Definitions

A febrile seizure is defined by the International League against Epilepsy as a seizure occurring in association with a febrile illness in the absence of a central nervous system (CNS) infection or acute electrolyte imbalance in children older than 1 month of age without prior afebrile seizures [Commission on Epidemiology and Prognosis, 1993]. This definition is similar to the one adopted earlier by the National Institutes of Health (NIH) Consensus Conference [1980]. The febrile illness must include a body temperature of more than 38.4°C, although the increased temperature may not occur until after the seizure. The child may be neurologically normal or abnormal. In the International League against Epilepsy guidelines for epidemiologic research [Commission on Epidemiology and Prognosis, 1993], febrile seizures and, by extension, febrile status epilepticus are further divided into those occurring in children without prior neonatal seizures and those in children with prior neonatal seizures. In earlier epidemiologic studies, the lower age range used was either 1 month [Annegers et al., 1979, 1987, 1990; Berg et al., 1990, 1992, 1997; Verity et al., 1985a] or 3 months [Nelson and Ellenberg, 1976, 1978]. No specific upper age limit is used. Febrile seizures are most common, however, between the ages of 6 months and 3 years, with peak incidence at approximately 18 months of age. Onset after the age of 7 years is uncommon.

Febrile seizures are further classified as simple or complex. A febrile seizure is complex if it is focal, prolonged (lasting for more than either 10 minutes [Annegers et al., 1979, 1987, 1990; Berg and Shinnar, 1996a; Berg et al., 1992, 1997] or 15 minutes [Berg and Shinnar, 1996a; Nelson and Ellenberg, 1976, 1978]), or multiple (occurrence of more than one seizure during the febrile illness). Conversely, it is simple if it is an isolated, brief, generalized seizure. Although neurologically abnormal children are more likely to experience complex febrile seizures and have a higher risk for subsequent afebrile seizures, the child’s prior neurologic condition is not used to classify the seizure as simple or complex [Commission on Epidemiology and Prognosis, 1993; NIH, 1980]. When a careful history is obtained, approximately 30 percent of patients with febrile seizures presenting to the emergency department are found to have complex features [Berg and Shinnar, 1996a].

Epidemiology

Febrile seizures are the most common form of childhood seizures. The peak incidence is at the age of approximately 18 months. In the United States and Western Europe, they occur in 2–4 percent of all children [AAP, 1996; Annegers et al., 1990; Berg, 1995; Nelson and Ellenberg, 1978; van den Berg and Yerushalmi, 1969; Verity et al., 1985a]. In Japan, however, 9–10 percent of all children experience at least one febrile seizure [Tsuboi, 1984], and rates as high as 14 percent have been reported from the Mariana Islands in Guam [Stanhope et al., 1972].

Traditionally, it was thought that febrile seizures most commonly occur as the first sign of a febrile illness. More recent studies, however, found that only 21 percent of the children experienced their seizure either before or within 1 hour of the onset of the fever, 57 percent had a seizure after 1–24 hours of fever, and 22 percent experienced their febrile seizure more than 24 hours after the onset of the fever [Berg et al., 1992, 1997].

Some children are at increased risk of experiencing a febrile seizure. In a case-control population-based study, Bethune et al. [1993] examined risk factors for a first febrile seizure and found that the following four factors were associated with an increased risk of febrile seizures:

Children with two of these factors had a 28 percent chance of experiencing at least one febrile seizure.

In another case-control study using febrile controls matched for age, site of routine pediatric care, and date of visit, Berg et al. [1995] examined the issue of which children with a febrile illness were most likely to experience a febrile seizure. On multivariate analysis, significant independent risk factors were the height of the peak temperature and a history of febrile seizures in an immediate relative. Gastroenteritis as the underlying illness had a significant inverse (i.e., protective) association with febrile seizures. Similar results on the importance of the peak temperature were reported from a hospital-based study [Rantala et al., 1995].

