Ophthalmology Question Quality Review
Executive Summary
This review covers a candidate sample of 100 validated non-gold questions drawn from a pool of 6,703 Ophthalmology items. The sample was evaluated against 8 benchmark questions and 12 recent PYQs as the quality bar.
The most pressing finding is a severe Blooms-level compression toward recall: 44 of 100 candidate questions sit at Blooms Level 1 and 45 at Level 2, leaving only 11 questions at Levels 3–5 combined. The benchmark and PYQ sets, by contrast, include meaningful proportions of Blooms 3–4 items that require clinical reasoning, differential diagnosis, and application. This mismatch means the candidate pool, if deployed as-is, will systematically under-challenge candidates and fail to replicate the cognitive demand of actual INICET/NEET-PG papers.
Beyond the Blooms problem, the reviewed set contains:
- At least 5–6 items with factual errors or unsafe keys that could actively mislead candidates
- Multiple broken-delivery items including one question with no correct answer marked and one with options that are not answers to the question asked
- Significant topic misclassification (cataract questions filed under Contact Lens, a lens-physiology question filed under Contact Lens, a uveitis question filed under Diseases of the Retina)
- A large cluster of low-value trivia — pure anatomy recall, normal values, and eponym-matching — that has no realistic exam precedent in recent PYQ sets
- Moderate repetition across conjunctival inflammation questions and glaucoma pharmacology questions
The overall disposition recommendation for the 100-item candidate sample is approximately: Keep ~30, Fix ~35, Disable ~35.
What Good Looks Like
The benchmark and PYQ sets establish a clear quality bar. The following features characterise the gold-standard items in this subject:
Clinical vignette framing at appropriate difficulty. Items like 56ec5ba8 (delayed post-cataract endophthalmitis with organism identification), b8d06c45 (superior quadrant loss + RHD → BRAO), and e12770bf (uveitic glaucoma + drug to avoid) embed the testable fact inside a realistic patient scenario. The candidate must reason through the clinical context, not just retrieve a label.
Discriminating distractors. In dd64e128 (corneal graft storage + suture), all four options are plausible combinations of real materials; the question cannot be answered by elimination. In 9c35af8a (pterygium recurrence rate), the incorrect option — 95% recurrence — is a common misconception, making it a genuine trap for under-prepared candidates.
Appropriate Blooms distribution. The gold set spans Blooms 1–4 within a single subject. Even the Blooms-1 items (a7d80cf3, 36c6211f) test facts that appear repeatedly in recent PYQs and carry high yield. Blooms-1 items that test obscure trivia with no PYQ precedent are absent from the gold set.
Image integration where clinically meaningful. Items dd3e6198 (phenylephrine blanching → episcleritis vs scleritis) and 2387f1f4 (surgical step identification) use images to test visual pattern recognition — a skill directly tested in INICET. These are not image-dependent for the sake of novelty; the image is the test.
Tight, unambiguous stems. Every gold-standard item has a single defensible correct answer. Distractors are wrong for a specific, teachable reason, not merely because they sound unfamiliar.
The candidate sample falls short on all five dimensions above in a substantial fraction of items.
Main Issue Categories
1. Wrong Key or Factually Unsafe
Why this pattern is bad
A wrong key is the most serious quality failure in a question bank. It actively teaches incorrect information, penalises well-prepared candidates, and — if deployed in a scored test — creates legal and reputational risk. These items must be resolved before any deployment.
How it shows up
In this sample, wrong-key problems cluster around three sub-types: (a) the correct answer is marked on a factually incorrect option, (b) the stem uses imprecise language that makes the intended key ambiguous, and (c) the key is defensible only under an outdated or minority-source interpretation.
Affected items
532a4a83 — Actions of the superior oblique muscle
The marked correct answer is "Extorsion." The superior oblique performs intorsion (incycloduction), depression, and abduction — not extorsion. Extorsion is the action of the inferior oblique. This is a straightforward factual inversion and is unsafe to deploy. Disable.
b4bdff6c — Wavelength of a YAG laser
The stem asks "What is the wavelength of a YAG laser?" The marked correct answer is "Colorless." Wavelength is a physical measurement in nanometres (Nd:YAG = 1064 nm, which is infrared and invisible to the human eye). "Colorless" is not a wavelength; it is a colloquial description of the beam's appearance. The question conflates two different properties and the key is not a valid answer to the question as asked. Disable.
