The Dermatology SCE is unusually image-dependent, and that produces a specific and avoidable failure. A photograph appears, your brain produces an instant impression, and the option list contains that impression plus four conditions that look very like it. If your first move is to search the options for the diagnosis you have already decided on, you have skipped the only step that would have discriminated between them. The examiners are explicit that everything you need is in the image and the stem, which is a promise and a warning: the discriminator is there, and it is your job to find it before you commit.
Key takeaways
- The exam is two papers of 100 best-of-five questions, three hours each, with a break, and no negative marking.
- Describe the lesion in a fixed sequence before you allow yourself to name a diagnosis.
- The discriminating feature is in the image or the stem, because the examiners put it there deliberately.
- Negative findings are information: what is absent frequently separates the lookalikes.
- Separate recognition from management, because they are different competencies and fail differently.
The fixed sequence
Run the same description every time, before you look at the options. It costs fifteen seconds and it is the entire technique.
Distribution and site. Where is it, and is the pattern flexural, extensor, photo-exposed, acral, dermatomal, symmetrical or asymmetrical, localised or generalised? Distribution alone eliminates several options in most questions.
Configuration. Are the lesions discrete, confluent, grouped, annular, targetoid, linear, reticulate or serpiginous? Grouping and shape are among the most powerful discriminators in dermatology and they are frequently the whole answer.
Primary lesion morphology. This is the core of it. Macule, patch, papule, plaque, nodule, vesicle, bulla, pustule, wheal. Naming the primary lesion correctly is the single most useful thing you do in the question, and it is the step candidates skip when they think they have already recognised the condition.
Secondary change. Scale, crust, erosion, ulceration, excoriation, lichenification, fissuring, atrophy. These tell you about chronicity and about what the patient has been doing to the lesion.
Colour, and whether it blanches. Erythematous, violaceous, hyperpigmented, hypopigmented, purpuric. And in the exam as in the clinic, purpura is a different category from erythema and it changes the differential entirely.
Only now look at the options.
Why this works
Naming a diagnosis before describing is pattern-matching, and pattern-matching is exactly what a well-written distractor set is designed to punish. The four wrong options are chosen because they resemble the right one, so a first impression will find several of them plausible and provide no basis for choosing between them.
A description, by contrast, produces constraints. Once you have committed to "grouped vesicles on an erythematous base in a dermatomal distribution", most of the option list has eliminated itself, and you have arrived by reasoning rather than by recognition. That reasoning is reproducible on the next question, whereas your first impression is a lottery.
The negatives are doing work
The examiners construct these stems carefully, and one of their favourite tools is the absent finding.
No itch. No systemic symptoms. No mucosal involvement. No nail changes. No preceding illness. A soft, non-tender lesion. These lines are not padding. Each one is doing the work of excluding an option, and candidates who read the stem for what is present and skim past what is explicitly absent are throwing away the discriminator they were handed.
Treat every stated negative as an instruction. If the examiners tell you the mucosa is spared, they are telling you which option to eliminate.
Recognition and management are separate competencies
The exam asks two distinct kinds of question about the same photograph, and candidates are frequently good at one and poor at the other.
Recognition questions ask what this is. They are won by the descriptive method above, plus a large visual vocabulary built from seeing many examples.
Management questions show you an image, assume you have recognised it, and ask what you do: which investigation, which treatment, which escalation, what to monitor, when to biopsy, when to refer, what to do when first-line therapy fails.
Track your performance on the two separately. A candidate who recognises everything and cannot manage it has a therapeutics gap, and no amount of looking at more photographs will help. A candidate who manages beautifully once told the diagnosis, and cannot get to the diagnosis, needs visual volume rather than more guidelines. These are opposite problems with opposite remedies, and a single accuracy figure conceals both.
Build comparison sets
The most efficient way to build discrimination is to study lookalikes together rather than separately.
Take the conditions that are genuinely confused in practice and in exams, put their images side by side, and write down the single feature that separates them. The scaly plaques that are psoriasis and the ones that are not. The annular lesions that are fungal and the ones that are granulomatous. The blistering conditions, which are separated by the level of the split and by the distribution rather than by how they look at a glance. The pigmented lesions where the discriminator is asymmetry, border, or a specific dermoscopic feature.
Studying a condition in isolation teaches you to recognise it when it is the only thing on the page. Studying it against its mimics teaches you to recognise it when four mimics are in the option list, which is the situation you will actually be in.
Do not neglect skin of colour
A specific and important point. Dermatological signs present differently across skin types, and erythema in particular is far less obvious in darker skin, where the same inflammation may appear violaceous, grey or simply as a change in texture rather than as redness.
If your visual vocabulary has been built predominantly on images of lighter skin, you have a real and consequential gap, and it is one the exam and clinical practice will both expose. Seek out image sets that span skin types deliberately, and when you learn a condition, learn what it looks like across the range rather than in a single presentation.
Where iatroX fits
iatroX's Dermatology SCE bank covers both halves of the exam, the recognition and the management, tracked separately so you can see which of the two is actually costing you marks rather than blending them into one figure. Explanations are grounded in current guidance so the management reasoning sits with the image, and missed questions can be opened in the Socratic Tutor, which asks you to describe and reason before it explains, which is precisely the discipline that image questions demand and that pattern-matching bypasses. Try it with free sample questions at iatroX. For the underlying descriptive vocabulary, see how to describe a rash like a dermatologist.
Frequently asked questions
How should I approach dermatology image questions? Describe before you diagnose. Work through distribution, configuration, primary lesion morphology, secondary change and colour, in that fixed order, before you look at the options. Naming the condition first is pattern-matching, which is exactly what the distractors are built to punish.
Why do I keep choosing the wrong lookalike condition? Because you recognised rather than reasoned. The four wrong options resemble the right one deliberately, so a first impression cannot discriminate between them. A systematic description produces constraints that eliminate most of the list.
Do negative findings in the stem matter? Enormously. Absent itch, spared mucosa, no systemic symptoms and similar statements are placed deliberately to exclude specific options. Skimming past them discards the discriminator you were given.
Should I revise dermatology recognition and management separately? Yes, and track them separately. They are different competencies with opposite remedies: a recognition gap needs visual volume, and a management gap needs therapeutics. A single accuracy figure hides which one you have.