The majority of febrile seizures are simple seizures. In a recent study of 428 children with a first febrile seizure, Berg and Shinnar [1996a] found that 35 percent had at least one complex feature, including focality in 16 percent, multiple seizures in 14 percent, and prolonged duration (longer than 10 minutes) in 13 percent. Approximately 6 percent of children had at least two complex features, and 1 percent had all three complex features. Of most concern have been prolonged febrile seizures. In that study, 14 percent of the children had seizures lasting longer than 10 minutes; 9 percent, longer than 15 minutes; and 5 percent, longer than 30 minutes, or febrile status epilepticus. Although febrile status epilepticus accounts for only 5 percent of febrile seizures, it accounts for approximately 25 percent of all cases of childhood status epilepticus [Aicardi and Chevrie, 1970; DeLorenzo et al., 1996; Dodson et al., 1993; Dunn, 1988; Maytal et al., 1989; Shinnar et al., 1997], and for more than two-thirds of cases of status epilepticus in the second year of life [Shinnar et al., 1997].

Initial Evaluation

As implied by the definition, to make the diagnosis of a febrile seizure, one must exclude meningitis, encephalitis, serious electrolyte imbalance, and other acute neurologic illnesses. These exclusions usually are feasible and are based on a detailed history and physical and neurologic examination. The most common issue in the emergency department is whether a lumbar puncture is necessary to exclude CNS infection, particularly meningitis or encephalitis. As reported in several studies, the incidence of meningitis in children with an apparent febrile seizure is between 2 and 5 percent [AAP, 1996; Heijbel et al., 1980; Jaffe et al., 1981; Joffe et al., 1983; Lorber and Sunderland, 1980; McIntyre et al., 1990; Rutter and Metcalf, 1978; Rutter and Smales, 1977; Wears et al., 1986]. In all of the reported series, however, a majority of the children with meningitis had identifiable risk factors. In one series, Joffe et al. [1983] reported that the children with meningitis had one of the following four features: a visit for medical care within the previous 48 hours, seizures on arrival to the emergency room, focal seizures, or suspicious findings on physical or neurologic examination. Other authors also have found a low yield on routine lumbar puncture in the absence of risk factors [AAP, 1996; Lorber and Sunderland, 1980; McIntyre et al., 1990; Rossi et al., 1986; Wears et al., 1986].

The AAP [1996] issued guidelines for the neurodiagnostic evaluation of the child with a simple febrile seizure between the ages of 6 months and 5 years. It recommended that a lumbar puncture be strongly considered in the infant younger than 12 months of age. The child between 12 and 18 months of age needs careful assessment, because the signs of meningitis may be subtle. In the absence of suspicious findings on history or examination, a lumbar puncture is not necessary in a child older than 18 months. A lumbar puncture is still recommended in children with a first complex febrile seizure, as well as in any child with persistent lethargy. It also should be strongly considered in a child who has already received prior antibiotic therapy. A lumbar puncture also should be considered in the child older than 5 years of age who presents with an apparent first febrile seizure, to exclude the possibility of encephalitis, as well as meningitis.

In the absence of suspicious findings in the history (e.g., vomiting, diarrhea) or on physical examination, routine blood cell counts and determination of serum electrolyte, calcium, phosphorus, magnesium, or blood glucose levels are of limited value in the evaluation of a child older than 6 months of age with a febrile seizure [AAP, 1996; Gerber and Berliner, 1981; Heijbel et al., 1980; Jaffe et al., 1981; Rutter and Smales, 1977].

Skull radiographs are of no value. Computed tomography (CT) scans also are of limited benefit in this clinical setting. Magnetic resonance imaging (MRI) scans are not indicated in children with a simple febrile seizure [AAP, 1996]. Whether an MRI study is indicated in the evaluation of a child with a prolonged or focal febrile seizure remains unclear [Lewis et al., 2002; Mitchell and Lewis, 2002; Rosman, 2002; VanLandingham et al., 1998]. MRIs often are clinically indicated in the diagnostic evaluation of the child who presents with status epilepticus, whether febrile or not [Dodson et al., 1993].

