A question bank can prove you know how to assess a child's development, when to raise a safeguarding concern, how to interpret a growth chart, and what good clinical communication contains. It cannot prove you can do any of those things with a real child, a worried parent and a role-player in an eight-station OSCE. That gap between knowing and performing is exactly what the DCH Clinical tests and what MCQ practice, including iatroX, cannot train. This article names the skills that fall in the gap and gives you a concrete way to build each one.
The DCH format map
The DCH (RCPCH) has two components. The Foundation of Practice (FOP) theory paper is 100 SBA questions in 2 hours, computer-based, no negative marking — the knowledge exam. The DCH Clinical is a separate circuit of eight OSCE-style stations, some using role-players as parents or adolescents, judged by experienced examiners across two annual exam periods (verify current structure and dates on the RCPCH pages; last checked 21 July 2026). An MCQ bank maps cleanly onto FOP. It does not map onto the Clinical — and that is the whole point of this article.
Knowledge is not performance
A correct selected answer proves one thing: that, given five options and a written stem, you could recognise the best one. It does not prove that you can elicit the relevant history from a parent under time pressure, structure a physical examination fluently, plot and interpret growth in real time, notice a safeguarding cue that no one has flagged for you, or convey a difficult message to an anxious family in plain language. The DCH Clinical assesses the doing; the bank assesses the recognising. Confusing the two is the most common reason strong theory candidates underperform in the OSCE. iatroX is honest about this: it is the FOP-level knowledge and unseen-MCQ layer, and it does not replace the DCH Clinical or any OSCE simulator.
The distinction is worth laying out item by item, because it shows exactly where a bank stops:
| The FOP paper can test that you… | The DCH Clinical tests whether you can… |
|---|---|
| recognise developmental red flags in a written stem | elicit a developmental history and interpret it live |
| know the safeguarding threshold and correct process | notice an unflagged cue and communicate a concern safely |
| interpret a growth chart already plotted on screen | plot, interpret and explain growth to a parent in real time |
| select the "best" phrasing in a communication MCQ | structure a consultation and respond to a family's emotion |
| identify the next investigation from a vignette | examine a child fluently while keeping them cooperative |
Every row on the left is bankable. Every row on the right needs a person, a clock and a rubric. No amount of practice in the left column moves you up the right.
The four skills MCQ practice cannot assess — and how to train each
For each skill: an observable behaviour (what an examiner sees), a deliberate-practice task, a feedback source, and an exit standard.
Development
- Observable behaviour: you take a focused developmental history across the four domains, use age-appropriate screening questions, and interpret findings against milestones — distinguishing delay from regression.
- Deliberate-practice task: run timed developmental-history stations with a colleague playing the parent of a child at a given age; then examine or interpret a developmental scenario.
- Feedback source: a paediatric clinician or trained examiner against the RCPCH station domains — not an app.
- Exit standard: you can complete a focused developmental assessment inside the station time and correctly classify normal, delayed or regressing development.
Safeguarding
- Observable behaviour: you recognise a safeguarding cue, respond without alarming or accusing the family, ask the right screening questions, and articulate the concern and the next step appropriately.
- Deliberate-practice task: rehearse safeguarding communication scenarios (an inconsistent injury history, a disclosure) with a role-player, then debrief on wording and escalation.
- Feedback source: a clinician experienced in child protection; sense-check the threshold and process against current UK statutory safeguarding guidance.
- Exit standard: you reliably identify the concern, communicate it safely, and state the correct escalation — without a script that collapses when the scenario shifts.
Growth
- Observable behaviour: you plot on a UK-WHO chart, interpret centiles and crossing, and explain faltering growth or a growth concern to a parent clearly.
- Deliberate-practice task: timed chart-plotting and interpretation stations, paired with an explanation to a role-player parent.
- Feedback source: a clinician checking both the technical plotting and the clarity of your explanation.
- Exit standard: accurate plotting and interpretation within the station time, plus a parent-facing explanation the role-player actually understands.
Clinical communication
- Observable behaviour: you structure the consultation, check understanding, respond to emotion, and share information (including consent and, with adolescents, confidentiality) at the right level.
- Deliberate-practice task: communication stations — breaking a diagnosis, explaining management, negotiating with an adolescent — recorded and reviewed.
