Introduction

COVID-19 is the infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first reported in China in December 2019 that later led to the COVID-19 pandemic. Since then, the pediatric population has not been spared with millions affected worldwide. Coronaviruses, including SARS-CoV-2, typically cause respiratory infections with fever and cough being the most common symptoms in the adult and pediatric populations. Other symptoms include rhinorrhea, sore throat, headache, fatigue, myalgia, diarrhea, and vomiting. Neurological manifestations including meningitis, encephalitis, seizures (febrile or afebrile), and stroke, although less common, have also been reported [1, 2].

Febrile seizures (FS) are common and account for the majority of seizure disorders in children with an estimated rate of between 2 and 5% in Western countries [3,4,5,6]. However, estimates vary and higher rates have been reported in Asian countries, ranging from 1 to 14% [7].

The International League Against Epilepsy (ILAE) defines FS as seizures occurring in childhood after the age of 1 month, associated with a febrile illness not caused by an infection of the central nervous system (CNS), without previous neonatal seizures or unprovoked seizures and not meeting the criteria for acute symptomatic seizures [7]. The infections that trigger FS are usually viral infections. In the EFES (viral etiological causes of febrile seizures for respiratory pathogens) study, the most frequently detected pathogens were adenovirus, influenza A and influenza B in children younger than 12 months, and coronavirus OC43 in older children [8]. FS can also occur outside of the typical age range (≥ 6 months to < 6 years), and this has been reported for COVID-19 patients. One study reported that a third of patients with COVID-19 due to the Omicron variant were older than 5 years [9].

To date, data on FS among children affected by COVID-19 remains scarce. There are few studies on the true prevalence, frequency, and characteristics of COVID-19-related FS [9,10,11, 21]. FS were also rarely reported during the earlier part of the pandemic and only became more common with the emergence of the Omicron variant. Hence, knowledge gaps remain. This study aims to assess the prevalence as well as characteristics and outcomes of FS in pediatric (< 12 years) patients with COVID-19 admitted to the National Isolation Centre (NIC) in Brunei Darussalam during the first three waves.

Materials and methods

Study design

This was an observational study looking at the prevalence of FS among paediatric patients with COVID-19 admitted to the NIC over the first three waves that affected Brunei Darussalam. The study periods included the first wave (due to the original variant; 9 March 2020 to 6 August 2021), second wave (due to the Delta variant B.1.617.2; 7 August 2021 to 31 January 2022), and third wave (due to the Omicron variant; 1 February to 9 March 2022; end of study period).

Setting

In Brunei Darussalam, the NIC was the only designated center for the management of COVID-19 during the first and earlier part of the second wave (7 August to 18 August 2021). During this period, all patients (adults and pediatric) with positive SARS-CoV-2 Reverse Transcription-Polymerase Chain Reaction (RT-PCR) tests were admitted to the NIC for isolation and treatment. Community isolation centers were opened on 18 August 2021 to cater for the increasing number of asymptomatic and mildly symptomatic patients (categories 1 and 2a: refer to “COVID-19 disease categorizations” section) who did not require any medical treatment.

Towards the latter part of the second wave, home isolation was introduced. These arrangements continued into the third wave. Throughout the three waves, symptomatic pediatric patients (category 2a needing treatment, 2b and 3) continued to be admitted to the NIC for isolation and management. Severe (category 4) to critically ill (category 5) pediatric patients were admitted and managed in the national tertiary center (RIPAS Hospital), as facilities for caring for critically ill pediatric patients were not available in the NIC.

COVID-19 disease categorizations

Disease severity has been categorized into five categories: asymptomatic (category 1), mildly symptomatic (category 2a: cough, rhinorrhea, sore throat, nausea, vomiting, loose stool, dizziness, or lethargy; category 2b: fever, shortness of breath, chest pain, loose stool or reduced appetite for 2 days or more), radiological changes of pneumonia (category 3), requirement of oxygen (category 4), and critical illness needing mechanical ventilation with or without other end organ injuries (category 5) [12].