Electroencephalograms (EEGs) are of limited value in the evaluation of the child with febrile seizures [AAP, 1996; Koyama et al., 1991; Maytal et al., 2000; Millichap, 1991; Sofianov et al., 1983, 1992; Stores, 1991]. They are more likely to be abnormal in the older child with febrile seizures and in children with a family history of febrile seizures, with a complex febrile seizure, or with pre-existing neurodevelopmental abnormalities [Doose et al., 1983; Frantzen et al., 1968; Koyama et al., 1991; Millichap et al., 1960; Sofianov et al., 1983, 1992; Tsuboi, 1978]. Although EEG abnormalities may be present in these children, their clinical significance is unclear. There is no evidence that they help predict either recurrence of febrile seizures or the development of subsequent epilepsy [AAP, 1996; Doose et al., 1983; Frantzen et al., 1968; Koyama et al., 1991; Kuturec et al., 1997; Maytal et al., 2000; Millichap et al., 1960; Sofianov et al., 1983, 1992; Tsuboi, 1978]. In most cases, they are of limited clinical value in the evaluation of the child with a febrile seizure. EEGs are indicated in the diagnostic evaluation of status epilepticus of all types, including febrile [Riviello et al., 2006], and may be of particular interest in the child with febrile status epilepticus, as they may have predictive value for the development of subsequent epilepsy [Frantzen et al., 1968; Nordli et al., 2009].

Pathophysiology

The pathophysiology of febrile seizures remains unclear. An age-specific increased susceptibility to seizures induced by fever is likely. Although it was thought that the rate of rise of the temperature was the key factor [Livingston et al., 1947], more recent data suggest that it is the actual peak temperature [Berg, 1993; Berg et al., 1995; Rantala et al., 1995]. Although, by definition, CNS infections are excluded, the nature of the illness does appear to play a role. Gastroenteritis is associated with a lower incidence of febrile seizures [Berg et al., 1995], and herpesvirus-6 and herpesvirus-7 infections have had a high reported rate of association with febrile seizures [Barone et al., 1995; Caserta et al., 1998; Hall et al., 1994; Kondo et al., 1993; Theodore et al., 2008]. There is increased interest in the role of herpesvirus-6 in the pathogenesis of prolonged febrile seizures and subsequent temporal lobe epilepsy [Theodore et al., 2008]. Animal models of febrile seizures also reveal an age-dependent effect [Baram et al., 1997; Germano et al., 1996; Holtzman et al., 1981; Olson et al., 1984]. In addition, in vitro preparations demonstrate induction of epileptiform activity by temperature elevation in hippocampal slices in young rats [Tancredi et al., 1992]. Animal data suggest that young rats with neuronal migration disorders are more susceptible to hyperthermia-induced seizures and also are more susceptible to hippocampal damage [Germano et al., 1996]. Of interest, in this model, hippocampal damage occurs with hyperthermia, even in the absence of seizures.

An animal model developed by Baram et al. [1997] mimics the age-specific human condition and has been providing many insights into the pathophysiology and physiologic consequences of febrile seizures. In this model, febrile seizures appear to be of limbic origin [Baram et al., 1997; Chen et al., 1999; Dube et al., 2000; Toth et al., 1998]. As in humans, brief febrile seizures have no detectable anatomic or physiologic sequelae [Chen et al., 1999; Dube et al., 2000; Toth et al., 1998]. Febrile seizures lasting longer than 20 minutes, however, although not causing cell death [Toth et al., 1998], are associated with long-lasting changes in h-channels, The h-channel is the hyperpolarization-activated cation channel, also known as the pacemaker channel, which can be either “excitatory” or “inhibitory” [Poolos, 2004]. These changes are associated with increased susceptibility to seizures, although not with spontaneous seizures [Chen et al., 1999; Dube et al., 2000]. A full review of the findings from this model is outside the scope of this chapter, and the interested reader is referred to a recent monograph on this rapidly evolving field [Baram and Shinnar, 2002].

