Febrile seizures (FS), the most prevalent neurological condition in early childhood, are generally categorized as either simple or complex based on seizure characteristics such as duration, focality, and recurrence within 24 hours [1,2]. Children with complex febrile seizures (CFS) are at increased risk of developing epilepsy [3]. Beyond these prognostic implications, these seizures also warrant careful diagnostic evaluation, as they may be associated with serious underlying neurological conditions. Accordingly, various clinical guidelines recommend additional testing for children with CFS but not for those with simple febrile seizures (SFS). These tests include lumbar puncture (LP), neuroimaging, and electroencephalography [1,2,4-8].

Exanthem subitum (ES), also known as roseola infantum, is a common acute febrile illness that primarily affects children under 2 years of age. ES is most frequently caused by human herpesvirus 6 (HHV-6), particularly the HHV-6B subtype [9-11]. FS is the most frequently reported neurological complication associated with ES, with estimates ranging from 1% to 50% [10-12]. The diagnosis of ES is typically based on clinical features [11]. In our study, we applied this clinical definition while acknowledging its well-established, though not exclusive, association with HHV-6B infection [12-15]. Prior research has characterized FS associated with HHV-6B infection, with some reports suggesting a higher occurrence of complex features in these cases [10,16-19]. Building on these findings, we hypothesized that FS occurring in the context of ES may, despite the benign and self-limiting course of ES, lead to more frequent diagnostic testing, empiric antibiotic use, and hospital admissions. To explore this hypothesis, we retrospectively compared the clinical presentation and management of FS in children with and without ES.

This retrospective observational study analyzed the medical records of children younger than 2 years who experienced their first FS at our institution between March 2013 and August 2024. An initial cohort of 147 patients was identified; six were excluded due to incomplete documentation, leaving 141 eligible participants for analysis. Given the well-established epidemiologic age distribution of ES [9-11], the study population was restricted to children younger than 2 years to focus on the clinical features and management of FS in the context of ES. During cohort screening, only one 30-month-old child met the clinical criteria for ES, supporting the decision to limit the analysis to this age group.

FS were defined as seizure episodes occurring in association with fever in the absence of an underlying central nervous system (CNS) infection, metabolic disturbance, or prior unprovoked seizures. CFS were identified based on any of the following features: focal onset, duration longer than 15 minutes, or recurrence within 24 hours [20]. All children classified as having ES were diagnosed clinically based on the characteristic sequence of several days of high fever followed by abrupt defervescence and the appearance of a maculopapular rash [11]. In our cohort, ES was diagnosed only when the post-defervescence rash was directly confirmed either during hospitalization or at outpatient follow-up visits by pediatric specialists. No patient was classified as having ES without documented rash confirmation. HHV-6B polymerase chain reaction (PCR) testing was not available at our institution during the study period and was therefore not included in the diagnostic evaluation.

All patients underwent initial blood tests upon arrival at the emergency department or at hospital admission. Rapid influenza testing was performed selectively during seasonal or epidemic outbreaks. LP and neuroimaging were undertaken in patients presenting with atypical features, such as CFS, prolonged postictal altered mental status, fever lasting more than 24 hours before seizure onset, recent antibiotic use, or clinical indicators of possible CNS infection, such as severe acute otitis media. At our institution, severe acute otitis media is defined by otoscopic findings accompanied by systemic signs and symptoms raising concern for CNS involvement. Children who underwent LP were generally admitted and treated for suspected CNS infection with empiric intravenous antibiotics, typically third-generation cephalosporins, with vancomycin added when indicated.

The collected demographic and clinical data included patient age, sex, history of developmental delay, family history of FS, seizure characteristics (focality, duration, and recurrence), peak body temperature, and fever duration prior to seizure. Laboratory findings were reviewed, including complete blood count, cerebrospinal fluid analysis, and influenza testing when performed. C-reactive protein levels were obtained as part of routine clinical care in some patients but were not included in the primary analysis. Neuroimaging results were also reviewed. These parameters were compared between children clinically diagnosed with ES and those without ES.

The study was approved by the relevant Institutional Review Board of Daegu Catholic University Medical Center (IRB No. DCUMC 2025-07-020). Informed consent was waived due to the retrospective design, and the study was conducted in accordance with the Declaration of Helsinki.

Statistical analyses were performed using GraphPad Prism version 10 (GraphPad Software Inc., San Diego, CA, USA). Continuous variables were summarized as medians with interquartile ranges (IQRs) and ranges, and categorical variables were presented as counts and percentages. Group comparisons were performed using the Mann–Whitney U test for continuous variables and the chi-square test or Fisher exact test for categorical variables, as appropriate. P values of less than 0.05 were considered to indicate statistical significance.

