Orthopaedics Question Quality Review

Executive Summary

The Orthopaedics candidate sample of 200 questions was reviewed across eight shards of 25 questions each. The subject pool of 4,052 questions is large, but this sample reveals a set of structural and content problems that are serious enough to affect the usability of a meaningful proportion of the bank.

The Blooms distribution is heavily skewed toward recall: 63 Blooms-1 and 90 Blooms-2 questions account for 76.5% of the candidate sample, against only 36 Blooms-3 and 11 Blooms-4 items. The benchmark and PYQ gold standard, by contrast, is dominated by Blooms-3 clinical vignettes with plausible distractors and genuine decision-making demand. The gap between what the bank contains and what the exam tests is the central problem.

Beyond the Blooms imbalance, the reviewed set contains six distinct, operationally separable problem types: questions that are structurally non-functional because images are missing; questions where the keyed answer is factually wrong or clinically outdated; questions that belong to a different subject entirely; questions that are pure recall trivia with no clinical anchor; questions with structural item-writing flaws that destroy discriminative value; and questions that are internally redundant within the same topic cluster. Each of these requires a different remediation path.

Across the 200-question sample, a rough disposition estimate is: approximately 40–50 questions are keep-quality as written or with minor polish; approximately 60–70 are fixable with targeted effort; and approximately 80–100 should be disabled, either because they are structurally broken, factually unsafe, out of scope, or so low-yield that strong gold-standard coverage already exists for the same concept.

What Good Looks Like

The benchmark and recent PYQ items establish a clear quality bar that the candidate sample only intermittently meets.

Clinical vignette with embedded reasoning demand. The best items in the gold set (bc66d83a, 9fc9e167, 4f8e32b2, b3dd8f1b) present a patient with age, mechanism, and findings, then ask for a diagnosis or management decision. The answer cannot be retrieved by pattern-matching a single keyword; it requires integrating two or three clinical facts. The candidate sample has items that reach this standard — Q-bc6b3589 (steroid-induced vertebral fracture in a patient on multiple medications), Q-b479bf15 (SCFE with limitation of abduction and internal rotation), Q-e7c4742b (AVN in an HIV patient on ART) — but they are outnumbered by items that do not.

Distractors that represent genuine clinical alternatives. In bc66d83a (intracapsular NOF fracture in a 75-year-old), the distractors — conservative management, THR, DHS — are all real treatment options that a candidate might plausibly choose if they misidentify the fracture type or the patient's age-fitness profile. In the candidate sample, many distractors are implausible, anatomically impossible, or near-identical to each other, allowing elimination without content knowledge.

Negative-format questions used only when the concept genuinely requires it. The PYQ item e807c756 (ranking external fixators by stability) uses a complex ordering format because the concept is inherently comparative. In the candidate sample, the NOT/EXCEPT format is frequently applied to simple recall lists, adding cognitive noise without analytical demand.

Factual precision and currency. The gold items are factually clean and reflect current practice. The candidate sample contains items where the keyed answer conflicts with standard Indian PG references or with post-2010 clinical guidelines.

Main Issue Categories

1. Structurally Non-Functional Questions: Missing Images

Why this pattern is bad

A question that references an X-ray, clinical photograph, or instrument image but does not embed that image is not a question — it is an unanswerable prompt. Candidates either guess randomly or, if the stem contains enough residual text, arrive at the answer through a process the question was never designed to test. In either case, the item provides no valid measurement of knowledge. This is the most severe structural defect in the sample because it cannot be partially mitigated; the item is either fixed or it is not.

How it shows up

The pattern appears consistently across shards. Stems contain phrases such as "X-ray is provided," "the deformity shown in the image," "identify the instrument shown in the image," "the lesion shown in the image," and "the test given in the image." In every case, no image is present in the question text. The affected questions span multiple topics — trauma, bone tumors, arthrology, and pediatric orthopaedics — suggesting this is not a topic-specific problem but a bank-wide content-loading failure.

