Recurrence of febrile convulsions in children treated with intermittent benzodiazepines: a retrospective study investigating predisposing factors

Abstract

Introduction

Febrile seizures are the most common neurologic disorder in infants and young children, affecting 2–4% of those under six years of age. Although their overall prognosis is benign, recurrence is frequent and a major concern for families. This study aimed to identify factors associated with recurrence despite intermittent benzodiazepine prophylaxis.

Method

We conducted a retrospective cohort study at Tehran Children’s Medical Center including 76 children aged 6–60 months with febrile seizures who were discharged on intermittent benzodiazepine therapy. The primary outcome was recurrence despite treatment. Clinical and laboratory variables were compared between recurrence and non-recurrence groups using logistic regression.

Results

Febrile seizures recurred in 20 of 76 children (26.3%). Age at first febrile seizure was significantly lower in the recurrence group (p-value = 0.001), and a history of previous febrile seizures was associated with recurrence (p = 0.018). Hyponatremia was also more frequent among recurrent cases (p-value = 0.003). In multivariable analysis, lower serum sodium (OR 0.713; 95% CI 0.518–0.981; p = 0.038) and younger age at first febrile seizure (OR 0.902; 95% CI 0.818–0.995; p = 0.039) remained independent predictors.

Conclusion

Younger age at first seizure and lower serum sodium were associated with recurrence despite intermittent benzodiazepine prophylaxis. These findings may help identify children at higher risk of treatment failure and guide counselling of families considering prophylaxis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12887-026-06742-5.

Introduction

Febrile convulsions are the most common seizures in children under five years of age [1]. They occur between 6 and 60 months of age with fever ≥ 38 °C, excluding cases due to central nervous system infection [2–5]. They are not classified as epileptic seizures [6]. The condition affects about 2–4% of children and is usually benign, but it often causes parental anxiety and repeated visits to emergency departments [1, 7, 8].

Febrile convulsions are classified as simple or complex based on seizure duration, clinical features, and recurrence within the same febrile illness [9–11]. Prevalence varies across regions, ranging from 2 to 4% in the United States and Western Europe, 9–10% in Japan, and up to 14% in Iceland [7, 12]. Rates close to 10% have also been reported in East Asia, including India [13]. Most seizures are brief, with 87% lasting less than 10 min and only 9% exceeding 15 min [8]. Despite their benign course, they remain a source of parental anxiety [14]. Preventive use of antipyretics such as acetaminophen or ibuprofen has not shown consistent benefit, though one study suggested a modest effect for acetaminophen [15].

Several mechanisms are thought to contribute to febrile seizures. Fever-induced cytokine release (e.g., IL-1β, IL-6, TNF-α), immature inhibitory neurotransmission, and hyperventilation leading to respiratory alkalosis with reduced ionized calcium can lower the seizure threshold. Electrolyte disturbances, such as hyponatremia, as well as short fever-to-seizure intervals and genetic susceptibility, These mechanisms help explain variability in recurrence risk among affected children [16].

Benzodiazepines, which act as sedatives, anxiolytics, muscle relaxants, and anticonvulsants [17], have been widely studied in febrile seizure management. Continuous prophylaxis is generally avoided because of side effects and limited benefit. In contrast, intermittent benzodiazepine prophylaxis at fever onset may be considered in children at high risk of recurrence or in families with significant anxiety [18–20]. The American Academy of Pediatrics allows for oral diazepam in such situations [8]. Japanese guidelines also recommend intermittent diazepam for children with specific recurrence risk factors, including young age at onset, positive family history, abnormal perinatal history, low fever at presentation, short interval between fever and seizure, recurrence within the same illness, and focal seizure features [4]. The Association of Child Neurology in India advises selective use of intermittent benzodiazepines in children with frequent recurrences, in families with marked anxiety, or where access to emergency care is limited [21].

Evidence supports this selective approach. The Cochrane Review by Offringa et al. (2021) reported that intermittent diazepam reduces recurrence risk over 6–24 months, with risk ratios of about 0.6–0.7 compared to placebo, but adverse effects such as drowsiness and ataxia occurred in about one-third of children. On this basis, the authors concluded that intermittent diazepam should not be used routinely, but may be considered in selected cases [18]. A clinical trial also showed that intermittent diazepam and clobazam were both effective when given during fever [19]. Johns Hopkins All Children’s Hospital recommends rectal or oral diazepam at fever onset, with a repeat dose after 8 h, for children with very frequent or prolonged febrile seizures [22]. Cohort data support these findings; in one study of 85 children with recurrent febrile seizures, intermittent benzodiazepine prophylaxis prevented about 90% of recurrences, though mild side effects occurred in 36% of cases [23].

