Articles

Incidence, cause, and short-term outcome of convulsive

status epilepticus in childhood: prospective population-
based study
Richard F M Chin, Brian G R Neville, Catherine Peckham, Helen Bedford, Angela Wade, Rod C Scott, for the NLSTEPSS Collaborative
Group*

Lancet 2006; 368: 222–29
See Comment page 184
*Members listed at end of paper
Neurosciences Unit, Institute
of Child Health, University
College London, and Great
Ormond Street Hospital for
Children NHS Trust, London,
UK (R F M Chin MRCPCH,
B G R Neville FRCPCH,
R C Scott MRCPCH); Centre for
Paediatric Epidemiology and
Biostatistics, Institute of Child
Health, University College
London, London, UK
(R F M Chin, C Peckham FRCP,
H Bedford PhD, A Wade PhD);
and Radiology and Physics
Unit, Institute of Child Health,
University College London,
London, UK (R C Scott)
Correspondence to:
Dr Richard F M Chin, The Wolfson
Centre, Mecklenburgh Square,
London WC1N 2AP, UK
r.chin@ich.ucl.ac.uk
Summary
Background Convulsive status epilepticus is the most common childhood medical neurological emergency, and is
associated with significant morbidity and mortality. Most data for this disorder are from mainly adult populations
and might not be relevant to childhood. Thus we undertook the North London Status Epilepticus in Childhood
Surveillance Study (NLSTEPSS): a prospective, population-based study of convulsive status epilepticus in
childhood, to obtain a uniquely paediatric perspective.
Methods Clinical and demographic data for episodes of childhood convulsive status epilepticus that took place in
north London were obtained through a clinical network that covered the target population. We obtained these data
from anonymised copies of a standardised admission proforma; accident and emergency, nursing, ambulance,
and intensive-care unit notes; and interviews with parents, medical, nursing, and paramedic staff. We investigated
ascertainment using capture-recapture modelling.
Findings Of 226 children enrolled, 176 had a first ever episode of convulsive status epilepticus. We estimated that
ascertainment was between 62% and 84%. The ascertainment-adjusted incidence was between 17 and 23 episodes
per 100 000 per year. 98 (56%, 95% CI 48–63) children were neurologically healthy before their first ever episode
and 56 (57%, 47–66) of those children had a prolonged febrile seizure. 11 (12%, 6–18) of children with first ever
febrile convulsive status epilepticus had acute bacterial meningitis. Conservative estimation of 1-year recurrence
of convulsive status epilepticus was 16% (10–24%). Case fatality was 3% (2–7%).
Interpretation Convulsive status epilepticus in childhood is more common, has a different range of causes, and a
lower risk of death than that in adults. These paediatric data will help inform management of convulsive status
epilepticus and appropriate allocation of resources to reduce the effects of this disorder in childhood.

Introduction
Convulsive status epilepticus is the most common
medical neurological emergency in childhood.1,2 Despite
treatment advances over the last two decades, it
continues to be associated with substantial morbidity
and mortality.3–6 However, since most of the published
data for convulsive status epilepticus in childhood are
from hospital-based studies, the size of the population
at risk, the range of causes, and the natural history in
the general population of children have not been well
defined.
Previously reported epidemiological studies of
convulsive status epilepticus have been mainly or
entirely based on adult populations and thus their
results might not provide a reliable characterisation of
this disorder in children.2,7–12 Furthermore, data from
adults might not be directly applicable to children
because the physical and neurochemical characteristics
of the developed brain differ from those of the
developing brain.13 Thus we undertook the North
London Convulsive Status Epilepticus in Childhood
Surveillance Study (NLSTEPSS), a prospective
population-based study, to investigate these issues in
an urban resource-rich setting.
Methods
Definitions
We defined convulsive status epilepticus as a tonic, clonic,
or tonic-clonic seizure (continuous convulsive status
epilepticus), or two or more such seizures between which
consciousness was not regained (intermittent convulsive
status epilepticus), which lasted for at least 30 min.14
Children with no previous convulsive status epilepticus
before enrolment into the study were defined as having
“first ever episodes” of convulsive status epilepticus. All
episodes of convulsive status epilepticus reported in the
study period, irrespective of whether they were first ever
episodes, were classified as occurrences. We classified as
“recurrences” all episodes after first ever episodes when
both took place in the study period.
Causes and seizure types were classified by two of the
authors. Disagreements about classification were resolved
by consensus or by third-party adjudication. The
classifications for causes are shown in the panel.15 We
classified seizure types as primary generalised, focal with
secondary generalisation, and focal. Generalised episodes
of convulsive status epilepticus were subtyped into tonic,
clonic, and tonic-clonic.14 We defined epilepsy as two or
more unprovoked seizures.

