing eyes. They do not apply automatically to every advanced or normal-tension glaucoma. 1 17

NICE NG 81 uses 24 mmHg for certain OHT referral and treatment decisions. Disc or field damage may also justify referral regardless of that value. 3

For OHT, NICE also considers the risk of visual impairment over the person's lifetime. Thus 24 mmHg is neither a universal disease boundary nor a target. This UK recommendation is not a DOG/BVA rule. 3

The DOG/BVA S 2 e guideline 045-015 of 3 March 2024 addresses risk factors and early detection. A screening interval for a healthy person at risk is not automatically suitable for established or progressing glaucoma. 2 3

Distinguish recommendation strength from supporting evidence. A strong recommendation may rest on limited data and expert consensus. NICE lists CCT measurement during assessment; this does not mean applying an IOP correction formula. 2 3

The AAO Glaucoma Summary Benchmarks dated December 2025 were read from the supplied nine-page PDF. They summarise PPP recommendations; they are not the full PPP with all rationale and supporting evidence. 17

At follow-up, ask about new symptoms, systemic disease, drop side effects and the last administration time. Check acuity, anterior segment, IOP, disc, RNFL/macula and visual fields. Repeat CCT after events that alter corneal thickness, such as refractive surgery. 17

Repeat gonioscopy with suspected closure, a shallower anterior chamber, angle abnormalities or unexplained IOP change. With progression despite reaching target, AAO also advises assessing missed pressure fluctuations and actual drop use before revising target and treatment. 17

The table gives approximate follow-up intervals for established POAG. It comes from the consensus-based AAO benchmark, not a trial proving better vision for each interval. 'Controlled for ≤6 months' refers to the duration of control so far, not the next appointment. 17

High lifetime risk or advanced damage may require shorter intervals. Acute findings and postoperative care follow their own urgency. 17

AAO 2025: approximate follow-up intervals for established POAG
SituationMild to moderateSevere
At target, no progression; controlled for ≤6 months3–6 months3–4 months
At target, no progression; controlled for >6 months6–12 months3–6 months
At target, but progressing1–3 months1–2 months
Not at target and progressing1–2 months1–2 months
Not at target, no progression1–6 months1–3 months

17

Study appraisal: OHTS 2002 – what pressure lowering demonstrates

OHTS 2002: 1,636 people aged 40–80 were randomly assigned to drops or observation. No glaucomatous damage was present at baseline. IOP was 24–32 mmHg in one eye and 21–32 mmHg in the other. The full text was read. 4

Result at 60 months: Estimated POAG incidence was 4.4% with drops and 9.5% with observation. HR 0.40; 95% CI 0.27–0.59. The absolute difference is 5.1 percentage points, giving an NNT of about 20 over five years for this study group. 4

Meaning: About 20 comparable people would need five years of treatment to prevent one additional POAG onset. The endpoint was disease onset, not blindness. A person with a different baseline risk has a different absolute benefit. 4

Masking: Participants and treating doctors knew the group. Reading centres and the endpoint committee did not. Intention-to-treat analysis kept participants in their original assigned groups. 4

Missing data: Inactive participants were 10.9% with medication and 10.2% with observation. Assignment deviations affected 4.9% and 5.1%. Similar attrition rates do not rule out missing-data bias. 4

The 9.5% comes from a Kaplan–Meier estimate of risk at 60 months. This method accounts for different observation times. It is not the crude proportion of all events across the whole observation period: 89/819 =10.9%. 4

In practice: Not every person with OHT has 9.5% risk. Pressure of 24 mmHg or more alone does not require treatment. 4

09Using findings to decide the next step

With a red OCT sector, first ask: Is the original image correct? Does the defect fit the disc or visual field? If not, check segmentation and reference data before diagnosing glaucoma. 1

With confirmed worsening despite 'normal' pressure, reassess measurement conditions, pressure over time, risks and target. The goal is to preserve vision and quality of life; a stable device number alone is insufficient. 1

OCT angiography, iris vessel density, implantable sensors and AI are being studied as additional tools. A good research result does not yet show that patients keep useful vision longer. 1

AUC describes how well a test separates groups in a study. It proves neither better treatment nor less vision loss. Association or more frequent measurements also cannot establish this. The separate research briefings are retained; each study still needs appraisal. 1

Connect the examination to a decision
StepWhat to checkConsequence
HistoryRisks, prior findings, symptoms, treatment and adherence.Determine pre-test probability and urgency.
DiscSize, rim, haemorrhage, myopia and comparison images.Substantiate and document structural suspicion.
OCTRaw images, segmentation, registration, normative data and floor.Distinguish artefact from a plausible damage pattern.
IOP / CCTTechnique, cornea, device, timing and medication.Interpret pressure in context; avoid spurious precision from correction formulae.
GonioscopyOpenness, pigment, PAS, vessels and dynamic response.Clarify mechanism; imaging complements clinical examination.
Visual fieldReliability, repeatability, topography and central function.Confirm functional damage and quantify trajectory.
Follow-upMultiple comparable tests, progression, lifetime risk and target pressure.Individualise monitoring intervals and treatment.

1 2 3

Why does an OCT colour code not establish progression?

The colour compares with a reference database. Progression needs comparison with earlier tests. Open the original images: same location, good quality, correct layer boundaries? Then check whether the defect fits the disc and field.

What should we do with a new disc haemorrhage despite an unremarkable single pressure?

