Of major concern in any febrile child presenting with a seizure is the possibility of bacterial meningitis (BM). We did a systematic review to estimate the risk of BM among various subgroups of young children with a first seizure in the context of fever, and to assess the utility of routine lumbar puncture (LP) in children with an apparent first FS.
MEDLINE, INIST, and the COCHRANE Library databases were searched from inception to December 2011 for published studies, supplemented by manual searches of bibliographies of potentially relevant articles and review articles. Studies reporting the prevalence of BM in young children presenting to emergency care with a first: i) “seizure and fever”, ii) apparent simple FS, and iii) apparent complex FS were included. Fourteen studies met the inclusion criteria. In children with a first “seizure and fever”, the pooled prevalence of BM was 2.6% (95% CI 0.9–5.1); the diagnosis of BM might be suspected from clinical examination in 95% of children >6 months. In children with an apparent simple FS, the average prevalence of BM was 0.2% (range 0 to 1%). The pooled prevalence of BM among children with an apparent complex FS was 0.6% (95% CI 0.2–1.4). The utility of routine LP for diagnosis of CNS infections requiring immediate treatment in children with an apparent first FS was low: the number of patients needed to test to identify one case of such infections was 1109 in children with an apparent first simple FS, and 180 in those with an apparent first complex FS.
Citation: Najaf-Zadeh A, Dubos F, Hue V, Pruvost I, Bennour A, Martinot A (2013) Risk of Bacterial Meningitis in Young Children with a First Seizure in the Context of Fever: A Systematic Review and Meta-Analysis. PLoS ONE 8(1): e55270. doi:10.1371/journal.pone.0055270
Editor: Robyn Klein, Washington University, United States of America
Received: September 4, 2012; Accepted: December 22, 2012; Published: January 28, 2013
Copyright: © 2013 Najaf-Zadeh et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Seizures occur in the context of many childhood illnesses, and account for approximately 1–5% of all emergency departments (ED) visits , . Febrile seizures (FS) represent the most common form of childhood seizures, affecting approximately 2–5% of infants and young children in Europe and North America, and 8% in Japan . It is broadly defined as a seizure accompanied by fever without evidence of central nervous system (CNS) infections, occurring in children 6 months to 5 years of age .
The association between seizure and bacterial meningitis (BM) is well established –. It is therefore imperative to rule out BM prior to making the diagnosis of FS. However, the diagnosis of FS in certain subgroups of children with an apparent FS is a challenge: FS may be the sole manifestation of BM in infants ; complex features of seizure can increase the risk of BM in others , . Accordingly, when these children present with an apparent FS, clinicians may remain uncertain about the risk of BM. In acute situation, the most challenging issue is to make a decision whether a lumbar puncture (LP) is necessary to rule out BM. Knowledge of the prevalence of BM among various subgroups of children with FS can assist clinicians to make appropriate clinical decisions in such challenging situations.
Numerous clinical studies reporting the prevalence of BM among children with an apparent FS have been conducted throughout the world –. There have been also four literature review articles (published in year 1980, 2001, 2003, and 2011) looking at the prevalence of BM in children with an apparent FS from developed countries –. However, the three first reviews – were subject to bias because of combining different patient groups in their review (i.e., children with a “seizure and fever” , , –, those with an apparent FS –, –, and those with an FS ). The 2011 review was limited to a subgroup of children <18 months of age with an apparent first simple FS who were enrolled during post-vaccine era (i.e., immunization against Haemophilus influenzae type b (Hib) and Streptococcus pneumoniae .
Consequently, by using a more rigorous methodology than the previous studies, and in light of the more recent publications, we undertook a systematic review and meta-analysis, where appropriate, of the relevant literature to provide more accurate estimates of BM prevalence in young children presenting with seizure in the context of fever. The main question informing this review was: what is the prevalence of BM in young children presenting with: (i) a first “seizure and fever”, (ii) an apparent first simple FS, and (iii) an apparent first complex FS? We also sought to evaluate the utility of routine LP for diagnosis of CNS infections requiring immediate treatment among children with an apparent first FS.