222 www.thelancet.com Vol 368 July 15, 2006


Procedures
Between May 1, 2002, and April 30, 2004, children aged
between 29 days and 15 years with episodes of convulsive
status epilepticus and who lived within a 494·29 km²
region of London, UK, north of the River Thames (north
London), were notified to NLSTEPSS. North London is
composed of 15 administrative boroughs to which UK
Census demographics relate, and has a population of
605 230 children (308 156 boys).
We enrolled children via a clinical network that covered
the target population. The network consisted of 18 hospitals
with 24 h accident and emergency services, five paediatric
intensive-care units, and a regional centralised paediatric
intensive-care unit retrieval service (the Children’s Acute
Transport Service). Medical, nursing, and administrative
staff reported potentially eligible patients by telephone to
the NLSTEPSS research-coordinating centre. Children
were also enrolled via an active regional surveillance-card
scheme developed in conjunction with the national British
Paediatric Surveillance Unit. Each month, we sent all
paediatricians in north London a reporting card, which
requested notification of children who had presented with
convulsive status epilepticus in the preceding month, or to
indicate if none had presented. For another source of
enrolment, daily telephone calls were made by members
of the research team to the Children’s Acute Transport
Service centre to identify children having presented with
convulsive status epilepticus the preceding day. At the time
of notification, we requested only the minimum
information required to verify the inclusion criteria for
enrolment. The research team assessed each reported case
of convulsive status epilepticus within 2 weeks of
notification to verify their eligibility for inclusion in the
study. Thus children who were notified only through the
surveillance-card scheme were assessed up to 6 weeks after
their episode of convulsive status epilepticus. Residence
within north London was validated by documentation of
each child’s home postcode and checking against an
electronic database of north London postcodes.
A standardised admission proforma, which had been
approved by the research collaborative group, provided
data for each episode in sufficient detail for characterisation
of the clinical aspects of episodes of convulsive status
epilepticus. Where necessary, additional information was
obtained from accident and emergency, nursing,
ambulance, and intensive-care records. Direct interview
with attending doctors, nurses, paramedics, and parents
further helped with accurate and detailed accounts of
clinical course and management. Data for each episode of
convulsive status epilepticus were entered into an
electronic database together with the child’s medical,
social, and family details. Follow-up data for recurrence of
convulsive status epilepticus were available up to 2 months
after the initial 2-year study period. Thus the possible
lengths of follow-up range from 2 to 26 months.
At the end of each year of the study, members of the
research team reviewed all cases of convulsive status
www.thelancet.com Vol 368 July 15, 2006
Articles
Panel: Classification system for the causes of convulsive
status epilepticus in childhood
1 Prolonged febrile seizure
Convulsive status epilepticus in a previously neurologically
healthy child aged between 6 months and 5 years during a
febrile (temperature >38°C) illness and in the absence of
defined CNS infection.
2 Acute symptomatic
Convulsive status epilepticus in a previously neurologically
normal child, within a week of an identified acute
neurological insult including bacterial meningitis, viral CNS
infection, metabolic derangements, drug-related effects,
head injury, hypoxia or anoxia, cerebrovascular disease.
3 Remote symptomatic
Convulsive status epilepticus in the absence of an identified
acute insult but with a history of a pre-existing CNS
abnormality more than 1 week previously.
4 Acute on remote symptomatic
Convulsive status epilepticus that occurred within a week of
an acute neurological insult or febrile illness and occurred in a
child with a history of previous neurological abnormality,
including epilepsy. This category included children with
cerebral palsy with a febrile illness not of CNS origin, and
children with obstructed ventriculoperitoneal shunts for
hydrocephalus.
5 Idiopathic epilepsy related
Convulsive status epilepticus that is not symptomatic and
occurred in children with a previous diagnosis of idiopathic
epilepsy or when the episode of convulsive status epilepticus
is the second unprovoked seizure that has led to a diagnosis
of idiopathic epilepsy.
6 Cryptogenic epilepsy related
Convulsive status epilepticus that is not symptomatic and
occurred in children with a previous diagnosis of cryptogenic
epilepsy or when the episode of convulsive status epilepticus
is the second unprovoked seizure that has led to a diagnosis
of cryptogenic epilepsy.
7 Unclassified
Convulsive status epilepticus that cannot be classified into
any other group.
Reproduced from reference 15 with permission from the BMJ Publishing Group.
epilepticus in that year to assess the degree of
ascertainment and to validate the database. Admissions
databases at six randomly selected north London hospitals
were searched for hospital International Classification of
Diseases (revision 10) codes for status epilepticus (G41)
and febrile seizures (R56·0) for the year, and the clinical
records of all identified patients were reviewed. All
referrals to Children’s Acute Transport Service for fits,
seizures, convulsions, or status epilepticus from the six
randomly selected hospitals for the year were also
223
 Articles
Telephone
9
12 21
13
11
6
1
Cards Reviews
Figure 1: Number of first ever episodes of convulsive status epilepticus
identified in six randomly selected hospitals according to source of
identification
Telephone=identified by telephone notification. Reviews=identified by yearly
reviews. Cards=identified by monthly surveillance card scheme.
reviewed. The labour-intensive nature of the review
required sampling of the hospitals in the research
network. For sample selection, all participating hospitals
were numbered and SPSS version 10.0 (Chicago, IL,
USA) was used by an independent observer to randomly
select six numbers. Patients identified by this method
who had not previously been identified were classified as
missed cases and their clinical data were not included in
the study. Our study was approved by the London
Multicentre Research Ethics Committee and local
research ethics committees of all health districts included
in the study.
Statistical analysis
We anonymised all data before analysis. Patients
identified from daily telephone calls to the Children’s
Acute Transport Service centre or the NLSTEPSS
research-coordinating centre were pooled for
ascertainment assessment. We investigated degree of
ascertainment via three-source capture-recapture log-
linear and sample-coverage modelling using CARE-1
software in S plus.16 Ascertainment of first ever episodes
of convulsive status epilepticus was initially estimated for
the six randomly selected hospitals included in the
database validation and the results were then extrapolated
to all hospitals in the network. We used the three-source
capture-recapture method to account for the likelihood
that being identified by one source was dependent on
being identified by others.17,18 We have presented
ascertainment estimates from the log-linear and sample-
coverage models with their associated CIs. We took the
224
lowest and highest limits obtained from these models’
CIs and used these to present the most conservative
95% CIs for the amount of ascertainment.
We estimated the incidence of convulsive status
epilepticus using north London population data from
UK Census 2001 as denominator and the average
number of first ever episodes per year as numerator.
We estimated occurrence rate using the exact
denominator as that used for incidence estimation, but
with average yearly occurrences of convulsive status
epilepticus as numerator. We calculated crude,
ascertainment-adjusted, age-adjusted, and age-specific
incidence rates by sex, seizure type, duration, and
cause. Ascertainment-adjusted rates were estimated by
dividing the crude rate by the lower and upper limits of
ascertainment obtained via the capture-recapture
modelling. We determined age-adjusted estimates by
directly adjusting crude estimates to the UK Census
2001 England and Wales paediatric population.19 We
used StatXact version 4.0.1 to calculate exact 95% CI
and p values for differences between incidence and
proportions within subgroups.
We used Kaplan-Meier survival analysis to investigate
the interval between first ever convulsive status epilepticus
and recurrence.
Role of the funding source
The funding source had no role in study design, data
collection, data analysis, data interpretation, or writing of
the report. The corresponding author had full access to
all the data in the study and had final responsibility for
the decision to submit for publication.
Results
226 children were enrolled into NLSTEPSS, of whom
176 had first ever episodes of convulsive status epilepticus
and 50 had had at least one previous episode. There was no
difference in ascertainment between years nor was there
any seasonal variation in incidence. However, in the first
year, 99 (69%) of children enrolled into NLSTEPSS had
first ever convulsive status epilepticus but in the second
year 77 (93%) of those enrolled had first ever convulsive
status epilepticus. Of the children with a first ever episode,
136 (77%) had convulsive status epilepticus that started out
of hospital (113 [83%] were taken to hospital by ambulance)
and 40 (23%) had an episode that started in hospital. A
total of 304 occurrences of convulsive status epilepticus
were reported in the study cohort.
There were 73 first ever episodes of convulsive status
epilepticus in the sample of six randomly selected hospitals
identified, of which 55 were identified by telephone
notifications, 32 by the surveillance-card system, and
46 by yearly reviews (figure 1). Case ascertainment by
capture-recapture log-linear modelling was 81% (95% CI
73–84) and by capture-recapture sample coverage
modelling was 74% (62–81). Thus a conservative estimate
of the 95% CI of case ascertainment is 62–84.
www.thelancet.com Vol 368 July 15, 2006
 Articles

