Status Epilepticus

Status epilepticus (SE) is the condition in which the mechanisms that normally terminate a seizure fail, producing either an abnormally prolonged seizure or a series of seizures without recovery of consciousness between them. The modern ILAE definition is explicitly operational and time-anchored, a deliberate move away from the older arbitrary thirty-minute rule toward a framework that ties intervention to pathophysiology. The point t1 marks the time beyond which a seizure is unlikely to stop on its own and emergency treatment should begin; for convulsive (tonic-clonic) SE this is 5 minutes. The point t2 marks the time beyond which ongoing seizure activity risks long-term neuronal injury and altered networks; for convulsive SE this is 30 minutes. These thresholds differ by seizure type. For focal SE with impaired awareness, t1 is taken as roughly 10 minutes and t2 as a longer interval on the order of an hour, and for absence SE the t1 estimate is longer still and the t2 for lasting injury is not well established, reflecting that not all status carries the same urgency or the same risk of harm.

Why status perpetuates itself

The self-sustaining nature of SE is rooted in activity-dependent receptor trafficking. As seizures continue, functional GABA-A receptors are internalized from the synaptic membrane into the cell interior, reducing inhibitory tone, while NMDA and AMPA glutamate receptors are mobilized to the surface, amplifying excitation. Over minutes to tens of minutes the synapse is thus pharmacologically remodeled away from inhibition and toward excitation. The pharmacologic corollary is decisive and time-critical: benzodiazepines, which act by enhancing GABA-A receptor function, become progressively less effective the longer SE continues because their molecular target is being withdrawn from the membrane, a phenomenon termed time-dependent pharmacoresistance. This is the mechanistic justification for treating SE as an emergency in which every minute of delay both worsens neuronal injury and erodes the efficacy of first-line therapy. The same biology explains why refractory SE so often requires agents that act outside the GABA-A system, including NMDA-receptor antagonists such as ketamine and anesthetic infusions, once benzodiazepines and standard antiseizure medications fail.

The treatment pathway follows directly from this biology and from high-quality trial evidence. First-line therapy is an adequately dosed benzodiazepine, given without delay. If seizures continue, a second-line intravenous antiseizure medication is given; the major comparative trial of fosphenytoin, valproate, and levetiracetam in established status found the three roughly equivalent in efficacy, so the choice is guided by patient factors and contraindications rather than a clear winner. Failure of first- and second-line agents defines refractory status epilepticus, which typically requires anesthetic infusions with continuous EEG to titrate to seizure suppression or burst-suppression, and continued seizures despite anesthesia, or recurrence on weaning it, define super-refractory status epilepticus. Continuous EEG is essential once a patient is sedated or paralyzed, because the clinical examination is then uninformative and only the EEG can confirm whether seizures persist or whether the targeted degree of suppression has been reached.

Treating the seizure is only half of management; the other half is finding and treating its cause, because status epilepticus is a syndrome with many etiologies rather than a single disease. The common precipitants in established epilepsy are subtherapeutic antiseizure medication levels and intercurrent illness, while in patients without prior epilepsy the differential spans acute symptomatic causes such as stroke, intracranial hemorrhage, central nervous system infection, traumatic brain injury, hypoxic-ischemic injury after cardiac arrest, and metabolic derangements including hypoglycemia, hyponatremia, hypocalcemia, and drug toxicity or withdrawal. A distinct and important category is new-onset refractory status epilepticus (NORSE), refractory status without a readily identifiable cause in a patient without active epilepsy, and its subset febrile infection-related epilepsy syndrome (FIRES), in which a febrile illness precedes the explosive onset; a substantial proportion of NORSE cases prove to be autoimmune or paraneoplastic encephalitis, which is why an aggressive search for antineuronal antibodies and consideration of early immunotherapy has become part of the modern workup. The practical point is that EEG confirmation of ongoing seizure runs in parallel with, and never replaces, the urgent etiologic evaluation that may itself be the key to stopping the seizures.

