Mohammad AskarPublic working copy · medical review pending

Visual fields in glaucoma · 167

Visual fields in glaucoma: quality, pattern and change

Check the test conditions before interpreting loss; combine repeated fields with optic nerve and OCT findings.

Public working copy · medical review pending. Source, translation and medical checks remain open. Read the limitations in the text and references. These educational materials do not replace individual medical assessment. Published: 2026-11-18T19:45:52+01:00. Internal project references lead to this notice. They are not publicly accessible primary sources; the source limitations remain stated in the text.

MD sign, grid, strategy, quality and repeatability determine how a change can be interpreted; one cutoff does not decide surgery.

Terminology and draft status

Terminology review is not complete. Selected actual WHO UMD queries and their limits are recorded in terminology.md; they do not approve the whole translation.

01What perimetry shows

Perimetry tests where a patient can detect light. The patient looks at a fixed target and presses a button when a light appears. The result therefore also depends on understanding the task and responding attentively. 7 19

In glaucoma, the test helps us identify field loss and follow its change. It does not measure every ability needed for reading, walking or driving. 7 19

Light sensitivity is highest at the fovea and decreases towards the periphery. The teaching images of a “hill of vision” or an “island of light in a sea of darkness” describe this pattern. 15

Rough monocular field extents are about 60° nasal, 50–60° superior, 70° inferior and up to 100° temporal. The commonly taught 60° superior extent is an approximation. Outer limits depend on anatomy, stimulus and method. They do not tell us whether a 24-2 is normal: that program tests only a central part of the field. 15

In kinetic perimetry, a light moves. We mark where the patient first detects it. Joining these locations for the same stimulus produces an isopter. Goldmann is the classic example. 7 19

In static perimetry, test locations stay fixed and brightness changes. Standard Automated Perimetry (SAP) is automated static threshold testing with white stimuli on a white background. It is the standard approach in glaucoma. Kinetic testing can help with advanced loss, neurological questions or difficulty performing SAP. Small early glaucoma defects may be missed between isopters. Humphrey and Octopus name instrument families, not one particular strategy. 7 19

A scotoma is a localized area of field loss. With a relative scotoma, the patient still detects a brighter light. With an absolute scotoma, the patient does not detect even the strongest appropriate stimulus in that test. This term applies to the testing conditions and does not establish absence of every form of vision there. 9

A positive scotoma is noticed as a disturbance. A negative scotoma may go unnoticed. Do not ask only about black patches: glaucoma loss need not appear as a black patch or black tunnel. 9

The threshold is the light intensity a patient can just detect at a test location. A threshold procedure approaches it with brighter and dimmer stimuli. Historical Humphrey Full Threshold starts with 4 dB steps and changes to 2 dB after the first reversal between seeing and not seeing. 6 7 22

4–2–1 dB is not a general rule for all instruments and strategies. SITA means Swedish Interactive Threshold Algorithm and uses a different algorithm to estimate the threshold. Strategy and grid answer different questions: the strategy determines how we measure; the grid determines where. 6 7 22

The visual field also matters for driving and some occupations. A glaucoma 24-2 alone is insufficient for a complete driving-fitness assessment. Check the required method and standards for the country, licence class and occupation. In Germany, visual requirements are specified, among other provisions, in FeV Annex 6. 14

02Preparing the test

Check name, eye, date, grid and strategy before reading the report. OD means right eye and OS left eye. The blind spot helps orientation, but rely on the eye label. A triangular border marker can have different meanings across reports; do not treat it as a universal nasal marker. 21

Explain the task before starting: “Look at the fixed target. Press as soon as you see a light. You do not have to see every light; some are deliberately very faint.” Demonstrate a few stimuli if needed. 7 21

The patient should not guess according to a rhythm. Offer a pause and remain available for questions. Record difficulties and observations. A useful test starts with clear instructions, not blame for poor cooperation. 7 21

Use correction for the testing distance, including the required near addition. Blur can reduce measured sensitivity. A poorly positioned trial-lens rim can create apparent field loss, sometimes with an arcuate shape. 19 21

Check posture, eyelid and lens distance. Correct the cause and repeat when needed before assuming new damage. Exact lens settings depend on the instrument and its instructions. 19 21

Document pupil size, medication and opacities such as cataract or vitreous opacity. These conditions can change sensitivity. Keep them as comparable as possible during follow-up. 16