We analyzed data from 141 children experiencing their first FS. The cohort included 79 boys and 62 girls, with a median age of 16 months (IQR, 12 to 19; range, 6 to 24). The median peak body temperature was 39.2°C (IQR, 38.8°C to 39.7°C; range, 38.0°C to 40.9°C). Seizures occurred more than 24 hours after fever onset in 17 children (12.1%). In total, 44 patients (31.2%) exhibited features consistent with CFS: focal seizures in seven children (5.0%), seizures lasting longer than 15 minutes in six (4.3%), and recurrent seizures within 24 hours in 35 (24.8%). A family history of FS was noted in 43 children (30.5%). LP was performed in 25 children (17.7%), and all cerebrospinal fluid results were normal. The main indication for LP was CFS (18/25, 72.0%); other indications included severe acute otitis media (n=4), seizure onset more than 24 hours after fever onset (n=2), and recent antibiotic use (n=1). Neuroimaging was conducted in 17 cases (12.1%), none of which showed abnormalities. Most patients (91.5%) initially presented to the emergency department, and 59 (41.8%) were admitted to the hospital. Empiric intravenous antibiotics targeting potential CNS infections were administered to 23 patients (16.3%). The underlying causes of fever included upper respiratory tract infection (58 cases, 41.1%), unknown origin (31 cases, 22.0%), ES (28 cases, 19.9%), influenza (12 cases, 8.5%), pneumonia (six cases, 4.3%), gastroenteritis (two cases, 1.4%), and urinary tract infection (two cases, 1.4%). Notably, none of the children who tested positive for influenza underwent LP, even when they presented with features of CFS.

We compared clinical characteristics between children diagnosed with ES (n=28) and those without ES (n=113). The results are shown in Table 1. No significant differences were observed in age, sex, peak body temperature, duration from fever onset to seizure, history of developmental delay, or family history of FS. Although the ES group had a higher rate of CFS, this difference did not reach statistical significance. Likewise, focality, seizure duration, and recurrence within 24 hours were similar between groups. However, the ES group displayed a trend toward more frequent recurrent seizures and delayed seizure onset (more than 24 hours after fever onset), possibly indicating a more complex or atypical seizure presentation. LP was performed significantly more often in the ES group (P<0.001), and empiric intravenous antibiotic use for suspected CNS infection was also more frequent (P<0.001). Although not statistically significant, the ES group also had higher rates of neuroimaging and hospital admission. Laboratory evaluations revealed lower median white blood cell (WBC) and platelet counts in the ES group than in the non-ES group (P<0.001 and P=0.035, respectively).

This study examined the clinical features and management patterns of FS in young children with and without ES. Accordingly, the findings should be interpreted as reflecting FS associated with clinically diagnosed ES, rather than virologically confirmed HHV-6B infection. Our results demonstrate that children diagnosed with ES were significantly more likely to undergo LP and receive empiric antibiotic treatment for presumed CNS infection, even though ES is typically a benign, self-resolving illness. These findings are consistent with previous research exploring FS in the context of HHV-6B infection, the most common cause of ES. For instance, Suga et al. [17] reported that 47.4% of HHV-6B–positive patients underwent LP compared with 10% of those without HHV-6B infection. Likewise, Laina et al. [21] noted that LP was performed in 30% of HHV-6B–positive children, whereas none of the children in the negative group underwent LP. Although both studies presented cerebrospinal fluid findings for HHV-6B–positive and –negative groups, neither explicitly reported LP rates. A review of their tabulated data revealed notable between-group differences in LP frequency, aligning with our findings.

The decision to perform LP in children with FS often hinges on clinical judgment, especially in cases of CFS, which are defined by features such as prolonged duration, focal onset, or recurrence [2,4,6]. In the pre-vaccine era, LP was commonly performed in nearly all children with CFS to rule out bacterial meningitis. However, in the current era, in which most children are immunized, the rate of bacterial meningitis in CFS cases has declined to 0%–1.5% [2,5]. Despite this decrease, many physicians still favor LP in the setting of CFS, even in the absence of definitive supporting evidence. Compared to children with SFS, these patients are more likely to undergo extensive testing, including blood work, urinalysis, and neuroimaging, and to be hospitalized [6]. Similarly, a previous study reported that the presence of complex features is a primary driver of both LP and empiric antibiotic use [4]. Consistent with these reports, our study found that CFS represented the most common indication for LP.