Convulsive, subtle, and nonconvulsive status

Convulsive status epilepticus (CSE) is the most recognizable form, presenting with sustained or repetitive generalized tonic-clonic activity and impaired consciousness, and it is the form for which the 5-minute t1 threshold and immediate treatment apply. Its danger lies partly in electromechanical dissociation: as CSE continues, the visible motor activity can wane even while the brain remains in electrographic seizure, because the spinal and muscular machinery fatigues faster than the cortical discharge resolves. This evolution produces subtle status epilepticus, in which only minor twitching of the face, eyelids, or distal limbs, or even no movement at all, accompanies ongoing ictal discharges and profound coma. Subtle SE is a treatment-failure state with a grave prognosis, and it can be diagnosed only by EEG, because the clinical examination has gone nearly silent while the cortex continues to seize. The patient who stops convulsing has not necessarily stopped seizing, and the only way to know is to record.

The systemic dimension of CSE is part of why it is an emergency beyond the brain. The early phase is marked by a massive sympathetic surge with hypertension, tachycardia, hyperglycemia, and a rise in cerebral blood flow that initially compensates for the enormous metabolic demand of seizing cortex. As the seizure continues into a later, decompensated phase, this compensation fails: cerebral autoregulation breaks down, blood flow no longer matches metabolic need, and the combination of hyperthermia, lactic acidosis, rhabdomyolysis, hypoxia, and excitotoxic calcium entry inflicts injury that the t2 time point is meant to capture. Prognosis in status epilepticus is driven less by any single EEG feature than by the triad of underlying etiology, the patient's age, and the duration and refractoriness of the seizure; an acute symptomatic cause such as anoxic injury after cardiac arrest carries a far worse outlook than a low antiseizure-drug level in a person with established epilepsy. This is why the dual mandate of stopping the seizure quickly and identifying its cause is not merely procedural but the principal determinant of whether the patient survives with intact function.

Nonconvulsive status epilepticus (NCSE) is ongoing electrographic seizure activity producing a change in awareness, behavior, or cognition without prominent motor manifestations. It spans a clinical spectrum from the relatively benign absence status and focal NCSE with preserved or fluctuating responsiveness to the malignant comatose NCSE of the critically ill patient, and these endpoints carry very different prognoses and very different treatment urgencies. NCSE is dramatically underrecognized: in cohorts of comatose patients in the intensive care unit, continuous EEG reveals nonconvulsive seizures in a substantial minority, and a meaningful fraction of patients who fail to wake after convulsive SE are found to be in ongoing nonconvulsive status. The clinical lesson is that unexplained altered consciousness is an indication for EEG, because NCSE is invisible without it and frequently treatable when found.

Detecting NCSE reliably usually requires continuous EEG (cEEG) monitoring rather than a brief routine study, because nonconvulsive seizures are often intermittent and may be missed by a twenty-minute snapshot. Consensus guidance recommends cEEG in patients with persistently altered mental status after convulsive status, in comatose critically ill patients in whom subclinical seizures are plausible, and to titrate anesthetic therapy in refractory status. The yield is time-dependent: many seizures are captured in the first day of recording, but a non-comatose patient may need roughly a day of monitoring and a comatose patient closer to two days to approach maximal sensitivity, because the probability of capturing an intermittent seizure rises with recording duration. Because reviewing days of multichannel EEG by eye is impractical, quantitative EEG (qEEG) trend tools such as color density spectral arrays, amplitude-integrated EEG, and rhythmicity spectrograms compress hours of data into a glance and help technologists and clinicians flag seizure-suspicious epochs for detailed review. These trends are screening aids, not diagnostic substitutes: every flagged epoch must be confirmed against the raw tracing for the evolution and reactivity that actually define a seizure, because artifacts and benign rhythms can produce trend signatures that mimic ictal activity.

