HKCA Final MCQ common mistakes
HKCA Final MCQ · Understand the traps and how to avoid them.
On single-best-answer stems, choosing the option that is true in general rather than the one that answers the exact question asked (e.g., picking a correct physiological statement when the stem asks for the most likely cause in the described scenario).
Under time pressure candidates pattern-match the topic and select a familiar true statement without re-reading the final clause of the stem; each distractor is deliberately a true but non-answering fact.
What to do instead
Before selecting, underline the interrogative phrase (most likely, initial step, contraindicated, best next) and test each option against that phrase, not against the topic. Discard any option that is true but does not answer the question.
Treating 'initial management' and 'definitive management' as interchangeable, so a correct definitive therapy (e.g., definitive surgical or interventional treatment) is chosen where the question asks for the first priority step.
Candidates recall the definitive treatment best because it is the most memorable fact, and sequence words in the stem are easily skimmed.
What to do instead
Classify every option as first-line/initial versus definitive/definitive-later before choosing. Build a habit of ranking options by order of action (resuscitate, diagnose, treat, refer) and choose the earliest appropriate step the stem requests.
Applying adult drug doses, doses per kg, or dilution volumes to paediatric and neonatal scenarios, or forgetting to recalculate when the stem gives a weight in a different unit.
Most revision is adult-focused; candidates default to remembered adult figures and do not perform the arithmetic because the option 'looks right'.
What to do instead
For every paediatric or neonatal item, write the weight, the per-kg dose, and the calculated total before looking at options. Check units (micrograms vs milligrams, per kg vs total) explicitly, and reject any option that matches an adult dose in a child.
Misreading arterial blood gas values as a single disorder and missing a mixed picture, or misclassifying compensation (e.g., labelling a partially compensated metabolic acidosis as fully compensated).
Candidates anchor on the pH or on one abnormal value and skip the systematic stepwise interpretation; compensation arithmetic is rarely rehearsed under time pressure.
What to do instead
Use a fixed sequence: pH, then primary disorder from PaCO2 and bicarbonate, then check whether compensation is appropriate using expected-value rules, then calculate the anion gap and delta ratio where relevant. Apply the same written sequence to every blood gas item.
Confusing analgesia and sedation endpoints, or choosing a sedative as an analgesic, in intensive care scenarios involving ventilated or delirious patients.
The drug classes overlap in memory and candidates recall a familiar drug name without matching it to the stated goal (pain versus anxiety versus delirium).
What to do instead
For each option, name its primary target (analgesia, anxiolysis, hypnosis, delirium control) and match it to the stated goal. Where the stem describes pain as the driver, prioritise an analgesic strategy over a sedative one.
Ignoring contraindications and cautions embedded in the stem (allergy, renal or hepatic impairment, pregnancy, raised intracranial pressure, local anaesthetic systemic toxicity risk) and selecting the otherwise standard drug.
Candidates focus on the diagnosis and treat the condition generically, treating patient-specific modifiers as background detail.
What to do instead
Do a second pass over the stem listing every patient-specific modifier, then screen each remaining option against that list. Any option that is contraindicated is wrong regardless of how well it fits the diagnosis.
Confusing local anaesthetic agents and adjuvants by maximum safe dose, onset, duration, or cardiotoxicity profile, and choosing an agent that does not match the required block characteristics.
Agent properties are learned as lists rather than as a comparison table, so similar-sounding agents blur together in recall.
What to do instead
Memorise a compact comparison grid (maximum dose with and without adrenaline, onset, duration, relative cardiotoxicity) and, for each item, first state the required onset and duration, then select the agent that matches both.
Selecting an investigation or monitoring modality that confirms a diagnosis when the stem asks for the test that should be done first, or that is most appropriate in an unstable patient.
Candidates equate 'best test' with 'most accurate test' and overlook that in an unstable patient the priority is a rapid bedside test or immediate resuscitation rather than a definitive investigation.
What to do instead
Ask two questions before choosing: is the patient stable, and is the question about ruling in, ruling out, or immediate management? Choose the fastest appropriate bedside option when unstable, and the definitive test only when the stem asks for confirmation.
Answering pain medicine items from a single-mechanism model, missing that the stem describes neuropathic or mixed pain and choosing a standard opioid or simple analgesic as the sole agent.
Acute pain algorithms are over-learned, and candidates apply them to chronic or neuropathic presentations without checking the pain mechanism described.
What to do instead
Classify the pain mechanism first (nociceptive, neuropathic, mixed, central) from the descriptors in the stem, then choose management consistent with that mechanism, including multimodal or adjuvant therapy where indicated.
Running out of time because flagged items are revisited repeatedly, or changing a first answer without a concrete reason found in the stem.
Anxiety about difficult critical care calculations encourages returning to the same items, and answer-changing is driven by doubt rather than by new evidence.
What to do instead
Answer every item on first pass, flag only genuine unknowns, and return once. Change an answer only when a specific word or value in the stem contradicts the original choice; otherwise keep the first reasoned answer.