These data indicate that intermittent benzodiazepines can reduce febrile seizure recurrence, but treatment failure still occurs. The present study was designed to identify factors associated with recurrence in children prescribed intermittent benzodiazepine therapy. While several factors have been reported as predictors of febrile seizure recurrence [24, 25], it is unclear whether these associations persist in children receiving prophylaxis. To the best of our knowledge, this is the first study to address this question, providing evidence on factors linked to prophylaxis failure and offering clinically relevant insights for risk stratification and parental counselling.

Materials and methods

Study design

This retrospective study was conducted at Tehran Children’s Medical Center, a pediatric tertiary care hospital. The study adhered to the Declaration of Helsinki and the Ethical Guidelines for Medical and Health Research [26].

Study population

Participants were children aged 6–60 months who presented with febrile seizures between March 20, 2020, and March 20, 2021. Eligible children had seizures associated with fever and no preexisting central nervous system (CNS) disorders or identifiable etiology. Only patients discharged with a prescription for intermittent benzodiazepines were included. The standard regimen was oral diazepam 0.3–0.33 mg/kg every 8 h for up to 48 h from fever onset, with clonazepam 0.01 mg/kg every 12 h used when diazepam was unavailable. Adequate adherence was defined as initiation of at least two doses within the first 24 h of fever onset; non-adherent families were excluded. Inclusion criteria were: [1] age 6–60 months; [2] simple or complex febrile seizures; [3] axillary temperature > 38 °C; [4] no neurological abnormalities, including altered mental status on examination; [5] discharge with intermittent benzodiazepine treatment; and [6] parental oral informed consent. Exclusion criteria were: [1] history of epilepsy, intracranial neoplasm, hydrocephalus, inborn errors of metabolism, hypoxic-ischemic encephalopathy, or multiple congenital anomalies, including chromosomal disorders; [2] ongoing long-term anticonvulsant therapy; or [3] prior use of intermittent benzodiazepines before the index febrile seizure.

Data collection

Data included demographic, epidemiological, and clinical variables: developmental status, fever-related factors, seizure characteristics, family history, maternal habits during pregnancy, hospitalization history, comorbid conditions (abstracted from discharge diagnoses and grouped into predefined categories), vaccination status, laboratory results, and medication history during the index episode. Blood samples were routinely obtained from all children with seizures immediately upon arrival, before any treatment or definitive diagnosis; thus, laboratory values reflect baseline status at presentation. Recurrence of febrile seizures despite benzodiazepine use was also recorded. Data were obtained from clinical records and by contacting families, on average two years after the initial episode. The study purpose was explained, confidentiality assured, and oral consent obtained. Children with incomplete data or inadequate benzodiazepine use within the first 24 h of fever onset were excluded.

Outcomes

The primary outcome was the recurrence of febrile seizures despite intermittent benzodiazepine therapy. Secondary aims were to identify factors associated with recurrence.

Statistical analysis

Patient characteristics, medical history, medications, seizure features, and laboratory results were summarized as mean ± SD, median (IQR), or n (%). Independent t-tests were used for normally distributed quantitative variables, Mann–Whitney U tests for non-normally distributed variables, and Chi-square tests for categorical variables. Children who did not adhere to the prescribed benzodiazepine regimen were excluded, leaving 76 patients for analysis, divided into recurrence and non-recurrence groups.

Binary logistic regression was used because the outcome variable, febrile seizure recurrence, was binary (0 = no recurrence, 1 = recurrence Univariate logistic regression tested each predictor individually. Categorical variables (e.g., gender, gestational age, prior febrile seizure) were coded as binary, and continuous predictors (e.g., serum sodium, age at first seizure, birth weight) were kept on their natural scales. Predictors with p < 0.05 in univariate analysis were entered into multivariable analysis.