The Salzburg criteria and the EEG diagnosis of NCSE

Because NCSE has no reliable motor signature, its diagnosis rests on the EEG, and the Salzburg consensus criteria provide the most widely used and validated framework. The criteria branch on whether the patient has a known epileptic encephalopathy, but the core logic is consistent. Unequivocal NCSE is established by repetitive epileptiform discharges at greater than 2.5 Hz, or by epileptiform discharges at 2.5 Hz or slower or rhythmic delta or theta activity that additionally show at least one of the following: spatiotemporal evolution, a clear clinical correlate time-locked to the pattern, or unequivocal clinical and electrographic improvement after intravenous benzodiazepine. Patterns that meet some but not all features are designated possible NCSE and demand further scrutiny, repeat or continuous recording, and integration with the clinical picture. The criteria thus operationalize the same principle that defines all ictal activity, evolution and reactivity, and adapt it to the bedside emergency where a static slow periodic pattern is common and the question is whether it has crossed into seizure.

Salzburg presupposes a shared vocabulary, and that vocabulary is the ACNS Standardized Critical Care EEG Terminology. This nomenclature names periodic and rhythmic patterns by a main term describing their localization (generalized, lateralized, bilateral independent, or multifocal) and their type (periodic discharges, rhythmic delta activity, or spike-and-wave), so that lateralized periodic discharges (LPDs), generalized periodic discharges (GPDs), lateralized rhythmic delta activity (LRDA), and generalized rhythmic delta activity (GRDA) mean the same thing to every reader. The terminology further attaches modifiers for prevalence, frequency, duration, and sharpness, and crucially a plus modifier that flags additional features such as superimposed fast activity, rhythmicity, or sharp morphology that nudge a pattern closer to the ictal end of the spectrum. A pattern annotated as a periodic discharge plus, at an intermediate frequency, carries a higher probability of being or becoming ictal than the same pattern without the plus, and this standardized language is what makes the Salzburg criteria reproducibly applicable across centers.

The recording below is a focal seizure whose evolution illustrates the very features the Salzburg criteria require. When such a pattern persists or recurs without recovery, or when a similar rhythmic discharge in an obtunded patient shows the spatiotemporal evolution that defines an electrographic seizure, the case for status epilepticus is made. Study the evolution here as the template you will look for in the critically ill patient, then imagine it failing to terminate: the same discharge, no longer self-limited, becomes the electrographic substrate of status.

The ictal-interictal continuum

Not every rhythmic or periodic pattern in a sick brain is clearly ictal or clearly benign. The ictal-interictal continuum (IIC) names the broad zone of patterns that share features of both, including lateralized periodic discharges, generalized periodic discharges, and lateralized or generalized rhythmic delta activity, especially when they occur at intermediate frequencies near 1 to 2.5 Hz, show fluctuating morphology, or carry sharp components or a plus modifier under the ACNS terminology. These patterns are not seizures by strict criteria, yet they are not innocent: they are associated with metabolic stress on cortex, may herald or accompany electrographic seizures, and in some patients are accompanied by markers of neuronal injury. They occupy a genuine middle ground that reflects the underlying biology, in which a stressed cortical network can hover near the seizure threshold without crossing it, rather than a failure of the classification to make up its mind.

Managing the IIC is an exercise in Bayesian reasoning rather than rule-following. The clinician weighs the discharge frequency, the presence or absence of any evolution, the plus modifiers, the clinical state and its trajectory, the neuroimaging, and ancillary markers, then often performs a diagnostic benzodiazepine trial: a definite and concordant resolution of both the EEG pattern and the clinical state supports an ictal interpretation, whereas no change argues against it. The trial is imperfect and must be interpreted with care, since benzodiazepines can suppress nonictal patterns and can sedate a patient into apparent improvement without clarifying whether the pattern was a seizure, so its result is incorporated as one piece of probabilistic evidence rather than a verdict. Increasingly, ancillary tools such as continuous EEG trend analysis, biomarkers of neuronal injury, and perfusion or metabolic imaging that can show regional hypermetabolism corresponding to the pattern are integrated to decide whether an IIC pattern warrants aggressive treatment, balancing the harm of undertreated seizures against the substantial harm of deep iatrogenic sedation, prolonged ventilation, and the complications of the intensive care unit.