66b1fba9 — Which instrument is used for fundus examination?
The stem asks for an instrument. The options are: "Examination of the optic disc," "Fundus Examination," "Corneal Examination," and "Examination of the Angle of Anterior Chamber." None of the options is an instrument. The marked correct answer is "Fundus Examination," which is a procedure, not an instrument. The question is structurally broken and factually incoherent. Disable.
92c5a370 — What is the refractive condition of the eye?
The stem asks "What is the refractive condition of the eye?" with no clinical data, no image reference, and no patient context. The marked answer is "Hypermetropia of 2 D." There is no basis in the stem for any answer. This appears to be a fragment of an image-dependent question where the image was lost. Disable.
43b35e0a — "Trucking of vessels"
The stem asks about "Trucking of vessels" — almost certainly a transcription error for "Trunking" or more likely "Boxcarring" (segmentation of blood column). The marked answer is CRAO, which is plausible for boxcarring. However, the misspelling "Trucking" makes the question ambiguous and potentially untestable. If the intended term is "boxcarring," the key is defensible; if it is a different phenomenon, the key may be wrong. Fix (verify intended term and correct spelling) or disable if the intended concept cannot be confirmed.
9e44b1d3 — Most common cause of neonatal conjunctivitis
The marked answer is Chlamydia. This is contested in the Indian PG literature. Older Indian textbooks (Khurana) list chemical irritation (silver nitrate) as the most common cause globally, while Chlamydia is the most common infectious cause. More recent sources and INICET-aligned teaching favour Chlamydia as the overall most common cause in the current era. The answer is defensible under current teaching but the question carries genuine ambiguity risk. Fix: add "infectious" qualifier to the stem, or add an explanatory note.
e2eabafe — "What does the given image show?"
All four options are marked as incorrect (no option has isCorrect: true). This is a broken key — no correct answer exists in the data. The question cannot be scored. Disable until corrected.
960caab3 — Papillae are seen in which conditions?
The marked answer is "All of the above" (Trachoma, Viral conjunctivitis, Spring catarrh). Papillae are a feature of trachoma and vernal/spring catarrh, but viral conjunctivitis characteristically produces follicles, not papillae. The "All of the above" key is factually incorrect. Fix: remove viral conjunctivitis from the options or change the key.
dd992460 — Ligneous conjunctivitis is caused by:
The marked answer is "Membranous conjunctivitis." Ligneous conjunctivitis is a type of membranous conjunctivitis (caused by plasminogen deficiency), not caused by it. The relationship is inverted. The question is also poorly framed — asking what a condition "is caused by" when the answer is a category it belongs to. Fix: reframe as "Ligneous conjunctivitis is a type of:" or disable.
Recommended disposition: Disable 532a4a83, b4bdff6c, 66b1fba9, 92c5a370, e2eabafe. Fix 43b35e0a, 9e44b1d3, 960caab3, dd992460.
2. Wrong Subject or Wrong Topic Placement
Why this pattern is bad
Misclassified questions corrupt topic-level analytics, cause incorrect questions to appear in topic-specific practice sets, and confuse candidates who are studying a specific area. At scale, systematic misclassification degrades the reliability of any topic-based performance reporting.
How it shows up
In this sample, the most common misclassification is cataract-related questions filed under "Contact Lens." There is also a uveitis question filed under "Diseases of the Retina."
Affected items
6333f470 — Rider's cataract / Anterior capsular cataract
Filed under: Contact Lens. Correct topic: Diseases of the Lens. Rider's cataract is a morphological variant of anterior capsular cataract. It has no connection to contact lens practice. Fix: reclassify to Diseases of the Lens.
23e05f2b — Surgery of choice for congenital cataract
Filed under: Contact Lens. Correct topic: Diseases of the Lens (or Ophthalmic Surgery). Congenital cataract surgery is a lens disease management question. Fix: reclassify.
