Recognition, reporting and analysis of adverse events — branching viva

Stem

The examiner opens with a bedside scenario, then probes the language you use, your detection logic, your systems analysis and your duty to the family and the distressed nurse. [1] [14]

Branch 1 — Definitions and classification

Examiner: A four-year-old's morphine infusion ran at ten times the prescribed rate. The nurse stopped it. Define your terms before you act. [5]

Strong answer: An adverse event is unintended injury caused by health care management rather than the disease alone — this child's opioid toxicity qualifies. A preventable adverse event is one that accepted practice could have avoided; a pump-programming error is preventable. A near miss is an event that could have harmed but did not, because someone caught it. An error is a failure of a planned action; not every error harms and not every harm is one person's slip. Separate disease progression from care-related harm, and separate active failure at the sharp end from latent conditions in the system. [1] [5]

Branch 2 — Immediate clinical and safety actions

Examiner: Walk me through what you do at the bedside over the next ten minutes. [2]

Strong answer: Rescue the child first — assess airway, breathing and circulation, keep the pump stopped, give naloxone if opioid toxicity is significant, call for senior and nursing help. Preserve the evidence: do not reset the pump, keep the original order, bag and label. Check whether any other child is exposed to the same pump setting, drug batch or protocol. Document a factual timeline without blame adjectives. Notify the consultant and the local safety pathway, and tell the family that an unexpected event has occurred and that the focus is on the child's safety. [1] [14]

Branch 3 — Detection and why voluntary reporting is weak

Examiner: Your unit boasts zero incidents this quarter. Why should I distrust that number? [11]

Strong answer: Voluntary reporting captures only what staff choose to file, and it shrinks under blame culture, time pressure and feedback that never returns. Classen showed that trigger-tool methods can reveal harm an order of magnitude greater than voluntary reporting implies. Trigger tools screen charts for signals — antidote use, abrupt stops, unplanned transfers, abnormal results — then confirm harm by review. "Zero reports" may describe silence, not safety. Pair voluntary reporting with trigger-tool sampling and surveillance so the system has more than one set of eyes. [11]

Branch 4 — Systems analysis and strong actions

Examiner: You lead the review. What does a good analysis look like, and what actions will you prefer? [1]

Strong answer: Build a timeline first, before any root-cause claim. Use systems thinking: map latent conditions — ambiguous protocol, look-alike packaging, understaffing, weak double-check — alongside the active failure of the pump programming. Run a structured method such as root-cause or systems analysis. Then prefer strong actions over weak ones: forcing functions, standardised concentrations, smart-pump libraries with hard limits, equipment redesign, and double-checks that genuinely block the error. Re-education posters alone are a weak action and tend to let the latent holes stay open. Assign owners and dates, implement, and measure recurrence. [1] [16]

Branch 5 — Disclosure and the second victim

Examiner: The family is at the bedside and the nurse is in tears in the tea room. Address both. [14] [15]

Strong answer: Disclose with open, plain language: what happened, what it means for the child, what is being done now, an apology as appropriate, and how the family can reach you. Gallagher showed families value honesty more than many clinicians fear. Do not speculate beyond the facts, and do not blame an individual. Then turn to the nurse as a second victim: name the phenomenon, relieve her from high-risk tasks if she is acutely distressed, offer peer support, and arrange follow-up. Unsupported second victims can impair the next shift's safety, so welfare here is a clinical governance act, not only kindness. [14] [15]

Examiner extras

• A cluster of similar incidents with no systems action is itself a red flag — escalate as a governance issue. [1]
• Structured handoff programmes such as I-PASS reduce medical errors; cite Starmer after a handoff-related event. [13]
• Pronovost's central-line bundle shows a system intervention can drive harm down — name it when asked for proof that bundles work. [16]