The crude incidence of convulsive status epilepticus was
14·5 per 100 000 per year (95% CI 11·5–17·6). After
adjustment for ascertainment, the incidence was 17–23 per
100 000 per year with 30–41 occurrences per 100 000 per
year. Although the paediatric population of north London
might not be representative of that of England and Wales,
extrapolation of these data could provide an approximation
of the national frequency of convulsive status epilepticus.
If these data are extrapolated to the UK Census 2001
paediatric population of England and Wales, about
1650–2240 children will have first ever episodes of
convulsive status epilepticus and about 3200–4300 episodes
of convulsive status epilepticus will occur each year.
The incidence of convulsive status epilepticus was
highest in children younger than 1 year (51 per 100 000 per
year, 95% CI 35–66) compared with those aged 1–4 years
(29 per 100 000 per year, 19–35), 5–9 years (9 per
100 000 per year, 5–11), and 10–15 years (2 per 100 000 per
year, 0·75–3·5), overall p=0·005). The high incidence of
convulsive status epilepticus in children younger than
1 year was particularly notable in the acute symptomatic
group. Within each age group, the incidence was similar
for boys and girls (table 1).
The age-adjusted incidence of convulsive status
epilepticus was also similar in boys and girls. Although
by definition, prolonged febrile seizure is restricted to
children younger than 6 years, the age-adjusted incidence
was still greater than that for each of the other causes of
convulsive status epilepticus across the whole of
childhood (table 1).