987e3416 — Oil drop cataract / Galactosemia
Filed under: Contact Lens. Correct topic: Diseases of the Lens. Fix: reclassify.
da78d5cc — MIP-26 function in the lens
Filed under: Contact Lens. Correct topic: Diseases of the Lens or Basic Sciences as Related to Eye. MIP-26 (Major Intrinsic Protein) is a lens fibre membrane channel protein. It has no relationship to contact lens practice. Fix: reclassify.
f02dea9e — Causes of panuveitis (Ankylosing spondylitis as exception)
Filed under: Diseases of the Retina. Correct topic: Uveitis (or Ocular Manifestations of Systemic Disorders). Panuveitis is a uveal tract disease classification question. Ankylosing spondylitis causes anterior uveitis, not panuveitis — this is a legitimate testable fact, but it belongs in a uveitis topic, not retinal disease. Fix: reclassify.
01981595 — Causes of vitreous haemorrhage (Glaucoma as exception)
Filed under: Diseases of the Retina. This is borderline acceptable since vitreous haemorrhage is a retinal/vitreous topic, but the question is primarily about differential causes and could reasonably sit in Vitreous Diseases or Ocular Trauma. The current placement is defensible; flag for review rather than mandatory reclassification.
Recommended disposition: Fix (reclassify) 6333f470, 23e05f2b, 987e3416, da78d5cc, f02dea9e.
3. Broken Delivery (Missing Image, Malformed Options, Incomplete Stem)
Why this pattern is bad
Broken delivery makes a question unscorable or unanswerable regardless of the candidate's knowledge. Image-dependent questions without images are particularly harmful because they appear in scored tests and generate candidate complaints and score disputes.
How it shows up
Two distinct sub-types appear in this sample: (a) image-dependent questions where the image is absent or the stem references visual data that is not present, and (b) questions where the options do not answer the question asked.
Affected items
e2eabafe — "What does the given image show?"
The stem explicitly references "the given image." No image is present in the data. Additionally, as noted in Category 1, no option is marked correct. This item is doubly broken. Disable.
92c5a370 — "What is the refractive condition of the eye?"
As noted above, the stem contains no clinical data and no image reference. The question is unanswerable without a missing image or clinical vignette. Disable.
66b1fba9 — "Which instrument is used for fundus examination?"
The options are all procedures or examination types, not instruments. The question asks for an instrument (ophthalmoscope, slit lamp, etc.) but none of the options provides one. This is a malformed option set. Disable.
b4bdff6c — "What is the wavelength of a YAG laser?"
Options are colours/descriptions ("Colorless," "Red," "Green," "Blue"), not wavelengths in nanometres. The options do not answer the question as asked. Disable.
Recommended disposition: Disable all four items above. e2eabafe and 92c5a370 are also flagged under Category 1; the broken delivery finding reinforces the disable call.
4. Low-Value But Correct (Too Simple, Low-Yield, Trivia-Heavy, Weak Exam Relevance)
Why this pattern is bad
Questions that are factually correct but test pure anatomy recall, normal numerical values, or eponym definitions with no clinical application do not prepare candidates for the reasoning demands of INICET or NEET-PG. They inflate the apparent size of the question bank without adding discriminatory value. When the Blooms-1 proportion is already 44% of the candidate sample, adding more recall items worsens the cognitive profile of any test built from this pool.
How it shows up
This is the largest single category in the reviewed set. It appears as: (a) pure anatomy recall with no clinical hook, (b) normal value questions, (c) eponym-to-definition matching with no differential, and (d) single-fact pharmacology recall that could be answered by any first-year student.
Affected items (representative, not exhaustive)
a33931ae — Posterior segment defined by posterior surface of lens and zonules
Pure anatomical definition. No PYQ precedent. No clinical application. Blooms 1. Disable.
f6a41815 — Normal depth of anterior chamber (2–3 mm)
Normal value recall. No clinical context. Blooms 1. Disable.
09c7fc64 — Normal pH of tears (7.5)
Normal value recall. No clinical relevance to any PG-level question. Blooms 1. Disable.
c8489725 — Colour of iris in complete albinism (Pink)
Trivia-level recall. No differential reasoning required. Blooms 1. Disable.