The literature search aimed to identify all studies looking at the prevalence of BM in young children presenting to emergency care with a first seizure in the context of fever. To identify eligible original articles, we searched the following electronic databases from inception to December 2011: MEDLINE via PUBMED, INIST (Scientific and Technical Information Institute) via article@inist, and COCHRANE library. In each electronic database, various combinations of the following search terms were used: “febrile seizure”, “febrile convulsion”, “fever”, “seizure”, “convulsion”, “meningitis”, and “central nervous system infections”. The reference lists of potentially relevant articles and review articles were also screened for additional articles of interest. The detailed search strategy for each electronic database can be found in Text S2.
Explicit a priori inclusion and exclusion criteria were applied. Cohort studies on consecutive, unselected children published before 30 December 2011 were eligible for inclusion in our review if: (1) they reported data on the prevalence of BM in young children who were admitted to the ED or inpatient ward for evaluation of a first “seizure and fever”, an apparent first simple FS, or an apparent first complex FS, (2) their definition of FS (simple or complex) was the same as or very similar to that used most commonly in literature , , (3) they were performed in high-resource countries (studies from low-resource countries were excluded because of the higher prevalence of BM and the different range of conditions (i.e., malaria, HIV related CNS infections, and CNS tuberculosis) that may present as seizure with fever in such countries –); the United Nations list was used to define high-resource countries, and (4) they were written in English or French. When multiple articles reported on a same study population, we included only the most detailed publication that met the inclusion criteria. When studies were identified as containing pertinent data not included in the published article (e.g., studies that included children with a first FS, but did not differentiate between simple FS and complex FS) we contacted the authors to obtain missing data. When a response was not provided and raw data from the original study allowed us to differentiate between two groups of patients, such articles were included in the study; otherwise, they were excluded. Because of the likelihood that small studies would have overestimated event outcome rates, studies including less than 20 patients were arbitrary excluded. Case reports, review articles, editorials, comments, and clinical guidelines were excluded.
Study selection process
Study selection was carried out independently by two of the reviewers (AN, AM) in two rounds. First, all titles and abstracts of the identified citations were screened. Second, potentially relevant articles were reviewed in their entirety. Each investigator made a recommendation for inclusion or exclusion of single articles. Any disagreements were resolved by consensus.
Outcome measures and definitions
The main outcome measure was the prevalence of BM among children with a first seizure in the context of fever. Other outcome measures were: (1) the overall prevalence of CNS infections (including meningitis (viral and bacterial), “possible BM”, Herpes simplex virus (HSV) encephalitis, and encephalitis of other aetiologies) among children with a first seizure in the context of fever, and (2) the utility of routine LP for diagnosis of CNS infections requiring immediate treatment in children with an apparent first FS. CNS infections requiring immediate treatment included BM and HSV encephalitis. BM was defined by positive cerebrospinal fluid (CSF) culture results for a relevant bacterial pathogen, positive Gram's staining of CSF with a negative CSF culture, CSF pleocytosis with positive blood culture of a relevant bacterial pathogen, or CSF pleocytosis with positive latex agglutination test of CSF. “Possible BM” was defined as CSF pleocytosis with negative Gram's staining and negative results of bacterial cultures of blood and CSF in a child pre-treated with antibiotics. Pleocytosis was defined by ≥5 white blood cells per µL . HSV encephalitis was defined by a positive CSF HSV polymerase chain reaction (PCR). The definition of FS was introduced in the introduction section (see above). “Seizure and fever” included any seizure in a child with fever of any cause . Apparent FS was defined as an event that fulfilled the criteria used to designate an FS, but for whom the possibility of CNS infections were not yet ruled out by LP or follow-up. Simple FS was defined as a primary generalized seizure lasting less than 15 minutes and not recurring within 24 hours . Complex FS was defined on one or more of the following features: a partial (focal) onset or showing focal features during the seizure, prolonged duration (greater than 10–15 minutes), and recurrence within 24 hours or within same febrile illness . For each study, the period preceding routine Hib and S. pneumoniae vaccine implementation was considered as the pre-vaccine era and the subsequent period as the post-vaccine era , –.