Example question IDs

5cd7a4b4: "A hip X-ray is provided" — no X-ray present. The intended topic (GCT of hip) is clinically relevant, but the question cannot function without the image.
668b5fa6: "A knee X-ray is shown" — no image. The correct answer (Rickets) cannot be derived from the text alone; a 12-year-old with chronic knee pain has multiple plausible diagnoses.
930d36e0: "X-ray humerus shows presence of" — no X-ray. Radiological descriptors (fallen fragment sign, central lytic lesion) that would make this self-contained are absent.
66eda231: "The test given in the image" — no image. Likely a knee ligament test (Lachman or anterior drawer), which is a high-yield topic, but the question is unanswerable as written.
af3a20e8: "What is the X-ray diagnosis?" — no image. Completely non-functional.
83035266: "Identify the instrument shown in the image" — no image. FMGE-PYQ tagged, so the item has provenance value, but it cannot be used without the visual.
2cda6ca2: "The deformity shown in the image" — no image. RA hand deformity is a recognisable topic but the question is structurally broken.
c3cf388a: "The lesion shown in the image" — no image. Intended as a GCT diagnosis question; distractors are appropriate but the stem is non-functional.
96c8dbe8: "79-year-old lady, X-ray, treatment" — no image. Also contains a spelling error ("Hemiahroplasty").
4819eb45: "X-ray film shown below" — absent. Additionally has an answer accuracy concern regarding antibiotic cement weight thresholds.

Recommended disposition

For each affected item: attempt image recovery first. If the original image can be reliably embedded and rendered, the question may be salvageable. If the image cannot be recovered, rewrite the stem using descriptive radiological or clinical language (e.g., "soap-bubble lytic lesion in the epiphysis of the distal femur" instead of "as shown in the X-ray"). If neither path is feasible within a reasonable effort threshold, disable. Do not leave image-dependent questions in the active pool in their current state.

2. Factually Incorrect or Clinically Unsafe Answer Keys

Why this pattern is bad

A question with a wrong keyed answer actively harms candidates by reinforcing incorrect knowledge. In a subject like Orthopaedics, where management decisions have direct patient-safety implications, this is not a minor quality issue. Items in this category are more dangerous than low-quality items because they appear functional — they have a stem, four options, and a marked answer — but they teach the wrong thing.

How it shows up

The pattern appears in two sub-forms. The first is a straightforward factual error: the keyed answer contradicts the standard teaching in major Indian PG references (Maheshwari, Campbell). The second is a clinical currency problem: the keyed answer was once correct but conflicts with post-2010 guidelines or current practice consensus.

A particularly serious sub-pattern is the direct internal contradiction: two questions in the same bank give opposite answers to the same clinical question, which means at least one is wrong and both are harmful.

Example question IDs

d2f7cde2: Marks "proximal tibia" as the most common site of osteosarcoma. The standard answer in every major reference is distal femur. This is a clear factual error. Fix: correct the answer key.
8cf1e4cd: Marks "humerus" as the most common birth fracture. Standard teaching and most references cite clavicle. If humerus is intended for a specific delivery context (breech), the stem must specify this. Fix: correct the answer key or add clinical context.
4043d12e vs. 4b7e5ae0: These two questions directly contradict each other on the most common complication of extracapsular femur fracture — one marks non-union, the other marks malunion. Standard teaching is that extracapsular (intertrochanteric) fractures have good blood supply and malunion is the complication; non-union is the complication of intracapsular fractures. Q-4043d12e has the wrong answer. Fix Q-4043d12e; reconcile both items.
132ff0c3: In a Bone Tumors topic context, marks "idiopathic growth pain" as the cause of night pain in an adolescent's knee. The classic high-yield answer for night pain relieved by NSAIDs in this context is osteoid osteoma. The question as marked is misleading. Fix: rewrite stem to specify growth pain context, or correct the answer and distractors.
951cddfc: Asks for the methylprednisolone dose in acute spinal cord injury as a straightforward factual recall. High-dose methylprednisolone for acute SCI is no longer recommended in current AANS/CNS guidelines (2013 onwards) and is associated with significant adverse effects. Presenting this as a clean factual answer without any clinical caveat is potentially harmful. Disable pending expert guideline review.
073a8536: Tuli's staging sequence for spinal TB has a marked answer that appears to invert stages 2a and 2b relative to the standard published classification. Requires expert clinical review before use. Hold for expert verification.
6cbbd47c: Associates hyperglycemia with osteosarcoma. This association is not established in standard Indian PG orthopaedics references. Risks teaching an incorrect fact. Disable pending expert content review with cited reference.
e8a90a6d: Marks CR&IF as the correct answer for a 50-year-old with displaced NOF fracture presenting after 3 days. In current practice, delayed presentation in a middle-aged patient with a displaced fracture often warrants hemiarthroplasty or THR. The question needs explicit displacement status and answer rationale review. Fix: add displacement status to stem; review answer against current guidelines.
ba065416: Marks "abduction and external rotation" as the mode of injury for Colles fracture. The standard mechanism is FOOSH with wrist dorsiflexion and pronation. The options presented do not map to any standard description of this injury. Fix: rewrite options entirely.
38722ecb: Marks "central dislocation" for a hip presentation that describes abduction and internal rotation, which is characteristic of anterior dislocation. Fix: verify answer key against standard references.

