Adverse drug reactions and pharmacovigilance — branching viva

Opening (warm-up)

Examiner. A 7-year-old is admitted with a rash that began two weeks after starting lamotrigine. Take me through your first thought. [5]

Candidate. My first thought is that any new symptom with a temporal relationship to a newly started medicine puts an adverse drug reaction on the differential. Lamotrigine is an aromatic anticonvulsant with a well-recognised risk of cutaneous reactions, including rare Stevens-Johnson syndrome, and the two-week latency fits. So I would assess the child immediately, stop the lamotrigine, and characterise the rash to separate a benign exanthem from a severe cutaneous adverse reaction. [1]

Branch 1 — Definitions and classification

Examiner. Define an adverse drug reaction. How does it differ from an adverse drug event and a medication error? [1]

Candidate. An adverse drug reaction is a noxious and unintended response to a medicine occurring at doses normally used in humans. The phrase "normally used" excludes overdose and error. An adverse drug event is broader: any harm from a medicine, whether the dose was correct or not, so it includes both true ADRs and harm from medication errors. A medication error is the process failure itself — wrong drug, dose, route, time or patient — that may or may not reach the child and may or may not cause harm. [1]

Examiner. How do you classify ADRs? [1]

Candidate. The classic Rawlins and Thompson scheme divides reactions into Type A (augmented) and Type B (bizarre). Type A is an exaggerated primary pharmacological effect: predictable, dose-related, common, usually low mortality — opioid respiratory depression, beta-agonist tachycardia. Type B is qualitatively abnormal: unpredictable, largely dose-independent, rare, potentially lethal — anaphylaxis, SJS/TEN, DRESS, or idiosyncratic haemolysis in G6PD deficiency. The alphabet extends this: Type C chronic, Type D delayed, Type E end-of-use, Type F failure of therapy. The modern DoTS extension classifies each reaction across Dose-relatedness, Time-relatedness and Susceptibility, which complements the Type A/B scheme by explaining mixed patterns. [1]

Branch 2 — Causality and severity

Examiner. How would you decide whether lamotrigine actually caused this rash? [1]

Candidate. I would apply the Naranjo algorithm, which scores ten items — previous conclusive reports, temporal relationship to the suspected drug, improvement on withdrawal, recurrence on rechallenge, alternative causes, placebo response, drug in blood, dose-dependence, prior exposure and objective confirmation. I would sum and band the score: 0 to 4 doubtful, 5 to 6 possible, 7 to 13 probable, 14 to 18 definite. The strong temporal relationship, the known reaction profile of lamotrigine, and improvement on withdrawal would usually place this in the probable range. I would grade severity with the Hartwig scale and assess avoidability with the Liverpool tool. [12]

Examiner. What about rechallenge? [1]

Candidate. Deliberate rechallenge is contraindicated after any serious immune-mediated reaction because it can be fatal. I would not rechallenge this child. Rechallenge belongs only in the algorithm as a historical or inadvertent observation, never as a planned test after SJS, TEN, DRESS or anaphylaxis. [1]

Branch 3 — Management

Examiner. The rash now shows blistering with oral and conjunctival involvement. Talk me through management. [1]

Candidate. I would now treat this as Stevens-Johnson syndrome. I would stop lamotrigine immediately and withdraw all non-essential medicines, secure airway, breathing and circulation, admit to PICU, and treat as a burn: fluids titrated to the area of detachment, meticulous wound care, ophthalmology review within 24 hours to prevent ocular adhesions, nutritional support and pain relief. I would avoid prophylactic antibiotics and seek specialist advice on immunomodulation. I would estimate the percentage of body surface area detachment to classify the reaction — under 10 percent SJS, 10 to 30 percent overlap, over 30 percent TEN — because that predicts severity. [1]

Examiner. Suppose instead the child had developed wheeze, urticaria and hypotension minutes after a dose of amoxicillin. First-line treatment? [1]

Candidate. That is anaphylaxis, and first-line treatment is intramuscular adrenaline into the anterolateral thigh, roughly 10 microgram per kilogram of 1:1000 solution, repeatable every five minutes. For a young child 7.5 to 20 kg the 150 microgram autoinjector approximates the dose, and for over 20 kg the 300 microgram device. Antihistamines, corticosteroids and bronchodilators are adjuncts, never substitutes. Lay the child flat with legs raised, give high-flow oxygen, a fluid bolus, and observe for at least six hours because biphasic reactions occur. [1]

Branch 4 — Pharmacovigilance and reporting

Examiner. What is your duty to report this reaction, and to whom? [1]

Candidate. Reporting is a professional duty and is independent of certainty. I would report to the national pharmacovigilance system — the TGA in ANZ, the MHRA Yellow Card in the UK, or the FDA MedWatch programme in the US — and the report reaches WHO VigiBase through the national centre. Serious reactions, paediatric ADRs, and reactions to Black Triangle medicines should all be reported. [7]

Examiner. Why does it matter? Is one report not lost in the noise? [1]

Candidate. It matters because under-reporting is the rule, not the exception — a systematic review found a median under-reporting around 94 percent. VigiBase pools tens of millions of reports and detects signals by disproportionality analysis, such as the proportional reporting ratio. Children are rarely in premarketing trials, so the first signs of harm often appear only after widespread use. A single report contributes to the signal that protects the next child. [9]

Closing probe

Examiner. Give me one sentence that captures your approach to a suspected adverse drug reaction. [1]

Candidate. Recognise it on the differential, stop the drug, treat life threats, apply Naranjo and Hartwig, choose a non-cross-reactive alternative, and report it to the national system so VigiBase can find the signal. [8]