OMSB Ophthalmology (OEOPTH) common mistakes
OMSB Ophthalmology (OEOPTH) · Understand the traps and how to avoid them.
Answering management questions with the single next test or treatment instead of the full staged plan the question stem asks for (e.g., giving only 'start topical steroids' when asked how to manage a suspected endophthalmitis).
Candidates train on single-best-answer habits and skim the stem's instruction word ('initial', 'next', 'most appropriate long-term management'), so they anchor on the first correct-sounding intervention and stop.
What to do instead
Before answering, underline the instruction word in the stem and classify it as next-step, best-initial, or definitive/long-term. For management items, draft the full sequence (resuscitate/urgent step, diagnostic confirmation, specific therapy, monitoring) and then select the option that matches the requested stage. Practise 20 stems where the same vignette is asked at three different stages and compare the correct answers.
Choosing a diagnosis from the most striking single sign rather than weighing the discriminating features; for example, labelling every painful red eye with a fixed dilated pupil as acute angle closure when the stem describes a hypopyon or a contact-lens wearer.
Pattern recognition from a few vivid keywords overrides systematic feature extraction, and candidates do not actively generate and exclude the two or three closest differentials.
What to do instead
Use a fixed five-feature extraction for every red-eye or vision-loss vignette: laterality, onset speed, pain, pupil/reflex findings, and associated systemic or exposure history. Then explicitly write the two closest differentials and the one finding that separates them before selecting. Build a personal one-page table of discriminating signs (e.g., ciliary flush vs diffuse injection, pupil behaviour, fluorescein uptake pattern).
Misreading visual field, OCT, or imaging descriptions as normal because the candidate does not translate the described defect into an anatomical lesion; e.g., a bitemporal defect described in words is not linked to the chiasm.
Candidates memorise field patterns as pictures rather than as rules, and because no image is shown they fail to reconstruct the pattern from the verbal description.
What to do instead
Learn the localisation rules as text: respect the vertical midline, respect the horizontal midline, congruity, and monocular versus binocular. Practise converting written descriptions ('left eye temporal loss plus right eye nasal loss') into a lesion site and a short differential list. For each site, list the top three causes and the confirmatory test.
Ignoring the paediatric or congenital dimension of a case: treating a child's leukocoria, squint, or watering eye with adult algorithms and missing retinoblastoma, congenital glaucoma, or amblyopia risk.
Adult ophthalmology dominates study time, and paediatric presentations are described subtly in stems (e.g., 'parents noticed a white reflex in photographs').
What to do instead
Create a paediatric safety checklist: white pupillary reflex, new-onset strabismus after age 6 months, epiphora with photophobia and corneal enlargement, and unexplained poor vision. For each, memorise the must-not-miss diagnosis, the first-line investigation, and the urgency. Rehearse the red-flag list until it is automatic.
Under-weighting the trauma or emergency vignette and selecting an outpatient workup when the stem describes a chemical injury, intraocular foreign body, or open globe.
The stem is long and the urgent clue is buried mid-paragraph, so candidates allocate attention to the later, more detailed history and miss the time-critical element.
What to do instead
Read the first and last sentence of every stem first, then scan for a trauma or acuity red flag. Memorise the emergency triad of actions for chemical injury (immediate copious irrigation before any examination), open globe (shield, no pressure, antibiotics, theatre), and intraocular foreign body (CT, no MRI). Time yourself answering ten trauma stems in under ten minutes to build the reflex.
Confusing the treatment ladder for common chronic disease: mixing up first-line, add-on, and surgical thresholds in glaucoma, diabetic macular oedema, or dry eye, and choosing an option that is one step too early or too late.
Guidelines are learned as lists without a decision rule, so the candidate cannot tell which rung of the ladder the stem's clinical details place the patient on.
What to do instead
For each major chronic condition, draw a ladder with the trigger for moving up (target intraocular pressure not met, persistence of oedema, failure of conservative measures) and the contraindication at each step. Learn the specific numbers that trigger escalation rather than vague phrases. Test yourself by writing the ladder from memory, then checking against a current guideline.
Selecting an investigation that is contraindicated or uninformative in the given context, such as requesting an MRI in a metallic intraocular foreign body, or ordering fluorescein angiography in a patient described as pregnant or in renal failure without noting the constraint.
Investigations are memorised by disease rather than by patient-specific safety and yield, and the stem's constraint (allergy, pregnancy, renal function, pacemaker) is treated as background detail.
What to do instead
For every investigation you revise, record three things: what it shows, its absolute contraindications, and the safer alternative. Build a short contraindication table (fluorescein, indocyanine green, MRI, contrast CT, mydriatics in narrow angles). In practice questions, highlight any comorbidity in the stem and check it against your table before answering.
Losing marks on the non-clinical components: ethics and consent scenarios, audit and research interpretation, or health-system resource questions, by treating them as common sense and not preparing at all.
Candidates assume the exam is purely clinical, so they spend no time on consent in an incompetent adult, confidentiality in a visually impaired driver, or interpreting sensitivity and specificity in a screening question.
What to do instead
Allocate a fixed block of revision to professionalism, consent, capacity, and driving or fitness-to-work rules, plus basic biostatistics (sensitivity, specificity, positive predictive value, number needed to treat). For ethics items, apply a four-step frame: identify the conflict, identify the decision-maker's capacity, consider the least restrictive option, and document. Practise ten ethics vignettes and write the reasoning explicitly.
Rushing the final third of the paper and changing correct answers on review without a stated reason, producing clusters of errors in the last items.
Time is unevenly distributed, difficult early items consume the budget, and anxiety drives answer-changing without new evidence.
What to do instead
Adopt a two-pass strategy: answer all items at a fixed pace, flag uncertain ones, and reserve the final minutes for flagged items only. Change an answer only when you can name the specific fact you misread. Simulate full-length papers under timed conditions at least three times before the exam and record your error pattern by section.
Failing to commit key numerical thresholds and classifications to memory, so options differing only by a number (intraocular pressure level, cup-to-disc ratio, duration of symptoms, grading of diabetic retinopathy or retinopathy of prematurity) cannot be separated.
Candidates understand concepts but rely on approximate recall, and exams deliberately offer near-miss numbers that punish approximation.
What to do instead
Build a one-page numerical sheet covering the thresholds you meet repeatedly (glaucoma pressure and disc criteria, retinopathy of prematurity staging and treatment criteria, diabetic retinopathy grading, keratoconus staging, paediatric acuity norms by age) and review it daily for the final two weeks. Convert each number into a flashcard with a clinical consequence attached so it is memorable rather than abstract.