Recommended disposition

Items with clear factual errors in the answer key: fix if the concept is high-yield and the stem is otherwise sound. Items where the answer conflicts with current guidelines and the question presents it as unambiguous fact (951cddfc): disable until expert review confirms the appropriate framing. Items with direct internal contradictions: fix the incorrect item and add a note to reconcile the pair. The content team should conduct a targeted accuracy audit of all "most common site," "most common complication," and "first-line treatment" items in Bone Tumors and Trauma, as these are the topic areas where factual errors cluster in this sample.

3. Subject Boundary Violations: Non-Orthopaedic Content in the Pool

Why this pattern is bad

Questions from Oral and Maxillofacial Surgery, Dental Surgery, and occasionally Rheumatology or Evolutionary Biology appear in the Orthopaedics pool. These items test content that Orthopaedics candidates are not expected to know at PG level, they displace high-quality orthopaedic content, and they create confusion about the subject's scope. When a candidate encounters a mandible fracture management question in an Orthopaedics test, it signals a quality control failure that undermines confidence in the entire question set.

How it shows up

The contamination is most visible in the Trauma topic tag, which appears to have been used as a catch-all for any fracture-related question regardless of anatomical region or clinical specialty. The Spine Disorders topic tag shows a similar problem, absorbing hemophilic arthropathy and foot joint anatomy questions that have no spine relevance.

Example question IDs

eb440c5e: Mandible fracture managed with interdental fixation. Pure OMFS content. Filed under Trauma in Orthopaedics. Disable from this subject.
3e3137fc: TMJ ankylosis cause. OMFS content. Blooms-1, no clinical stem. Disable from this subject.
dce4636e: Subcondylar mandibular fracture management in a 7-year-old. Options include "cap splint with circumferential wiring" and "intermaxillary fixation." Dental/maxillofacial surgery. Disable from this subject or reassign.
b4f62ce2: Symphyseal fracture in an 8-year-old with options including "cap splint with circumferential wiring." Dental surgery content. Disable from this subject.
1a3ebc3b: Tietze syndrome (costochondral junction). Filed under Trauma. Marginal orthopaedic relevance; more appropriate in Rheumatology or General Medicine. Disable or reassign.
5c8a4bd9: "Bird-face appearance — NOT associated" — drifts into maxillofacial territory. Review for subject relevance.
acca9222: Talocalcaneonavicular joint type filed under Spine Disorders. Correct subject (Orthopaedics) but wrong topic tag. Fix topic tag.
98622158: Hemophilic arthropathy filed under Spine Disorders. Correct subject but wrong topic tag. Fix topic tag.
e91691f9: High crural index in jumping athletes. Crural index is an anthropometric ratio from evolutionary biology, not standard orthopaedic clinical teaching. Disable.
7bbf235c: Post-polio hip deformity tagged under Arthroscopy. Correct subject, wrong topic. Fix topic tag.

Recommended disposition

All confirmed OMFS/dental items: disable from the Orthopaedics pool (they may be valid in their own subject pool). Items with wrong topic tags but correct subject: fix the topic tag. The content team should run a systematic subject-boundary filter across the full 4,052-question Orthopaedics set, specifically targeting the Trauma and Spine Disorders topic buckets, which appear to be the primary entry points for misclassified content.

4. Pure Recall Trivia with No Clinical Anchor

Why this pattern is bad

This is the largest single category by volume in the reviewed sample. Questions that ask for a named association, an eponym definition, a single anatomical fact, or a historical attribution at Blooms-1 with no patient context do not test the clinical reasoning that Indian PG examinations assess. They inflate the easy-question proportion of the bank, reduce the average discriminative index, and crowd out space for higher-quality items. When a candidate can answer a question correctly by having seen the word "Brodie's abscess" once in a textbook, the question is not measuring PG-level competence.

How it shows up

This pattern is pervasive across topics. It is most concentrated in Bone Tumors (location recall), Pediatric Orthopaedics (eponym definitions), Basic Science (laws and principles), and Trauma (single-fact associations). The EXCEPT format is frequently used to dress up recall as analysis, but when the underlying content is a simple list, the negative format adds no genuine cognitive demand.