2–6 mm is sometimes taught as a favorable pupil range, but it is not a universal cutoff for accepting or rejecting a field. Dilation can also change the measurement: in a small study of 18 healthy participants, MD worsened by a mean 0.83 dB after tropicamide. The study does not define an optimal range for glaucoma patients. Do not change pupil size or medication solely because of the 2–6 mm figure. 16

The task is unfamiliar during the first test. Sensitivity and reliability may improve over subsequent tests: this is the learning effect. Use several interpretable fields to establish baseline. Do not automatically discard the first two or three tests. 7 8 19

Learning: Heijl/Bengtsson 1996: Prospective: 25 inexperienced glaucoma patients, five 30-2 Full Threshold sessions. Mean MD improved 2.81 dB between sessions 1 and 2 (p < 0.001). Small historical study, not a modern SITA cutoff. 8

03Reading the report

Start with numerical thresholds and deviation maps, not just grayscale. Raw thresholds show sensitivity at each location. Total Deviation (TD) shows the difference from the age-adjusted normal value. 7

Pattern Deviation (PD) also adjusts for a general shift in sensitivity. Local defects often become more apparent. PD is not simply a “cataract filter.” Compare TD, PD and raw values: the maps answer different questions. 7

Fixation losses mean the patient responded during blind-spot testing. This may reflect poor fixation, but it can also result from an incorrectly located blind spot. 4 7

False-positive responses (FP) are responses without a detectable stimulus. False-negative responses (FN) mean a brighter stimulus is missed at a previously responsive location. The often quoted 20% fixation-loss or 15–30% FP figures are not a universal pass/fail test. Also review gaze monitoring, unusually high thresholds, the pattern and observations during testing. 4 7

Many FP responses can inflate sensitivity, making the field look better than it is. An unusually bright report is sometimes called a “white scotoma.” That appearance alone does not establish that the entire test is unusable. 4 7 10

Many FN responses do not prove inattention. Detection fluctuates more at damaged locations; FN can therefore be high, particularly in moderate or severe disease. Decide whether to repeat from the whole report. A doubtful field should not trigger unchecked escalation of treatment. 4 7 10

The Glaucoma Hemifield Test (GHT) compares matching zones in the upper and lower field. It primarily looks for asymmetry. It can report “within normal limits” even with approximately symmetric glaucomatous defects. 7

Pattern Standard Deviation (PSD) describes how unevenly sensitivity is distributed. GHT and PSD help interpret the field. Neither diagnoses disease by itself. 7

Three adjacent abnormal points are more suspicious than three scattered ones. Common Anderson-type teaching criteria are: GHT outside normal limits; or at least three adjacent non-edge PD points at p <5%, including one at p <1%; or abnormal PSD at p <5%. 7 18 19 20

These p-values describe how unusual the result is compared with the normal database. They are not the probability that the patient has glaucoma. Assess the pattern, repeatability, disc/OCT and alternative causes. Confirmation is especially important for a small doubtful defect; a clear defect with matching clinical findings is a different situation. 7 18 19 20

Reliability: Yohannan 2017: Retrospective modeling: 10,262 fields, 1,538 eyes, 909 people. Error effects varied by severity. Limitations: observational data and modeled MD, not a universal quality boundary for all instruments. 4

SITA-Faster: Phu/Kalloniatis 2022: Analysis of 2,320 tests/1,160 eyes within a visit. MD rose by 0.09 dB per additional FP percentage point above 12%. This modeled finding is not a new automatic 12% rejection cutoff; it does not validate longitudinal treatment decisions. 10

0424-2, 24-2C or 10-2?

The grid determines where sensitivity is measured. 24-2 tests a wider central area with relatively widely spaced points. 10-2 samples the central 10° more densely. 7 19

Add 10-2 when you need to examine a paracentral defect, threatened fixation or a corresponding macular finding more closely. It can also help with a small residual central field. Do not routinely replace wider testing with 10-2. 7 19

24-2C adds ten central points to 24-2, giving 62 points in total. It combines wider coverage with more information about the macula. EGS describes it as a possible compromise and notes limited supporting literature. 3 7 17

24-2C is not a complete replacement for 10-2. When the exact extent of a central defect matters, the denser 10-2 may provide additional information. 3 7 17

SITA Faster is a measurement strategy; 24-2C is a grid. Use the same instrument, grid and strategy during follow-up where possible. After a change, check which results the software can compare. An added 10-2 starts a separate central comparison series; earlier wider fields remain important. 7 19