Diagnosis of FS

We have used the ILAE definition of FS, where the seizures are associated with a febrile illness not caused by an infection of the CNS, without previous neonatal seizures or unprovoked seizures and not meeting the criteria for acute symptomatic seizures [7].Simple FS are defined as generalized seizures lasting < 15 min without recurrence during a 24-h period, in children aged ≥ 6 months to < 6 years and in the absence of any neurological deficit [13,14,15]. Complex FS are defined as focal seizures, seizures lasting longer than 15 min, which have a neurological deficit and which occur multiple times in 24 h [16]. Complex FS can also be defined as prolonged seizures that have been stopped with anticonvulsant therapy before the 15th minute [13, 17, 18]. Febrile status epilepticus is defined as a febrile seizure lasting 30 min or more or a series of FS without full return to consciousness during that period [18, 19].

In this study, FS have been attributed to COVID-19 based on the temporal link between COVID-19 and the development of FS, and in the absence of other causes (significant electrolyte abnormalities and possible CNS infections).

Management

Basic laboratory investigations such as complete blood count and serum electrolyte tests were carried out. Blood culture, urinalysis, and urine culture were carried out especially if there was concern for other infections. However, in some cases, laboratory investigations were not done when patients were clinically well with simple FS and where there was difficulty obtaining specimens. Chest radiographs were not part of the routine investigation protocol and were only performed if there were concerns, especially in patients with any respiratory distress. Patients were assessed on admission and on a daily basis to monitor progress, and their disease category was recorded. The highest category achieved was taken as overall disease severity.

Patients were managed according to their symptoms; if necessary, acetaminophen, fluid replacement, and antimicrobials were prescribed according to their signs and symptoms. Patients were only discharged after they had recovered or improved (i.e., when they had been without fever and seizures over the previous 24 h), and well enough to continue home isolation. All patients with symptomatic COVID-19 were advised to seek medical attention if any issue arose after discharge. Patients with complex FS who were 6 years and older were followed up in the pediatric clinic.

Inclusion and exclusion criteria

All patients aged < 12 years admitted to the NIC with COVID-19 in the first wave, second wave and third wave from 9 March 2020 to 9 March 2022 were included in the study. The age cutoff for pediatric patients in Brunei Darussalam is under 12 years, and hence, the age of 12 years has been used in this study. Children with COVID-19 that were not admitted to the NIC for management of COVID-19 were excluded from the study.

Data collections

Demographic and clinical data was extracted from a prospectively maintained database used for COVID-19 management that had been established in the first wave. The database captured clinical demographic, comorbidities, previous medical history including any history of epilepsy, neonatal seizures or unprovoked seizures, family history of FS, symptom manifestations at presentation and during hospitalizations, COVID-19 progress (daily disease categorizations), dates of admission, and discharge or transfers. Laboratory data not included in the above database was retrieved from the Ministry of Health electronic health record system (Brunei Darussalam Healthcare Information and Management Systems, Bru-HIMS). This national health record links all the government health institutions in the country. For follow up, a review of patients’ Bru-HIMS data was conducted on 5 May 2022 to check for recurrence of FS. Extracted data was anonymized prior to statistical analyses. Data will be made available upon request.

Statistical analysis

For descriptive statistics, data was presented as frequency and percentage for categorical variable, mean, and standard deviations for continuous variable for patients with FS. For analysis, age was also categorized into three groups: < 6 months, ≥ 6 months to < 6 years, and ≥ 6 years. For comparative analysis, age groups < 6 months and ≥ 6 years were grouped together (atypical age group) for comparison with the ≥ 6 months to < 6 years group (typical age group) for FS. We also converted the continuous variables (laboratory and clinical variables; number of symptoms on presentation, peak temperature, fever duration, and length of hospitalization) to categorical variables using the mean of the respective variables as cut-offs. A comparison of the trend of COVID-19 disease severity between the three waves was conducted using the chi-squared test. For univariate comparisons between patients with and without FS (dependent variable) in the third wave, Pearson’s correlation was used for nominal (categorical variables) data, and the Mann–Whitney test was used for non-normally distributed data (age, number of symptoms on presentation, laboratory variables, fever duration and length of stay) and the Student t-test for normally distributed data (peak temperature) for continuous variables. The normality of continuous variables was assessed using the Shapiro–Wilk test. A p value of < 0.05 was taken as significant. Categorical variables with a p value of < 0.05 on univariate analyses were entered into multivariate analysis. For multivariate analyses, significant laboratory variables were not included in the analysis as the dataset was incomplete as not all patients had undergone investigations. Results of the multivariate analysis were presented as odds ratio (OR), p value, and 95% confidence interval (95% CI). The Hosmer–Lemeshow goodness-of-fit test was used to assess model fitness.