Example question IDs

1f026d88: "Who coined the term orthopaedics?" → Nicholas Andry. Pure historical trivia. No PYQ precedent. No clinical relevance. Disable.
a7fed176: "Brodie's abscess = subacute osteomyelitis." Pure eponym recall, Blooms-1, no clinical scenario. Equivalent items exist in every question bank. Disable.
01ad6ea3: "Posterior iliac horn → nail-patella syndrome." Blooms-1 eponym association, no clinical stem. Disable.
2de634eb: "Perthes disease etiology = Unknown." Single-word answer, no clinical context, zero discriminatory value. Disable.
ff7d6b8a: "Muscle in congenital muscular torticollis → SCM." Blooms-1, easy-flagged, single-word recall. Disable or upgrade.
1f722417: "Sprengel's deformity → congenital elevation of scapula." Definition-level recall. Disable or reframe as clinical scenario.
989f6949: "Kebab treatment → osteogenesis imperfecta." Eponym recall, Blooms-1, no PYQ precedent. Disable.
3b276a13: "Heberden's nodes at DIP joint." First-year anatomy/pathology fact. No clinical context, no discriminative value at PG level. Disable.
1a047298: "Commonest site of fat embolism = femur." Blooms-1, easy-flagged, no clinical context. Disable.
e0d2bacd: "Maximum weight in skeletal traction = 20 kg." Pure numerical recall, no clinical scenario. Disable.
129499d0: "Greenstick fracture = incomplete fracture." Blooms-1, NEET-PG 2012 vintage, trivially easy. Disable.
bc7fc58b: "Commonest ligament injured in ankle = ATFL." Blooms-1, easy-flagged, NEET-PG 2012 vintage. Disable.
a9beacf7: "Which test is used to assess ACL injury?" → Lachman test. Blooms-1, no clinical context. Disable.
9dce8b46: "Most common type of shoulder dislocation?" → Subcoracoid. Blooms-1, no stem. Disable.
ad0301d0: "Watson-Jones approach is indicated for?" → Hip replacement. Blooms-1, no clinical reasoning. Disable.
b6339939: "Apert syndrome is a disease of?" → Bones and joints. Trivially easy, no discriminatory value. Disable.
2655d2e3: "Von-Rosen's sign in DDH." Blooms-1, easy, PYQ from 2013. Outdated recall item. Disable or replace with DDH management scenario.
57ab3f20: "Caffey's disease is defined as" → infantile cortical hyperostosis. No clinical scenario. Disable or convert to vignette.
f4a6dcab: "Pauwels classification is used for which fracture?" Pure eponym-recall. Disable or upgrade to ask what Pauwels Type III implies for management.
10b4f8ef: "Commonest level in lower cervical spine injury = C5-C6." Blooms-1, easy-flagged, no clinical scenario. Disable.

Recommended disposition

For items that are pure recall with no clinical anchor and where strong gold-standard coverage of the same concept already exists: disable. For items where the underlying concept is genuinely high-yield (e.g., Lachman test, ATFL injury, DDH signs) but the question is at Blooms-1: fix by upgrading to a clinical vignette only if the rewrite effort is proportionate to the concept's exam frequency. Do not speculatively rewrite low-yield trivia items (historical attributions, obscure eponyms, crural index) — disable them.

5. Item-Writing Structural Flaws That Destroy Discriminative Value

Why this pattern is bad

This category covers questions where the factual content may be correct but the construction of the item allows test-wise candidates to identify the answer without using the knowledge being tested. These flaws are well-documented in item-writing literature and are specifically problematic in high-stakes PG examinations where candidates are sophisticated test-takers. The flaws observed in this sample fall into four distinct sub-types, each with a different fix path.

How it shows up

Sub-type A: "All of the above" or "None of the above" as the keyed answer. When the correct answer is "all of the above," a candidate who recognises any two options as correct can select the right answer without evaluating the third. This rewards partial knowledge and test-taking strategy over complete understanding.

e06ca1b0 (Osteoid osteoma): "All of the above" is the correct answer. Options A and C are individually verifiable as correct, making the question guessable with partial knowledge. Fix: rewrite as single best-answer.
481072aa (Gorham's disease): "All of the above" as correct answer. Same structural problem. Fix.
40ddd7a3 (Colles' fracture statements): "All of the options" is the correct answer, but option A contains a factual error (radial styloid height), making the "all of the above" answer itself factually wrong. Fix: rewrite as single best-answer and correct the factual error in option A.
5f0d77f6 (Wolff's law): "None of the above" as a distractor. Fix: replace with a substantive distractor.