In the multivariable model, serum sodium, age at first febrile seizure, and prior febrile seizures were introduced sequentially. Prior febrile seizures significant in univariate analysis (p = 0.018), lost significance when adjusted for sodium and age (p = 0.218). Serum sodium remained significant (p < 0.05), and age at first febrile seizure showed borderline significance (p = 0.059). Collinearity diagnostics confirmed low correlation (VIF < 2). Model fit was assessed using the − 2 log-likelihood, Nagelkerke R², the likelihood ratio test (LRT), and the Akaike Information Criterion (AIC). These measures supported excluding prior febrile seizures as a confounder and retaining sodium and age at first febrile seizure in the model.

A final multivariable model, adjusted for demographic, clinical, and laboratory covariates, confirmed that lower sodium levels and younger age at first febrile seizure were independent predictors of recurrence. Analyses were performed using SPSS version 27, with two-sided p < 0.05 considered statistically significant.

Power analysis

A Monte-Carlo power analysis for logistic regression was conducted to assess whether the sample size was adequate to detect the observed effect sizes. Results are reported in the Results section.

Results

Between March 2020 and March 2021, 190 patients with fever and seizures were referred to the Children’s Medical Center; 76 met study criteria. Of these, 34 (44.7%) were girls and 42 (55.3%) were boys (female-to-male ratio: 0.81). Febrile seizure recurrence occurred in 20 children (26.3%), including 7 girls (35%) and 13 boys (65%), despite intermittent benzodiazepine treatment. Recurrence rates were 20.6% in girls and 31% in boys, a difference that was not significant (p = 0.433).

Birth weights did not differ between recurrence and non-recurrence groups (2782 ± 489 g vs. 2864 ± 597 g, P = 0.549). Recurrence was also not associated with gestational age (preterm vs. term: 27% vs. 23.1%; p = 1.0) or complications at childbirth (15.4% vs. 28.6%; p = 0.494). Children with neurodevelopmental disorders showed a higher recurrence rate (33.3% vs. 25.7%), but this was not significant (p = 0.65), likely due to the small number of affected children.

Comorbid conditions were present in 27 children (35.5%). Recurrence was more frequent in this group compared with children without comorbidities (40.7% vs. 18.4%), though the difference did not reach statistical significance (p = 0.055; Supplementary Table S1).

A history of at least one hospitalization was documented in 47.4% of the cohort. Recurrence occurred in 36.1% of these patients versus 17.5% of those without prior hospitalization (p = 0.075). Details of the children’s characteristics are presented in Table 1.

Table 1.

Patient characteristics

Characteristic		Total (n = 76)	Without Febrile Convulsions recurrence (FC)	With Recurrent Febrile Convulsions (FC)	p-value
Sex	Male (n, %)	42 (55.3%)	29 (69%)	13 (31%)	0.433
	Female (n, %)	34 (44.7%)	27 (79.4%)	7 (20.6%)
Gestational age	Term	63 (82.9%)	46 (73%)	17 (27%)	1
	Pre-term	13 (17.1%)	10 (76.9%)	3 (23.1%)
Birth weight			2864 ± 597 g	2782 ± 489 g	0.549
Insult during labor	Have	13 (17.1%)	11 (84.6%)	2 (15.4%)	0.494
	Don’t have	63 (82.9%)	45 (71.4%)	18 (28.6%)
Neurodevelopment	Normal	70 (92.1%)	52 (74.3%)	18 (25.7%)	0.651
	NDD	6 (7.9%)	4 (66.7%)	2 (33.3%)
Comorbid conditions	Have	27 (35.5%)	16 (59.3%)	11 (40.7%)	0.055
	Don’t have	49 (64.5%)	40 (81.6%)	9 (18.4%)
Previous hospitalization	Have	20 (26.3%)	23 (41.1%)	13 (65%)	0.75
	Don’t have	56 (73.7%)	33 (58.9%)	7 (35%)
FC Prior to the Current Session	Have	21 (27.6%)	11 (52.4%)	10 (47.6%)	0.018
	Don’t have	55 (72.4%)	45 (81.8%)	10 (18.2%)
Family history of epilepsy, febrile seizure and etc.	Have	34 (44.7%)	25 (73.5%)	9 (26.5%)	1
	Don’t have	42 (55.3%)	31 (73.8%)	11 (26.2%)
Use of antipyretics during fever episode	Yes	48 (63.2%)	33 (68.6%)	15 (31.3%)	0.282
	No	28 (36.8%)	23 (82.1%)	5 (17.9%)
Use of antihistamines during fever episode	Yes	6 (7.9%)	4 (66,7%)	2 (33.3%)	0.651
	No	70 (92.1%)	52 (74.3%)	18 (25.7%)