Results

The disease categories for pediatric patients admitted to the NIC in the three waves are shown in Table 1. There was a larger proportion of patients in categories 1 and 2a in the first and second waves. A significantly higher number of category 2b patients were admitted in the third wave (p < 0.001 for trend). There were no patients in the category 3 disease and above.

Table 1 Disease severity of COVID-19 infection among pediatric patients admitted to the National Isolation Center (NIC)
Full size table

The age group breakdown of the three waves and proportion of patients with FS in age groups are shown in Table 2. All 29 cases of FS were encountered in the third wave, and none in the first and second waves, giving an overall prevalence of 4.5% and prevalence of 13% in the third wave. The majority of patients with FS were in the typical age group (≥ 6 months to < 6 years) with the highest frequency in the 1–2-year (n = 13, 45%) and 2–3-year (n = 6, 21%) groups (Supplementary Table 1). Five cases (17%) occurred in the ≥ 6 years group; 2 patients aged 6 years and 3 patients aged between 7 and 9 years. Two of the five cases had FS previously and one had a background of speech delay and learning disability.

Table 2 The number of pediatric patients with COVID-19 infection and febrile seizures (FS) in each wave and age group
Full size table

Twenty-six (90%) of the FS were generalized tonic–clonic seizures and the types of FS were not specified in the remaining (n = 3) patients. The majority of cases were simple FS (n = 20, 69%) and others (n = 9, 31%) were classified as complex FS. FS occurred while sponging (n = 4, 14%), feeding (n = 2, 6.9%), after waking up (n = 1, 3.4%), on the way to hospital (n = 1, 3.4%), and not stated in the rest of patients (n = 21, 72%). The rest of the characteristics of FS patients in our study are shown in Table 3. FS (n = 18, 62%) occurred mainly between evening and early morning hours (19:00 to 24:00 h and 00:00 to 06:00 h). The median time from onset of FS to presentation was 48 min (range 10–167 min). Ninety percent (n = 26) of the cases had only one febrile seizure. There were three peak intervals of FS occurring after onset of fever: 6 h or less (n = 8, 28%), with most occurring between 13 and 24 h (n = 12, 41%) and > 48 h (n = 7, 24%). The majority (n = 23, 79%) of the seizures had stopped at presentation and did not require any treatment.

Table 3 Characteristics of febrile seizures (FS) among pediatric patients (N = 29) with FS
Full size table

Six (21%) patients needed treatment to stop the seizure, including administration of rectal diazepam (n = 4), initial administration of rectal diazepam later supplemented with intravenous midazolam (n = 1), and two doses of intravenous midazolam and phenytoin loading dose (n = 1). The majority of seizures lasted between 1 and 5 min (n = 18, 62%), followed by < 1 min (n = 6, 21%). One patient had febrile status epilepticus and three patients had more than one seizure within one hour, one of whom received rectal diazepam. Four patients were given emergency anticonvulsant medications due to prolonged seizures of more than 5 min. Another patient had two seizures within 24 h but did not receive any anticonvulsant medication.

Among patients with FS, 14% (n = 4) had a previous history of FS and 17% (n = 5) had a family history of FS. No patients reported both previous personal and family history of FS. Overall, for both febrile seizure and non-febrile seizure patients, the mean number of symptoms on presentation was 3 ± 1, and the five most common symptoms were fever (n = 212, 97%), reduced appetite (n = 94, 43%), cough (n = 89, 41%), vomiting (n = 74, 34%), and rhinorrhea (n = 35, 16%) (Supplementary Table 1). Patients with FS reported fewer symptoms on presentation than those without FS, with mean number of symptoms of 2 ± 1. The mean peak temperature for FS patients was 39.2 ± 0.7 °C and the mean duration of fever was 3.1 ± 1.0 days. The mean hospitalization period for patients with FS was 3.0 ± 0.8 days. The comparisons between patients with FS and without FS in the third wave are summarized in Table 4.

