CAQ-EM common mistakes
CAQ-EM · Understand the traps and how to avoid them.
Committing to a single diagnosis in the first minute and then interpreting every subsequent finding as confirmation, rather than deliberately listing and actively excluding the dangerous alternatives (aortic dissection, pulmonary embolism, ectopic pregnancy, meningitis) before closing the case.
Emergency scenarios in the item set are written with a plausible anchor diagnosis and one or two quiet discriminators. Candidates under time pressure latch onto the most familiar pattern and stop generating a differential, so they miss the finding that was placed specifically to break the anchor.
What to do instead
Before answering, write a two-line differential: the leading diagnosis plus the can't-miss alternative for that presentation. Then scan the stem once more asking only 'which detail argues against my leading diagnosis?' Practise with a fixed can't-miss list per chief complaint (chest pain, abdominal pain, headache, dyspnoea, altered mental status) until it is automatic.
Treating the first set of vital signs as the patient's baseline and not re-triaging when a repeat set is given later in the stem, so the answer reflects the stable patient rather than the deteriorating one.
Stems often present an initial stable period followed by a subtle trend (rising respiratory rate, narrowing pulse pressure, falling saturation, rising lactate). Candidates read top-down once, form an impression, and never compare the two time points.
What to do instead
When a stem contains more than one set of observations or a time marker such as 'one hour later', build a small trend table before reading the options. Answer for the patient's current physiology, not their arrival physiology. If the trend is worsening, the correct option is almost always the one that escalates care or reassesses, not the one that continues the original plan.
Choosing the most definitive intervention (intubation, central line, thrombolysis, immediate surgery) when the question is really asking for the next step in a sequence, and skipping the stabilising or diagnostic step that must come first.
Definitive options feel decisive and clinically impressive, and candidates equate 'best answer' with 'most aggressive answer'. Many items test sequencing: oxygen and monitoring before sedation, confirmation before lysis, resuscitation before imaging in an unstable patient.
What to do instead
Classify each option by stage: resuscitate, diagnose, treat definitively, disposition. Ask which stage the stem has actually reached. If the patient is unstable, the correct answer is nearly always the resuscitative or bedside step; definitive therapy belongs to the stable or confirmed patient. Read the final sentence of the stem carefully, because it usually names the stage being tested.
Applying adult dosing, adult vital-sign thresholds, and adult ECG or laboratory cut-offs to paediatric or elderly patients, or using a single normal range across the whole age spectrum.
Candidates memorise one reference range during preparation and apply it uniformly. Paediatric heart and respiratory rates, weight-based drug doses, and elderly presentations (afebrile sepsis, atypical myocardial infarction, delirium instead of pain) all break the adult template.
What to do instead
For any patient under about 12 years or over about 65 years, pause and re-derive the expected physiology from age before evaluating options. Memorise weight-based formulae for the drugs that appear repeatedly (adrenaline in arrest, fluid boluses, paracetamol, opioid dosing) and the age-specific upper limits for heart and respiratory rate. Treat afebrile hypotension in an older patient as sepsis until proven otherwise.
Reading the investigation results as a checklist of abnormalities and picking the option that names the most abnormal value, instead of synthesising the pattern (for example anion gap, osmolality, or a combination of ECG and electrolyte findings) that defines the diagnosis.
Stems supply several abnormal numbers, only some of which are relevant. Candidates anchor on the most visually striking value, which is frequently a distractor, and miss the derived value (corrected sodium, anion gap, delta gap, shock index) that the item is actually testing.
What to do instead
Before looking at the options, calculate the derived indices the presentation invites: anion gap and delta ratio for metabolic acidosis, corrected sodium for hyperglycaemia, shock index for occult shock, Wells or PERC components for pulmonary embolism. Let the calculated pattern select the diagnosis, then read the options.
Confusing the indication for a time-critical reperfusion or antidote strategy with the contraindication list, and selecting thrombolysis, anticoagulation, or an antidote in a patient who has an explicit absolute contraindication buried in the history.
Candidates rehearse indications far more than contraindications, and the contraindication is usually a single clause in a long past medical history (recent intracranial surgery, prior haemorrhagic stroke, active bleeding, known allergy, pregnancy). Under time pressure that clause is skimmed.
What to do instead
When an option involves thrombolysis, anticoagulation, or an antidote, re-read the entire history specifically hunting for contraindications before committing. Keep a short absolute-contraindication list for thrombolysis and for anticoagulation and rehearse it weekly. If an absolute contraindication is present, the correct answer is the alternative pathway (primary percutaneous intervention, reversal, supportive care), not a reduced dose.
Missing the environmental or exposure clue that changes management entirely (carbon monoxide in a house fire or faulty heater, cyanide in industrial smoke, organophosphate in a farm worker, methaemoglobinaemia after nitrites or dapsone) and treating the presentation as a generic respiratory or neurological case.
Toxicology items are written so the toxidrome is disguised as a common complaint, and the discriminating detail is a single occupational or social clause. Candidates who scan only the vital signs and examination never connect the exposure to the diagnosis.
What to do instead
For every undifferentiated hypoxia, acidosis, or altered mental status, ask explicitly 'what was the patient doing and where?' before answering. Learn the small set of exposure-toxidrome pairings that change immediate treatment (high-flow oxygen and hyperbaric referral for carbon monoxide, hydroxocobalamin for cyanide, atropine and pralidoxime for organophosphate, methylene blue for methaemoglobinaemia).
Answering the disposition question with the most cautious option (admit everyone, or discharge everyone with a safety-net) without matching disposition to the risk stratification tool the stem implies, such as a validated chest pain, syncope, or head injury rule.
Disposition items look like judgement calls, so candidates answer from temperament rather than from the score. Both defensive over-admission and premature discharge are marked wrong when the stem has supplied the components of a validated decision rule.
What to do instead
Identify whether the stem supplies the variables of a recognised rule (for example the components of a syncope risk score, a head injury rule, or a pulmonary embolism rule-out pathway). If it does, apply the rule literally and let it dictate disposition. If it does not, justify the disposition with the specific risk factors present and document the follow-up plan in the option you choose.
Managing the airway or the monitor rather than the patient during a resuscitation item, choosing an intervention that addresses a number on the screen while the reversible cause of the arrest or shock remains untreated.
Advanced life support teaching emphasises algorithms and rhythms, so candidates fixate on the trace and forget the reversible-cause checklist. Items frequently embed a cause (tension pneumothorax, tamponade, hypovolaemia, hyperkalaemia, hypothermia, toxins) that must be named for the answer to be correct.
What to do instead
During any arrest or peri-arrest item, state the rhythm, then run the reversible-cause list aloud before reading options. If a reversible cause is identifiable from the stem, the correct option treats that cause, not the rhythm. Rehearse the list in a fixed order so it is retrieved under pressure.
Misreading the question's framing, for example answering 'which is the most likely diagnosis' with a management step, or answering 'which is contraindicated' with the indicated treatment, because the final line was skimmed after the clinical stem consumed attention.
Stems are long and cognitively demanding, so the interrogative sentence at the end receives less attention than the clinical narrative. This produces errors on items where the candidate's knowledge was actually sufficient.
What to do instead
Cover or ignore the options until you have read the final sentence and restated it in your own words as a task ('name the diagnosis', 'name the next step', 'name the contraindication'). Then predict an answer before revealing the options. If the predicted answer is absent, re-read the final sentence before re-reading the stem.