Bold p-values indicate statistical significance at p < 0.05

Children with recurrent febrile seizures experienced their first episode at a younger age than those without recurrence did (14.65 ± 2.78 months vs. 21.52 ± 2.78 months, P = 0.001). Twenty-one children (27.6%) had a prior history of febrile seizures; among them, 47.6% experienced recurrence compared with 18.2% of those without a prior history (P = 0.018).

The maximum recorded body temperature during febrile seizures was 38.81 ± 0.74 °C and was not associated with recurrence (P = 0.239). The most common fever etiologies were viral infections (67.1%), shigellosis (13.2%), and otitis media (6.6%). Neither antipyretic nor antihistamine use was associated with recurrence (p = 0.282 and p = 0.651, respectively). Recurrence did not differ between children with complex versus simple febrile seizures (p = 0.722). Average seizure duration was 3.45 min overall, 4.1 min in the recurrence group, and 3.21 min in the non-recurrence group, with no significant difference (p = 0.957). Fever-to-convulsion interval also showed no association with recurrence (Table 2).

Table 2.

Seizure characteristics

Seizure Characteristics		Total	Without Febrile Convulsions recurrence (FC)	With Recurrent Febrile Convulsions (FC)	p-value
Age at first episode of FC		19.72 ± 11.8 months	21.52 ± 2.78 months	14.65 ± 2.78 months	0.001
Type of FC	Simple	64 (84.2%)	48 (75%)	16 (25%)	0.722
	complex	12 (15.8%)	8 (66.7%)	4 (33.3%)
Duration of FC		3.45 ± 3 min	3.21 ± 2.35 min	4.1 ± 4.35 min	0.957
FC-T duration	Under 24 h.	54 (71.1%)	39 (72.2%)	15 (27.8%)	0.778
	Over 24 h.	22 (28.9%)	17 (77.3%)	5 (22.7%)
Maximum fever degree		38.80 ± 0.74 °C	38.72 ± 0.7 °C	39.02 ± 0.81 °C	0.239

Bold p-values indicate statistical significance at p < 0.05

Primary hyponatremia was significantly more common in the recurrence group (p = 0.003). In contrast, potassium and calcium levels did not differ between groups (p = 0.511 and p = 0.456, respectively). Initial blood glucose was higher in the recurrence group but not significantly so (124 ± 44 vs. 114 ± 42 mg/dL, p = 0.389). No significant differences were observed in hematologic parameters (WBC, RBC, MCV, HCT, Hb, PLT) (Table 3).

Table 3.

Laboratory findings

Laboratory factor	Without Febrile Convulsions recurrence (FC)	With Recurrent Febrile Convulsions (FC)	p-value
Na	134.93 ± 2.9	132.65 ± 2.66	0.003
K	4.22 ± 0.41	4.30 ± 0.45	0.511
Ca	9.68 ± 0.56	9.85 ± 0.97	0.338
Bs	114.59 ± 42.14	124.60 ± 44.65	0.389
WBC	9.85 ± 5.49	9.54 ± 4.60	0.807
RBC	4.99 ± 1.36	4.92 ± 0.69	0.767
Hb	12.24 ± 0.89	12.53 ± 1.68	0.465
MCV	78.44 ± 5.18	78.92 ± 4.01	0.675
HCT	37.48 ± 3.13	38.6 ± 5.91	0.429
PLT	266.05 ± 98.40	286.1 ± 107.02	0.447

Bold p-values indicate statistical significance at p < 0.05

Multivariable logistic regression confirmed that, after adjustment for demographic, clinical, and laboratory variables, lower serum sodium and younger age at first seizure remained independent predictors of recurrence. Lower sodium was associated with greater recurrence risk (p = 0.038), while older age at first seizure reduced recurrence risk (p = 0.039) (Table 4).

Table 4.

