HKMLE Part III common mistakes
HKMLE Part III · Understand the traps and how to avoid them.
Writing management plans that jump straight to a definitive treatment (e.g., surgery or a specific drug) without first stating the immediate resuscitation, monitoring, and stabilization steps for an acutely unwell patient.
Candidates study diseases as textbook entities and rehearse the 'correct' final diagnosis-treatment pair, so under time pressure they skip the ABCDE/early-management layer that Hong Kong examiners expect in emergency and acute medical scenarios.
What to do instead
For every acute scenario, practise a fixed scaffold: stabilize (ABCDE, IV access, oxygen, fluids, monitoring) → focused investigation → empiric/targeted treatment → disposition and safety-netting. Write the scaffold out even in practice answers so it becomes automatic before you name the definitive therapy.
Giving a single diagnosis for a presentation and ignoring the differential, the 'can't-miss' diagnoses, and the features that would discriminate between them.
Exam conditioning rewards one right answer; candidates treat the vignette as a puzzle to solve rather than a clinical problem to reason through, so they commit early and stop thinking.
What to do instead
End every case answer with a short differential ranked by likelihood and by danger, plus the two or three investigations or clinical features that would separate them. Practise stating 'most likely X, but I must exclude Y because…' explicitly.
Ordering investigations as an undifferentiated list (full blood panel, multiple imaging modalities) without stating the question each test answers or the sequence and urgency.
Rote memorization of 'workup for chest pain' produces lists; candidates have not practised justifying each test against a clinical question, and fear of missing something drives over-investigation.
What to do instead
For each test, write one line: what it will show, how it changes management, and how urgently it is needed. Delete any test you cannot justify. Rehearse common presentations with a first-line, second-line, and only-if structure.
Prescribing without checking dose, route, renal/hepatic adjustment, allergies, pregnancy status, or interactions, and omitting monitoring for the drug's toxicity.
Candidates focus on drug choice as the intellectual step and treat dosing as clerical detail; in a written exam there is no pharmacist to catch errors, and real-world habits of relying on electronic prescribing do not transfer.
What to do instead
Build a prescribing checklist you apply every time: indication, dose, route, frequency, duration, adjustments, interactions, monitoring, and what to do if it fails. Drill high-risk drugs (anticoagulants, insulin, opioids, aminoglycosides, digoxin, methotrexate) with their monitoring parameters.
Managing only the biomedical problem and omitting psychosocial, functional, and carer context - discharge planning, community resources, capacity, and follow-up arrangements.
Hospital-centric training and exam revision notes emphasize pathology; candidates underestimate how heavily Hong Kong assessments weight holistic and community care, especially in elderly and chronic disease cases.
What to do instead
Add a standard 'whole patient' paragraph to every plan: social situation and carer support, functional status and rehabilitation, mental state and capacity, adherence barriers, community nursing or social services referral, and a named follow-up route with timeframe.
Misjudging the urgency or disposition of a patient - sending home someone who needs admission, or failing to specify the level of care (ward, high dependency, ICU) and the timeframe for review.
Candidates concentrate on diagnosis and treatment and treat disposition as an afterthought, and they lack practice in triaging severity from vital signs and trajectory rather than from the final diagnosis.
What to do instead
Make disposition an explicit required line in every answer: admit or discharge, level of care, observation frequency, criteria for escalation and for safe discharge, and who reviews and when. Practise classifying vignettes by physiological derangement to calibrate urgency.
Giving vague, non-committal answers such as 'consider referral' or 'monitor closely' instead of naming the specialty, the timeframe, and the specific parameter to monitor.
Fear of being wrong pushes candidates toward hedging language, and revision notes themselves are often written at a general level, so the habit is copied uncritically.
What to do instead
Ban the words 'consider' and 'closely' from your practice answers. Replace each with a named action, a named destination, and a number: 'refer to cardiology within 24 hours', 'check potassium daily', 'repeat ECG in 6 hours'.
Losing marks on communication and ethics items by writing what the doctor will do to the patient rather than demonstrating consent, confidentiality, breaking bad news structure, or conflict-of-interest reasoning.
Candidates treat communication as common sense and do not rehearse frameworks, so under exam conditions they produce generic reassurance instead of a structured, patient-centred response.
What to do instead
Learn and practise explicit frameworks (e.g., a stepwise breaking-bad-news structure, a consent model covering capacity, voluntariness, and information, and a confidentiality model covering necessity and proportionality). Write model scripts for common scenarios and check each against the framework.
Failing to interpret data - dismissing or misreading abnormal ranges, trends, or reference intervals, and not acting on a markedly abnormal result embedded in the vignette.
Candidates skim numbers to get to the narrative and are used to laboratory flags doing the interpretation; they also forget that reference ranges vary with age, sex, pregnancy, and assay.
What to do instead
In every practice question, circle each abnormal value and state its clinical significance and required action before reading the options or writing the plan. Rehearse common calculations (e.g., anion gap, corrected calcium, eGFR, fluid and electrolyte deficits) until they are error-free.
Running out of time or misallocating it - over-writing early cases, leaving later questions blank, and not returning to flagged items.
Candidates write essay-length answers to short-mark questions because they equate length with thoroughness, and they do not rehearse under timed conditions, so pacing collapses under pressure.
What to do instead
Practise full papers with a timer and a mark-per-minute budget; before writing, note the marks available and allocate roughly one distinct point per mark. Set a hard cutoff per question, move on, and reserve the final minutes for flagged items and checking that every question has an attempt.