Infective Keratitis: Workup Algorithm and Exam-Ready Answers

The best answer to an infective-keratitis stem is description before diagnosis. State the visual status and the corneal findings in a reproducible order, identify the risk context, then explain why sampling, escalation and treatment decisions depend on severity and local protocol. A feathery infiltrate, a contact-lens history or severe pain can change the differential; none independently proves an organism.

This is a theory-and-viva framework for postgraduate doctors. It is not a treatment prescription, a substitute for local microbiology guidance, or advice for patients with a painful red eye. Microbial keratitis can threaten sight and requires clinical assessment. The AAO Bacterial Keratitis Preferred Practice Pattern is the primary guideline source used for the general assessment and culture principles below.

The 30-second answer framework

If an examiner presents a corneal ulcer, answer in this sequence:

1. Assess and document severity: visual acuity, location, size in two dimensions, depth, epithelial defect, infiltrate, thinning, perforation risk and anterior-chamber reaction.
2. Add context: contact lenses, trauma including vegetative matter, ocular-surface disease, previous surgery, topical medication exposure, immune status and water exposure are history points, not diagnoses.
3. Describe, then rank a differential: use morphology to say what is more likely and what must be excluded.
4. State investigation logic: significant, central, deep, atypical, sight-threatening or treatment-unresponsive disease may need corneal sampling and culture according to local protocol and senior supervision.
5. State escalation: worsening disease, thinning, perforation concern, severe inflammation or diagnostic uncertainty raises urgency.

The order matters. A candidate who opens with “This is Pseudomonas” has ignored the possibility that the question is testing how to describe, culture or triage a potentially serious ulcer.

What to document at the slit lamp

Use a fixed grid. It stops you from omitting the detail that differentiates a peripheral small infiltrate from a central progressive ulcer.

An answer such as “There is a 3 × 2 mm paracentral stromal infiltrate with overlying epithelial defect, surrounding oedema, anterior-chamber reaction and no obvious perforation” is more useful than “There is a corneal ulcer.” It is also easier for an examiner to mark because every descriptor can be checked.

A workup algorithm for exams, not a regimen

Step 1: decide whether the stem is infectious, sterile or uncertain

Pain, epithelial loss and infiltrate raise concern for infectious keratitis, but a question may include peripheral sterile inflammatory disease, marginal keratitis, a healing abrasion, toxic epitheliopathy or an immune corneal process as distractors. Do not diagnose infection solely from redness. The history, corneal morphology, distribution and inflammation must fit together.

Use language that reflects uncertainty: “The appearances are concerning for microbial keratitis; I would document severity and seek microbiological correlation where indicated.” This is stronger than a false binary because real morphology overlaps.

Step 2: identify risk context without turning it into proof

Contact-lens use is an important context for bacterial keratitis and is frequently paired with Pseudomonas in exams. It should make you consider bacterial disease and ask about lens wear, hygiene, overnight wear and water exposure. It does not prove Pseudomonas from history alone.

Vegetative trauma and a dry, raised infiltrate with feathery edges may raise suspicion for fungal keratitis. Satellite lesions are another classic clue. The fungal keratitis review makes the essential point: clinical appearance helps, but overlap and diagnostic delay are real. In a viva, call it a fungal pattern to consider and say that investigation supports the diagnosis.

Contact lens or water exposure with severe pain and a ring infiltrate is a classic prompt for Acanthamoeba keratitis. However, “pain out of proportion” and ring infiltrate are not mandatory early signs. The Acanthamoeba review is useful precisely because it shows why one remembered sign should not dictate certainty.

For HSV epithelial disease, a branching epithelial lesion with terminal bulbs is a classic teaching clue. A pseudodendrite has different morphology and context. In an MCQ, the distinction may be enough; in a clinical answer, avoid treating it as an excuse to prescribe without diagnosis-specific supervision.

Step 3: say when microbiology enters the answer

The AAO guideline supports microbiological work-up for selected ulcers rather than treating culture as a ritual for every minor peripheral lesion. In an exam, use condition-based language: “I would arrange appropriate corneal sampling in a central, large, deep, atypical, sight-threatening, contact-lens-associated or non-responsive ulcer, according to local protocol.” A question may use a different institutional threshold; do not invent an exact size cutoff if the stem does not supply one.

Sampling has a purpose. Smear and culture can guide organism-directed decisions, identify unusual pathogens and provide a basis for changing a plan when the clinical course is discordant. It does not guarantee that every culture will be positive, and a negative result does not erase the clinical picture. This is the conceptual point behind many examination questions.

Step 4: distinguish progression from a static label

When asked about follow-up, name the variables that show whether the situation is improving or worsening: symptoms, visual acuity, infiltrate dimensions/density, epithelial defect, thinning, anterior-chamber reaction and any new extension. “The ulcer looks worse” is not an adequate record. A measured change is.