Related Morbidity and Mortality

The mortality associated with febrile seizures is extremely low. No deaths were reported from the National Collaborative Perinatal Project [Nelson and Ellenberg, 1978, 1976] or the British Cohort Study [Verity et al., 1985a, 1985b, 1993]. Even in cases of febrile status epilepticus, which represents the extreme end of complex febrile seizures, the mortality rates in recent series are extremely low [DeLorenzo et al., 1996; Dodson et al., 1993; Dunn, 1988; Maytal and Shinnar, 1990; Maytal et al., 1989; Shinnar et al., 2001; Towne et al., 1994; Verity et al., 1993]. In one recent report of 172 cases of febrile status epilepticus, only one death occurred, which in retrospect likely was related to Shigella encephalopathy, rather than to the febrile seizure [Shinnar et al., 2001]. Neither the National Perinatal Project nor the British studies found any evidence of permanent motor deficits after febrile seizures. This finding coincides with a recent series of febrile status epilepticus studies [DeLorenzo et al., 1996; Dunn, 1988; Maytal et al., 1989; Shinnar et al., 2001; Towne et al., 1994; Verity et al., 1993].

The cognitive abilities of children with febrile seizures have been extensively studied. No reports describe acute deterioration of cognitive abilities after febrile seizures, even when studies limited to status epilepticus are included [Dunn, 1988; Ellenberg and Nelson, 1978; Hirtz, 2002; Maytal and Shinnar, 1990; Maytal et al., 1989; Shinnar et al., 2001; Verity et al., 1993]. Cognitive abilities and school performance of children with febrile seizures were found to be similar to those of controls in three large studies [Ellenberg and Nelson, 1978; Ross et al., 1980; Verity et al., 1985b, 1993]. The Collaborative Perinatal Project found no difference in IQ scores or performance on the Wide Range Achievement Test at the age of 7 years between children with febrile seizures and their siblings [Ellenberg and Nelson, 1978]. The British National Child Development Study reported that children with febrile seizures performed as well in school at 7 and 11 years of age as their peers without a history of febrile seizures [Ross et al., 1980]. The more recent British Cohort Study also found no difference between 5-year-olds with febrile seizures and 5-year-olds without a history of febrile seizures on a variety of performance tasks [Verity et al., 1985b, 1993]. A recent study from Taiwan, in addition to comparing intelligence and behavior, also found no difference in memory between children with febrile seizures, including complex ones, and population-based controls [Chang et al., 2000, 2001]. This finding is of particular interest because febrile seizures appear to be of limbic origin [Baram et al., 1997; Dube et al., 2000], and memory is subserved by the hippocampus.

Even prolonged febrile seizures do not appear to be associated with adverse cognitive outcomes [Ellenberg and Nelson, 1978; Shinnar and Babb, 1997; Shinnar et al., 2001; Verity et al., 1993]. In the British Cohort Study, no differences were found between 5-year-olds with and those without febrile seizures, even when the analysis was limited to complex febrile seizures [Verity et al., 1985b, 1993]. Ellenberg and Nelson [1978] examined 27 children with febrile convulsions lasting more than 30 minutes and found no differences in cognitive function at 7 years of age between them and their siblings. Insufficient data are available on memory function in children who experienced prolonged febrile seizures, although memory is known to be impaired in those with chronic temporal lobe epilepsy [Bell and Davies, 1998].

Recurrent Febrile Seizures

Approximately one-third of children with a first febrile seizure will experience a recurrence, and 10 percent will have three or more febrile seizures [AAP, 1996; Annegers et al., 1990; Berg, 2002; Berg et al., 1990, 1992, 1997; Knudsen, 1985b; Nelson and Ellenberg, 1976, 1978; Offringa et al., 1992, 1994; van den Berg, 1974; Verity et al., 1985a]. Factors associated with a differential risk of recurrent febrile seizures are summarized in Table 57-1. The most consistent risk factors reported are a family history of febrile seizures and age at first febrile seizure (before age 18 months) [AAP, 1996; Annegers et al., 1990; Berg, 2002; Berg et al., 1990, 1992, 1997; Knudsen, 1985b; Nelson and Ellenberg, 1978; 1976; Offringa et al., 1992, 1994; van den Berg, 1974]. This relation appears to be due to the longer period during which a child with a younger age at onset will be in the age group at risk for febrile seizures, rather than to a greater tendency to have seizures with each specific illness [Offringa et al., 1992, 1994; Shirts et al., 1987].