Multivariable logistic regression analysis

Variable	OR	CI (95%)	p-value
Na	0.713	0.518–0.981	0.038
Age at first episode of FC	0.902	0.818–0.995	0.039
Gender	2.369	0.465–12.056	0.299
Gestational Age	0.666	0.034–12.961	0.789
Birth Weight	1.000	0.998–1.002	0.780
Insult during labor	0.545	0.043–6.883	0.639
Neurodevelopment	1.715	0.063–47.003	0.749
Underlying disease	1.188	0.118–12.016	0.884
Previous hospitalization	2.210	0.159–30.770	0.555
Family history of epilepsy, febrile seizure and etc.	2.198	0.383–12.598	0.377
Use of antipyretics during fever episode	2.876	0.367–22.526	0.314
Use of antihistamines during fever episode	1.067	0.041–27.606	0.969
Duration of FC	1.191	0.876–1.620	0.265
FC-T duration	0.715	0.085–6.002	0.758
Maximum fever degree	1.961	0.580–6.629	0.278
K	1.395	0.198–9.808	0.738
Ca	1.393	0.420–4.616	0.588
Bs	0.993	0.975–1.012	0.470
WBC	0.883	0.686–1.135	0.332
RBC	0.237	0.004–12.942	0.481
Hb	0.551	0.054–5.646	0.616
MCV	1.107	0.770–1.593	0.583
HCT	1.440	0.621–3.341	0.395
PLT	1.010	0.997–1.023	0.118

Bold p-values indicate statistical significance at p < 0.05

Monte-Carlo power analysis confirmed adequate sample size to detect observed effects. With an event rate of 26.3% (20/76), achieved power was 0.95 for sodium (OR = 0.713) and 0.99 for age (OR = 0.902). Simulations showed that ≥ 80% power would be achieved with approximately 50 participants for sodium and 39 for age (Fig. 1).

Fig. 1.

Achieved power for age at first febrile seizure and serum sodium in predicting febrile seizure recurrence

ROC analysis showed good predictive ability for both sodium and age. For sodium, the AUC was 0.729 (95% CI: 0.594–0.864; SE = 0.069; p = 0.002) (Fig. 2). For age at first seizure, the AUC was 0.777 (95% CI: 0.645–0.909; SE = 0.067; p < 0.0001) (Fig. 3).

Fig. 2.

ROC curve for sodium levels in predicting febrile seizure recurrence

Fig. 3.

ROC curve for age at first febrile seizure in predicting recurrence

Discussion

Several studies have examined the effect of intermittent and continuous benzodiazepine use in preventing recurrent febrile seizures. In a 2022 study of 306 children, the combination of acetaminophen and intermittent diazepam significantly reduced recurrence compared with acetaminophen alone [27]. A 2021 meta-analysis by Offringa et al. concluded that, among preventive drugs, only continuous phenobarbital and intermittent diazepam were effective [18]. A 2023 meta-analysis further supported the efficacy of continuous phenobarbital and intermittent diazepam in reducing the risk of recurrent febrile seizures [28]. Despite these results, preventive treatment is not uniformly effective. Our study examined 76 children prescribed intermittent benzodiazepines to identify factors associated with treatment failure and explore potential reasons behind the lack of efficacy of this preventive approach. We found that younger age at first seizure and lower serum sodium levels were associated with recurrence even in children who received intermittent benzodiazepine prophylaxis.

In a 2006 study by Vestergaard et al., febrile seizure incidence was greater in boys than in girls, yet there was no statistically significant difference in the risk of recurrent febrile seizures between the two genders. Similarly, in our study, the number of affected boys (n = 42) exceeded that of girls (n = 34), and despite a higher recurrence rate among boys (31%) than among girls (20.6%) with intermittent benzodiazepine use, this difference did not reach statistical significance [29].

Age at first seizure has consistently been reported as an important predictor. Berg and Shinnar (1997) found recurrence risks of 50% in children younger than 12 months and 30% in those older than 12 months [30]. Consistent with these findings, our study showed that younger age at first febrile seizure was significantly associated with recurrence despite intermittent treatment.

Previous studies have also shown associations between recurrence and low birth weight or preterm birth [31]. In our study, recurrence rates did not differ significantly between term and preterm children, although the number of preterm patients was small (n = 13). Mean birth weight was slightly lower in the recurrence group, but this difference was not significant. Larger studies are needed to clarify these associations.

