Most seizures stop. Failure to do so implies that the mechanisms needed to stop seizures are ineffective or treatment simply fails. Defining status epilepticus (SE) clinically is virtually impossible, and many recommendations have included time periods, but the time periods are artificial and not clinically useful. In epidemiologic studies, a definition of ongoing or recurrent seizure without return to baseline level of consciousness for a period of 30 minutes or longer has been us.¹ In clinical practice, however, deferring treatment for 30 minutes is unacceptable, and most physicians have adopted an operational definition of seizures persisting beyond 5 to 10 minutes.² Generally, treatment of convulsive SE should be well on its way before the results of the first electroencephalogram (EEG) become available.

Generalized convulsive status epilepticus (GCSE) is the most common and potentially harmful type of SE. GCSE can evolve from primary generalized or partial seizures with secondary generalization. Convulsive and nonconvulsive SE is the leading indication for emergency EEG and for prolonged video-EEG monitoring (VEM) in the intensive care unit.^3–5 Whereas GCSE is diagnosed on clinical grounds, EEG plays a crucial role in determining progression to subtle or nonconvulsive SE, guiding pharmacologic treatment, and in prognostication. The causes, outcome, and management of GCSE in children and adults are quite distinct, and this chapter will focus on adults.

The true frequency of GCSE is somewhat difficult to determine. Case series from tertiary care hospitals are misleading because of selection bias, whereas population-based studies have generally combined GCSE with partial and nonconvulsive SE types. Prospective and retrospective population-based studies from the United States and Europe have estimated an incidence for SE in general of 8.0 to 41.0 cases per 100,000 persons per year.^6–12 The incidence is greatest in the first year of life and after the age of 60.^6–9,11 GCSE represents 34 to 75% of the SE cases in these reports.^6–9 A review of GCSE identified from a California hospital discharge diagnosis database reported an incidence rate of 4.58 per 100,000 for adults ages 20 to 54, increasing to 22.32 per 100,000 for patients older than 75.¹² The incidence of SE appears to be greater in men, for reasons that are unclear.^6–8,13,14 The impact of race, socioeconomic status, and rural versus urban environment remains uncertain. Most cases (75%) will resolve with treatment in <24 hours.^7,9 SE persisting beyond 24 hours is more often seen in elderly patients and in those with acute underlying pathology.⁷

More than half of all patients with SE have no prior history of epilepsy.^6,9,15 This is particularly the case in the elderly. For 70% of those 60 years and older, SE is the initial seizure.⁶ Among persons of all ages with established epilepsy, 20% will have an episode of SE within the first 5 years of diagnosis, and 12% will present with SE.^7,16 One episode of SE increases the risk of future SE. In the population-based epidemiologic studies as of 2001, recurrent SE represents 13 to 18% of the total.^6,7,9 Hesdorffer et al reported a recurrence risk for nonfebrile SE of 13% at 1 year, 20% at 2 years, and 30% at 10 years.¹⁷

Causes of SE can be broadly classified into four main groups. The first is SE arising from an acute symptomatic insult to the cerebrum, such as stroke, hemorrhage, hypoxic-ischemic injury, infection, primary or metastatic brain tumors, or toxic/metabolic derangements. The acute symptomatic group comprises the majority of SE cases in adults.^6–9,18 The second is SE resulting from a remote symptomatic cause, such as chronic cerebrovascular injury, prior traumatic brain injury, or a congenital malformation. GCSE also occurs as a complication of preexisting epilepsy, usually as a result of subtherapeutic serum antiepileptic drug (AED) levels. Finally, GCSE can have no identifiable cause, which in some may be the initial presentation of idiopathic or cryptogenic epilepsy.

In adults, the most common etiology of SE is acute and chronic cerebrovascular disease, followed by subtherapeutic levels of AEDs and hypoxic-ischemic injury. In a prospective population-based study of SE in Richmond, Virginia, these causes accounted for 40%, 30%, and 10% of all SE cases, respectively.^6,8,19 In elderly patients, 40% of GCSE is attributable to acute or chronic stroke.²⁰ Other less common symptomatic etiologies are metabolic derangements, toxic effects of drugs or alcohol, brain tumors, degenerative disorders, and infection, each of which accounts for <10% of all SE.⁶ Idiopathic cases are in the minority, with at most 15% of SE having no identifiable cause.^6–9 Currently available case series have not differentiated between SE occurring in preexisting partial versus generalized epilepsy, but clinical experience suggests that GCSE is a rare occurrence in idiopathic generalized epilepsy.

Anyone caring for patients in an inpatient VEM unit must be aware that GCSE is a known potential adverse outcome of this procedure. It should come as no surprise that when AEDs are withdrawn, and additional procedures such as sleep deprivation are used to induce seizure for diagnostic purposes, SE will sometimes result. In a report from five VEM units at tertiary care hospitals, 48.5% of patients had seizure clusters, and 3% developed SE.²¹ It is therefore critical that patients be closely monitored and that protocols are in place for early therapeutic intervention at all times to avoid unwarranted morbidity and mortality.