24-2C versus 10-2: Phu/Kalloniatis 2021: Prospective cross-section: 188 eyes, 131 glaucoma and 57 suspect. 10-2 detected more defect clusters. Median 201 versus 154 seconds; different SITA strategies limit the timing comparison. No longitudinal or quality-of-life evaluation. 3

05Assessing pattern and structure together

Damage to superior nerve-fiber bundles appears in the inferior field, and vice versa. A nasal step is a sensitivity difference above and below the horizontal meridian on the nasal side. An arcuate scotoma follows a curved nerve-fiber pattern. A paracentral defect lies near fixation. 7 19

These patterns fit glaucoma but do not prove it alone. With advanced loss, a small central island and temporal remnant may remain. Not every eye follows exactly this pattern. 7 19

Compare the field with the disc and OCT. OCT shows structure; the field shows measured function. RNFL means retinal nerve fiber layer; macular ganglion-cell measurements add information about the center. 7 19

For example, superior RNFL thinning fits an inferior field defect. If they do not match, check scan quality, field quality and exact location. Structure and function need not correspond perfectly at every time point. 7 19

If the defect does not fit glaucoma, consider other causes: retinal disease, high myopia, another optic neuropathy or a visual-pathway lesion. A defect aligned with the vertical meridian or acute field loss deserves particular attention. 19

Clarify onset and accompanying symptoms. Sudden new loss or new neurological symptoms require urgent assessment; do not reflexively treat them as routine glaucoma progression. 19

06Assessing change over time

Event analysis compares a new field with baseline: is the change larger than usual test-to-test variation? Trend analysis evaluates several fields over time: how quickly is function declining? 7 19

An event therefore identifies unusual change; a trend estimates its speed. For both, first check quality and comparable testing conditions. A new event signal usually needs confirmation. 7 19

Humphrey Mean Deviation (MD) summarizes the average sensitivity difference from the age-adjusted norm. MD becomes more negative as loss increases. For example, −4 to −6 dB over two years gives an arithmetic rate of −1 dB/year. 5 7

This is initially just the calculation between two values. More fields are needed for a reliable rate. Read the estimate with its confidence interval and measurement variability: how precise is it? 5 7

In Octopus, MD can mean Mean Defect: a larger positive value then describes greater loss. Other displays also exist. Check the index name, legend and software view before comparing instruments. 7 11

PSD or Octopus Loss Variance (LV) should not be used as a global progression rate: they can rise initially and fall again as damage advances. A smaller value then does not mean recovery. 7 11

0.5 and 1.0 dB/year are useful examples of the magnitude of a loss rate. They are not universal thresholds for surgery. The same rate is more threatening with little remaining field or a long expected lifespan than with substantial remaining function and low lifetime risk. 7

Ask: where is the defect, how much function remains, how certain is the rate, and will this speed substantially impair vision during the patient’s lifetime? 7

With confirmed progression, reassess target pressure, pressure over time and actual treatment use. Discuss the next intervention’s benefits and risks. Further pressure reduction may be needed even when the previous target has been reached. 7 19

Clarify an uncertain measurement. Credible rapid deterioration calls for timely action; repeated waiting must not delay necessary treatment. 7 19

Testing frequency: Chauhan 2008: Modeling with empirical variability: three tests yearly can detect an overall 4 dB MD loss in two years with average variability. Six tests are thus an explained example, not a DOG mandate or universally optimal schedule. 5

07Stage and remaining field

For HFA Mean Deviation, −15 dB indicates more global loss than −5 dB. A value below −12 dB therefore belongs to the severe category in the MD classification used below. 12 7

The table states its boundaries explicitly. It is a simplified MD-based guide. A global average does not adequately describe the location of a central defect. 12 7

A small defect near fixation can matter even when MD is only mildly negative. Assess location, depth and extent, and consider both eyes’ function. A stage label alone does not determine a change of test program. 7 19

With widespread loss, PD can show less damage than TD. Adjustment for general sensitivity can mask part of the loss. A better-looking PD does not prove recovery; this is sometimes called “pseudo-improvement.” 7 13

Read TD, raw thresholds and the remaining central points together. Neither “PD only” nor “TD only” is a safe universal rule. In advanced loss, 10-2, a larger Goldmann size V stimulus or kinetic testing may help. A change may require a new comparison baseline. 7 13

