Status epilepticus (SE) is a life-threatening neurological emergency, one of the most frequently found in pediatric settings, with an annual incidence of 10 to 73 per 100,000 children [1]. It is a major cause of admission to pediatric intensive care units (PICUs) globally [1].

Our understanding of the pathophysiology of SE has advanced significantly in recent years. However, this has not straightforwardly translated to clinical practice and guidelines [2]. In 2015, the International League Against Epilepsy (ILAE) reviewed some clinical aspects and proposed a new definition and classification for SE [3]. The definition states that SE is a condition “resulting either from the failure of the mechanisms responsible for seizure termination or from the initiation of mechanisms which lead to abnormally prolonged seizures (after time point 1)... that can have long term consequences (after time point 2)” [3]. For convulsive status epilepticus (CSE), it was suggested that time point 1 is 5 minutes, while time point 2 is 30 minutes [3]. This is extremely important because the timing of antiseizure medication (ASM) is a major determinant for preventing brain injury from excitotoxicity and cell death [4]. The new classification of SE has clinical and epidemiological meaning, and it involves four axes: semiology, etiology, correlates in electroencephalography (EEG), and age [3]. Globally, several research groups have applied these updates by ILAE in their observational studies of SE in children [5-7].

There is not a complete consensus regarding the treatment of SE, and it may vary according to local guidelines [8]. In our country, Portugal, a review of the approach to CSE in children and adolescents was published in 2020 by the Portuguese Society of Neuropediatrics [9]. To our knowledge, no observational studies from Portugal have previously applied the latest definition and classification by ILAE.

Therefore, the aim of our study was to characterize clinically and demographically all pediatric cases of SE admitted to our tertiary hospital in an 11-year period. We also aimed to describe the classification of all cases according to the 2015 ILAE report.

To our knowledge, the present study is the first observational study in Portugal that summarizes the epidemiological and clinical characteristics of SE in a pediatric cohort while applying the most recent ILAE guidelines. It represents the real-life decisions of general pediatricians and pediatric intensivists in the difficult management of pediatric SE during the past 11 years. Furthermore, our results are not limited to common practices in our hospital, since we also reviewed data from many patients in whom treatment was initiated before admission or who were transferred from other national health institutions. Most of our findings are in accord with the current literature.

Concerning demographic features, previous studies reported a higher prevalence of SE in boys and preschool children, with median ages ranging from 1.9 to 4.5 years, similar to our results (median age, 2.5 years; IQR, 1.3 to 5.0) [1,5,11-13].

In clinical aspects such as etiology and semiology, we also obtained some expected findings, which were more easily compared to the existing literature due to the standardization provided by the 2015 ILAE guidelines. According to several papers, acute symptomatic etiology is the most common cause of pediatric SE [1,11,14]. Unlike a study performed by Chiarello et al. [5] in a unit with primarily hematological-oncological disorders, our data set included only three patients with a personal history of brain tumor, even though our hospital is a regional referral center, and only one patient displayed an episode of SE on account of progressive etiology. This may be due to two factors: prophylactic ASM and early treatment were given to patients with brain tumors, which stopped seizures from evolving to SE, and some patients with brain tumors and SE could have lacked a discharge ICD-10 code for SE and therefore were missed in our research.

The prevalence of unknown (cryptogenic) etiology has mixed results in the literature, which may be due to the characteristics of the populations studied, namely the percentage of patients with a previous diagnosis of epilepsy. Interestingly, some researchers have noticed that unknown causes of SE were more prevalent in children older than 2 years, which we also verified [5,15]. The association between etiology and age was detailed by Shinnar et al. [15], who documented significant differences in the distribution of causes between the first and second year of life, with a rapid decrease in the frequency of acute symptomatic cases and a higher proportion of cryptogenic and remote symptomatic cases in the older children.

CSE, particularly the generalized subtype, was the most frequent semiological classification, which also agrees with many published works [1,5,12,16]. Nonetheless, recent studies have focused on NCSE, especially in the PICU. NCSE is the result of a prolonged electrographic seizure leading to non-convulsive clinical symptoms [5]. In the literature, it accounts for only 6% of the cases, but diagnosis requires a high index of suspicion, since cognitive or behavioral changes are often difficult to notice [1]. In the present study, we were able to confirm the diagnosis of NCSE in 7.8% of our patients, and it was associated with prolonged SE (>60 minutes). Focal motor SE was also associated with episodes >60 minutes. These findings reinforce the importance of semiology and its standardization in pediatric SE.

