UKOM Sp.N (Neurology) common mistakes
UKOM Sp.N (Neurology) · Understand the traps and how to avoid them.
When asked to localize a lesion, the candidate names a disease (for example 'stroke' or 'GBS') instead of stating the anatomical level and side, and never walks through the long-tract, cranial nerve, and cortical signs in a fixed order.
Undergraduate and early postgraduate training rewards pattern-recognition of syndromes, so candidates jump from a few findings straight to a diagnosis label; the exam explicitly tests the reasoning chain from findings to level to differential, and skipping the chain loses the marks even when the final diagnosis is right.
What to do instead
Before answering any localization question, write one line per system in a fixed template: mental status/cortex, cranial nerves, motor (tone, power, reflexes, plantar), sensory level and modality, cerebellar, gait, plus a one-sentence statement of the level and side. Practise 20 vignettes aloud using only this template until the order is automatic, then attach the differential as a separate final sentence.
The candidate interprets every acute focal deficit as ischaemic stroke and proceeds to thrombolysis reasoning without first checking glucose, seizure activity, migraine aura, or functional overlay, and without stating the time last known well and the CT/CTA findings that change management.
Stroke is over-represented in teaching, so 'sudden focal deficit equals stroke' becomes an unexamined reflex; candidates also conflate the thrombolysis window, thrombectomy window, and wake-up stroke pathways, and forget that the imaging decision precedes the drug decision.
What to do instead
For every acute deficit vignette, state in order: onset time and last known well, glucose, NIHSS-relevant findings, non-contrast CT result (haemorrhage excluded, early ischaemic change), vessel imaging, contraindications to lysis, and then the treatment decision with its time window. Build a one-page table of the 4.5-hour lysis criteria, the extended-window and wake-up imaging selection criteria, and the thrombectomy eligibility criteria, and recite it before each practice block.
On EEG and EMG questions the candidate describes the clinical picture and guesses the diagnosis, rather than reading the tracing: not distinguishing epileptiform discharges from artefacts, not identifying the ictal rhythm's frequency and evolution, and not separating axonal from demyelinating patterns on nerve conduction studies.
Neurophysiology is taught briefly and examined heavily, and candidates memorise associations ('3 Hz spike-wave equals absence') without learning to describe what is on the page; artefacts such as eye blink, muscle, and electrode pop are rarely drilled, so they are misread as pathology.
What to do instead
Learn a fixed descriptive vocabulary for each modality. For EEG: background frequency and symmetry, presence of alpha rhythm, focal slowing, epileptiform discharges (spike, sharp wave, spike-wave), and the ictal pattern with onset, frequency, and evolution. For NCS/EMG: sensory versus motor involvement, amplitude, conduction velocity, distal latency, F-wave, and the presence of conduction block or temporal dispersion, then the axonal versus demyelinating conclusion. Practise describing 10 tracings per week before naming any diagnosis.
The candidate answers a movement disorder question by naming the syndrome but cannot state the defining phenomenology (for example the difference between resting and action tremor, or between rigidity and spasticity), and does not apply the diagnostic criteria that separate Parkinson's disease from atypical parkinsonism.
Movement disorders are visually taught, and without video the candidate falls back on labels; the exam tests the descriptive vocabulary and the red flags (early falls, vertical gaze palsy, cerebellar signs, autonomic failure, poor levodopa response) that redirect the diagnosis.
What to do instead
Write out definitions in your own words for tremor (rest, postural, kinetic, intention), rigidity (lead-pipe, cogwheel), bradykinesia (sequence effect, decrement), dystonia, chorea, myoclonus, and tics, with one distinguishing bedside test each. Then memorise the red flags that argue against idiopathic Parkinson's disease and rehearse three atypical parkinsonism profiles against them.
When a vignette involves weakness, the candidate does not separate upper from lower motor neuron signs, and does not use the pattern (proximal versus distal, symmetric versus asymmetric, sensory involvement, reflex changes, fasciculation) to place the lesion in anterior horn cell, root, plexus, peripheral nerve, neuromuscular junction, or muscle.
The motor examination is often recorded as a single power grade without tone, reflexes, or sensory correlation, so the data needed for anatomical separation are never collected; candidates also forget that neuromuscular junction disorders show fatigability without sensory loss and that myopathy spares reflexes until late.