Age (years) Age-adjusted incidence*

(95% CI)
<1 1–4 5–15 0–15
number (incidence)* number (incidence)* number (incidence)* number (incidence, 95% CI)*
Cause
PFS 15 (18·1)· 41 (12·7) 0 (0·0) 56 (4·6, 2·8 to 6·4) 4·1 (3·7 to 4·8)
Acute symptomatic 14 (16·9) 8 (2·5) 7 (0·9) 30 (2·5, 1·3 to 3·6) 2·2 (1·9 to 2·5)
Remote symptomatic 4 (4·8) 12 (3·7) 13 (1·6) 29 (2·4, 1·2 to 3·6) 2·3 (2·0 to 2·6)
Acute on remote 5 (6·0) 17 (5·3) 6 (0·7) 28 (2·3, 1·1 to 3·5) 2·1 (1·8 to 2·4)
Idiopathic 1 (1·2) 8 (2·5) 9 (1·1) 18 (1·5, 0·5 to 2·5) 1·4 (1·2 to 1·7)
Cryptogenic 2 (2·4) 0 (0·0) 1 (0·1) 3 (0·2, –0·1 to 0·6) 0·2 (0·1 to 0·3)
Unclassified 1 (1·2) 8 (2·5) 4 (0·5) 12 (1·1, 0·2 to 1·9) 1·0 (0·8 to 1·2)
Epilepsy
No epilepsy 39 (47·1) 86 (26·7) 30 (3·7) 155 (12·8, 10·0 to 15·7) 11·7 (11·0 to 12·3)
Before CSE 3 (3·6) 2 (0·6) 8 (1·0) 13 (1·1, 0·2 to 1·9) 1·1 (0·9 to 1·2)
At CSE 0 (0·0) 6 (1·9) 0 (0·0) 6 (0·5, –0·1 to 1·1) 0·4 (0·3 to 0·6)
After CSE 0 (0·0) 0 (0·0) 2 (0·2) 2 (0·2, –0·2 to 0·5) 0·2 (0·1 to 0·3)
Sex
Male 24 (56·6) 51 (31·2) 20 (4·9) 95 (15·4, 11·0 to 19·8) 14·0 (13·0 to 15·0)
Female 18 (44·5) 43 (27·1) 20 (5·1) 81 (13·6, 9·4 to 17·8) 12·5 (11·5 to 13·4)
Character
Continuous 19 (22·9) 47 (14·6) 18 (2·2) 84 (6·9, 4·8 to 9·0) 6·3 (5·8 to 6·8)
Intermittent 23 (27·8) 47 (14·6) 22 (2·7) 92 ( 7·6, 5·4 to 9·8) 6·9 (6·4 to 7·4)
Final seizure type
Focal 2 (2·4) 5 (1·6) 2 (0·2) 9 (0·7, 0·1 to 1·4) 0·7 (0·5 to 0·8)
Secondary generalised 14 (16·9) 26 (8·1) 12 (1·5) 52 (4·3, 2·6 to 5·9) 3·9 (3·6 to 4·3)
Primary generalised 26 (31·4) 63 (19·6) 26 (3·2) 115 (9·5, 7·0 to 12·0) 8·7 (8·2 to 9·3)
Motor type
Tonic 8 (9·7) 10 (3·1) 5 (0·6) 23 (2·0, 0·8 to3·2) 1·6 (1·2 to 1·8)
Clonic 0 (0·0) 2 (0·6) 0 (0·0) 2 (0·2, –0·2 to 0·5) 0·1 (0·1 to 0·2)
Tonic-clonic 34 (41·1) 82 (25·5) 35 (4·3) 151 (12·5, 9·7 to 15·3) 11·5 (10·8 to 12·1)
Duration (min)
30–60 17 (20·5) 36 (11·2) 17 (2·2) 71 (5·9, 3·9 to 7·8) 5·4 (4·9 to 5·8)
>60 25 (30·2) 58 (18·0) 22 (2·7) 105 (8·7, 6·3 to11·0) 7·3 (6·9 to 7·7)
All 42 (50·7) 94 (29·2) 31 (5·0) 176 (14·5, 11·5 to 17·6) 13·3 (12·6 to 14·0)

PFS=prolonged febrile seizure. Acute on remote=acute on remote symptomatic. Idiopathic=idiopathic epilepsy related. Cryptogenic=cryptogenic epilepsy related.
CSE=convulsive status epilepticus. *Per 100 000 children per year. Age-adjusted to the 2001 England and Wales childhood population.