41855fcb — Definition of Euryblepharon
Rare eponym definition. No PYQ precedent in the reviewed gold set. Blooms 1. Disable.
c76375a4 — Inferior oblique supplied by CN III
Basic anatomy. Covered in every first-year anatomy course. No ophthalmology-specific reasoning required. Blooms 1. Disable.
eec156c7 — Pilocarpine is a miotic used in glaucoma
Extremely basic pharmacology. Any candidate who has opened a pharmacology textbook knows this. No discriminatory value at PG level. Disable.
a66c9088 — Thickening of eyelashes is a side effect of latanoprost
This is a legitimate PG-level fact (hypertrichosis as a prostaglandin analogue side effect), but the question is framed as pure recall with no clinical context. It could be upgraded to a vignette. Fix: embed in a clinical scenario (e.g., patient on latanoprost for glaucoma notices eyelash changes).
8c552a1d — Soft contact lenses made of HEMA
Material science recall. Low yield for PG exams. Blooms 1. Disable.
21591838 — Vossius ring on anterior capsule of lens
This is a legitimate PYQ-level fact (post-traumatic iris pigment imprint on lens). However, the question is pure recall with no clinical context. It could be kept if the pool is thin on ocular trauma, but given the low-yield framing, Fix: add a trauma vignette context, or disable if gold coverage exists.
c8bf6a67 — Tonometry is commonly performed during ophthalmological examination
The question asks which investigation is "commonly performed during an ophthalmological examination." This is not a meaningful discriminator — all four options (pachymetry, tonometry, biometry, LASER interferometry) are performed in ophthalmological practice. The question tests nothing beyond the most superficial familiarity with the word "tonometry." Disable.
ac422e05 — Primary colours (Red, Green, Blue — "None of the above" is correct)
The marked answer is "None of the above" because all three listed options are primary colours. This is a trick question with no clinical relevance. The "None of the above" key is also a poor distractor format. Disable.
5b38cd02 — Examining eye made myopic in indirect ophthalmoscopy
This is a legitimate optics fact but is framed as pure recall. It has some PG relevance. Keep if the pool is thin on optics; otherwise low priority.
422b4baa — Retinoscopy = objective assessment of refractive state
Pure definition recall. Blooms 1. Disable.
Recommended disposition: Disable the majority of items in this category. Items with some PG relevance but poor framing (e.g., a66c9088, 21591838) should be fixed rather than disabled.
5. Repetitive or Duplicative Coverage
Why this pattern is bad
When multiple questions test the same narrow fact at the same cognitive level, they crowd out coverage of other high-yield topics and reduce the effective diversity of any test built from the pool. In a bank of 6,703 items, repetition at the sample level suggests the problem is likely more pronounced in the full pool.
How it shows up
In this sample, repetition is observed in three clusters: (a) vernal keratoconjunctivitis features, (b) glaucoma pharmacology, and (c) ERG wave origins.
Affected items
Vernal keratoconjunctivitis / Spring catarrh cluster:
062f867f— Trantas spots in vernal conjunctivitisc32b432c— Cobblestone appearance in spring catarrh77297f61— Features of VKC (Tantra's Spot — note misspelling of Trantas)0dba68b6— Spring catarrh = Type I hypersensitivityd85b7a76— Follicle formation absent in VKC
Five questions on VKC/spring catarrh in a 100-item sample is excessive. The misspelling in 77297f61 ("Tantra's Spot" instead of "Trantas' spots") also introduces a factual presentation error. The best single item in this cluster is d85b7a76 (follicle vs papilla distinction), which tests a clinically meaningful differential. The others are largely redundant. Keep d85b7a76; review 062f847f for retention; disable or consolidate the rest.
Glaucoma pharmacology cluster:
eec156c7— Pilocarpine is a miotic in glaucomad159954e— Miotics used in angle-closure glaucoma1f0da52d— First-line treatment for open-angle glaucoma (prostaglandin analogs)83866c72— Beta-blockers decrease aqueous productionbc12942a— Fluorometholone causes least IOP rise
These five items cover overlapping pharmacology territory. 1f0da52d and bc12942a are the strongest (clinical scenario and comparative pharmacology respectively). eec156c7 and d159954e are redundant with each other and with gold-standard coverage. Keep 1f0da52d and bc12942a; disable eec156c7; review d159954e and 83866c72.
ERG wave origins:
The gold PYQ set already contains cccf836d (A-wave = rods and cones). The candidate sample contains c03d428e (C-wave is a positive wave in ERG). These two items together provide reasonable ERG coverage. However, if additional ERG items exist in the broader pool, the topic may be over-represented. Keep c03d428e as it tests a different and less obvious ERG fact; flag for pool-level deduplication.
