Phenomenology of Prolonged Febrile Seizures: Results of the FEBSTAT study
DETAILED METHODS(all references are to the references in the main manuscript)
As this is the first FEBSTAT manuscript, we will describe the cohort and the methods in some detail.
Population: Children, aged one month through 5 years, were included if they presented with an episode of FSE, defined as a seizure lasting a total of 30 minutes or more without fully regaining consciousness [14,15] that also met the definition of a febrile seizure [14,16]. A febrile seizure was defined as a provoked seizure where the sole acute provocation was fever (temperature greater than 38.4oC, 101.0 F) without prior history of afebrile seizures and with no evidence of an acute CNS infection or insult [14,16]. As the study was designed to address the relationship between FSE and subsequent MTS and TLE in otherwise normal children, those with severe neurological disability were excluded. However, once a child was entered, subsequent information about developmental status did not exclude them from the study. The 5 recruiting sites were Montefiore Medical center and JacobiHospital in the Bronx, Children’s MemorialHospital in Chicago, Duke University Medical Center, Virginia Commonwealth University Hospital, and EasternVirginiaMedicalSchool. In addition, the International Epilepsy Consortium at VirginiaCommonwealthUniversitywas the DataCoordinatingCenter and the Epidemiology/Biostatistics Core was based at ColumbiaUniversity.
Recruitment: Emergency department and hospitalization records were reviewed on a daily basis to identify potential subjects. To maximize the chance of ascertaining all cases of FSE, the trigger to screen was any child evaluated for a febrile seizure reportedly lasting 15 minutes or more. The pediatric staff and the pediatric neurology residents and attendings at all the recruiting sites were alerted to help identify potential subjects. Once a potential subject was identified, the emergency department record, admission notes and ambulance call sheet, if available, were reviewed by the study team. If a subject was deemed potentially appropriate for the study, the family was approached for willingness to participate. Among families indicating an interest, additional screening questions were asked, regarding seizure duration and presence of prior afebrile seizures to clarify eligibility. If eligible, the full study procedures, risks and benefits, voluntariness of participation, alternatives and other elements of the informed consent were reviewed and written informed consent was obtained. The procedures were approved by the Institutional Review Boards for the Protection of Human Subjects at all the participating institutions.
Procedures: All subjects were recruited within 72 hours of the episode of FSE. As part of the diagnostic evaluation during the initial hospitalization the following procedures were performed:
1. Detailed history of the circumstances of the seizure with a focus on duration and focality;
2. A structured interview which covered the reported age of developmental milestones, history of illness (including seizures) and disability, prenatal and perinatal history, ethnicity, indicators of SES, family history of febrile seizure and epilepsy in first degree relatives and characteristics of the febrile seizure;
3. Complete neurological and physical examination;
4. Laboratory data summarized including CSF results if lumbar puncture was done. More than 20WBCs in the CSF was considered an exclusionary criterion. However, lumbar punctures were being done for clinical indication and the absence of a lumbar puncture was not an exclusionary criterion;
5. Magnetic resonance imaging (MRI) examination - MRI examinations were done within 72 hours of the FSE whenever possible or very shortly thereafter. The MRIs included sequences that focused on the hippocampus and were designed to allow both visual and quantitative measurements of volume and of T2 signal as well as ADC. Exams were standardized through the use of phantoms;
6. Electroencephalograms (EEG) – a routine EEG without sedation was performed when feasible in all subjects within 72 hours;
7. Specimens for acute human herpes virus (HHV)-6 and HHV-7 assays. – Serum specimens at baseline and at one month were sent to a laboratory at Children’s Memorial Hospital (LE) for analysis of HHV-6a, HHV-6B and HHV-7 status using quantitative polymerase chain reaction (PCR) for viral specific DNA and RNA and using indirect immunofluorescence to determine viral antibody titers.
At one month, the children received a follow-up neurological examination, age-appropriate developmental testing including measures of memory in children over age 2 ½, years, quality of life and behavioral measures, and a convalescent specimen for HHV-6 and HHV-7. The entire evaluation including MRI, EEG, neurological examination, developmental and quality of life battery, but not the viral studies, was repeated again at one year. The intent is to try and repeat these examinations again at 5 years. Additionally, if the child had another episode of FSE or developed epilepsy then the baseline, one-month and one-year evaluations were repeated.
Families are approached for genetic studies during the study, typically at the one month visit. Participation in this part is optional and a separate consent is obtained. Serum samples from probands and both parents are obtained and sent to the NINDS repository at Coriell Institute for Medical Research [17] for future analysis once sufficient number of specimens has been collected.