In the simplified HFA Mean Deviation convention, mild is MD ≥−6 dB, moderate −12≤MD<−6 dB and severe MD<−12 dB. The original “severe MD >−12” sign is reversed. These global bands do not replace fixation/location criteria or require a universal change to 10-2. 23

08Using guidelines accurately

DOG/BVA guideline 045-015, dated 2024-03-03, is an S2e guideline on risk factors and early detection. It is not comprehensive S3 guidance for monitoring established glaucoma. Attribute testing-frequency recommendations to their actual source. 2

The AAO PPP distinguish a follow-up visit from a field test. A visit reviews pressure, symptoms and treatment, among other findings. Field testing has its own schedule. 19 20

For example, in mild to moderate glaucoma with target pressure reached, no progression and control for no more than six months, AAO lists 3–6 months for visits but 6–12 months for field testing. These tables are consensus-based. A new suspicious finding, baseline establishment or rapid progression may require earlier tests. 19 20

EGS 6 presents three SAP tests yearly during the first two years as a proposed approach to identify rapid progression early. Its treatment chapter particularly places this in patients with significant damage. Not every patient needs such a dense schedule. Frequency can be adjusted after estimating the rate. 5 7 19

For ocular hypertension without glaucoma damage, EGS states that testing every one to two years is often sufficient. New suspicious findings or other risks can change the plan. Six tests in two years are thus a targeted approach to estimating change, not a universal German requirement. 5 7 19

09Practical workflow and revision

Read the report in this order: are patient, eye and test correct; are conditions and responses plausible; where is the defect; does it fit the disc and OCT; is it new or changing? 7 19

Compare several earlier fields. Record why you repeat, add another grid or reassess treatment. A software alert prompts assessment; it is not a finished treatment plan. 7 19

Avoid these common mistakes: reading only grayscale; dismissing many FN as poor cooperation; assuming 24-2C answers every central question; interpreting better PD as recovery; reading MD without its sign convention; inferring surgery directly from one alert. 7 19

Practical workflow and revision
TopicOriginal teaching pointEvidence-based interpretationPractical next stepCaveat
ReliabilityFP 15–30%, fixation loss 20% examplesNot universal pass/fail thresholdsReview gaze, thresholds and repeatabilityFN rises with advanced damage; white scotoma alone does not reject the whole test.
Early lossAnderson PD/GHT/PSD criteriaSuspicion requiring clinical correlationRepeat uncertain clusters and compare RNFLMyopia/retinal disease can confound; p-value is not diagnostic probability.
ArtefactsRefraction, pupil 2–6 mmDocument testing conditions; not a universal pupil cutoffCheck lens rim, blur, lid and mediaLens rim may mimic an arcuate defect.
Strategy24-2C adds 10 central pointsNot a complete replacement for 10-2Choose central density for fixation riskMD <−12 dB alone does not mandate abandoning wider testing.
ProgressionSix tests in 2 years; 0.5–1.0 dB/yearSelected baseline/risk approach, not universal DOG S3 mandateCompare event/trend, precision and lifetime riskLearning can imitate improvement; a single alert does not decide surgery.
PD pseudo-improvementDiffuse loss can distort PD adjustmentNot evidence of recovery; not exclusively terminalRead TD, raw thresholds and remaining central functionNeither PD-only nor TD-only is universally sufficient.
How do you distinguish real progression from artefact?

MD sign, grid, strategy, quality and repeatability determine how a change can be interpreted; one cutoff does not decide surgery.

When is assessment urgent?

Seek immediate medical help for sudden new field loss, rapid visual deterioration, severe eye pain or new neurological symptoms. Do not wait for planned glaucoma follow-up.

References and access status

  1. 1
    AAO · Glaucoma Summary Benchmarks · December 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: Owner-supplied PDF, pp3–6 read. Consensus intervals; complete PPP not read.

  2. 2
    DOG/BVA · S2e 045-015 · 2024-03-03

    Historical Review 006 documented access of 7–10 October 2026 reused: Official text: early-detection risks, not comprehensive follow-up guidance.

  3. 3
    Phu/Kalloniatis · 24-2C versus 10-2 · 2021

    Historical Review 006 documented access of 7–10 October 2026 reused: Journal text/abstract checked; cross-section and different strategies.

  4. 4
    Yohannan et al. · Visual-field reliability · 2017

    Historical Review 006 documented access of 7–10 October 2026 reused: Indexed original text checked; retrospective modeling.

