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 names the interacting pair, the shared enzyme, the direction and timing of the effect, and the planned action. [1] [7]

Candidate brief

You are the general paediatric registrar caring for a 14-year-old girl who is four years post-liver-transplant. Speak as you would on a ward round. Anticipate interactions before you prescribe, name the enzyme and the direction, and reconcile every medicine at every transition. Each escalation branch leads to the next update. [7]

Question 1 — The new prescription

Stimulus update. Priya is on twice-daily oral tacrolimus with trough levels in target range, plus an SSRI for anxiety and a combined oral contraceptive. She now has oral candidiasis and a lower-lobe pneumonia. The on-call team plans to start fluconazole and clarithromycin. Question: What do you do before the prescription is signed? [1]

Consultant-level model answer. "Before signing I run an interaction check against her list. Tacrolimus is a CYP3A4 substrate with a narrow therapeutic window. Both fluconazole and clarithromycin are potent CYP3A4 inhibitors, so together they will rapidly raise her tacrolimus level and risk nephrotoxicity and neurotoxicity. I would avoid this combination: for the candidiasis I use a non-azole such as nystatin, and for the pneumonia I substitute azithromycin, which is a weak CYP3A4 inhibitor and the safer macrolide. If an azole were unavoidable, I would reduce the tacrolimus dose preemptively and check the level within a few days." [1] [2]

Probing follow-up. "How quickly would the tacrolimus level rise, and how does that differ from induction?" A strong answer is: "Inhibition acts on existing enzyme, so the rise begins within hours to days and toxicity can appear within a week. Induction requires new enzyme synthesis, so an inducer such as rifampicin lowers the substrate level over one to two weeks. I state both time courses whenever I discuss an interaction." [1] [2]

Common weak answer. "I will start both drugs and check the tacrolimus level if she becomes unwell." This ignores the predictable interaction until harm has occurred — anticipation is the safety habit. [1]

Escalation branch. If the candidate anticipates the interaction and substitutes, release in Question 2 that she is also on an SSRI and a contraceptive. If they prescribe blindly, ask which enzyme each drug shares with tacrolimus. [1]

Question 2 — The serotonergic and contraceptive layers

Stimulus update. Her full list also includes sertraline and a combined oral contraceptive. Question: What other interactions must you now consider, and does anything change? [6]

Consultant-level model answer. "Two further interactions. First, a macrolide with her SSRI adds a small QT risk, and if tramadol were added for pain the serotonergic combination could cause serotonin syndrome — I would avoid tramadol and use paracetamol or an NSAID if renal function allows. Second, if her pneumonia treatment were later changed to include rifampicin, that would induce CYP3A4 and lower her contraceptive oestrogen, risking failure — but rifampicin is not part of this regimen. The polypharmacy here is the core risk: twelve medicines create combinatorial interacting pairs, so a full reconciliation and a single shared list matter." [6]

Probing follow-up. "Why is polypharmacy more than a linear increase in risk?" A strong answer is: "Because each added medicine creates new pairs, the number of potential interactions rises combinatorially — a child on twelve medicines has far more than twelve times the interacting pairs of a child on one." [6]

Common weak answer. "She only needs antibiotics, so the other drugs are irrelevant." This ignores that every concurrent drug is a candidate substrate for the new prescription. [6]

Escalation branch. If the candidate works through the layers, move to Question 3 on reconciliation. [7]

Question 3 — Reconciling the list

Stimulus update. On admission her medicines list was copied from the referral letter. Question: How do you build a reliable list, and why does the single source matter? [7]

Consultant-level model answer. "A referral letter alone is not a reconciliation. I build the Best Possible Medication History from at least two sources — the family interview, the community pharmacy record, the GP summary, and any prior discharge letter. For every medicine I record name, dose, route, frequency, last dose, and adherence, and I ask specifically about over-the-counter and herbal products, because St John's wort is a potent CYP3A4 inducer easily missed. I compare this list against every admission order and flag omissions, duplications, wrong doses, and interactions for resolution with the prescriber." [7]

Probing follow-up. "Why ask about herbal products in a transplant patient?" A strong answer is: "Because unrecorded St John's wort would induce CYP3A4 and lower her tacrolimus toward rejection, and the only way to catch it is a structured reconciliation that explicitly asks." [7]

Common weak answer. "The referral letter has her list, so I will use that." This propagates any error already in the letter and misses unrecorded medicines. [7]

Escalation branch. Move to Question 4 on discharge. [10]

Question 4 — Safe discharge

Stimulus update. She improves and is ready for discharge on a changed regimen. Question: Describe your discharge reconciliation and communication. [10]

Consultant-level model answer. "I produce a verified, reconciled discharge list that records and explains every change made during the admission. I explain it to Priya and her family in plain language, give it to them in writing, and confirm it has reached her GP and community pharmacy before she leaves. The evidence shows plain-language, health-literacy-informed discharge communication reduces medication errors in hospitalised children. I set explicit tacrolimus monitoring review points with the transplant team and document the indication, dose, and duration for each new or changed medicine." [10]

Probing follow-up. "What is the single principle you carry forward?" A strong answer is: "Anticipate the interaction before you prescribe, and reconcile every medicine at every transition — collect, compare, communicate, and monitor." [1] [10]

Common weak answer. "I will print the discharge script and the family will collect it from the pharmacy." This omits plain-language explanation and confirmation that the GP and community pharmacy received the reconciled plan. [10]