Photograph the haemorrhage. Check other causes and compare OCT and fields with earlier results. Reassess pressure over time and risks. Then decide whether an earlier review or a different target is needed. One normal reading does not exclude damage.

10References and original lectures

Numbers link to sources. A guideline recommendation differs from a single study result. Each entry states which text was read and which limits remain.

  1. 1
    EGS. Terminology and Guidelines for Glaucoma, 6 th edition (2025). DOI: 10.1136/bjophthalmol-2025-egsguidelines.

    Full guideline inspected: clinical diagnosis, tonometry, disc, OCT, gonioscopy, visual field and individual target pressure. Published 8 September 2025.

  2. 2
    DOG/BVA. S 2 e Bewertung von Risikofaktoren für das Auftreten des Offenwinkelglaukoms, AWMF 045-015 (03.03.2024).

    Full text inspected. Risk factors and early detection; not a comprehensive treatment guideline. Distinguish recommendation strength from evidence level.

  3. 3
    NICE NG 81. Glaucoma: diagnosis and management (updated 26.01.2022).

    Current recommendations inspected, sections 1.1.4–1.1.5, 1.2.1 and 1.4.4. 24 mmHg is a context-dependent threshold, not a universal diagnostic or target pressure.

  4. 4
    Kass et al. Ocular Hypertension Treatment Study (2002). PMID: 12049574.

    Original full text, including methods and Table 3, inspected on 3 October 2026. Open treatment with masked endpoint assessment; distinguish Kaplan–Meier risk from crude event proportions.

  5. 5
    Brandt et al. Adjusting IOP for CCT does not improve OHTS prediction models (2012). PMID: 21705084.

    Original full text on the trial website inspected on 3 October 2026, methods and Table 3. Model reanalysis, not a randomised comparison of correction strategies; CCT remains clinically informative.

  6. 6
    Bowd et al. Estimating OCT structural measurement floors (2017). PMID: 27914978.

    Original full text inspected 04.10.2026: 41 eyes had five measurements over five weeks; an independent 87-eye group had only two measurements over a mean 1.8 years. Spectralis/SALSA; floors and area proportions are not universal thresholds or diagnostic sensitivities.

  7. 7
    Leung et al. Retinal nerve fibre layer imaging: impact of normal ageing (2012). PMID: 22264886.

    Primary abstract accessible: 100 healthy participants cross-sectionally, 35 longitudinally, mean 30 months. Full text unavailable; do not derive an individual ageing correction.

  8. 8
    Jonsson et al. Optic disc grey crescent in the Reykjavik Eye Study (2005). PMID: 15615743; PMCID: PMC 1772477.

    Full text inspected; population-based cross-sectional study, 1,012 right eyes. Morphological variant, not a progression study.

  9. 9
    Budenz et al. Detection and prognostic significance of optic disc haemorrhages during 13 years of OHTS (2017). PMID: 27832941.

    Prospective OHTS cohort analysis; original full text inspected 05.10.2026: annual photographs, median 13 years and a time-dependent association. Baseline haemorrhage was excluded, transient events were probably underestimated and haemorrhage was the only intercurrent factor modelled; not a randomised post-haemorrhage treatment strategy.

  10. 10
    FDA. SENSIMED Triggerfish De Novo review, DEN 140017.

    Original regulatory assessment inspected. Pressure-related dimensional patterns; no direct IOP output in mmHg.

  11. 11
    Influence of body weight on IOP in the water drinking test (2018). PMID: 30067521.

    Original protocol: 800 mL, 1,000 mL or 10 mL/kg. Does not support a four-litre rule or self-testing; not a universal routine substitute.

  12. 12
    EGS. Terminology and Guidelines, 4 th edition; Figure 1.11, disc-size assessment.

    Historical full text inspected, Figure 1.11: 90-D lens correction factors 1.36–1.59 depending on model. Historical clinical calibration, not a universal current device specification.

  13. 13
    Volk. Slit lamp lenses: manufacturer specifications, 90 D.

    Original technical document visually inspected. Image magnification and laser spot factor are different quantities; not a diagnostic outcome study.

  14. 14
    Liu et al. Artifacts in Spectralis OCT RNFL imaging (2015). PMID: 25498118; PMCID: PMC 4423408.

    Retrospective artefact analysis. Quality score 15 belongs to the studied Spectralis context; raw-image review remains necessary.

  15. 15
    A Critical Discussion of the Structure–Function Relationship in Glaucoma (2019). PMID: 31688445.

    Methodological discussion; abstract inspected, not an original randomised diagnostic study. Supports distinguishing histological cell counts, OCT layers and function.

  16. 16
    AAO. Primary Open-Angle Glaucoma Preferred Practice Pattern 2025; journal version online 09.02.2026.

    Current PPP journal article identified; its full text remains unread. The official AAO Summary Benchmarks of December 2025 were read separately (reference 17); recommendations from them are attributed to the benchmark.

  17. 17
    AAO. Glaucoma Summary Benchmarks for Preferred Practice Pattern Guidelines (December 2025), pp. 1–9.

    Supplied original PDF read in full on 6 October 2026: diagnosis pp. 3/6, follow-up and consensus interval table p. 5, angle closure pp. 8–9. Official summary, not the full PPP. The 2020 in the stable AAO URL is not the edition date of the linked 2025 benchmark.

Lectures as the starting point

The original three teaching videos are retained. They do not replace verified primary evidence or guidelines.