- Feedback source: a clinician or communication tutor against the examiner-judged communication domain.
- Exit standard: the family or adolescent leaves the station informed and heard, and your structure holds when the scenario is unfamiliar.
The four-week modality ladder
Train performance in stages, not all at once:
- Week 1 — isolated skill. Drill each component alone: chart-plotting, a developmental-history opening, a safeguarding sentence, a consent explanation. Low pressure, high repetition.
- Week 2 — coached case. Run full stations with a colleague and stop for feedback mid-station. Fix structure before speed.
- Week 3 — timed integrated case. Full stations at real station timing, no stopping, feedback afterwards. Add the clock.
- Week 4 — unseen simulation. A fresh circuit of stations you have never rehearsed, ideally scored by someone using the official domain rubric. This is the closest thing to the exam and the truest readiness signal.
Throughout, keep an FOP-level knowledge base ticking over — an unseen, timed knowledge baseline in iatroX confirms the theory underneath the performance is current — but do not let question volume crowd out station rehearsal. The knowledge layer and the performance layer are different jobs.
How much station practice is enough?
There is no official minimum, and anyone quoting one is inventing it — the useful answer is defined by coverage and calibration, not by a count. You have done enough when your case matrix has no empty columns, when you have completed at least one full unseen circuit marked by a clinician against the official domains, and when your marks are stable rather than swinging from station to station. For most candidates that means practising well into double figures of distinct stations across the age bands and skill domains, with the uncomfortable combinations rehearsed more often than the comfortable ones. Quality and coverage beat volume: ten varied, honestly marked stations tell you far more than thirty repeats of the scenarios you already pass.
When AI feedback helps, when it misleads, and when a clinician is essential
AI feedback is useful for the mechanical and the factual: checking your knowledge on an FOP-style item, generating practice scenarios, drafting a structure to rehearse, or explaining why a written answer is wrong. It becomes unreliable the moment scoring depends on human judgement — the warmth of your safeguarding conversation, whether a parent felt heard, whether your examination was fluent. Automated scores of communication and examination performance are easy to game and hard to calibrate, so treat any number they produce with suspicion. A clinician or trained examiner is required for the performance domains: safeguarding thresholds, developmental interpretation, and communication quality all need a human who knows the DCH standard. If you use any automated feedback on performance, calibrate it against a clinician's mark first — the method is set out in how to calibrate automated feedback before you trust the score.
Build a balanced case matrix
The trap in station practice is rehearsing only the scenarios you find comfortable. Build a matrix so you cover the range: across age (infant, toddler, school-age, adolescent), domain (development, safeguarding, growth, acute presentation, communication), and station type (history, examination, explanation, data interpretation). Tick a cell only when you have practised it under time and had it reviewed. Blank cells are your next sessions — not the scenarios you keep re-running because they go well.
| Age band | Development | Safeguarding | Growth | Acute presentation | Communication |
|---|---|---|---|---|---|
| Infant | — | — | — | — | — |
| Toddler / pre-school | — | — | — | — | — |
| School-age | — | — | — | — | — |
| Adolescent | — | — | — | — | — |
Fill a cell only when that combination has been rehearsed under station time and marked by a clinician. Most candidates find, honestly completing this, that whole columns — safeguarding across the ages, or adolescent communication — are empty, while one comfortable box has been practised five times. The matrix turns "I've done lots of stations" into a specific, testable coverage claim.
Red flags your OSCE preparation is failing
- Memorised scripts that fall apart when the role-player deviates.
- Repeated cases — high scores because you have seen the scenario, not because you can perform.
- Generic feedback ("good rapport") that does not tell you what to change.
- Uncalibrated scoring — an automated or peer mark never checked against the official rubric.
- No official-rubric check — practising to your own standard instead of the examiner's domains.
Any of these means your practice is measuring familiarity, not competence.