Data extraction and quality assessment
Data extraction and quality assessment of the included studies were performed independently by two of the reviewers (AN, AM) using a standardized data collection form. Any disagreement was resolved unanimously by discussion. The following information was extracted from each study: first author, country of origin, dates of enrollment, type (prospective versus retrospective), clinical setting, number of patients, inclusion and exclusion criteria, method of outcome ascertainment (clinical grounds, clinical grounds with follow-up, LP), number of outcome events (CNS infections, BM), definition of FS (for studies on simple FS, and complex FS), number and outcome of patients pre-treated with antibiotics, and enrollment era (pre-vaccine versus post-vaccine era). We adapted a quality assessment system for prevalence articles . Each article was reviewed to determine whether: (1) study design was appropriate for obtaining prevalence estimates, (2) sample was representative of the population of interest on key characteristics (age, medical condition), (3) a clear and acceptable definition of the outcome of interest was provided, and (4) the outcome ascertainment methods were well defined and adequate. We assessed these quality indicators separately for each article; a total quality score was not calculated .
Data synthesis and analysis
All analyses were performed separately for each group of patients: children with a “seizure and fever”, those with an apparent first simple FS, and those with an apparent first complex FS. Studies, patients, outcome ascertainment methods, and outcome data were summarized using basic descriptive statistics (simple counts and proportions). The average prevalences of CNS infections and BM were calculated by dividing the number of children with target outcome by the number of children included in the study. As the prevalence of BM was the main focus of our study, the pooled proportion of BM was synthesized by using meta-analytic techniques. Proportions were first transformed into a quantity based upon the Freeman-Tukey variant of the arcsine square root transformed proportion  suitable for the usual fixed and random effects summaries . The pooled proportion was calculated as the back-transform of the weighted mean of the transformed proportions, using inverse arcsine variance weights for the fixed effects model and DerSimonian-Laird weights for the random effects model. Statistical heterogeneity across studies was measured using the Cochran chi-square test (p<0.1 considered significant) and assessed visually using the Galbraith plot of heterogeneity. To determine the percentage of heterogeneity across studies, the I-squared (I2) statistic was calculated . To evaluate the weight of particular articles on the pooled estimates, we performed influence analysis. This method recalculates the pooled prevalence estimate omitting one study at a time. Meta-regression analysis and publication bias assessment were not performed because of the small number of studies. The utility of routine LP for diagnosis of CNS infections requiring immediate treatment was estimated as “number needed to test” (NNT), denoting the number of patients who need to undergo an LP to detect one case of such infections. NNT was calculated by dividing the number of patients included in the study by the number of patients with CNS infections requiring immediate treatment who were diagnosed after routine LP. All statistical tests were performed using STATA version 11.1 (Stata Corp, College Station, Texas) and StatsDirect version 2.7.9 (StatsDirect, Ltd, UK).
Seven hundred thirty four articles were identified through electronic database search, of which 20 were deemed relevant for full text review. Seventeen additional relevant articles were identified by screening the reference lists of relevant articles and review articles. Of these 37 articles, 14 met all criteria for inclusion. The flow of articles through the literature search and screening process, and the reasons for exclusion of the identified studies are illustrated in Figure 1.
Children with “seizure and fever”
Outcome data for children with a “seizure and fever” were derived from 5 studies evaluating 1996 patients , , –. Characteristics, results, and methodological details of the studies are summarized in Table 1 and Table 2. In all studies, samples were representative of the population of interest on key characteristics (Table 2). Outcomes were well documented by LP or clinical examination with follow-up in 100% of patients in all studies (Table 2). Out of 1996 children, 77 were diagnosed with CNS infections, of whom 41 with BM. Of the 41 children with BM, 4 were <6 months and 37>6 months of age. The diagnosis of BM might be suspected from clinical examination in 95% (n = 35) of children >6 months (Table 1). The overall average prevalence of CNS infections was 3.9% (range 2.3 to 7.4%). The pooled prevalence of BM using a random effects model was 2.6% (95% CI 0.9–5.1) (Figure 2). When the individual studies were combined in a meta-analysis, there was significant heterogeneity among the estimates for the prevalence of BM from the studies (I2 = 87%, p<0.001). The Galbraith plot identified the populations studied by Offringa et al. and Joffe et al. as the sources of this heterogeneity , ; however, exclusion of one or both of these studies did not significantly alter the results (p = 0.4). Furthermore, influence analysis showed that no study, including these two, significantly impacted the pooled prevalence estimate.
