ABPN Epilepsy common mistakes
ABPN Epilepsy · Understand the traps and how to avoid them.
Classifying a spell as epileptic versus nonepileptic without first committing to a semiology-based localization hypothesis, then choosing treatment before the diagnosis is settled.
Candidates anchor on the referral diagnosis or the word 'seizure' and jump to drug selection, skipping the step where the event is characterized (onset, awareness, automatisms, postictal features, provocation).
What to do instead
For every spell question, write a one-line event description (onset, awareness, motor features, duration, postictal state) before answering; use that description to assign focal/generalized and epileptic/nonepileptic, and only then address management.
Treating a single unprovoked seizure or a first seizure with normal EEG/imaging as requiring immediate long-term antiseizure medication without weighing recurrence risk and patient-specific factors.
Exam stems reward action, so candidates default to starting a drug; they forget that recurrence risk after a single unprovoked seizure is roughly 40-50% and that treatment decisions are individualized.
What to do instead
Recall the recurrence-risk framework (abnormal EEG, abnormal imaging, nocturnal event, prior neurologic insult raise risk) and select 'discuss risks/benefits and defer or start based on shared decision-making' when the stem describes a low-risk first event.
Misreading the EEG: calling a normal variant (wicket spikes, 14-and-6 positive bursts, benign epileptiform transients of sleep, rhythmic midtemporal theta) epileptiform, or missing that a normal interictal EEG does not exclude epilepsy.
Candidates pattern-match sharp-looking waveforms to epilepsy and treat EEG as a rule-in/rule-out test rather than a probabilistic adjunct.
What to do instead
Learn the named benign variants and their distinguishing features (morphology, location, state dependence, lack of field), and state explicitly that a single normal EEG has limited negative predictive value; use sleep-deprived or prolonged recording as the answer when sensitivity is the issue.
Choosing an antiseizure medication without matching mechanism and adverse-effect profile to the patient's syndrome, age, sex, and comorbidities (e.g., valproate in a woman of childbearing age, enzyme inducers in porphyria or with interacting drugs, sodium-channel blockers in generalized epilepsy).
Candidates memorize a list of 'drugs for focal' and 'drugs for generalized' but do not run the patient-specific filter that board items are built to test.
What to do instead
For every drug question, run a fixed checklist: seizure type/syndrome, age, pregnancy potential, hepatic/renal status, psychiatric comorbidity, weight, and drug interactions; eliminate options that fail any box before choosing.
Failing to recognize and manage status epilepticus in the correct time-ordered sequence, including when to move from benzodiazepines to a second-line agent and when to escalate to anesthetic infusion.
Candidates know the drugs but not the clock, and they hesitate to redose benzodiazepines or to escalate when seizures persist.
What to do instead
Memorize the staged algorithm (benzodiazepine at 0-5 min, repeat once, second-line IV agent at 10-20 min, anesthetic infusion for refractory at ~30-40 min) and pair each stage with the monitoring and airway considerations the stem mentions.
Missing the presurgical evaluation logic: not recognizing when a patient is a surgical candidate, and confusing the roles of phase I vs invasive phase II monitoring, Wada testing, and the various imaging modalities.
Candidates treat surgery as a last resort and do not know the criteria for drug-resistant epilepsy or which test answers which question (lateralization of language/memory vs seizure onset zone vs structural lesion).
What to do instead
Anchor on drug-resistant epilepsy (failure of two appropriately chosen, tolerated drugs) as the trigger for referral, and build a table mapping each presurgical test to the specific question it answers before answering.
Applying adult dosing, dosing schedules, and drug choices directly to neonates, infants, and the elderly without adjusting for altered pharmacokinetics and age-specific syndromes.
Candidates study adult epilepsy most and under-weight the developmental and geriatric pharmacology that special-population items target.
What to do instead
For special-population items, explicitly adjust for age: neonatal dosing per kg and immature clearance, infantile spasms as an emergency with specific first-line agents, and elderly patients' lower clearance, higher interaction burden, and vascular etiology.
Overlooking psychiatric and psychosocial comorbidity (depression, anxiety, suicidality, cognitive effects, driving, pregnancy counseling, stigma) when the question is framed around 'best next step'.
Candidates focus on seizure control as the only outcome and forget that screening for depression and counseling on driving/pregnancy are frequently the correct answer in a stem that mentions mood, isolation, or life events.
What to do instead
When a stem mentions mood, behavior, employment, driving, or pregnancy, screen the answer choices for the psychosocial intervention (screen/treat depression, counsel on driving laws, preconception folate and drug review) before selecting a seizure-drug change.
Confusing the basic science of epileptogenesis and channel/neurotransmitter physiology with clinical management, e.g., attributing all epilepsy to GABA failure or misassigning mechanisms to specific drugs.
Candidates memorize isolated facts (NMDA, GABA-A, sodium channels) without linking them to the drug classes and syndromes the exam pairs them with.
What to do instead
Build a mechanism-to-drug map (sodium-channel blockade, SV2A binding, GABAergic enhancement, calcium-channel T-type in absence) and a syndrome-to-mechanism map, and answer basic-science items by tracing the pathway rather than recalling a single fact.
Ignoring the time course and postictal/peri-ictal features when localizing or when distinguishing seizure from syncope, sleep disorder, or psychogenic nonepileptic events.
Candidates fixate on the most dramatic feature (convulsion, injury) and neglect duration, recovery, provocation, and ictal/postictal behavior that carry the discriminating information.
What to do instead
For each event, record duration, provocation, eye/mouth behavior, incontinence, tongue biting, and recovery time; use these to separate epileptic from nonepileptic and to localize onset before choosing an answer.