Table 4 Comparisons between patients with febrile seizures (FS) and those without FS in the third wave
Full size table

Univariate analysis showed a typical age group (≥ 6 months to < 6 years), history of FS, family history of FS, fewer reported symptoms on presentation (Supplementary Table 1), peak temperature during the seizure and serum sodium levels were associated with FS (all p < 0.05). However, only typical age group (OR 4.7; 95% CI [1.0, 21.4], p = 0.047), family history (OR 33.8; 95% CI [3.0, 379.5], p = 0.004) and fewer symptoms on presentation (OR 4.0; 95% CI [1.4, 11.7], p = 0.011) remained significant on multivariate analyses. The model fitness was good (Hosmer–Lemeshow goodness-of-fit test; chi-square 5.62, df 7 and p = 0.564).

At a median follow-up of 65 days (range 56–80 days), none of the patients who had had FS presented with recurrence of seizure, and all had remained well without needing any treatment.

Discussion

Although it has been 4 years since the start of the COVID-19 pandemic, studies on FS due to COVID-19 are scarce. Our study showed that all cases of FS occurred only during the third wave of the COVID-19 in which the Omicron variant was dominant. There was also a larger proportion of fever cases admitted in the third wave.

The increase in the proportion of category 2b cases in the third wave was partly due to the change in our admitting criteria. Our admission criteria to NIC evolved as the outbreak progressed and to cope with the increasing number of patients. Moreover, the increasing infectivity of the subsequent SARS-CoV-2 variants resulted in a higher number of patients being admitted. The number of children with COVID-19, particularly in the first wave, was very small with only 12 children out of a total of 380 COVID-19 patients. This small number was due to the readiness and strict measures implemented in the country as well as the low infectivity of the original variant.

It is interesting that cases of FS only started to appear in the third wave with the Omicron variant [9, 20, 21]. This has been reported in other literature, with few to no reports of FS during the periods with the original and Delta variants. The occurrence of FS in children with COVID-19 has been reported as 0.5 to 5.5% in the USA and the UK [11, 22]. However, these countries report lower rates of FS in general, with cumulative incidence of FS prior to COVID-19 pandemic being estimated as between 2 and 5% in the USA and Western Europe [3,4,5,6].

The prevalence rates of FS prior to the COVID-19 pandemic were variable in Asian countries, ranging from 5 to 10% in India, 6 to 9% in Japan, and 14% in Guam [23, 24]. Our rate of FS was high (13%) in children with COVID-19 associated with the Omicron variant when compared to Western-published reports of COVID-19-related FS. However, it is comparable to rates reported for COVID-19-related FS in certain Asian countries according to studies that also look at FS when the Omicron variant was prevalent [9, 21]. This may indicate that viral makeup plays a part in clinical manifestations. Pre-COVID-19 studies have reported that FS are common with some viral illnesses (adenovirus, influenza A and influenza B) as reported by the EFES study [8]. In different age groups, the viral etiologies of FS differ.

This study has shown that FS due to COVID-19 have many similarities to FS due to non-COVID-19 causes. FS usually peak in occurrence between 12 and 18 months of age [5, 16, 25] and incidence then decreases with age. Our study showed similar findings with most cases in the 1–3-year age group. Pediatric patients aged ≥ 6 months to < 6 years admitted with COVID-19 were more likely to have FS compared to the other group (< 6 months and ≥ 6 years), with an OR of 4.7. This reinforces the data showing that FS occurs more in the typical age group (≥ 6 months to < 6 years) [26].

However, there were five cases outside the defined age for simple FS; two patients aged 6 and the other three were aged between 7 and 9 years. All five have recovered completely and follow up after hospital discharge has not shown any recurrence of FS. One study looking at FS during Omicron wave showed a higher mean age of 33 months in the COVID-19-related FS group than that of the COVID-19 negative group. In that study, 15 out of 61 FS patients were above 6 years of age [21]. Iijima et al. also found that 36% (8/22) of the patients with seizures in the Omicron era were older than the typical age for FS [9].

In our study, FS were liable to occur at any time during the day but they occurred mainly between the evening and early morning hours (19:00 to 24:00 h and 00:00 to 06:00 h). There were three peaks (6 h or less, 13–24 h and > 48 h) after onset of fever. The significance of these findings is unknown and will require further study. The majority of FS had stopped on presentation to the clinic or hospital without needing treatment but more importantly, none of the patients in the study required critical care input, which is reassuring for parents and carers. These findings are similar to another study which looked at COVID-19 patients with FS and found that only 25% required hospitalization when seen in the emergency department and even fewer required critical care services [11].