Children with an apparent simple FS
Seven studies with an aggregate of 1869 patients reported outcome data for children with a first simple FS –. Characteristics, results, and methodological details of the studies are summarized in Table 1 and Table 2. In all studies, samples were representative of the population of interest on key characteristics (Table 2). Outcomes were well documented by LP or clinical examination with follow-up in 100% of patients in 3 studies , , . A formal meta-analysis was not done because of the excess number of zero events in the sample. The overall average prevalence of CNS infections in children 6 to 72 months of age (4 studies, n = 911) was 0.2% (range 0.0 to 1.4%), and that of BM (5 studies, n = 1109) was 0.2% (range 0.0 to 1.0%). The NNT to detect one case of CNS infection requiring immediate treatment in children 6 to 72 months of age was 1109 (of the two BM cases, one was diagnosed after routine LP and the other had a normal CSF after routine LP, but a repeated LP 24 hours later showed BM) (Table 1).
Children with an apparent complex FS
Outcome data for children with a first complex FS were derived from two studies with an aggregate of 718 patients , . Characteristics, results and methodological details of the studies are summarized in Table 1 and Table 2. In all studies, samples were representative of the population of interest on key characteristics (Table 2). The overall average prevalence of CNS infections was 2.2% (range 0.5 to 2.9%). The pooled prevalence of BM using a fixed effects model was 0.6% (95% CI 0.2–1.5). When the individual studies were combined in a meta-analysis, there was no significant heterogeneity among the estimates for the prevalence of BM from the studies (I2 = 0.0%, p = 0.9). The NNT to detect one case of CNS infection requiring immediate treatment was 180 (Table 1).
Our study is the first systematic review and meta-analysis that attempts to quantify the risk of BM in different subgroups of children with seizure in the context of fever, and to evaluate the utility of routine LP in children with an apparent first FS. The overall risk of BM was low, ranging from 0.2% in children with an apparent first simple FS to 2.6% in those with a first “seizure and fever”. The utility of routine LP for diagnosis of CNS infections requiring immediate treatment among children with an apparent first FS was low: the NNT was 1109 in children with an apparent first simple FS, and 180 in those with an apparent first complex FS.
BM may present as a seizure associated with fever –. In our study, 2.6% of children with “seizure and fever” were found to have BM; the diagnosis of BM might be suspected from clinical examination in 95% of children (Table 1). This figure illustrates the ability of clinical examination to identify almost all children with a first “seizure and fever” who are most likely to benefit from LP, thereby avoiding unnecessary routine LP. However, given the retrospective nature of the majority of studies, this finding will need further clinical validation.
The fear of missing BM has led some authors to advocate routine LP in infants presenting with an apparent simple FS , , . In our study, we were not able to assess the utility of routine LP in infants, because 4 of the 7 studies (totalizing 654 children) did not specify the number of such children – (the three other studies included 303 infants –) (Table 1). In Trainor et al.'s study, 13% of the included children were infants . Extrapolation of these data suggests that approximately 85 of the 654 children were infants. Therefore, the total number of infants with an apparent first simple FS in our study could be estimated to be 388 (of whom 223 were enrolled during pre-vaccine era and 165 during post-vaccine era) (Table 1). As only one infant with an apparent first simple FS was found to have CNS infection requiring immediate treatment in our study (Table 1), the NNT to detect one case of such infections would be 388 in these children. This finding challenges the utility of routine LP in infants with an apparent simple FS. However, BM is a rapidly progressive disease and will show itself within a short time, when an LP can be done. A careful clinical observation is necessary in these children during the first few hours after the seizure.
