A symptom can belong to several diagnoses. Vomiting, poor feeding, wheeze and fatigue are useful starting points, but none should determine an answer on its own. In an AKP question, the strongest diagnosis usually explains the combination of findings and their time course with fewer unsupported assumptions than its nearest alternative.
The practical skill is comparison. Instead of asking only whether a diagnosis could cause one symptom, ask which details it explains well, which it leaves unexplained and what evidence would distinguish it from another possibility. The original teaching cases below show how this approach changes a decision.
Build a short problem statement
Compress the case into age, dominant problem, time course and discriminating features. Leave out details that do not currently affect the comparison. This is not an invitation to ignore information; it is a way to prevent a long stem from becoming a collection of equally weighted facts.
A useful problem statement might be: "A school-age child has acute vomiting after two weeks of thirst, frequent urination and weight loss." The order matters. Vomiting is recent, while the metabolic symptoms preceded it. That timeline immediately changes how you compare gastroenteritis with another cause.
Then name two or three plausible explanations. A focused comparison is usually more productive than producing a large differential without prioritisation. Keep an alternative available if a major feature remains unexplained.
Case one: vomiting after a longer metabolic story
An eight-year-old presents with vomiting and abdominal discomfort. A sibling recently had a brief vomiting illness. The child has also been unusually thirsty, passing urine frequently and losing weight for two weeks. Breathing is deep and rapid. In this illustrative case, capillary glucose is 24 mmol/L, blood ketones are 5 mmol/L and venous pH is 7.18.
The family contact makes gastroenteritis an understandable initial consideration. It does not explain the preceding polyuria, polydipsia and weight loss as well as new diabetes, and the biochemical findings support diabetic ketoacidosis. The combination requires urgent hospital management, rather than treatment as uncomplicated gastroenteritis.
NICE guidance on diabetes in children describes the relevant presentation and assessment of suspected ketoacidosis. For examination reasoning, the central point is to integrate the laboratory pattern with the longer history. A recent contact should not outweigh several findings that support a different, urgent explanation.
Explain why the alternative is weaker
The phrase "gastroenteritis is wrong" is a poor revision note. A better explanation is that it does not adequately account for the preceding metabolic symptoms and the supplied biochemical pattern. This makes the lesson transferable to another stem without the same wording.
It is also possible for an infection and metabolic illness to coexist. Choosing the best answer does not require claiming that the sibling's illness is irrelevant or that a child with diabetes cannot have gastroenteritis. It requires recognising which diagnosis explains the decision-critical findings and determines the immediate response.
This distinction prevents an overly rigid version of diagnostic reasoning. A unifying explanation is helpful, but the aim is clinical coherence, not forcing every detail into one label at any cost.
Case two: wheeze with a pattern of recurrence
Consider a twenty-month-old with wheeze during a respiratory illness. They have had several similar episodes, are well between them, and have eczema. On examination there is wheeze without crackles. The question asks which diagnosis best fits the overall pattern.
Bronchiolitis is a plausible consideration in a young child with respiratory symptoms. However, the age, recurrent episodic wheeze and atopic history support considering viral-induced wheeze or early-onset asthma. NICE bronchiolitis guidance specifically identifies these features as reasons to consider those alternatives in older infants and young children.
No single feature should be treated as a universal diagnostic switch. Eczema alone does not prove asthma, and the absence of crackles does not independently settle every case. The strength comes from the combined pattern. Severity assessment remains essential whichever label is most likely.
Compare patterns, not isolated keywords
Place the two leading diagnoses beside one another and ask which observations would be expected under each. Some features are shared and therefore contribute little to the distinction. Others change the balance substantially.
| Finding | How it helps the comparison |
|---|---|
| A common symptom shared by both diagnoses | Establishes plausibility but may not discriminate |
| A preceding history with a different time course | Can reveal a process that began before the apparent trigger |
| A coherent cluster of associated findings | Strengthens an explanation beyond a single keyword |
| An objective test matching the clinical pattern | Can substantially change the working diagnosis |
| A major unexplained feature | Requires reconsideration or a second process |
This table is a reasoning framework rather than a scoring system. Avoid assigning invented numerical weights to findings unless a validated tool or explicit calculation is part of the question.
Use negative findings carefully
A missing symptom can be helpful when it would normally be expected, but many conditions have variable presentations. An option that says a diagnosis is impossible because one classical feature is absent may overstate the evidence.
Ask whether the negative finding was actually assessed. "No history is provided" differs from "the history explicitly denies it". Likewise, a normal test may have limited relevance if it was performed before the current deterioration or does not measure the process being considered.
During review, identify whether you rejected a diagnosis because of a meaningful contradiction or because it did not resemble the textbook example you had memorised. That distinction often reveals the next learning task.
Keep diagnosis connected to the question's purpose
Sometimes the most important decision is an action that cannot wait for a complete explanation. If the child is unstable, the next step may concern immediate assessment or treatment. A lengthy comparison between two diagnoses should not distract from the urgent problem described in the stem.
In other questions, the task is to select the investigation most likely to distinguish competing explanations. State what result would shift your interpretation and whether it would change management. This keeps testing connected to a clinical purpose.
If the evidence is incomplete, acknowledge the missing discriminator in your reasoning rather than invent it. A question bank explanation should show how the available information supports the intended answer, not require assumptions absent from the case.
Practise with a case that changes one detail
Take the wheeze case and remove the recurrent episodes and atopic history. Then alter the child's age and auscultatory findings. Explain how each version changes the comparison. Repeat the exercise with the vomiting case by separating the acute illness from the preceding symptoms.
iatroX AKP questions, adaptive practice and Socratic Tutor can support this kind of structured review. Ask why the closest alternative is less satisfactory, then return later to a different case testing the same distinction. The useful result is a clearer explanation of the whole pattern, rather than a stronger memory of one original answer.
Frequently asked questions
Should I always choose the diagnosis that explains every finding?
Choose the explanation best supported by the case, while recognising that conditions can coexist. Do not force unrelated findings into one diagnosis or ignore an urgent problem because the whole picture is not yet complete.
How can I identify the most important discriminator?
Compare the leading alternatives directly. Look for a finding, time course or test result that one explains substantially better than the other, rather than a symptom common to both.
What should I do when a classical feature is absent?
Check how strongly its absence argues against the diagnosis and whether it was explicitly assessed. Many presentations vary, so a missing textbook feature does not automatically exclude a condition.
Practise comparing clinical explanations with the iatroX MRCPCH AKP question bank.