The mortality rate for SE varies with underlying etiology, age, and duration. In population-based studies from Olmsted County, Minnesota, and Richmond, Virginia, the overall mortality for SE was ~20% at 30 days.^22,23 In French-speaking Switzerland, the rate was 7.6%, a difference that may be explained in part by exclusion of anoxic injury in the European study.⁸ In a retrospective review of 11,580 adult and pediatric cases of GCSE from the U.S. National Inpatient Sample database, the in-hospital mortality rate was 3.5% overall and 7.4% in those requiring mechanical ventilation.¹⁴ The National Inpatient Sample database includes discharge data collected from approximately 1000 hospitals across the United States, including small, nonteaching, and rural providers. In contrast, a review of 96 adults treated for SE at two tertiary care hospitals in Boston, Massachusetts, found a mortality rate at discharge of 16% overall and 40% for GCSE when cases arising from anoxic injury were excluded.¹⁸ In the California study of GCSE based on hospital discharge coding, the in-hospital mortality rate was 10.7%.¹³

It is difficult to determine the degree to which morbidity and mortality result from GCSE rather than the underlying symptomatic cause. The highest mortality is in cases related to anoxic brain injury or acute cerebrovascular disease.^6,14,23 SE resulting from subtherapeutic AED levels has a relatively low mortality (4%), but it may be more responsive to pharmacologic treatment than symptomatic GCSE.⁶ It has also been shown that ischemic stroke and intracerebral and subarachnoid hemorrhage have worse outcomes when complicated by GCSE.^24,25

Mortality in SE is greater in elderly patients. The population-based analysis of SE cases in Richmond, Virginia, found that the overall mortality rate for adults was 26%, increasing to 38% in patients age 60 or older and to 50% after age 80.^6,22 Similarly, in-hospital mortality increases with age, with an odds ratio for death of 7.37 for adults ages 61 to 70 and 12.59 for adults age 80 or older compared with children younger than 10.¹⁴

Longer duration of SE has been linked to poorer outcome.^15,26,27 In one report, the 30-day mortality for SE lasting >1 hour rose from 3 to 32%.²⁷ The presence of continuous rather than intermittent clinical seizure activity is associated with greater mortality in adults with a relative risk of 1.79.²⁸

Long-term morbidity from GCSE often includes future seizures. In earlier data from a tertiary epilepsy clinic, it was estimated that epilepsy develops following SE ~25% of the time, a number similar to that seen after a single unprovoked seizure.^29,30 However, in a larger population-based study from Olmsted County, Minnesota, 40% of adults presenting with SE of acute symptomatic cause had epilepsy 10 years later, compared with only 13% following an acute symptomatic seizure.³¹ Additional adverse sequelae include cognitive impairment, aphasia, and motor deficits, affecting 3 to 8% of survivors.^15,26,32

Ictal Behavior

Convulsive SE often begins with repetitive partial or generalized tonic-clonic seizures occurring in increasingly rapid succession with progressively greater interictal obtundation (Videos 24-1a, 24-1b, and 24-1c). The tonic phase will typically shorten, and clonic movements will increasingly diminish and often become asynchronous. For those with idiopathic generalized epilepsy, GCSE may be heralded by increasingly frequent myoclonic seizures. As seizures continue, the motor manifestations become more subtle and may be limited to rhythmic muscle twitching in the face or eyes, or to low-amplitude multifocal myoclonus. This state has been referred to as “subtle generalized convulsive status.”³³ In a prospective study of 518 patients presenting with SE, 384 had clinically overt and 134 had subtle GCSE, with a median duration of seizure activity of 2.8 and 5.8 hours, respectively.³⁴ If SE persists beyond this stage, a state of electromechanical dissociation can follow in which no motor activity is observed, but continued electrographic seizure activity is recorded on the EEG (nonconvulsive SE).³⁵

Ictal EEG

EEG confirms the clinical impression of SE and assists in judging the effects of AEDs. The initial findings are no different than would be seen with any primary or secondary generalized seizure, with the differentiating feature being that seizure discharges cluster or persist. As motor movements become more subtle, or when patients are pharmacologically paralyzed, the EEG will usually demonstrate a pattern of generalized rhythmic discharges, which may be continuous, waxing/waning in frequency and amplitude, or intermittent (Figure 24-1).^33,35–39 EEG patterns of periodic epileptiform discharges or of interruption of ictal discharges by periods of flattening also occur.^33,36,38,39 It is important for those interpreting EEG to recognize that these discontinuous patterns can still represent persistent electrographic SE. The EEG of nonconvulsive SE is discussed in greater detail in Chapter 29.