Sub-type B: Explanatory parenthetical text embedded in option stems. This telegraphs the answer and converts the question into a reading comprehension exercise.

aff89251 (Blount's disease): Options read "External tibial torsion (can occur in some cases)" and "Internal tibial torsion (common in Blount's disease)." The parenthetical text identifies the correct answer. Fix: remove all parenthetical text from options.

Sub-type C: Near-identical or overlapping distractors. When two options say essentially the same thing, the question effectively has three options, reducing difficulty artificially and signalling poor construction.

8a1a3c77 (Barton's fracture): Options C and D both describe intra-articular fracture with carpal subluxation; option D adds "joint involvement," which is redundant. Fix: collapse into one option and add a genuinely distinct distractor.
a1b41e74 (Perthes disease "not true"): Option D is "All are true," which is logically incompatible with a "not true" stem. Fix: replace option D with a substantive fourth distractor.

Sub-type D: Stem gives away the answer. The question stem contains enough information to identify the correct answer without using the options.

484725b1 (Spondylolisthesis): The stem states "anterior displacement of vertebral body" — this is the definition of spondylolisthesis, making the question a definition-matching exercise. Fix: upgrade stem to include neurological findings or a management decision.
9a8fd11f (Dashboard injury → posterior hip dislocation): The mechanism directly implies the diagnosis; no reasoning is required. Fix: add clinical findings or ask about management/complications.

Recommended disposition

All "all of the above" / "none of the above" keyed-answer items: fix by rewriting as single best-answer format. Items with parenthetical hints in options: fix by removing the parenthetical text. Items with overlapping distractors: fix by replacing the redundant option. Items where the stem gives away the answer: fix by adding clinical complexity to the stem. These are all fixable with moderate effort and the underlying concepts are generally worth retaining.

6. Bone Tumor Topic: Redundancy, Recall Inflation, and Accuracy Concerns

Why this pattern is bad

Bone Tumors is the topic with the most concentrated quality problems in this sample. It shows three overlapping issues that together make it the highest-priority topic for a dedicated audit: internal redundancy (multiple questions testing the same narrow fact), Blooms-1 location recall dominating the topic, and factual accuracy concerns in the answer keys. These problems compound each other — the topic is simultaneously over-represented in low-quality recall items and under-represented in high-quality clinical vignettes.

How it shows up

The chondroblastoma cluster is the clearest example of redundancy. Three questions — Q-bfa032fb (most common site of chondroblastoma), Q-3daf3c95 (which tumor arises near the epiphyseal plate), and Q-020352db (epiphyseal tumor before fusion) — all resolve to chondroblastoma with near-identical reasoning. None of them presents a clinical vignette. The topic also contains multiple pure location-recall items (enchondroma → phalanges, chondroblastoma → epiphysis) that add no discriminative value.

The factual accuracy concerns are separate from the redundancy. Q-d2f7cde2 has the wrong site for osteosarcoma (proximal tibia instead of distal femur). Q-6cbbd47c associates hyperglycemia with osteosarcoma without a credible reference. Q-132ff0c3 gives a misleading answer for night pain in an adolescent in a Bone Tumors context.

Example question IDs

bfa032fb: Most common site of chondroblastoma → proximal tibia. Blooms-1, easy-flagged, redundant with 020352db and 3daf3c95. The answer is also debatable (proximal humerus and distal femur are more classically cited). Disable.
3daf3c95: Which tumor arises near the epiphyseal plate → chondroblastoma. Near-duplicate of bfa032fb and 020352db. Disable one of the three; retain the best-constructed version.
020352db: Epiphyseal tumor before fusion → chondroblastoma. NEET-PG tagged, reasonable distractor set including GCT as a plausible foil. The strongest of the three. Keep; disable the other two.
d4f7c03e: Chondroblastoma → epiphysis. Pure Blooms-1 location recall, no vignette. Disable or upgrade.
febb6af3: Enchondroma location → phalanges. Pure Blooms-1 anatomical recall. Disable or upgrade to clinical vignette.
d2f7cde2: Most common site of osteosarcoma → proximal tibia (wrong). Fix: correct answer key to distal femur.
6cbbd47c: Hyperglycemia associated with osteosarcoma. Unverified association. Disable.
132ff0c3: Night pain in adolescent knee → idiopathic growth pain (misleading in Bone Tumors context). Fix or disable.
5466d952: Parosteal osteosarcoma — TRUE statement (may involve medullary cavity). Well-constructed Blooms-2 item with genuine discriminating content. Keep.
b0246951: GCT false statement (chemotherapy not mainstay). Blooms-4, PYQ-tagged. Keep.