Table 57-1 Risk Factors for Recurrent Febrile Seizures and for Epilepsy after a Febrile Seizure

(Data from Annegers et al., 1979, 1987, 1990; Berg and Shinnar, 1996b; Berg et al., 1990, 1992, 1997; El-Rahdi and Banajeh, 1989; Knudsen, 1985b; Nelson and Ellenberg, 1976, 1978; Offringa et al., 1992, 1994; van den Berg, 1974; Verity and Golding, 1991.)

In studies that examined features of the acute illness, the peak temperature [Berg et al., 1990, 1992, 1997; El-Rahdi and Banajeh, 1989; Offringa et al., 1992, 1994] and also the duration of the fever before the seizure [Berg et al., 1992, 1997] were associated with a differential risk of recurrent febrile seizures. The higher the peak temperature, the lower the chance of recurrence. In the studies by Berg et al. [1992, 1997], patients with a peak temperature of 101°F had a 42 percent recurrence risk at 1 year, compared with 29 percent for those with a peak temperature of 103°F and only 12 percent for those with a peak temperature of 105°F or greater. The shorter the duration of recognized fever, the higher the chance of recurrence. For those with a febrile seizure within an hour of recognized onset of fever, the recurrence risk at 1 year was 46 percent, compared with 25 percent for those with prior fever lasting 1–24 hours and 15 percent for those with more than 24 hours of recognized fever before the febrile seizure. The fact that those infants in whom the febrile seizure occurs at the onset of fever have the highest risk of recurrence has implications for prophylactic strategies that rely on giving medications at the onset of the febrile illness.

The data regarding a family history of unprovoked seizures or epilepsy are conflicting. A large study in Rochester, Minnesota, found no difference in recurrence risk between children with a family history of epilepsy (25 percent) and those with no such family history (23 percent) [Annegers et al., 1990]. Other studies have found more equivocal results [Berg et al., 1992, 1997; Offringa et al., 1992, 1994]. However, even those studies that report an increased risk of recurrence in children with a family history of unprovoked seizures found only a modest increase in risk. A complex febrile seizure is not associated with an increased risk of recurrence in most studies [Annegers et al., 1990; Berg et al., 1990, 1992, 1997; Berg and Shinnar, 1996a; Nelson and Ellenberg, 1978; Offringa, 1992, 1994]. If the initial febrile seizure is prolonged, however, a recurrent febrile seizure also is more likely to be prolonged [Berg and Shinnar, 1996a; Offringa et al., 1994]. The presence of a neurodevelopmental abnormality in the child also has not been demonstrated to be significantly associated with an increased risk of subsequent febrile seizures [Annegers et al., 1990; Berg et al., 1990, 1992, 1997; Offringa et al., 1992, 1994]. Ethnicity and gender also have not been associated with a clear differential risk of recurrent febrile seizures.

Children with multiple risk factors are at highest risk for recurrence [Berg, 2002; Berg et al., 1997; Offringa et al., 1994]. A child with two or more risk factors has a greater than 30 percent recurrence risk at 2 years, and the child with three or more risk factors has a greater than 60 percent recurrence risk [Berg et al., 1997]. By contrast, the child older than 18 months with no family history of febrile seizures who experiences a first febrile seizure associated with a peak temperature higher than 40°C after a recognized fever longer than 1 hour in duration (i.e., no risk factors) has a 2-year recurrence risk of greater than 15 percent [Berg et al., 1997; Offringa et al., 1994].

Febrile Seizures and Subsequent Epilepsy

Data from five large cohorts of children with febrile seizures indicate that epilepsy subsequently develops in 2–10 percent of children who experience febrile seizures [Annegers et al., 1979, 1987; Berg and Shinnar, 1996b; Nelson and Ellenberg, 1978; van den Berg and Yerushalmi, 1969; Verity and Golding, 1991]. The higher number comes from the study by Annegers et al. [1987], which had the longest follow-up period. In addition, in population- and community-based studies, 15–20 percent of children and adults with epilepsy have a history of prior febrile seizures [Berg et al., 1999; Camfield et al., 1994; Hamati-Haddad and Abou-Khalil, 1998; Rocca et al., 1987a, 1987b, 1987c; Sofianov et al., 1983