What to do instead
For every weakness vignette, tabulate: distribution, symmetry, tone, power by muscle group, reflexes, plantar response, sensory loss, fasciculation, and fatigability. Then apply a decision ladder: upper motor neuron pattern, lower motor neuron pattern, both, or neither, and within lower motor neuron use sensory and reflex findings to choose among root, plexus, nerve, junction, and muscle. Do this on paper for 15 vignettes.
The candidate treats headache questions as a test of naming a primary headache type and misses the red flags and the indicated next step, or over-investigates a stable primary headache; in pain questions the candidate does not distinguish nociceptive from neuropathic mechanisms or state the first-line agent and its evidence.
Headache is perceived as a low-yield descriptive topic, so the danger signs (thunderclap onset, positional change, papilloedema, fever, focal deficit, new onset over 50, pregnancy) are under-rehearsed; candidates also default to whichever drug they have seen used rather than the guideline first-line choice.
What to do instead
Memorise a red-flag checklist for headache and attach the correct next investigation to each flag. For each common headache syndrome, write the diagnostic criteria and the first-line acute and preventive treatment with one supporting trial or guideline. For neuropathic pain, state the mechanism, the validated screening tool, and the first-line agents with their numbers needed to treat.
The candidate answers neuroimmunology questions by listing all available tests rather than reasoning about specificity and timing: not knowing which antibody is associated with which syndrome, not recognising that oligoclonal bands and IgG index reflect intrathecal synthesis rather than disease activity, and not applying the dissemination-in-space-and-time criteria correctly.
Antibody panels have expanded rapidly and candidates learn names without the associated phenotype or the caveat that a positive result in a low-probability patient is often false positive; the diagnostic criteria are memorised as a list without understanding which elements are clinical and which are paraclinical.
What to do instead
Build a two-column table: antibody or test on the left, associated syndrome and typical clinical/imaging features on the right, with a note on when the test is indicated and its main pitfall. For multiple sclerosis, write out the current dissemination-in-space and dissemination-in-time criteria in your own words and apply them to five vignettes, stating explicitly which criterion each finding satisfies.
In neuroinfectious disease vignettes the candidate orders a lumbar puncture without first addressing contraindications (mass lesion, coagulopathy, raised intracranial pressure) or the need for imaging, and interprets cerebrospinal fluid by a single cell count instead of the full pattern (opening pressure, cell differential, protein, glucose ratio, Gram stain and culture, PCR, and paired serum glucose).
Meningitis and encephalitis are taught as emergencies, so the reflex is 'tap immediately', and the safety checklist is forgotten under time pressure; candidates also memorise a single 'typical' CSF profile per organism and then misclassify atypical but classic presentations.
What to do instead
For every CSF vignette, first state whether imaging or coagulation correction is needed before the tap, then present the CSF as a structured panel: opening pressure, white cell count and differential, protein, CSF-to-serum glucose ratio, Gram stain, culture, and PCR. Learn the classic patterns for bacterial, viral, tuberculous, and fungal meningitis and the empiric regimen for each, including dexamethasone timing where indicated.
The candidate answers neuropharmacology and therapeutics questions from drug-class memory alone, without linking mechanism to the specific indication, monitoring requirement, or adverse effect that the question is actually testing, and without citing the level of evidence behind a recommendation.
Pharmacology is revised as lists of side effects, so candidates cannot connect a mechanism (sodium channel blockade, dopamine agonism, GABA potentiation) to the clinical scenario; they also ignore that the exam distinguishes standard of care from emerging or unproven therapy, and that some questions test deprescribing and monitoring rather than initiation.
What to do instead
For each drug class on the syllabus, write one line each for mechanism, main indications, one serious adverse effect with its monitoring plan, and one interaction. Add a separate short list of common 'do not do' practices with the evidence that refutes them. When a question asks for management, state the recommendation, the evidence level or guideline source, and the monitoring or follow-up step.
The candidate treats ethics, consent, capacity, and patient-safety questions as common sense and answers with a general principle, without applying the legal test (for example the functional test of capacity, the distinction between best-interests and substituted-judgement decisions, or the duty of candour after a harm event) or the required documentation and escalation steps.
Professionalism is perceived as unexaminable, so candidates do not learn the statutory framework or the order of steps in an incident; they also conflate confidentiality exceptions and the rules on disclosing to the driver licensing authority or to public health.
What to do instead
Learn the four-part functional capacity assessment and apply it to three vignettes, stating which element fails. Write out the steps for an unexpected harm event (identify, disclose to patient, document, report, analyse, change practice) and the grounds on which confidentiality may be breached. For each, name the relevant legal or regulatory instrument in your jurisdiction and the person responsible for the decision.