Recommended disposition: Consolidate VKC cluster to 1–2 items. Consolidate glaucoma pharmacology to 2–3 items. Flag ERG items for pool-level deduplication review.
6. Worthwhile Concept, Weak Execution (Keep the Concept, Fix the Stem/Options/Vignette)
Why this pattern is bad
These questions test high-yield, exam-relevant concepts but are written in a way that reduces their discriminatory power — typically because the stem is too bare, the distractors are implausible, or the clinical context is absent. Discarding these items entirely would waste the underlying concept; fixing them is the right call.
How it shows up
The most common pattern is a clinically important fact presented as a one-line recall question when a short vignette would elevate it to Blooms 3–4. A secondary pattern is implausible distractors that make the correct answer guessable without knowledge.
Affected items
a576aab3 — Acute angle-closure glaucoma vignette
This is the best-structured item in the candidate sample. A 66-year-old with sudden severe pain, shallow hazy cornea, and stony-hard eye is a classic AACG presentation. The distractors (keratitis, uveitis, conjunctivitis) are plausible. However, the correct answer is simply "Glaucoma" — not specifying the type. For a Blooms-3 item, the answer should be "Acute angle-closure glaucoma" to test the specific diagnosis. Fix: change the correct option to "Acute angle-closure glaucoma" and adjust distractors accordingly.
9c6532c7 — Sympathetic ophthalmitis after gunshot injury
Good clinical vignette (fellow eye involvement 3 weeks after penetrating injury). The timing and mechanism are classic. However, the distractor "Macular edema" is implausible in this context, and "Hyphaema" would be an acute finding, not a 3-week delayed presentation. The distractors could be strengthened. Fix: replace weak distractors with more plausible alternatives (e.g., Vogt-Koyanagi-Harada disease, endophthalmitis of the fellow eye).
f8dab676 — Buphthalmos in a 1-month-old with tearing and enlarged cornea
Reasonable vignette. The distractors include Hurler syndrome and galactosemia, which are legitimate differentials for corneal clouding in infancy. However, the question asks for "most likely diagnosis" and the clinical picture (tearing + enlarged cornea) is pathognomonic for congenital glaucoma/buphthalmos. The item is good but could be strengthened by adding IOP or Haab's striae to the stem to make it more discriminating. Keep as-is or minor fix.
2919ac5f — Treatment of advanced PDR with vitreoretinal fibrosis (Exophotocoagulation as exception)
The concept is high-yield (vitrectomy, membrane peeling, retinal reattachment in advanced PDR). However, "Exophotocoagulation" is not a standard term in current ophthalmology literature — it appears to be a non-standard or invented term. If the intended distractor is "external photocoagulation" or "transpupillary photocoagulation," the terminology needs correction. Fix: verify and standardise the terminology of the correct answer option.
9abab06d — CRVO vs Ocular Ischemic Syndrome differentiation
The concept is high-yield and the question format (EXCEPT) is appropriate. However, the marked correct answer is "Dilated Retinal Vein" — meaning dilated retinal veins do NOT differentiate CRVO from OIS. This is factually correct (both can have dilated veins), but the question is tagged as Blooms 1 and easy, which undersells its difficulty. The distractors are appropriate. Keep, but upgrade difficulty and Blooms tagging to reflect the actual cognitive demand.
f0dac4d6 — NOT an early sign of glaucoma (Horizontal cup enlargement)
The concept is important — vertical cup enlargement is the early change, not horizontal. However, "Peripheral halo" is listed as an early sign of glaucoma, which is debatable (halos are a symptom of acute angle-closure, not a sign of early chronic glaucoma). The distractor set needs review. Fix: replace "Peripheral halo" with a more appropriate distractor (e.g., nasal step, paracentral scotoma).
ba4405f4 — Trachoma characteristics (Follicle formation is a late stage)
The marked answer is "Follicle formation is a late stage." This is incorrect — follicle formation is an early feature of trachoma (Stage I and II). The question appears to be asking which statement is TRUE, and the marked answer is factually wrong. This item may belong in Category 1 (Wrong Key). Disable pending fact-check; if the intent was "which is NOT a feature," the stem needs rewriting.
b1384014 — Windshield wiper syndrome / IOL malposition
The concept (IOL malposition causing unpredictable movement) is legitimate and the correct answer "Malposition of lens" is defensible. However, the stem provides the definition in the question itself ("unpredictable movement of an IOL during head motion"), making the question self-answering. A candidate who has never heard of windshield wiper syndrome can answer correctly by reading the stem. Fix: remove the definition from the stem; ask "Windshield wiper syndrome is best described as:" without the explanatory clause.