Since introduction of Hib and S. pneumoniae vaccines, the incidence of BM has dramatically decreased among young children , . As such, the AAP revised its guideline for the neurodiagnostic evaluation of the child with a simple FS in 2011 . The updated AAP guideline no longer supports the routine LP in fully immunized against both Hib and S. Pneumoniae infants who present with a simple FS, but recommends an LP as optional in any infants with a simple FS who are missing immunizations or have an indeterminate immunization status . In our study, 223 infants with an apparent first simple FS were enrolled during pre-vaccine era, of whom one was found to have BM (Table 1). The NNT to detect one case of CNS infection requiring immediate treatment could be estimated to be 223 in these children. Accordingly, a careful clinical observation before deciding to perform an LP appears also to be acceptable in infants who are missing immunizations or have an indeterminate immunization status.
Complex FS has been cited as a risk factor for BM , . Accordingly, routine LP has been often recommended in the evaluation of such children , . However, in our study, the utility of routine LP in children with an apparent first complex FS was low (NNT = 180). In addition, routine LP following an FS is not devoid of risk. It is sometimes associated with post-LP syndrome which occurs often when LPs have been unnecessary with unpleasant headaches and sometimes vomiting , and occasionally may result in fatal cerebellar coning or introduction of organisms from the bloodstream into the CSF . Theses facts challenge the utility of this strategy in such children. Thus routine LP, based solely on complex features of seizure, seems not to be necessary. Short hospital admission for close observation could also be a reasonable strategy in such children.
Our study has some limitations that need to be considered. First, the majority of studies included in our review were retrospective and subject to the biases inherent in that method. Second, in the cases of simple FS and complex FS, BM was ruled out based on clinical grounds alone without CSF examination or clinical examination with follow-up in 43% and 18% of patients, respectively. It is possible that a substantial number of these children who were sent home but who developed meningitis and went to a different hospital may have been missed. However, limiting the analysis to cases in which BM was ruled out by CSF examination or clinical examination with follow-up did not alter the results significantly. Third, the risk of BM in children pre-treated with antibiotics is a diagnostic challenge. Such antimicrobial pre-treatment has been reported to obscure the initial signs and symptoms of BM, and to decrease the rate of CSF culture positivity . As the majority of studies included in this review did not provide data separately on children pre-treated with antibiotics, we calculated the prevalence of BM regardless of absence or presence of antimicrobial pre-treatment. This limits applicability of our results to patients pre-treated with antibiotics. Fourth, while the majority of FS are benign and safely managed at home by the general practitioner , more complicated cases of FS with a higher risk of serious infections are referred to hospital. As our study included children seen in the ED or admitted to inpatient ward, our findings may only be applicable to children from similar settings. Finally, the present study was restricted to children from high resource countries, so our findings may not be generalizable to those from low-resource countries.
The risk of BM presenting solely as an apparent FS is very low, whatever age or features of seizure (simple or complex). Hence, performing routine LP in the absence of any other signs and symptoms suggestive of BM is likely to be of low utility in febrile, young children presenting with a first seizure. Our findings challenge the utility of routine LP in children with an apparent complex FS, and in infants, fully immunized or not against both Hib and S. pneumoniae, with an apparent simple FS. A careful clinical observation during the first few hours after the seizure could be an acceptable strategy in such children.
PRISMA 2009 Checklist.
Detailed search strategy for each electronic database.
Agreed with manuscript results and conclusions: AN FD VH IP AB AM. Contributed to data interpretation and critical revision of the report: AN FD VH IP AB AM. Conceived and designed the experiments: AN FD AM. Performed the experiments: AN VH AM. Analyzed the data: AN VH AB AM. Contributed reagents/materials/analysis tools: AN IP AM. Wrote the paper: AN FD AM.
- 1. Pallin DJ, Goldstein JN, Moussally JS, Pelletier AJ, Green AR, et al. (2008) Seizure visits in US emergency departments: epidemiology and potential disparities in care. Int J Emerg Med 1: 97–105.
- 2. Armon K, Stephenson T, Gabriel V, MacFaul R, Eccleston P, et al. (2001) Determining the common medical presenting problems to an accident and emergency department. Arch Dis Child 84: 390–392.