Recommended disposition

Conduct a dedicated accuracy audit of all Bone Tumors answer keys against Maheshwari and Campbell before any items in this topic go live. Disable the chondroblastoma redundancy cluster, retaining only the best-constructed item. Disable pure location-recall items. Prioritise development of clinical vignette items for this topic (age + location + radiological description + diagnosis/management) to replace the disabled recall items.

7. Clinically Thin Vignettes: Correct Facts, Insufficient Reasoning Demand

Why this pattern is bad

This category is distinct from pure recall trivia (Category 4) because the questions in this group are not wrong and are not trivial — they test genuinely high-yield concepts. The problem is that they present those concepts as bare factual questions rather than as clinical scenarios, which means they measure recognition rather than application. A candidate who has memorised a list can answer these correctly; a candidate who understands the clinical reasoning but has not memorised the specific fact cannot. This inverts the intended measurement.

The benchmark standard (bc66d83a, 9fc9e167, b3dd8f1b) consistently embeds the same facts in a patient scenario. The candidate sample has many questions that test the same concepts but strip out the clinical context.

How it shows up

The pattern is most common in Trauma, Spine Disorders, and Pediatric Orthopaedics. Questions ask "what is the most common complication of X" or "what nerve is injured in Y" without providing a patient. The Blooms tag is often inflated to 2 or 3, but the actual cognitive demand is Blooms-1 because the stem contains no information that requires integration.

Example question IDs

95ff367e: "What is the leg position in NOF fracture?" — bare factual recall. The benchmark equivalent (bc66d83a) presents an elderly patient post-fall. Fix: add clinical scenario (elderly patient, post-fall, shortened externally rotated limb).
2156e4f5: "Most common site of osteomyelitis in children" — no patient context. Fix: add a child with fever, bone pain, and local tenderness.
e5815b6f: "RA spinal involvement" — pure recall. Fix: embed in a clinical scenario (RA patient with neck pain and myelopathy) or disable as low-priority filler.
9272fc2f: "Absence of lamina dura → hyperparathyroidism" — decontextualised. Fix: add clinical scenario (patient with renal failure or hypercalcaemia).
207f5d6f: "Congenital pseudoarthrosis associated with neurofibromatosis" — bare fact. Fix: add vignette (child with café-au-lait spots and tibial bowing).
777dde66: "Dennis 3-column concept" — bare-fact recall with no clinical anchor. Fix: add a clinical scenario (burst fracture with middle column involvement).
9ab66c7e: "NOT in Gurd's criteria" — pure list recall in EXCEPT format. Also has an ambiguous option (tachycardia threshold direction). Fix: add clinical context and clarify option text.
56862f41: "Sudeck's atrophy → osteopenia" — does not distinguish CRPS from simple disuse osteoporosis. Fix: add clinical context (post-fracture burning pain, skin changes) and improve distractors.
2c4afca0: "Looser's zones except hypoparathyroidism" — factually sound but no clinical stem. Fix: add radiological/clinical scenario.

Recommended disposition

These items are fix candidates, not disable candidates. The underlying concepts are high-yield and the factual content is generally correct. The fix path is consistent: add a brief clinical scenario (2–3 sentences: age, presentation, relevant finding) and verify that the distractors represent genuine clinical alternatives. The effort per item is moderate. Prioritise the highest-frequency exam topics (NOF fracture, osteomyelitis, spinal TB, CRPS) for the first fix pass.

Prioritization

The six issue categories are not equal in urgency or in the effort required to address them. The following ordering reflects both the severity of the problem and the operational path to resolution.

Immediate action required (before any live deployment)

Structurally Non-Functional Questions (Missing Images) — These items are unanswerable as written. They must be triaged now: attempt image recovery, rewrite with descriptive stems, or disable. No image-dependent question should be in an active test until this is resolved.
Factually Incorrect or Clinically Unsafe Answer Keys — Items with wrong keyed answers (d2f7cde2, 8cf1e4cd, 4043d12e, ba065416) and the clinically outdated methylprednisolone item (951cddfc) must be corrected or disabled before use. The direct contradiction pair (4043d12e vs. 4b7e5ae0) needs immediate reconciliation.

