When a child does not improve, the next question is not automatically which treatment to add. Reconstruct what was expected, what actually happened and which new information changes the working explanation. MRCPCH AKP revision should include that reassessment step rather than repeatedly rewarding the first plausible diagnosis.
This article is published by iatroX and discusses using Pastest and iatroX alongside official paediatric preparation material. The case below is an original learning exercise, not a recalled examination question or a replacement for assessing a child in practice.
Write down the expected course before seeing the update
A sequential case works best when the learner commits to an expectation. After the first description, write the leading explanation, its supporting evidence and the change that would make you reconsider. Do not wait until the new finding appears and then pretend it was part of your original plan.
This distinguishes reasonable uncertainty from hindsight. A working diagnosis can be sensible at one point yet need revision later. Equally, a later diagnosis does not prove that every earlier clinician had enough information to reach it. Your review should follow the information available at each stage.
Ask separately whether the child is failing to follow the expected course, becoming more unwell, or developing features that point elsewhere. These are overlapping but different reasons to reassess. A longer list of investigations is not automatically a better response to any of them.
An evolving fever case
At the beginning of this fictional exercise, a preschool child has fever and a short history of non-specific symptoms. The learner is asked to identify the information needed to assess severity and formulate a provisional differential. The stem does not provide enough evidence for a definitive diagnosis.
At review, the fever has persisted for at least five days. New information includes red eyes, changes affecting the mouth and a rash. The parent says the child remains unusually irritable. The question is now what needs reconsideration, not whether the original label can be defended indefinitely.
The NHS description of Kawasaki disease, checked on 13 September 2026, identifies prolonged fever with features including eye, mouth and skin changes. It also explains that diagnosis is not established by a single test. These findings warrant appropriate assessment; the abbreviated exercise should not be used as a diagnostic checklist that excludes all alternatives.
A useful answer explains why the updated pattern changes the differential and urgency of review. An inadequate answer simply repeats the original reassurance because it was reasonable earlier. Another inadequate answer announces a definitive diagnosis without acknowledging the missing examination and assessment information.
Change the follow-up, not just the disease name
Run a second version in which the fever settles, the child is drinking normally again and activity is returning towards baseline. Ask what supports improvement and what information would still be needed. This stops the exercise teaching the simplistic rule that every reattendance proves a missed rare condition.
In a third version, the parent reports that the child never received the proposed intervention because the instructions were misunderstood. Now the question concerns what actually happened, not an assumed biological failure of treatment. Establish the facts before describing the condition as unresponsive.
These variations test different reasoning errors: anchoring on the first explanation, overcorrecting towards a dramatic alternative, and judging response to an intervention that was not delivered. They should not all produce the same revision note.
A reassessment worksheet that preserves the timeline
Complete the first two columns before receiving the follow-up. A partner can reveal the update only after you have committed to an initial account.
| Stage | Record |
|---|---|
| Initial representation | Age, presentation, severity information and uncertainty |
| Expected course | The change you expected and why |
| New evidence | What is genuinely new, rather than reworded old information |
| Delivery check | What actually occurred after the initial plan |
| Revised interpretation | Which explanation gains or loses support |
| Immediate concern | What now needs assessment or escalation |
| Next review question | What you need to know to judge the revised plan |
Do not erase the original entry. Draw a line beneath it and add the update. Being able to show why your view changed is more educational than leaving only a polished final answer.
Use disagreement in a study group properly
Give one learner the first stage and another the follow-up. Ask the second learner to identify what they would need from the first clinician before taking over. This introduces a useful distinction between a short case summary and a complete history of every thought considered.
Observers should record the evidence used to change the plan. "Needs to be more confident" is poor feedback if the real issue is failure to recognise a new feature. "You continued the old explanation without addressing the changed course" gives the learner something specific to practise.
The exercise does not require a consensus diagnosis in every incomplete case. Agreement that further assessment is needed can be the defensible conclusion. What matters is whether the uncertainty has been identified and translated into an appropriate next task.
What to use after a missed AKP question
The RCPCH examination resources, checked on 13 September 2026, remain the reference for official preparation material. Pastest advertises MRCPCH resources, and BMJ's AKP bank is another publicly described option. This article has not compared their private case collections.
For a missing disease concept, return to a reliable explanation before attempting another complex sequence. For a failure to revise your initial view, use a changed case and explain the transition aloud. More questions on the same static presentation may miss the actual weakness.
Per the September 2026 iatroX specification, targeted question practice can be followed by Socratic Tutor questions about the attempted answer and planned revisiting through the study planner. This is supplementary applied-knowledge practice, not a substitute for supervised paediatric experience.
The September 2026 combined subscription costs £99 upfront annually, equivalent to £8.25 a month billed annually, or £29 monthly, with banks, Tutor, planner, simulations and CPD included. Retain a resource that teaches the topic well; add guided reasoning when your difficulty is updating the decision after new information rather than remembering the original explanation.
Frequently asked questions
Does a child failing to improve always mean the original diagnosis was wrong?
No: reassess the diagnosis, severity, expected course and what actually happened after the initial plan. Non-improvement is a reason to investigate the explanation, not proof of one particular alternative.
Should every sequential case end with a rare diagnosis?
No: use improving and unchanged versions as well as deteriorating ones. Otherwise the exercise teaches anticipation of a twist rather than attention to the evidence.
Can iatroX replace supervised assessment of children?
No: question practice and tutoring support learning. Examination, observation and clinical decisions require the appropriate real-world supervision and training.