Table 1: Number, age-specific and age-adjusted incidence per 100 000 children per year according to features of first ever episodes of convulsive status
epilepticus in north London

www.thelancet.com Vol 368 July 15, 2006 225

 Articles

convulsive status epilepticus each accounted for 16%

35
(29 and 28 respectively, 12–23%) of episodes, and 21 (12%,
30 8–18) children had idiopathic or cryptogenic epilepsy.
Proportion of first-ever eposodes of
convulsive status epilepticus

25 When convulsive status epilepticus was associated with

idiopathic or cryptogenic epilepsy, 13 (62%, 41–79) had a
20
history of epilepsy, 6 (29%, 14–50) had had one previous
15 seizure, and 2 (10%, 3–29) had their first seizure leading
to the diagnosis of epilepsy. Convulsive status epilepticus
10 Subcategories of acute symptomatic
convulsive status epilepticus was attributable to low antiepileptic drug concentrations
5 in only one child. In 12 (7%, 4–12) of children, the cause of
0 convulsive status epilepticus could not be identified.
In a subgroup analysis of 95 children with first ever
ia
S

NS

ic

ic

ed
ni
ur
ut

ot

ot
PF

te

CV
ox
ol

th
AB

ifi
episodes of convulsive status epilepticus associated with
lC

ge
j
Ac

ela

em
ab

pa
in

an

ss
ra

to
Re
et

gr

io
ad

cla
/

nr
Vi

xia

yp
em

Id
He
u

fever (temperature >38°C at presentation), 11 children

Un
eo

Cr
Dr

po
ut

ut
Hy
Ac

Ac

(12%, 95% CI 6–18) had acute bacterial meningitis and

Figure 2: Causes of first ever episodes of convulsive status epilepticus
PFS=prolonged febrile seizure. Acute=acute symptomatic. ABM=acute bacterial meningitis. Viral CNS=acute viral
CNS infection. Acute metabolic=acute metabolic disturbance. CVA=cerebrovascular accident. Remote=remote
symptomatic. Acute on remote=acute on remote symptomatic. Idiopathic=ideopathic epilepsy related.
Cryptogenic=cryptogenic epilepsy related.
Causes of first ever episodes of convulsive status
epilepticus are shown in figure 2. 56% (98, 95% CI 48–63)
of children with first-ever convulsive status epilepticus
were previously neurologically healthy (healthy
neurodevelopment, no history of epilepsy, and no
neurological deficits). A third (56, 25–39) of episodes were
prolonged febrile seizures, and acute symptomatic
convulsive status epilepticus occurred in 17% (30, 12–23).
Most children with acute symptomatic convulsive status
epilepticus had either acute metabolic derangement
(electrolyte imbalance, hypoglycaemia, hypocalcaemia, or
hypomagnesaemia) or an acute CNS infection (figure 2).
Remote symptomatic and acute on remote symptomatic
seven (8%, 2–13) had a viral CNS infection. The remaining
children had prolonged febrile seizure (n=56 [59%]) or
had a previous neurological abnormality with a febrile
intercurrent illness (21 [22%]).
Tables 1 and 2 show features of first ever episodes of
convulsive status epilepticus. Although a third had a
focal onset only, 5% (95% CI 3–9) remained focal. The
incidence of primary generalised convulsive status
epilepticus was twice that of any other seizure type and
the incidence of intermittent and continuous convulsive
status epilepticus was similar across all age-groups.
Most seizures were tonic-clonic in nature (86%,
80–90%) and the incidence of tonic convulsive status
epilepticus was greatest in children younger than 1 year
(10 per 100 000 per year). All children with tonic attacks
had intermittent convulsive status epilepticus. In 60%
(52–67) convulsive status epilepticus lasted longer than
1 h.

Prolonged Acute Remote Acute on Idiopathic Cryptogenic Unclassified All

febrile seizures symptomatic symptomatic remote
Sex
Male 34 15 13 17 10 1 5 95
Female 22 15 16 11 8 2 7 81
Duration (min)
30–60 27 11 12 8 8 0 5 71
>60 29 19 17 20 10 3 12 105
Character
Continuous 32 13 14 12 8 0 5 84
Intermittent 24 17 15 16 10 3 7 92
Final seizure type
Focal 1 3 4 1 0 0 0 9
Second generalised 11 11 5 14 4 1 6 52
Primary generalised 44 16 20 13 14 2 6 115
Motor type
Tonic 3 7 2 5 1 1 4 23
Clonic 1 1 0 0 0 0 0 2
Tonic-clonic 52 22 27 23 17 2 8 151
All 56 30 29 28 18 3 12 176

Table 2: Numbers of children with first ever episodes of convulsive status epilepticus, according to cause

