Read this topic

A fresh doctor might hear "seizure" and reach straight for an antiepileptic drug list. That skips the single most useful skill in this topic: semiology (the pattern of signs and symptoms during and around an event) tells you, often without any scan or blood test, roughly where in the brain a seizure started. Hold the phrase WHERE IT STARTS through every section below, because localisation is what the whole ILAE (International League Against Epilepsy) classification system is built around.

A seizure is a sudden, excessive, synchronised electrical discharge of neurons in the brain, producing a transient change in movement, sensation, behaviour, or awareness. Epilepsy is the tendency to have recurrent, unprovoked seizures, not a single seizure in isolation: one seizure after a head injury or during severe hypoglycaemia (low blood sugar) is provoked and does not, by itself, mean epilepsy. The ILAE classifies seizures first by ONSET: focal (starting in one part of one hemisphere) or generalised (starting simultaneously across both hemispheres).

A common mistake is assuming any dramatic convulsive event is epileptic. Dissociative (non-epileptic) seizures typically have a gradual onset, a waxing-and-waning course, eyes held closed with resistance to opening, side-to-side head or pelvic thrusting movements, and a prolonged duration, often many minutes rather than the one-to-three minutes typical of a true tonic-clonic seizure, with no true post-ictal confusion. A raised serum prolactin taken shortly after the event is sometimes checked but is unreliable in isolation, since it can be normal after a genuine epileptic seizure and mildly raised after some non-epileptic events; it should never be used alone to make or exclude the diagnosis.

Imagine two patients having a focal seizure that starts in the same spot: one stays fully aware throughout and can describe exactly what they felt, the other loses awareness partway through and later cannot recall it at all. The first is a focal aware seizure (previously called "simple partial"), the second a focal impaired-awareness seizure (previously called "complex partial"). Either can spread to become a focal-to-bilateral tonic-clonic seizure, in which the electrical activity crosses to involve both hemispheres and produces a generalised convulsion.

Temporal lobe seizures are the classic focal seizure of exam vignettes. They often begin with an aura (a warning sensation that is itself the first part of the seizure, not a separate premonition) such as an epigastric rising sensation, déjà vu, jamais vu, or an olfactory or gustatory hallucination (a false smell or taste with no real stimulus). This is followed by impaired awareness with automatisms (small, purposeless repeated movements such as lip-smacking, fumbling, or picking at clothing) and then post-ictal (after-seizure) confusion that can last minutes to hours.

Frontal lobe seizures behave almost oppositely: they are typically brief, often occur at night, and produce bizarre, hyperkinetic (violent, thrashing, bicycling-type) motor activity with rapid recovery afterwards: a pattern frequently misdiagnosed as a non-epileptic or behavioural event because it looks so unlike the textbook tonic-clonic seizure. Generalised-onset seizures include absence seizures (brief, blank staring episodes in children with instant recovery and no post-ictal confusion) and generalised tonic-clonic seizures (loss of consciousness with stiffening then rhythmic jerking of all four limbs, often with tongue-biting and incontinence).

An absence seizure should not be confused with a focal impaired-awareness (temporal lobe) seizure: the absence seizure starts and stops abruptly with no aura and no post-ictal confusion, while the temporal lobe seizure has a preceding aura and leaves the patient confused afterwards. A frontal lobe seizure's brief, bizarre, nocturnal, hyperkinetic pattern should not be dismissed as a parasomnia (a sleep-related behavioural disturbance) or a psychiatric event purely because it looks unusual. A dissociative (non-epileptic) seizure should not be confused with any of the above based on convulsive movements alone: duration, gradual onset, eye closure with resistance, and the absence of true post-ictal confusion are the discriminating features, not the drama of the movements themselves.

Video-EEG (electroencephalogram) telemetry, which continuously records both video and brain electrical activity over an admission, is the gold-standard investigation when the seizure type, or the epileptic-versus-non-epileptic distinction, is unclear, since it can capture an actual event and match the clinical semiology to (or against) an EEG change in real time. A clear semiology history (what the patient felt first, what happened next, and how they recovered, ideally corroborated by a witness) is often more diagnostically useful than a single routine interictal EEG (a recording taken between seizures), which can be normal even in confirmed epilepsy, since it only catches abnormal activity if a discharge happens to occur during the recording. Structural imaging, usually MRI brain, is used to look for an underlying focal lesion (such as hippocampal sclerosis or a tumour) once the seizure type itself has been established from the history and EEG pattern.

Diagnosis and management follow NICE epilepsy guidance, with antiepileptic drug choice guided specifically by seizure type, focal versus generalised, rather than treating "epilepsy" as one single condition, since the wrong drug for a given seizure type can worsen control (sodium valproate or lamotrigine are typical first-line choices for generalised seizures, while lamotrigine or carbamazepine are typical first-line choices for focal seizures, though the exact choice always follows current NICE and specialist guidance rather than a fixed rule). Non-epileptic seizures are not treated with antiepileptic drugs at all once correctly identified; management instead centres on psychological therapy and addressing the underlying psychological driver, which is exactly why getting the diagnosis right with video-EEG telemetry matters so much before committing a patient to lifelong medication.

Vignettes often describe a rising stomach sensation and déjà vu before an episode of impaired awareness and expect recognition of a temporal lobe (focal impaired-awareness) seizure rather than a generalised seizure. Another recurring trap offers a normal prolactin level as supposed proof an event was non-epileptic: the correct answer is that prolactin is unreliable in isolation and video-EEG telemetry is the definitive investigation, not a single blood result.

Seizure semiology localises the seizure: temporal lobe brings an aura and post-ictal confusion, frontal lobe is brief and bizarrely nocturnal, absence starts and stops abruptly, and non-epileptic events are gradual and prolonged with no true post-ictal state: and video-EEG telemetry, not prolactin, is what actually confirms which one you're looking at.