  5. 5
    Chauhan et al. · Rates of visual-field change · 2008

    Historical Review 006 documented access of 7–10 October 2026 reused: Modeling/practical perspective; not a randomized schedule trial.

  6. 6
    Glass et al. · Historical 4–2 dB full-threshold simulations · 1995

    Historical Review 006 documented access of 7–10 October 2026 reused: Title/indexed information checked; historical procedure.

  7. 7
    European Glaucoma Society · 6th edition · 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: EGS 6: complete perimetry section pp103–109 and testing-frequency passage p160 read. The staging table has problematic labels/signs; our HFA table explicitly uses negative MD.

  8. 8
    Heijl/Bengtsson · Perimetric experience · 1996

    Historical Review 006 documented access of 7–10 October 2026 reused: Full abstract read; small historical prospective study.

  9. 9
    Crabb et al. · Patient perception of visual-field loss · 2013

    Historical Review 006 documented access of 7–10 October 2026 reused: Indexed original description read; perception, not a treatment study.

  10. 10
    Phu/Kalloniatis · SITA-Faster reliability · 2022

    Historical Review 006 documented access of 7–10 October 2026 reused: Full abstract and figure legends read; within-visit comparison, not longitudinal validation.

  11. 11
    Haag-Streit · EyeSuite glaucoma analysis

    Historical Review 006 documented access of 7–10 October 2026 reused: Manufacturer description read; index/software, not independent efficacy appraisal.

  12. 12
    MD categories in original clinical research · 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: Indexed methods: explicit MD bands. Not complete HPA staging.

  13. 13
    World Glaucoma Association · Consensus 8 · 2011

    Historical Review 006 documented access of 7–10 October 2026 reused: Official consensus text read; historical guidance, not EGS 6.

  14. 14
    Fahrerlaubnis-Verordnung (FeV), Anlage 6: Anforderungen an das Sehvermögen

    Historical Review 006 documented access of 7–10 October 2026 reused: Official regulatory text checked on 7 October 2026; cited for driving-fitness context, not as a complete fitness assessment.

  15. 15
    Phu et al. · The value of visual-field testing in the era of advanced imaging · 2017

    Historical Review 006 documented access of 7–10 October 2026 reused: Indexed review checked; outer extents are rough monocular reference values and depend on anatomy, stimulus and method.

  16. 16
    Lindenmuth et al. · Effect of pupillary dilation on automated perimetry · 1990

    Historical Review 006 documented access of 7–10 October 2026 reused: Original study in 18 healthy participants: MD worsened by a mean 0.83 dB after tropicamide. Small normal sample; not a universal pupil-size limit.

  17. 17
    ZEISS · Humphrey Field Analyzer 3 · 24-2C

    Historical Review 006 documented access of 7–10 October 2026 reused: Official manufacturer page checked: 24-2C adds ten central locations and contains 62 points. Product description, not an independent effectiveness assessment.

  18. 18
    Visual-field cluster repeatability in healthy, suspect and early-glaucoma eyes · 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: Retrospective cohort of 197 eyes/103 people; examines repeatability of initial clusters across two subsequent fields. Not an independent diagnostic rule.

  19. 19
    AAO · Primary Open-Angle Glaucoma PPP · approved 6 September 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: Supplied full text: pp26–27 and 57–60 read; Tables 7/8 checked separately. Approval date taken from the document, not its “2026” filename.

  20. 20
    AAO · Primary Open-Angle Glaucoma Suspect PPP · approved 6 September 2025

    Historical Review 006 documented access of 7–10 October 2026 reused: Supplied full text: p21 (printed P124) read; confirmation of new defects and central testing.

  21. 21
    Haag-Streit · EyeSuite i9 static examination quick guide · 2019-11

    Historical Review 006 documented access of 7–10 October 2026 reused: Official quick guide read: eye, trial lens, positioning, parameters and monitoring. Instrument-specific; not a substitute for full instructions.

  22. 22
    Development and evaluation of a linear staircase strategy · 2008

    Historical Review 006 documented access of 7–10 October 2026 reused: Indexed methods: historical HFA Full Threshold simulation with 4 dB then 2 dB steps. Direct full-text access blocked; not a current instrument manual.

  23. 23
    Immutable original export; visual-field examination guide

    Complete export read; original attribution and any recording/transcript or appointment not independently verified.

Original lecture link