226 www.thelancet.com Vol 368 July 15, 2006


23 children with first ever convulsive status epilepticus
(13%, 95% CI 9–19) had at least one further episode
during the follow-up. Median interval between first ever
convulsive status epilepticus and first recurrence was
25 days (95% CI 0–78, range 0–463). There was no
difference in the mean interval from first ever convulsive
status epilepticus to first recurrence between children
with previous neurological abnormality (n=9, 63 days,
95% CI 17–109, range 1–174) and previously healthy
children (n=14, 88 days, 22–153, 0–463 days) (figure 3).
18 children who had first ever convulsive status
epilepticus had a recurrence within a year (1-year
recurrence=16%, 10–24). Eight children had one further
episode, seven had two further episodes, two had four
further episodes and one child had nine further
episodes within 1 year after their first ever episode of
convulsive status epilepticus. The child with nine
further episodes had a focal area of cortical dysplasia.
Children with a pre-existing neurological abnormality
were 2·9 times (95% CI 1·01–8·45, p=0·048) more
likely to have a recurrence within a year of first ever
convulsive status epilepticus (12 of 51) than previously
neurologically healthy children (six of 63). In 30 children
with first ever episodes of prolonged febrile seizure,
five (17%, 95% CI 7–34) had a recurrence within a year
Seven children died during their stay in hospital. Six
had first ever convulsive status epilepticus (three boys;
median age 1·1 years, range 0·18–5·15 years) and the
other, a boy aged 13·4 years, had had an episode of
convulsive status epilepticus before enrolment into
NLSTEPSS. Mortality rates were higher for children with
intermittent convulsive status epilepticus (six of
148 children) than for those with continuous convulsive
status epilepticus (one of 156 children), but this difference
was not significant (difference in proportion=3·4%,
95% CI –0·2 to 8%, p=0·06). Case fatality for first ever
episodes of convulsive status epilepticus was 3% (2–7%).
Three children had acute bacterial meningitis, one had
glutaric aciduria type I, and three (including the child
who had had a previous episode of convulsive status
epilepticus) had progressive neurodegenerative disorders
that, despite investigations, did not have syndromic
diagnoses.
Discussion
Our results show that the frequency, range of causes, and
mortality of convulsive status epilepticus in childhood
differ from published population-based data for this
disorder in adults. In view of the large size of the study
cohort and the high ascertainment (74–81%), our study
is likely to provide a comprehensive and reliable charac-
terisation of the epidemiology of childhood convulsive
status epilepticus.
The incidence of convulsive status epilepticus in
childhood in north London is 18–20 per 100 000 per year
(95% CI 17–23), which is higher than the 4–6 per
100 000 per year which can be estimated from published
www.thelancet.com Vol 368 July 15, 2006
Articles
1·0
Previous neurological abnormality
Previously healthy
0·8
Cumulative probability of recurrence
0·6
0·4
0·2
0
0 50 100 150 200
Days from initial case to recurrence
Numbers at risk
Previous neurological 78 33 28 22 6
abnormality
Previously healthy 98 44 33 22 0
Figure 3: Probability of recurrence according to neurological state before first ever episode of convulsive
status epilepticus
epidemiological studies of convulsive status epilepticus
in adult populations.7–9 Previous population-based
studies have reported an incidence of status epilepticus
in childhood of 21–38 per 100 000 per year but those
estimates included episodes of non-convulsive status
epilepticus.7,8,20 The diagnostic criteria for non-convulsive
status epilepticus remain controversial21 but complex
partial, partial, absence, and subtle status epilepticus
have all been used synonymously with non-convulsive
status epilepticus.22,23 When these are excluded, the
incidence of convulsive status epilepticus in childhood
reported in other population-based studies ranges from
10 to 27 per 100 000 children per year (95% CI 8–43)7,8,20
and is similar to the incidence estimated in this study.
Our study was restricted to convulsive status epilepticus
because immediate electroencephalogram investigation
was not universally available in north London.
In the first year of NLSTEPSS, 69% of participants
had first ever convulsive status epilepticus. This figure
was 93% in the second year. These results might indicate
that, in the second year, children with first-ever
convulsive status epilepticus were more likely to be
notified to NLSTEPSS than children who had had
previous episodes of convulsive status epilepticus.
Consequently, although we are confident of our
incidence estimates, our reported occurrence and
recurrence rates of convulsive status epilepticus might
be minimum estimates.
227
 Articles
The most common cause of convulsive status epilepticus
in our study cohort was prolonged febrile seizure, which is
probably associated with low morbidity and mortality.7,24
Suggestions have been made that prolonged febrile seizure
should not be regarded as a separate category to acute
symptomatic convulsive status epilepticus.14 However,
results of studies that use this single classification might
erroneously amplify the severity of outcome of prolonged
febrile seizure and conversely, dilute the severity of acute
neurological insults. Thus in NLSTEPSS we used a
classification that included prolonged febrile seizure as a
distinct category. Longitudinal studies of cohorts with
prolonged febrile seizure will allow a more comprehensive
characterisation of outcomes.
There are several differences between the range of causes
of acute symptomatic convulsive status epilepticus in
children and adults. There is an increased likelihood of
acute bacterial meningitis in children with first ever
