CSHK Part 2 EMQ common mistakes
CSHK Part 2 EMQ · Understand the traps and how to avoid them.
Treating the EMQ option list as a matching exercise and assigning each option to exactly one stem, so that options are 'used up' and later stems are forced onto leftover options rather than the correct one.
Candidates carry the habit of single-best-answer or true matching formats where each option is used once; EMQs in this exam deliberately reuse options across multiple stems and include near-identical distractors, so an 'elimination' strategy systematically misassigns answers.
What to do instead
Before answering, write down which options are genuinely plausible for each stem independently, then check whether any option is the right answer for more than one stem. Practise sets where 2-3 stems share the same option so the reuse pattern becomes familiar. Never strike out an option because you already used it.
Answering the stem that was asked rather than the one intended: e.g. giving the single most likely diagnosis when the stem asks for the next investigation, or giving the investigation when it asks for initial management.
Under time pressure candidates pattern-match the clinical vignette to a diagnosis and select that option, ignoring the final directive word (diagnosis / investigation / management / complication). The vignette is designed so the diagnosis is obvious but the correct next step is not.
What to do instead
Underline or mentally box the final directive in every stem before looking at the options. If the option you favour is the diagnosis but the stem asks for management, discard it. Do a final pass re-reading only the directive words.
Choosing the definitive or 'best possible' treatment when the stem describes an unstable or undifferentiated patient, skipping resuscitation and immediate life-saving steps.
Textbook knowledge is organised around definitive management, and candidates default to the elegant answer (e.g. definitive operative repair) over the ATLS/ABCDE priority step (e.g. haemorrhage control, airway, damage-control surgery).
What to do instead
For any emergency or trauma stem, run ABCDE mentally and ask 'what kills this patient in the next 10 minutes?' before considering definitive care. Learn the damage-control and resuscitation-first sequences as discrete option sets.
Selecting an investigation or intervention that is contraindicated or unsafe in the specific physiological state described (e.g. contrast study in renal impairment, early definitive fixation in an unstable trauma patient, neuraxial block in coagulopathy) because it is the usual workup.
Candidates recall the standard pathway for a condition but do not cross-check it against the comorbidities, haemodynamics, or anticoagulation stated in the stem; the distractor list includes the 'usual' test as a trap.
What to do instead
Before committing, scan the stem for modifiers: renal function, anticoagulants, haemodynamic stability, pregnancy, sepsis, allergy. Reject any option that conflicts with a stated modifier even if it is the textbook default.
Ignoring the timing and trajectory in the vignette (onset, duration, interval since surgery, day of admission) and choosing an answer appropriate to a different time point, such as selecting anastomotic leak for day 2 or wound infection for day 7.
Candidates focus on the symptom pattern and forget that postoperative and post-injury complications have characteristic time windows; the same symptoms map to different diagnoses depending on day.
What to do instead
Build a mental timeline: 0-24h (haemorrhage, anaesthetic complications), 24-72h (leak, ischaemia, ARDS), day 5-8 (infection, abscess), weeks (stricture, incisional hernia). Note the day explicitly before choosing.
Confusing the operative principle or anatomical structure being tested (e.g. wrong plane, wrong vessel, wrong nerve at risk) when the stem describes a specific operation, so the option chosen is the correct fact for a different procedure.
Options are drawn from related procedures and the candidate recognises the fact but not its correct operative context; the exam exploits proximity of similar operations (e.g. laparoscopic vs open, right vs left).
What to do instead
For each stem, first name the exact operation and its key steps, then check each candidate option against those steps. Revise by procedure with a one-line 'structures at risk / critical step / common complication' summary for each.
Applying adult dosing, adult physiology, or adult scoring to a paediatric or elderly patient, or using a paediatric threshold for an adult.
Most revision material is adult-centric; candidates default to adult norms and miss that the stem specifies a neonate, child, or frail elderly patient with different fluid, drug, or resuscitation parameters.
What to do instead
Flag the patient's age group at the start of the stem. Recall the key differences: paediatric maintenance fluids and weight-based dosing, elderly physiology, altered thresholds for tachycardia and hypotension. Practise mixed-age stems.
Misreading the negative or exclusion wording in a stem ('least likely', 'except', 'not appropriate', 'most unlikely') and selecting the true statement instead of the exception.
The brain auto-completes familiar question formats; negative stems are a known source of avoidable error, and the option list contains several true statements so the 'true' answer feels correct.
What to do instead
Circle negative words before reading options. If a stem is negative, mark the true options with a tick and the false ones with a cross, then choose the cross. Re-read the stem after selecting to confirm polarity.
Failing to manage time and answer distribution: spending excessive time on a few difficult stems, leaving later stems blank, or changing correct answers late in the exam without a clear reason.
EMQ stems vary in difficulty and the exam is long; anxiety and perfectionism cause candidates to over-invest in one stem, and fatigue-driven answer changing reduces scores.
What to do instead
Set a per-stem time budget, answer every stem (no blank answers), flag difficult ones and return once, and only change an answer if you can state a specific reason. Rehearse with timed full-length mock papers.
Choosing an option based on a single keyword association (e.g. 'fever + rigors = antibiotics') without integrating the whole clinical picture and excluding the competing options.
Pattern recognition from revision is fast but shallow; the exam deliberately includes overlapping features so that keyword matching selects a plausible but wrong option.
What to do instead
For each stem, generate a two-line problem list (key findings, key negatives) before looking at options, then test each candidate against both lines. Practise stems with deliberately overlapping presentations.