Worked example: a strong theory candidate, four weeks out
Consider a GP registrar who scores well on FOP-style questions — comfortably above the pass mark on unseen, timed blocks — but has never rehearsed a station. On paper she looks ready; in a mock circuit she is not. In week one she isolates the components and finds the specifics: her chart-plotting is slow, her safeguarding opening is a memorised script that stalls when the role-player pushes back, and she runs over time explaining a diagnosis. Week two she runs coached stations and stops mid-station for feedback, rebuilding her safeguarding conversation so it survives an unexpected disclosure. Week three she adds the clock and practises full stations at real timing, discovering that her growth explanation is technically correct but loses the parent. Week four she sits an unseen circuit scored by a colleague against the official domains — the first honest readiness signal she has had. Her knowledge never changed; her performance did. A question bank would have told her, every day for four weeks, that she was already ready.
Reading your results
The lesson generalises. A high theory score and a low station score are not a contradiction — they are two different measurements, and only one of them predicts the DCH Clinical. If your bank percentage is strong but you have never been marked on a station by a clinician, you do not yet know whether you can pass the Clinical. Treat the two components as separate projects with separate evidence, and do not let comfort in one paper stand in for readiness in the other.
Bottom line
MCQ banks, iatroX included, do one job well: they build and measure the FOP-level knowledge under the DCH. They cannot assess whether you can perform — take a developmental history, hold a safeguarding conversation, interpret growth with a parent watching, or communicate clearly under examiner scrutiny. Those skills are trained by climbing the modality ladder from isolated drill to unseen simulation, with a clinician marking against the official rubric, not by another block of questions. Keep an unseen knowledge baseline ticking over to confirm the theory is current, but if the DCH Clinical is ahead of you, the marginal gains are almost entirely in coached, timed, human-marked station practice. Know which component you are actually preparing, and use the right tool for each.
FAQ
How do I know whether I have covered the full DCH blueprint? For the FOP theory paper, you cover the blueprint by auditing your SBA practice against every published syllabus area — but for the DCH Clinical, "coverage" means something different: it means every performance skill (development, safeguarding, growth, communication) and every station type has been practised under time and reviewed by a clinician. A complete question bank tells you nothing about the second kind of coverage, which is why theory-only candidates are so often caught out.
Can one question bank be enough for DCH? For the FOP theory component, a current well-matched bank can carry most of your knowledge preparation. For the DCH as a whole, no bank of any kind is enough, because the Clinical assesses performance an MCQ cannot reach. A bank is a necessary tool for one component and the wrong tool entirely for the other; treat "enough for the theory" and "enough for the diploma" as separate questions.
What should I measure instead of my overall Q-bank percentage for DCH? For the theory, measure per-domain accuracy, unseen-timed accuracy and high-confidence errors rather than a headline percentage. For the Clinical, the meaningful measures are performance-based: can you complete each station type within the time, and does a clinician marking against the official domains judge you competent? A percentage on a bank predicts neither of these, so do not let it stand in for OSCE readiness.
When should I stop doing new DCH questions? Stop adding new theory questions once your FOP blueprint is covered with even accuracy, your errors are worked through and your timing is controlled — after that, more questions add fatigue, not readiness. Crucially, stopping new questions does not mean you have stopped preparing: if the Clinical is still ahead of you, your remaining effort should move almost entirely to coached, timed and unseen station practice, which is where the marginal gains now are.
Which DCH resource should I use for my weakest component? Match the resource to the modality. If your weakness is FOP-level knowledge, use targeted unseen SBA practice and a Socratic tutor to interrogate the misses. If your weakness is performance — safeguarding communication, developmental assessment, growth explanation, or consultation structure — a question bank cannot help; you need coached station practice with a clinician against the official DCH Clinical rubric, climbing the four-week modality ladder from isolated skill to unseen simulation.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 21 July 2026. DCH structure and station details are RCPCH-reported and can change between diets — verify the current DCH Clinical format and dates on the RCPCH pages before relying on them. Disclosure: iatroX operates a competing UK question bank; this article positions it strictly as the FOP-level knowledge and unseen-MCQ measurement layer and states plainly that it does not train or replace the DCH Clinical OSCE — those performance skills need coached, clinician-marked station practice. Corrections via the feedback route on iatrox.com.
References: RCPCH — Diploma in Child Health (about), DCH Clinical exam (how to apply) and Theory exams (structure and syllabi). Internal: the DCH content-gap checklist, how to calibrate automated feedback before you trust the score and why your Q-bank percentage is not your exam score.