convulsive status epilepticus associated with fever, whereas
in adults, convulsive status epilepticus is rarely associated
with acute bacterial meningitis.2 Studies of convulsive
status epilepticus associated with fever might exclude
patients with acute CNS infections to comply with the
definition of febrile seizures, but such a retrospective
judgment is not appropriate in the emergency management
of these patients. In the first 6 months of NLSTEPSS, we
identified four children (three died, one had severe
neurological sequelae) with acute bacterial meningitis who
presented to their local accident and emergency department
with convulsive status epilepticus associated with fever,
and were initially managed as children with prolonged
febrile seizure. These data prompted us to report the case
series and a discussion of optimum management
strategies.25 The likelihood of acute bacterial meningitis
(12%, 95% CI 6–18) in children with first ever convulsive
status epilepticus associated with fever is much higher
than that reported for children with short febrile seizures
(1·2%).26 Thus there should be a high index of suspicion
for acute bacterial meningitis in a child with first ever
convulsive status epilepticus associated with fever. Other
notable differences are that low antiepileptic drug
concentrations and cerebrovascular disease are common
causes of acute symptomatic convulsive status epilepticus
in adults,2 whereas, only two children in our study had
acute symptomatic convulsive status epilepticus
attributable to these causes.
Our results also provide useful prognostic data. A
substantial proportion (16%) of children with first ever
convulsive status epilepticus will have a recurrence within
a year, and those with previous neurological abnormality
are more likely to have a recurrence (p=0·04). The 3%
mortality associated with convulsive status epilepticus
noted in this study is similar to that previously reported10
but less than the 13% reported in young adults and the
38% reported in elderly people.2 The 3% mortality is lower
than the 11% reported in the landmark study of Aicardi
and Chevrie,27 but remains clinically significant. For
228
comparison, mortality from bronchiolitis in children is
less than 1%28 and in childhood, mortality from diabetes
mellitus is 1–2%.29 Our results are consistent with those of
other studies that suggest that cause is a major factor
affecting short-term mortality.4,30 In this study, children
with acute or remote symptomatic convulsive status
epilepticus, in similar proportions, had the highest in-
hospital mortality. Our sample size, however, might be too
small to accurately establish the risk factors for mortality.
Our study did not include short-term morbidity associated
with convulsive status epilepticus.
Our findings show that convulsive status epilepticus in
childhood is common and differs from convulsive status
epilepticus in adulthood. Prolonged febrile seizure is the
single most common cause, there should be a high index
of suspicion for acute bacterial meningitis in children with
first ever episodes of convulsive status epilepticus
associated with fever, one in six children will have a
recurrence within a year of a first ever convulsive status
epilepticus, and 3% of children with convulsive status
epilepticus will die during their stay in hospital.
Contributors
Richard Chin, Rodney Scott, Brian Neville, Helen Bedford,
Catherine Peckham, and Angie Wade participated in the design and
conduct of the study and the writing of the report. Richard Chin and
Angela Wade did the statistical analysis.
Members of the NLSTEPSS Collaborative Group
Richard Chin, Rodney Scott, Brian Neville, Helen Bedford,
Catherine Peckham, Angela Wade (Institute of Child Health), Simon Roth
(Barnet General Hospital), Ruby Schwartz (Central Middlesex Hospital),
Jacqueline Taylor (Chase Farm Hospital), Edwin Abrahamson (Chelsea
and Westminster Hospital), Mark Kenny, Mark Peters (Great Ormond
Street Hospital for Children and Children’s Acute Transport Service),
Shane Tibby (Guy’s Hospital), Adnan Manzur (Hammersmith Hospital),
Rajiv Sood (Homerton Hospital), Elaine Hughes (King’s College London),
MAS Ahmed (King George Hospital), Satheesh Mathew (Newham General
Hospital), Arvind Shah (North Middlesex Hospital), Warren Hyer
(Northwick Park Hospital), Michael Greenberg (Royal Free Hospital),
Adelaida Martinez (Royal London Hospital), Sophie Skellet (St George’s
Hospital), Simon Nadel, Ian Maconochie (St Mary’s Hospital),
Adrian Goudie, John Jackman (St Thomas’ Hospital), Mark Gardiner
(University College London), Anthony Cohn (Watford General Hospital),
Corina O’Neill (Whipps Cross Hospital), Andrew Robins (Whittington
Hospital); London, UK.
Conflict of interest statement
We declare that we have no conflict of interest.
Acknowledgments
We thank all medical, nursing, and administrative staff who reported to
NLSTEPSS, and Richard Lynn from the British Paediatric Surveillance
Unit whose help was invaluable in the development and implementation
of the active regional surveillance card scheme. The Wellcome Trust
supported Rod Scott. NLSTEPSS was funded by an anonymous donor to
the Institute of Child Health. Research at the Institute of Child Health
and Great Ormond Street Hospital for Children NHS Trust benefits
from research and development funding from the NHS Executive.
References
1 Leppik IE. Status epilepticus. Clin Ther 1985; 7: 272–78.
2 DeLorenzo RJ, Pellock JM, Towne AR, Boggs JG. Epidemiology of
status epilepticus. J Clin Neurophysiol 1995; 12: 316–25.
3 Yager JY, Cheang M, Seshia SS. Status epilepticus in children.
Can J Neurol Sci 1988; 15: 402–05.
4 Logroscino G, Hesdorffer DC, Cascino G, Annegers JF,
Hauser WA. Short-term mortality after a first episode of status
epilepticus, Epilepsia 1997; 38: 1344–49.
www.thelancet.com Vol 368 July 15, 2006