Rectal diazepam was the preferred first-line BZD, both before arrival at the hospital and at the hospital. Some studies performed in pediatric cohorts have not shown the superiority of any BZD [14]. In a Cochrane review, intravenous lorazepam and intravenous diazepam led to more rapid seizure cessation, but in the absence of intravenous access, buccal midazolam or rectal diazepam were considered acceptable first-line ASMs [17]. Although phenytoin was the most frequently selected second-line ASM, as in previously published studies [12], it was followed by propofol as the second most used. Although propofol is not usually selected as a second-line ASM because of its safety profile, it is very effective in seizure cessation, particularly in refractory cases, and it can be used in the emergency room, especially with support from the PICU team [8]. Midazolam in continuous infusion was the preferred third-line ASM and was most frequently responsible for the termination of SE, as has also been reported in the literature [4].

Despite the recognition of the need for urgent ASM administration, the timing of drug administration remains a significant barrier in pediatric SE. The 2015 ILAE guidelines indicate that the minimum time for starting emergency treatment of SE is 5 minutes for CSE [3]. In our study, the timing of the start of ASM was rarely listed in the clinical records. From the available data, we obtained a median interval of 17.5 minutes between the beginning of SE and ASM in CSE, which is longer than the recommended timing. The circumstances of the occurrence of SE, such as the place of the event (particularly in new-onset SE without at home rescue diazepam), may interfere with the urgent administration of an ASM, and studies on this topic should systematically review this issue.

The longer SE goes untreated, the more likely it is to become BZD-resistant and require anesthetic doses for cessation (refractory SE) [4]. In animal models, this seems to be associated with reuptake/decreased surface expression of the BZD-specific gamma-aminobutyric acid receptor subunit γ2 [4]. Refractory SE occurs in 29% to 43% of SE cases, according to the literature [6]. In our study, we found a very high percentage of refractory SE (79.4%). A more recent standardized protocol for the management of pediatric SE has been available in our hospital since 2019, and it suggests using midazolam or diazepam as first-line drugs (the administration routes differ between prehospital or hospital initiation of treatment); levetiracetam, valproate sodium, phenytoin, or phenobarbital as second-line drugs; and midazolam, thiopental, or propofol in continuous intravenous infusion as third-line drugs. It would be interesting to compare the therapeutic management of SE and outcomes in our hospital before and after the implementation of this protocol. In fact, we suspect that some changes in choices of ASM may have occurred after protocol implementation, such as levetiracetam replacing phenytoin as the preferred second-line drug. However, we should note that very prolonged SE (>60 minutes) was significantly associated with a diagnosis of epilepsy. It is known that patients with more severe epilepsy may not respond as well to the standard dosing/dose sequence of ASMs [18]. This raises a question regarding whether management is inappropriately homogeneous when protocols are applied, as has been discussed in previous studies [18].

The main limitation of this study is its retrospective and mainly descriptive nature. Patients were identified based on ICD-10 codes and confirmed by a clinical review; therefore, diagnoses could have been missed. ICD-10 codes are recorded by pediatricians at discharge, and sometimes the cause of SE is coded as the final diagnosis (e.g., infection) and the associated diagnoses are omitted. Intensive care pediatricians may be more likely to systematically codify SE compared to general pediatricians in the ED, which might explain why we obtained a higher proportion of cases admitted in the PICU than we excepted. For that reason, refractory cases may have been overestimated, but we also received many patients transferred from other hospitals, which are usually more severe cases. Temporal variables may lack accuracy due to post-management documentation in electronic records by medical doctors. Our results may have been influenced by treatment prior to admission, and we lacked information on the timing of ambulance arrival, duration of transport, and other prehospital data. Semiology was analyzed based on medical descriptions, which may have been influenced by experience. We did not obtain complete information about EEG in seizure monitoring, only for diagnostic purposes in NCSE. It would be interesting to evaluate EEG patterns in future studies to better characterize NCSE. Outcomes were not fully explored in this article. The observed in-hospital mortality rate was consistent with mortality rates reported in pediatric cohorts in developed countries [7]; however, neurodevelopmental outcomes were not reviewed, which could be particularly important given the high percentage of prolonged SE in our sample.

In summary, most of our findings are in accord with the current literature, including the most frequent demographic features (boys and preschool children), etiological and semiological classification (acute symptomatic etiology and CSE), and the selection of ASMs (the preferred ASMs were rectal diazepam as first BZD, phenytoin as second-line therapy, and midazolam with a continuous infusion as third-line treatment). The standardization provided by the 2015 ILAE guidelines allowed an easier comparison with other recent studies. Epilepsy, NCSE, and focal motor SE were associated with prolonged SE (>60 minutes), which reinforces the importance of the underlying neurological diseases and semiological standardization in pediatric SE.

This paper reports high percentages of patients with refractory SE. A standardized protocol for the management of pediatric SE has been available in our hospital since 2019, and hopefully, it will allow a more rapid recognition and initial treatment.