Although earlier research suggested a greater frequency of CFS in children with primary HHV-6B infection [10,12,17,18], more recent studies have not consistently confirmed this association. For example, Miyake et al. [22] reported no significant differences in seizure duration, focality, or recurrence between HHV-6B–positive and –negative patients. In our cohort, although the overall frequency of CFS did not differ significantly between groups, we observed a higher rate of recurrent seizures in the ES group (39.3% vs. 21.2%, P=0.055), suggesting a possible trend toward greater seizure complexity.

Another noteworthy finding was the longer duration of fever prior to seizure onset in children with ES. Miyake et al. [22] found that in HHV-6B–positive children, seizures occurred more than 24 hours after fever onset in 50% of cases, compared to 6.3% in HHV-6B–negative peers (P<0.001). Most FS occur within the first 24 hours of fever onset, as described by Berg et al. [23], and later-onset seizures have been associated with greater concern for serious underlying conditions, including CNS infections [24,25]. In our study, children with ES more frequently exhibited delayed seizure onset as well as recurrence within 24 hours. These atypical presentations can heighten clinical suspicion for CNS infection at the initial encounter, particularly during the early febrile phase, when the etiology remains unclear and the characteristic ES rash has not yet appeared. Together, these atypical seizure features and the associated diagnostic uncertainty likely lowered clinicians’ thresholds for invasive evaluation at the time of initial assessment, which may help explain the higher rate of LP in the ES group. Overall, the higher use of invasive management likely reflects presentation-driven clinical decision-making rather than an intrinsic effect of ES.

Patient age is another factor that influences clinical decisions regarding LP. Some studies have found that HHV-6B–associated FS tends to occur at a younger age [12,22], although others report no significant age difference [21,26]. In our study, the similar ages observed in both groups may be attributable to the study’s restriction to children under 2 years, which encompasses the typical age range for ES. Therefore, age likely had minimal influence on LP decisions in our study population.

Rapid viral diagnostics have been shown to reduce unnecessary interventions in febrile children with respiratory illnesses, such as influenza or respiratory syncytial virus infection [27,28]. In our cohort, no influenza-positive child underwent LP, even when presenting with features of CFS, likely reflecting the influence of rapid diagnostic testing. However, no rapid test currently exists for HHV-6B. Current PCR methods require blood or cerebrospinal fluid samples, limiting their value in acute care. As a result, etiologic uncertainty often persists during the early febrile phase, which may partly explain why children later diagnosed with ES underwent more extensive diagnostic and therapeutic procedures in our cohort. Therefore, the development of a rapid, noninvasive test for HHV-6B, similar to point-of-care influenza assays, may help reduce unnecessary LPs, improve clinical decision-making, and alleviate caregiver anxiety.

We also observed significantly lower WBC and platelet counts in the ES group, consistent with previous reports of transient bone marrow suppression during acute HHV-6B infection [10,11,26]. The increased use of empiric antibiotics in ES likely reflects the same concern for possible CNS infection that contributed to increased LP use, despite low WBC counts.

While this study does not provide new diagnostic or virologic insights, it underscores the clinical relevance of recognizing ES in the management of FS. Clinical suspicion may be raised by seizures occurring more than 24 hours after fever onset, by recurrent episodes, or by transient leukopenia or thrombocytopenia. Awareness of these features may help physicians make appropriate decisions and reduce unnecessary LP, empiric antibiotics, and hospitalizations in children with otherwise benign illness.

This study has several limitations. First, it was a single-center retrospective review in which diagnostic and management decisions were not standardized; rather, they relied on the judgment of individual physicians. Second, ES was diagnosed clinically without virologic confirmation of HHV-6B. Although HHV-6B is the most common cause, HHV-7 can also produce this syndrome, and not all HHV-6B infections manifest as the classic ES presentation. Accordingly, clinically diagnosed ES is not equivalent to virologically confirmed HHV-6B infection, which may limit direct comparability with studies based on laboratory-confirmed cases. Third, detailed fever trajectories could not be reliably assessed due to the study’s retrospective design. Despite these limitations, our findings remain clinically relevant for decision-making in similar settings.

In conclusion, children with ES-associated FS were more likely to undergo LP and receive empiric antibiotics, despite the typically benign course of ES. Our findings underscore the need for rapid, noninvasive diagnostic tools for HHV-6B testing and updated clinical guidelines for FS to minimize unnecessary interventions. Prospective multicenter studies are needed to confirm these observations and to establish evidence-based strategies that optimize clinical decision-making in young children with FS.