Pattern-recognition table: useful, but confidence-labelled

The key phrase is “consider.” It tells the examiner that you recognise the morphology and understand the limit of morphology. It also protects your revision from the common error of converting a classic association into a rule.

Questions examiners ask, with answer structures

“How would you examine and document this case?”

“I would first record visual acuity and examine both eyes. At the slit lamp I would document ulcer site, infiltrate and epithelial-defect dimensions, depth, edges, stromal oedema, thinning or perforation risk, corneal sensation if relevant, anterior-chamber reaction including hypopyon, and lid/conjunctival findings. I would then obtain risk history, including contact lenses, trauma, surface disease, prior surgery and topical medication exposure.”

This answer is chronological and objective. It is better than reciting a list of organisms.

“When would you scrape or culture?”

“I would follow local ophthalmology and microbiology protocol. In a clinically significant, central, deep, large, atypical, sight-threatening, contact-lens-associated or treatment-unresponsive ulcer, I would consider corneal scraping for smear and culture before or alongside directed management as locally indicated. The purpose is to support organism identification and guide subsequent decisions.”

The phrase “as locally indicated” matters because testing methods, availability and thresholds vary. It is neither evasive nor unsafe.

“How do you differentiate bacterial from fungal keratitis?”

“I would not claim definitive differentiation by appearance alone. A suppurative ulcer in a contact-lens context favours bacterial disease; feathery margins, a dry raised infiltrate, satellite lesions and vegetative trauma support a fungal differential. I would use the full examination and microbiological correlation where indicated.”

“Why is topical steroid history important?”

“It may alter the clinical appearance and course of infectious disease, so I would ask specifically and make diagnosis-specific decisions with senior/local-protocol input. I would not make a blanket treatment statement from the viva stem.”

That is a better answer than declaring steroids always right or always wrong. The point is diagnostic and safety awareness.

High-yield MCQ traps

A hypopyon is not an organism label. It demonstrates significant anterior-chamber inflammation. The stem still needs morphology, context and investigation.

Feathery edges are a clue, not a culture result. Fungal keratitis belongs in the differential; the sign is not a licence to ignore bacteria or mixed disease.

Contact lens history is not equal to Acanthamoeba. It is relevant to several microbial risks. Water exposure and disproportionate pain may raise the question, but no single sign is definitive.

A dendrite is not every branching lesion. Terminal bulbs, staining pattern and context matter. Be ready to distinguish a true dendrite from a pseudodendrite.

“No growth” does not mean “not infection.” Sampling quality, prior treatment and organism characteristics influence results. The exam lesson is to interpret a result in clinical context.

Do not invent a drop schedule. A paper may ask the broad principle, but regimen choice belongs to the clinical setting, diagnosis, severity, local resistance pattern and supervision. A source-led exam answer recognises this boundary.

A one-minute viva close

If you need to summarise, use this:

“This is a potentially sight-threatening corneal process. I would document visual function and the ulcer objectively, identify risk factors and morphology, and create a differential rather than assume an organism. For a significant or atypical ulcer, I would arrange appropriate microbiological sampling and urgent ophthalmology management according to local protocol. I would monitor measured changes in infiltrate, epithelial defect, thinning, anterior-chamber inflammation and vision.”

That response answers assessment, investigation, urgency and monitoring without pretending that a viva image permits a complete clinical plan.

For the wider cluster, review the cornea study guide. To rehearse concise presentation sequences, the site’s OSCE, Practical & Viva Voce Ready bundle is the relevant product page. You can also use general ophthalmology MCQs for retrieval practice. These links are educational resources; OphthaMCQ does not provide patient-specific management advice.

Build an error log that improves your next answer

After every keratitis question, record more than the correct organism. Use four fields: morphology missed, risk-context missed, investigation principle missed, and overclaim made. For example, “I recognised a ring infiltrate but treated it as diagnostic of Acanthamoeba” is an overclaim error. The repair is to learn the full pattern and its limitation, then answer two fresh questions in which the same sign has a different context. “I forgot to document thinning” is a process error; the repair is to rehearse the slit-lamp grid aloud. This turns a single wrong MCQ into a reusable clinical-reasoning habit without converting an educational article into a treatment protocol.

At the end of a mixed block, take one stem from each group—bacterial, fungal, Acanthamoeba, viral and sterile peripheral differential—and give a 30-second answer without naming an organism until you have described the ulcer. That constraint is a simple way to practise the discipline examiners reward.

Sources

1. American Academy of Ophthalmology: Bacterial Keratitis Preferred Practice Pattern — guideline source for assessment, culture and severity principles; checked 18 August 2026.
2. Bacterial Keratitis, StatPearls — examination and documentation cross-check; checked 18 August 2026.
3. Fungal keratitis review — fungal morphology and diagnostic limits; checked 18 August 2026.
4. Acanthamoeba keratitis review — clinical features and limitations of classic signs; checked 18 August 2026.