As expected, generalized tonic–clonic seizures were more common, occurring in 26 out of 29 (90%) patients. Most had simple FS (69%) and 31% were classified as having complex FS. One patient had febrile status epilepticus, and three patients had more than one seizure within 1 h. Four patients were given emergency anticonvulsant medications due to prolonged seizures of more than 5 min. Another patient had two seizures within 24 h. There is a higher risk of epilepsy in subjects with complex FS, which is estimated at between 4 and 15% [27,28,29]. Therefore, these patients would need follow up care to monitor for any progression to epilepsy. At a median follow-up of 65 days, there were no subsequent presentations to any government health institutions with recurrent seizure disorders. However, our follow up was too soon after initial care and a longer period between initial care and follow-up will be required.

Family history of FS is important and having a first or second degree relative with febrile or afebrile seizures is a well-known risk factor for develo** FS [30]. In our study, the OR of FS in those with a family history of FS was high (OR 33.8) with a very wide confidence interval (95% CI [3.0, 379.5]), and this was due to the overall small sample size of patients with a family history of FS.

Interestingly, patients with FS reported fewer symptoms on presentation than those without FS. Patients with FS were 4 times more likely to have fewer than 3 symptoms compared to those without FS. When we compared the individual symptoms (Supplementary Table 1), there was no significant difference between the two groups. To our knowledge, association with fewer symptoms on presentation has not been reported before in the literature. The exact reason for this is not known and need to be studied further. This could be due to majority of patients with FS presented within 24 h of onset of fever and other symptoms might appear later. As reporting of symptoms was done by the parents or carers, it is possible that the development of FS may have resulted in recall bias and milder symptoms may not have been reported.

Among the laboratory variables assessed, serum sodium levels showed significant correlation with FS on univariate analyses, indicating that lower sodium levels may predispose individuals to FS. Lower serum sodium levels have been previously shown to be associated with FS [31, 32]. Interestingly, the method of testing has also been reported to be important. In our setting, serum sodium levels were analyzed using the indirect potentiometry. A study has reported that there can be a discrepancy between direct and indirect potentiometry [33]. The discrepancy may be > 10 mmol/L with the direct method being more accurate. Therefore, it is possible that the true serum sodium level may be higher. Even though serum sodium levels were significant on univariate analysis, we did not include serum sodium levels in the multivariate analysis as a large proportion of our patients had recovered by the time of presentation and did not undergo laboratory investigations. Therefore, the dataset was not complete and this may have impacted the robustness of the analysis. Incomplete datasets were expected as the study utilized data collected from a real-world situation. The other laboratory parameters assessed were not significant.

Patients with FS were seen to have higher peak temperatures. This is in kee** with recent evidence that suggests the risk of FS is related to height of temperature elevation and not the rate of temperature rise [34].

In our study, there are several strengths and limitations that need to be considered when interpreting our findings. Most of the data was collected prospectively and entered into a database used for the management of COVID-19 during the outbreak. Some data, in particular laboratory data, was retrospectively retrieved from Bru-HIMS. Hence the data is comprehensive and there was no missing data on demographics, comorbidities, and relevant risk factors for FS. All patients were verified SARS-CoV-2 RT-PCR positive cases and diagnosis of FS followed the ILAE definition.

Potential limitations may include the small number of patients in the first wave and possible missed cases of FS that were not admitted to the NIC. Our national response protocol for COVID-19 over the study period was very strict, which most likely contributed to the low number of COVID-19 cases in the first wave. Furthermore, our threshold for admission was low and all patients with FS were admitted. Therefore, it is very likely that the number of patients missed would have been minimal, given FS are considered an alarming manifestation and parents would most likely have brought their child to hospital. Another limitation of our study is that it took place in a single center and did not include other hospitals in the country, who may also have admitted patients with COVID-19 and seizures.

Conclusion

In Brunei Darussalam, FS only occurred in the third wave that has been associated with Omicron variant. The overall prevalence over the outbreak is comparable to the rates reported for FS in previous literature. The characteristics of the FS in our study are similar to what is already known, most being simple FS and patients having recovered without any need for intervention. However, there were instances where complex FS and febrile status epilepticus occurred. Therefore, health professionals need to be aware of FS being a presentation of COVID-19 in the pediatric population. Despite the majority of patients being in the typical age group for FS, seizures can also occur in older children. Long term follow up is needed particularly in those with complex FS, given the higher risk of epilepsy.