5 Cascino GD, Hesdorffer D, Logroscino G, Hauser WA. Morbidity of
nonfebrile status epilepticus in Rochester, Minnesota, 1965–1984.
Epilepsia 1998; 39: 829–32.
6 Logroscino G, Hesdorffer D, Cascino G, Annegers J F, Bagiella E,
Hauser AW. Long-term mortality after a first episode of status
epilepticus. Neurology 2002; 58: 537–41.
7 Hesdorffer DC, Logroscino G, Cascino G, Annegers JF, Hauser WA.
Incidence of status epilepticus in Rochester, Minnesota, 1965–1984.
Neurology 1998; 50: 735–41.
8 Coeytaux A, Jallon P, Galobardes B, Morabia A. Incidence of status
epilepticus in French-speaking Switzerland: (EPISTAR). Neurology
2000; 55: 693–97.
9 Knake S, Rosenow F, Vescovi M, et al. Incidence of status
epilepticus in adults in Germany: a prospective, population-based
study. Epilepsia 2001; 42: 714–18.
10 Wu YW, Shek DW, Garcia PA, Zhao S, Johnston SC. Incidence and
mortality of generalized convulsive status epilepticus in California.
Neurology 2002; 58: 1070–76.
11 Trevathan E, Fitzgerald R, Wang D. The impact of convulsive status
epilepticus on the risk of death varies by age. Epilepsia 2002; 43: 75.
12 Chin RF, Neville BG, Scott RC. A systematic review of the
epidemiology of status epilepticus. Eur J Neurol 2004; 11: 800–10.
13 Holmes GL. Epilepsy in the developing brain: lessons from the
laboratory and clinic. Epilepsia 1997; 38: 12–30.
14 International League Against Epilepsy, Commission on
Epidemiology and Prognosis. Guidelines for epidemiologic studies
on epilepsy. Epilepsia 1993; 34: 592–96.
15 Chin RFM, Verhulst L, Neville BGR, Peters MJ, Scott RC.
Inappropriate emergency management of status epilepticus in
children contributes to need for intensive care.
J Neurol Neurosurg Psychiatry 2004; 75: 1584–88.
16 Chao A, Tsay PK, Lin SH, Shau WY, Chao DY. The applications of
capture-recapture models to epidemiological data. Stat Med 2001;
20: 3123–57.
17 Hook EB, Regal RR. Capture-recapture methods in epidemiology:
methods and limitations. Epidemiol Rev 1995; 17: 243–64.
www.thelancet.com Vol 368 July 15, 2006
Articles
18 Simondon F, Khodja H. Capture-recapture method for estimating
misclassification errors: application to the measurement of vaccine
efficacy in randomized controlled trials. Int J Epidemiol 1999; 28:
113–16.
19 UK Office of National Statistics. Census 2001 Standard tables for
wards in England and Wales (DVD) (2004).
20 DeLorenzo RJ, Hauser WA, Towne AR, et al. A prospective,
population-based epidemiologic study of status epilepticus in
Richmond, Virginia. Neurology 1996; 46: 1029–35.
21 Treiman DM. Status epilepticus. Baillieres Clin Neurol 1996; 5:
821–39.
22 Drislane FW. Presentation, evaluation, and treatment of
nonconvulsive status epilepticus. Epilepsy Behav 2000; 1: 301–14.
23 Chang JW, Chang JH, Park SC, Kim TS, Park YG, Chung SS.
Radiologically confirmed de novo glioblastoma multiforme and
hippocampal sclerosis associated with the first onset of
nonconvulsive simple partial status epilepticus.
Acta Neurochir (Wien) 2001; 143: 297–300.
24 Nelson KB, Ellenberg JH. Prognosis in children with febrile
seizures. Pediatrics 1978; 61: 720–27.
25 Chin RFM, Neville BGR, Scott RC. Meningitis is a common cause
of convulsive status epilepticus with fever. Arch Dis Child 2005; 90:
66–69.
26 Offringa M, Moyer VA. Evidence based paediatrics: evidence based
management of seizures associated with fever. BMJ 2001; 323:
1111–14.
27 Aicardi J, Chevrie JJ. Convulsive status epilepticus in infants and
children: a study of 239 cases. Epilepsia 1970; 11: 187–97.
28 Leung AK, Kellner JD, Davies HD. Respiratory syncytial virus
bronchiolitis. J Natl Med Assoc 2005; 97: 1708–13.
29 Rosenbloom AL, Schatz DA. Diabetic ketoacidosis in childhood.
Pediatr Ann 1994; 23: 284–88.
30 Garzon E, Fernandes RM, Sakamoto AC. Analysis of clinical
characteristics and risk factors for mortality in human status
epilepticus. Seizure 2003; 12: 337–45.
229