A history of febrile seizures prior to the index episode was significantly associated with recurrence in our cohort. This aligns with earlier reports by Berg and Shinnar [30], who observed that children with a second febrile seizure episode have an approximately 50% chance of further recurrence.

Other studies have identified additional risk factors, including age < 1 year at first seizure, male sex, lower peak temperature, positive family history, complex seizure type, seizure duration > 15 min, and maternal smoking during pregnancy [24, 25]. In our study, no significant associations were found with family history, seizure type, seizure duration, or fever height. None of the mothers reported smoking or alcohol use, so these variables could not be assessed. Complex seizures showed a higher recurrence rate (33.3% vs. 25% for simple seizures), but the small number of cases limits interpretation.

In a 2022 study, children were divided into three groups: febrile without seizures, simple febrile seizures, and recurrent febrile seizures. Recurrent seizures were significantly associated with lower serum sodium levels [32]. A meta-analysis also confirmed the link between hyponatremia and recurrent febrile seizures during the same illness, reporting an optimal cutoff of 134.72 mmol/L [33]. Our findings are consistent: children with recurrence had significantly lower serum sodium at admission despite benzodiazepine treatment.

The mechanisms linking hyponatremia to febrile seizures are well described. Low extracellular sodium reduces osmolality, causing water influx into brain cells and neuronal swelling, which disrupts ionic gradients and increases excitability [34]. Fever-induced hyperventilation may further produce respiratory alkalosis, lowering CO₂ and ionized calcium, thereby reducing seizure threshold [35]. Together, these mechanisms create a pro-convulsive state. Our results support these pathophysiological links and show that hyponatremia remains a predictor of recurrence even with intermittent benzodiazepine prophylaxis.

Finally, EEG abnormalities have been reported in some children with febrile seizures. In our study, two patients had epileptiform discharges on EEG, both of whom experienced recurrence despite prophylaxis. Similar findings were reported in a 2004 study, which concluded that intermittent benzodiazepines could not prevent recurrence in children with epileptiform EEGs, although treatment was often continued due to parental anxiety [36].

Limitations and future directions

This study has several limitations. It was conducted at a single center with a retrospective design, which may reduce generalizability. The sample size was modest, and some clinical files were incomplete, while several families could not be contacted despite repeated attempts. Adherence to intermittent benzodiazepine therapy was assessed by parental report obtained about two years after the index event; although a standardized script was used and clearly inadequate dosing was excluded, recall bias and misclassification remain possible. The study was adequately powered for the main predictors of interest but underpowered to assess less common variables such as neurodevelopmental disorders, family history of epilepsy, and specific viral etiologies.

Future studies should include larger, prospective, multi-center cohorts with longer follow-up to improve reliability and generalizability. Additional factors such as parental education, treatment cost and availability, and other influences on adherence should be examined. Controlled interventional studies may help clarify the causes of treatment failure, and investigations into biological contributors, including iron status and zinc levels, could further define their role in seizure outcomes.

By focusing on children treated with intermittent benzodiazepines, this study highlights factors linked to prophylaxis failure, providing context-specific evidence to guide risk stratification and parental counselling.

Authors’ contributions

S.A. contributed to selecting the title, conducting the literature review, designing the questionnaire, obtaining informed consent from parents, overseeing the data collection process, designing the methodology, performing statistical analysis, drafting the manuscript (all sections), and revising the final script. N.M. contributed to data collection, methodology design, statistical analysis, drafting the manuscript (methods section), and revising the final script. A.A. was responsible for statistical analysis and drafting the manuscript (results section). A.Z. contributed to data collection and drafting the manuscript (introduction section). S.A. contributed to data collection and drafting the manuscript (discussion section). E.H.E.M. served as the corresponding author, assisted in selecting the title, contributed to the methodology design, and revised the manuscript. All authors have reviewed and approved the final manuscript and take full responsibility for the accuracy and integrity of the work.

Funding

This research received no external funding.

Data availability

The data that support the findings of this study are not publicly available but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Informed oral consent was obtained from the parents of all participants before their inclusion in the study. This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of Tehran University of Medical Sciences (Approval ID: IR.TUMS.CHMC.REC.1400.070).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.