Consensus Process for seizure semiology and duration: In addition to local coding features of FSE, FSE was classified by 3 independent raters (SS, DN, JMP). Materials available for the review included the medical record, including the emergency department records and the ambulance call sheets, if available, the structured interview and the written comments from the local study nurse. Seizure classification was based upon phenomenology and not upon EEG and MRI results, even for determination of focality. Each rater coded independently. Disagreements across classifications were discussed in a consensus meeting where materials were reviewed to determine if discrepancies could be resolved; consensus was reached in all cases after review. The consensus classifications are the final semiology determinations for analysis, thus avoiding site to site variability in the interpretation of data.
Consensus classifications address aspects of the phenomenology of the FSE and circumstances surrounding the seizure. First, each rater determined whether the seizure qualified as a febrile seizure and then if the seizure met criteria for FSE. Episodes of FSE were further classified as continuous versus behaviorally intermittent. Total duration of FSE was classified based on ambulance call sheets, emergency department notes and the structured interview. Because the vast majority of these children were still having persistent seizure activity on arrival to the emergency department, there was objective data on seizure start and stop times. For intermittent seizures, in addition to total duration, we also coded the duration of the longest convulsion and total convulsive time.
Focality was rated on a 4 point scale as definitely focal, probably focal, possibly focal or definitely generalized. Definite focality was coded when there was focal convulsive activity, clearly lateralized head turning at onset, a Todd’s paresis or a clear behavior arrest or staring with or without automatisms. Probable focality was coded when there was clear lateralized eye movement at onset, a stare or other feature suggestive of a complex partial phase prior to the generalized convulsion. Possibly focal was when there was a vague description of features such as eye movements and stare but they were not clear cut or lateralized. For purposes of analysis those considered definitely or probably focal were classified as focal and those possibly focal and definitely generalized were considered as generalized. Seizures classified as definitely focal, were rated for evidence of definite cerebral lateralization. Definite cerebral lateralization was based on focal convulsive activity or a Todd’s paresis postictally. Clearly lateralized eye movements or head turning at onset were not considered definitive evidence of cerebral lateralization. Seizures were then classified in accordance with the International League Against Epilepsy (ILAE) seizure classification as generalized tonic clonic, simple partial, complex partial or partial with secondary generalization [18]. Seizure classification was based exclusively on clinical features of the seizures and did not rely on imaging or subsequent EEG abnormalities.
Finally, the raters classified whether the treating clinicians in the emergency department recognized that the child had SE as definitely, probably, possibly and definitely not. Definite recognition of SE was based upon emergency department physician notes and diagnosis, indicating either status epilepticus or seizure 30 minutes. Probably recognized SE was based upon recognition that the seizure lasted 30minutes in the flow sheet in the absence of similar information in the physician notes. Possible recognition of SE was based upon recognition of a prolonged seizure without notation in the flow sheets or the physician notes that the seizure was 30 minutes or longer. SE was considered unrecognized when it was clear from both the flow sheets and the clinician notes that there was no recognition of the fact that this was a prolonged (>20 minutes) seizure. For children transferred from a local emergency department to one of the study hospitals, rating of SE recognition was based on the first emergency department as this is where initial treatment would occur. Recognized episodes of SE included those categorized as definitely and probably; unrecognized episodes included those categorized as possibly and definitely not.
Statistics: Descriptive analysis was conducted through the determination of a percentage, mean, or median. Standard deviations are presented for means and the interquartile range (IQR) for the median. For the comparison of intermittent versus continuous seizures, comparison of means was performed using the t-test, comparison of medians using the Wilcoxon Rank Sum test, and comparison of frequencies using the Chi-square test [19]. For the comparison of each rater to the consensus determination, inter-rater reliability was calculated using the kappa statistic for categorical variables and a 95% confidence interval was constructed [20]. A kappa of 0.75 or more was considered to be excellent agreement beyond chance, a kappa of 0.41 to 0.75 was considered to be good agreement, and values of 0.40 or lower were considered to be poor agreement beyond chance. For continuous variables, inter-rater reliability was determined using the intraclass correlation coefficient comparing each of the three raters.
Seizure duration was modeled as a Weibull distribution with two parameters, α or the scale parameter, and β or the shape parameter [21]. A best fit was obtained for the function
S(t) = e−(t/α)βFit of the Weibull model was assessed with the Log Likelihood statistic [21]. The predicted survival curve based upon the computed shape and scale parameters was compared to the Kaplan Meier survival curve [21].