High priority (next content operations cycle)

Subject Boundary Violations — OMFS/dental items (eb440c5e, 3e3137fc, dce4636e, b4f62ce2) should be disabled from the Orthopaedics pool in a single pass. Wrong topic tags (acca9222, 98622158, 7bbf235c) should be corrected. This is a relatively low-effort, high-impact cleanup.
Bone Tumor Topic Audit — The combination of redundancy, recall inflation, and accuracy concerns in this topic makes it the highest-priority topic for a dedicated review. Disable the chondroblastoma cluster duplicates, correct the osteosarcoma site error, and commission replacement vignette items.

Medium priority (systematic improvement pass)

Item-Writing Structural Flaws — "All of the above" keyed answers, parenthetical hints in options, and overlapping distractors are fixable with moderate effort. A systematic search for "all of the above" as the correct answer across the full 4,052-question set should be run and each instance reviewed.
Pure Recall Trivia — The volume of Blooms-1 items with no clinical anchor is large. A bulk disable pass on items flagged as easy + Blooms-1 + no PYQ tag + no clinical stem would address the majority of this category efficiently. Upgrade only the highest-yield concepts (Lachman test, ATFL, DDH) to vignette format.

Ongoing improvement (lower urgency)

Clinically Thin Vignettes — These are correct but underperforming items. A fix pass to add clinical context is worthwhile but not urgent, as these items are not harmful in the way that wrong answer keys or missing images are. Prioritise the most frequently tested concepts first.

Example Keep / Fix / Disable Calls

The following table summarises representative disposition calls drawn from across the reviewed sample. These are intended as concrete operational examples for the content team.

KEEP — as written or with minor polish

Question ID	Topic	Reason
157899bc	Trauma	Compartment syndrome management vignette; correct answer (fasciotomy); Blooms-3; plausible distractors
52e3b038	Hand Surgery	CTS provocative tests; Finkelstein's correctly identified as not applicable; Blooms-3 application
3217e410	Spine	C6–C7 disc prolapse from clinical findings; clean vignette; Blooms-3
f1a7d618	Shoulder	Diabetic female with frozen shoulder; Blooms-4; good foil (rotator cuff tear)
b479bf15	Pediatric Ortho	SCFE clinical vignette; age, weight, signs correctly assembled; Blooms-3
98104bda	Pediatric Ortho	Lateral condyle fracture → delayed ulnar palsy; classic high-yield scenario; Blooms-3
e7c4742b	Degenerative	AVN in HIV patient on ART; high-yield association; Blooms-3
5466d952	Bone Tumors	Parosteal osteosarcoma true statement; genuine discriminator; Blooms-2
b0246951	Bone Tumors	GCT false statement; chemotherapy not mainstay; Blooms-4; PYQ-tagged
abd2c822	Trauma	Compartment syndrome NOT a feature; "normal sensation distally" is a genuine clinical misconception test
d4a339e8	Spine	Atlas fracture EXCEPT; quadriplegia uncommon in Jefferson fracture; Blooms-4
360a4e73	Trauma	Gustilo-Anderson Type IIIB classification; clinical stem; medium difficulty
52713dfe	Spine	EHL paralysis → L5 root; clean clinical reasoning; Blooms-2
8a137929	Hand Surgery	Saturday night palsy; knuckle bender splint; applied reasoning; Blooms-3
020352db	Bone Tumors	Epiphyseal tumor before fusion → chondroblastoma; NEET-PG tagged; best of the redundant cluster

FIX — retain concept, specific remediation required

Question ID	Topic	Required Fix
d2f7cde2	Bone Tumors	Correct answer key: distal femur, not proximal tibia
8cf1e4cd	Pediatric Ortho	Correct answer key: clavicle, not humerus, as most common birth fracture
4043d12e	Trauma	Correct answer key: malunion (not non-union) for extracapsular NOF fracture; reconcile with 4b7e5ae0
ba065416	Trauma	Rewrite all four options; mechanism of Colles fracture is FOOSH, not rotation
aff89251	Pediatric Ortho	Remove all parenthetical explanatory text from option stems
e06ca1b0	Bone Tumors	Rewrite as single best-answer; remove "all of the above" as correct answer
40ddd7a3	Trauma	Rewrite as single best-answer; correct factual error in option A; remove "all of the above"
96c8dbe8	Trauma	Embed image or rewrite with descriptive radiological stem; fix spelling error
83035266	Trauma	Embed image or rewrite as descriptive question about Bohler-Braun splint
c3cf388a	Bone Tumors	Embed image or rewrite with soap-bubble lytic lesion description
12e5f4ca	Biomechanics	Rewrite options as complete statements; options are currently bare chemical symbols
073a8536	Spine	Expert clinical review of Tuli's staging sequence before use
e8a90a6d	Trauma	Add displacement status to stem; review answer against current NOF fracture guidelines
95ff367e	Trauma	Add clinical scenario (elderly patient, post-fall, shortened externally rotated limb)
56862f41	Trauma	Add CRPS clinical context; improve distractors
9272fc2f	Metabolic Bone	Add clinical scenario (renal failure or hypercalcaemia patient)
7bbf235c	Pediatric Ortho	Fix topic tag from Arthroscopy to Pediatric Orthopaedics; add clinical vignette
a1b41e74	Pediatric Ortho	Replace "All are true" option D with substantive fourth distractor
e267e0e6	Spine	Rewrite option B to be clearly false; current option B is plausibly correct
1c80682e	Arthroplasty	Revise stem to remove acute inflammatory features, or reconsider answer key
bb92e3d0	Spine	Differentiate from near-duplicate 51e0d845; add specific neurological deficit or change level
51cd266d	Spine	Expert review: osteomyelitis is a recognised cause of pseudoarthrosis; answer key may be wrong