4e416e1c — NOT an inherited retinal disorder (Butterfly macular dystrophy)
The marked answer is "Butterfly macular dystrophy." However, butterfly-shaped pigment dystrophy of the fovea (Pattern dystrophy) is in fact an inherited (autosomal dominant) retinal dystrophy caused by PRPH2 mutations. This may be a wrong key. "Goldmann syndrome" (listed as a distractor) is not a standard ophthalmology eponym — this may be a reference to Goldmann-Favre syndrome, which is also inherited. The question requires fact-checking before any disposition. Flag as potential wrong key; disable pending expert review.
319953ed — Hassall-Henle bodies (drop-like excrescences in Descemet's membrane periphery)
The concept is legitimate and the question tests a specific corneal degeneration. However, the distractors are all obscure eponyms with no clinical context, making this a pure eponym-matching exercise. The question could be improved by adding a clinical scenario (elderly patient, peripheral corneal changes on slit lamp). Fix: add clinical context, or keep as-is if the pool is thin on corneal degeneration topics.
Recommended disposition: Fix a576aab3, 9c6532c7, 2919ac5f, f0dac4d6, b1384014, 319953ed. Keep f8dab676, 9abab06d, c03d428e. Disable ba4405f4 (potential wrong key). Flag 4e416e1c for expert fact-check before any deployment.
Prioritization
The table below ranks action categories by urgency and impact.
| Priority | Category | Rationale | Approximate Item Count in Sample |
|---|---|---|---|
| P1 — Immediate | Wrong Key or Factually Unsafe | Active harm to candidates; legal/reputational risk | 8–10 items |
| P1 — Immediate | Broken Delivery | Unscorable; generates candidate complaints | 4 items |
| P2 — High | Wrong Subject / Topic Misclassification | Corrupts topic analytics and practice set integrity | 5 items |
| P2 — High | Worthwhile Concept, Weak Execution | High-yield concepts being wasted; fixable with moderate effort | ~15 items |
| P3 — Medium | Low-Value But Correct | Degrades test cognitive profile; large volume | ~25 items |
| P3 — Medium | Repetitive / Duplicative Coverage | Reduces topic diversity; pool-level deduplication needed | ~8 items in sample |
Immediate actions (before next deployment cycle):
- Disable all items with no correct answer marked (
e2eabafe) and all items with confirmed wrong keys (532a4a83,b4bdff6c,66b1fba9,92c5a370). - Disable
ba4405f4and flag4e416e1cfor expert review. - Reclassify the five misclassified items (
6333f470,23e05f2b,987e3416,da78d5cc,f02dea9e).
Short-term actions (within current sprint): 4. Fix the 8–10 items in the "Worthwhile Concept, Weak Execution" category that require stem or distractor revision. 5. Disable the ~15 pure-recall trivia items with no PYQ precedent. 6. Consolidate the VKC and glaucoma pharmacology clusters.
Medium-term actions (pool-level): 7. Conduct a pool-level Blooms audit. The candidate sample shows 89% of items at Blooms 1–2. The full pool of 6,703 items likely has a similar distribution. A targeted commissioning effort for Blooms 3–4 clinical vignettes is needed to bring the pool into alignment with the cognitive demands of INICET and NEET-PG. 8. Conduct a pool-level topic-balance audit. The candidate sample is heavily weighted toward Diseases of the Retina and Conjunctiva; topics like Oculoplasty, Orbital Diseases, and Refractive Surgery appear underrepresented.
Example Keep / Fix / Disable Calls
KEEP
9c6532c7 — Sympathetic ophthalmitis after gunshot injury
Clinical vignette, correct key, appropriate Blooms 3. Minor distractor weakness noted but does not compromise the item. Keep with optional distractor improvement.
f8dab676 — Buphthalmos in a 1-month-old
Good vignette, correct key, plausible distractors (Hurler, galactosemia are real differentials for corneal clouding). Keep.
