Branching cross-examination

This is a MedVellum formative viva. It is not an official RACP, MRCPCH, ABP, ACGME or RCPSC station, mark scheme, duration or pass standard. Release each update only after the candidate states the frame, immediate action and residual risk. [14] [20]

Candidate brief

You are the paediatric registrar in a mixed ED and short-stay service. Speak as you would at the bedside. Keep known diagnoses on the table without letting them close the case. [1] [10]

Question 1 — Frame the problem

Stimulus update. An 11-year-old with intellectual disability, epilepsy, GORD and chronic constipation arrives with two days of hand-biting and poorer sleep. The triage note says “behavioural — known disability.” Question: What is wrong with that note, and how will you reframe the encounter? [1] [3]

Consultant-level model answer. “The note risks diagnostic overshadowing. A known disability label is being used as a full explanation for a new change. I reframe this as a multimorbid child with a new behavioural expression of possible pain or illness. I will establish personal baseline, open a ranked problem list and keep a threat-first differential. I will not discharge on ‘behavioural baseline’ alone.” [1] [3] [14]

Probing follow-up. “Define multimorbidity versus comorbidity in one sentence each.” Strong answer: “Multimorbidity means two or more concurrent chronic conditions without requiring one index disease. Comorbidity usually means a secondary condition attached to an index disease.” [6]

Common weak answer. “This is just her disability, so psychology can see her later.” That closes medical search too early. [1] [13]

Escalation branch. If the candidate names overshadowing and reopens the differential, release caregiver detail in Question 2. If they accept the triage label, ask what treatable pains could present as self-injury. [13] [24]

Question 2 — Baseline and caregiver concern

Stimulus update. The usual carer says, “She usually greets me with eye contact and a smile. For two nights she is quiet, biting her hand, and her bowels have not opened.” First observations are near her usual band. Question: How do baseline and concern change your plan? [10] [17]

Consultant-level model answer. “I treat change from personal baseline as core data. Near-usual numbers do not cancel caregiver concern. I document the exact words, compare interaction and bowel pattern with usual, and raise diagnostic caution. I examine for constipation, retention, dental and ENT disease, abdomen, hips and skin. I keep infection and medicine effects open.” [13] [17]

Probing follow-up. “Does caregiver concern have one universal likelihood ratio?” Strong answer: “No. Prospective work links concern with critical illness risk, but I do not invent a global LR. I investigate the concern and provide an escalation route.” [17]

Common weak answer. “Observations are fine, so I can reassure and discharge.” This discards baseline and concern. [17] [18]

Escalation branch. If baseline is used correctly, reveal polypharmacy in Question 3. [11]

Question 3 — Polypharmacy

Stimulus update. The medicine list includes two antiseizure medicines, a PPI, a laxative often missed at home, an antihistamine and a recent new sedating agent for “behaviour.” Question: What are your medication safety actions? [11]

Consultant-level model answer. “I reconcile every medicine, dose, timing and recent change. I check for anticholinergic burden, sedation and interactions. I ask about missed laxative doses and home administration technique. I consider whether the new sedating agent is a prescribing cascade. I plan simplification with the usual prescribers when safe, and I teach a clear home plan.” [11]

Probing follow-up. “Why can polypharmacy mimic new disease?” Strong answer: “Adverse effects and interactions can look like infection, encephalopathy, constipation crises or behavioural escalation, then trigger more medicines.” [11]

Common weak answer. “Add another medicine for agitation tonight.” That deepens the cascade. [11]

Escalation branch. If reconciliation is named, reveal examination findings in Question 4. [13]

Question 4 — Examination and pain search

Stimulus update. The abdomen is soft but the lower abdomen is full. There is dental tenderness on the left and marked stool loading on a care-plan review. Interaction improves briefly after positioning and a toilet trial plan. Question: What is your working synthesis now? [13] [24]

Consultant-level model answer. “I synthesise possible dual medical drivers: constipation with retention risk and dental pain, on a multimorbid background. I treat both while watching for infection or medicine toxicity. I do not call this pure behaviour. I use a structured pain approach because self-report is limited.” [13] [24]

Probing follow-up. “Which AAP clinical report anchors pain assessment in significant CNS impairment?” Strong answer: “Hauer and colleagues’ AAP clinical report on pain assessment and treatment in children with significant CNS impairment.” [13]

Common weak answer. “Dentistry can wait; this is behavioural.” Delayed pain care is overshadowing. [24]

Escalation branch. If medical drivers are treated as primary, move to communication and handover in Question 5. [20]

Question 5 — Residual risk and handover

Stimulus update. After disimpaction planning, analgesia and dental review, the child is more settled but not fully back to baseline. Night staff want a one-line handover. Question: What will you say? [14] [20]

Consultant-level model answer. “Eleven-year-old with multimorbidity — intellectual disability, epilepsy, GORD, constipation — presenting with new self-injury. Working drivers: constipation and dental pain. Residual risks still open: infection, medicine effect, seizure change and incomplete pain control. Meds reconciled; missed laxative doses noted. Review at set time or sooner for caregiver concern, fever, reduced interaction or breathing change.” [11] [20]

Probing follow-up. “Why is ‘stable complex child’ unsafe?” Strong answer: “It closes residual risks, erases baseline comparison and invites the next shift to overshadow again.” [18] [20]

Common weak answer. “Behavioural — known disability — no medical issues.” Critical fail pattern. [1] [14]

Escalation branch. End with synthesis if residual risks are explicit. If not, require rewrite of the handover sentence. [20]

Terminal synthesis

A strong candidate defines multimorbidity and overshadowing, uses baseline and caregiver concern, runs a medical pain/illness search, reconciles polypharmacy, and hands over open residual risks without discrimination against complexity. [1] [10] [11] [13] [17]