DISABLE — remove from active pool

Question ID	Topic	Reason
668b5fa6	Trauma	Image missing; correct answer (Rickets) not derivable from text
4819eb45	Trauma	Image missing + answer accuracy concern on antibiotic cement threshold
af3a20e8	Trauma	Image missing; entirely non-functional as text-only item
951cddfc	Spine	Methylprednisolone in acute SCI no longer standard of care per current guidelines
6cbbd47c	Bone Tumors	Hyperglycemia/osteosarcoma association not established in standard references
eb440c5e	Trauma	OMFS content (mandible fracture); not Orthopaedics
3e3137fc	Arthrology	OMFS content (TMJ ankylosis); not Orthopaedics
dce4636e	Trauma	OMFS content (subcondylar mandibular fracture); not Orthopaedics
b4f62ce2	Trauma	Dental/maxillofacial content (symphyseal fracture, cap splint); not Orthopaedics
1f026d88	Basic Science	Historical trivia (who coined "orthopaedics"); no clinical relevance
e91691f9	Biomechanics	Crural index is evolutionary biology, not clinical orthopaedics; answer contestable
bfa032fb	Bone Tumors	Blooms-1 location recall; redundant with 020352db; answer debatable
3daf3c95	Bone Tumors	Near-duplicate of 020352db and bfa032fb; disable to reduce redundancy
2de634eb	Pediatric Ortho	"Perthes disease etiology = Unknown"; zero discriminatory value
a7fed176	Arthrology	Brodie's abscess eponym recall; Blooms-1; no clinical scenario
01ad6ea3	Basic Science	Nail-patella syndrome eponym; Blooms-1; no clinical stem
1a047298	Trauma	Fat embolism site; Blooms-1; easy-flagged; no clinical context
e0d2bacd	Trauma	Maximum traction weight; pure numerical recall; no clinical scenario
129499d0	Trauma	Greenstick = incomplete fracture; Blooms-1; NEET-PG 2012 vintage
bc7fc58b	Sports Medicine	ATFL commonest ankle ligament; Blooms-1; NEET-PG 2012 vintage
a9beacf7	Sports Medicine	"Which test assesses ACL?" → Lachman; Blooms-1; no clinical context
9dce8b46	Trauma	Most common shoulder dislocation type; Blooms-1; no stem
2655d2e3	Pediatric Ortho	Von-Rosen's sign; Blooms-1; PYQ from 2013; outdated recall
f4a6dcab	Trauma	Pauwels classification eponym recall; Blooms-1; no clinical reasoning
57ab3f20	Pediatric Ortho	Caffey's disease definition; Blooms-1; no clinical scenario
e260e5e1	Arthroplasty	Computer-assisted navigation accuracy; narrow numerical trivia; no PYQ precedent
f8e61854	Prosthetics	Jaipur foot trivia; template membership in JKBOSE Physics indicates tagging error
989f6949	Pediatric Ortho	Sofield-Millar "kebab" eponym recall; Blooms-1; no PYQ precedent
6df47fcb	Trauma	Distraction rate for small bones; factual basis unclear; conflicts with Ilizarov standard
10b4f8ef	Spine	Commonest cervical injury level; Blooms-1; easy-flagged; no clinical scenario
144a0898	Trauma	"Salter Harris classification used for" — what the classification is for, not how to apply it; redundant with benchmark item 4cd377d9