9abab06d — CRVO vs OIS differentiation
High-yield concept, correct key, appropriate EXCEPT format. Upgrade Blooms tag from 1 to 3. Keep.
080a6f3d — "Umbrella" configuration on FFA in central serous retinopathy
Correct key, PYQ-tagged (NEET-PG 2013), tests a specific and memorable imaging finding. Keep.
db6cba22 — False statement about Eale's disease (Optic neuritis)
Correct key, Blooms 4, PYQ-tagged. Tests a nuanced negative fact about a high-yield condition. Keep.
9bb9792b — Birdshot retinopathy (Unilateral is false)
Correct key, Blooms 4, PYQ-tagged. Tests a specific and discriminating feature. Keep.
1f658cb7 — Kayser-Fleischer ring = copper in Descemet's membrane
Correct key, clinically relevant, appropriate difficulty. Keep.
fcd6f153 — Hutchinson's rule in herpes zoster ophthalmicus
Correct key, clinically important (nasociliary nerve involvement predicts ocular disease). Keep.
FIX
a576aab3 — Acute angle-closure glaucoma vignette
Fix: Change correct option from "Glaucoma" to "Acute angle-closure glaucoma." Adjust distractors to include open-angle glaucoma as a foil.
b1384014 — Windshield wiper syndrome
Fix: Remove the definition from the stem. The question should test recall of the syndrome, not reading comprehension of the definition provided in the stem.
f0dac4d6 — NOT an early sign of glaucoma
Fix: Replace "Peripheral halo" with a more appropriate distractor (e.g., "Nasal step on perimetry" or "Paracentral scotoma") to avoid introducing a misleading association between halos and early glaucoma.
77297f61 — Features of VKC
Fix: Correct "Tantra's Spot" to "Trantas' spots" throughout. Also note that "Papillary hypertrophy" is a feature of VKC and should not be listed as a distractor without clarification — the question needs a full stem and option review.
a66c9088 — Thickening of eyelashes / latanoprost
Fix: Embed in a clinical scenario. Example: "A 55-year-old woman on latanoprost for open-angle glaucoma notices increased length and thickness of her eyelashes. Which property of this drug is responsible?" This elevates the item to Blooms 2–3.
2919ac5f — Treatment of advanced PDR (Exophotocoagulation)
Fix: Verify whether "Exophotocoagulation" is the intended term. If not, replace with the correct terminology. If the term is non-standard, replace with a clearly defined distractor.
DISABLE
532a4a83 — Superior oblique actions (Extorsion marked correct)
Wrong key. Superior oblique performs intorsion, not extorsion. Unsafe to deploy.
e2eabafe — "What does the given image show?"
No correct answer marked. Image absent. Doubly broken.
66b1fba9 — Which instrument is used for fundus examination?
Options are not instruments. Question is structurally incoherent.
b4bdff6c — Wavelength of YAG laser
Options are colours, not wavelengths. Question asks for a physical measurement but provides descriptive terms as options.
92c5a370 — What is the refractive condition of the eye?
No clinical data or image. Unanswerable as written.
a33931ae — Posterior segment defined by posterior surface of lens
Pure anatomy definition. No PYQ precedent. No clinical application. Blooms 1 trivia.
09c7fc64 — Normal pH of tears
Normal value with no clinical relevance at PG level.
c8489725 — Colour of iris in complete albinism
Trivia-level recall. No differential reasoning. No PYQ precedent.
ac422e05 — Primary colours (None of the above)
Trick question with no clinical relevance. "None of the above" key format is poor practice.
c8bf6a67 — Tonometry is commonly performed in ophthalmology
Does not discriminate between any meaningful knowledge levels. Any layperson could answer this.
ba4405f4 — Trachoma: follicle formation is a late stage (marked correct)
Follicle formation is an early feature of trachoma, not late. Likely wrong key. Disable pending fact-check.
6333f470, 23e05f2b, 987e3416, da78d5cc — Cataract/lens questions filed under Contact Lens
Reclassify rather than disable, but remove from Contact Lens topic immediately to prevent incorrect topic-set deployment.