- 3. Fetveit A (2008) Assessment of febrile seizures in children. Eur J Pediatr 167: 17–27.
- 4. American Academy of Pediatrics. Provisional Committee on Quality Improvement, Subcommittee on Febrile Seizures (1996) Practice parameter: the neurodiagnostic evaluation of the child with a first simple febrile seizure. Pediatrics 97: 769–772; discussion 773–765.
- 5. Dubos F, De la Rocque F, Levy C, Bingen E, Aujard Y, et al. (2008) Sensitivity of the bacterial meningitis score in 889 children with bacterial meningitis. J Pediatr 152: 378–382.
- 6. Ratcliffe JC, Wolf SM (1977) Febrile convulsions caused by meningitis in young children. Ann Neurol 1: 285–286.
- 7. Samson JH, Apthorp J, Finley A (1969) Febrile seizures and purulent meningitis. JAMA 210: 1918–1919.
- 8. Joffe A, McCormick M, DeAngelis C (1983) Which children with febrile seizures need lumbar puncture? A decision analysis approach. Am J Dis Child 137: 1153–1156.
- 9. Offringa M, Beishuizen A, Derksen-Lubsen G, Lubsen J (1992) Seizures and fever: can we rule out meningitis on clinical grounds alone? Clin Pediatr (Phila) 31: 514–522.
- 10. Jaffe M, Bar-Joseph G, Tirosh E (1981) Fever and convulsions–indications for laboratory investigations. Pediatrics 67: 729–731.
- 11. Gerber MA, Berliner BC (1981) The child with a ‘simple’ febrile seizure. Appropriate diagnostic evaluation. Am J Dis Child 135: 431–433.
- 12. McIntyre PB, Gray SV, Vance JC (1990) Unsuspected bacterial infections in febrile convulsions. Med J Aust 152: 183–186.
- 13. Kinsella JB, O'Sullivan P, McShane DP (1995) The role of the middle ear and tonsil in the etiology of febrile convulsions. Int J Pediatr Otorhinolaryngol 32: 153–157.
- 14. Trainor JL, Hampers LC, Krug SE, Listernick R (2001) Children with first-time simple febrile seizures are at low risk of serious bacterial illness. Acad Emerg Med 8: 781–787.
- 15. Kimia AA, Capraro AJ, Hummel D, Johnston P, Harper MB (2009) Utility of lumbar puncture for first simple febrile seizure among children 6 to 18 months of age. Pediatrics 123: 6–12.
- 16. Shaked O, Pena BM, Linares MY, Baker RL (2009) Simple febrile seizures: are the AAP guidelines regarding lumbar puncture being followed? Pediatr Emerg Care 25: 8–11.
- 17. Seltz LB, Cohen E, Weinstein M (2009) Risk of bacterial or herpes simplex virus meningitis/encephalitis in children with complex febrile seizures. Pediatr Emerg Care 25: 494–497.
- 18. Kimia A, Ben-Joseph EP, Rudloe T, Capraro A, Sarco D, et al. (2010) Yield of lumbar puncture among children who present with their first complex febrile seizure. Pediatrics 126: 62–69.
- 19. Wears RL, Luten RC, Lyons RG (1986) Which laboratory tests should be performed on children with apparent febrile convulsions? An analysis and review of the literature. Pediatr Emerg Care 2: 191–196.
- 20. Offringa M, Moyer VA (2001) An evidence-based approach to managing seizures associated with fever in children. West J Med 175: 254–259.
- 21. Carroll W, Brookfield D (2002) Lumbar puncture following febrile convulsion. Arch Dis Child 87: 238–240.
- 22. Hom J, Medwid K (2011) The low rate of bacterial meningitis in children, ages 6 to 18 months, with simple febrile seizures. Acad Emerg Med 18: 1114–1120.
- 23. Heijbel J, Blom S, Bergfors PG (1980) Simple febrile convulsions. A prospective incidence study and an evaluation of investigations initially needed. Neuropadiatrie 11: 45–56.
- 24. Lorber J, Sunderland R (1980) Lumbar puncture in children with convulsions associated with fever. Lancet 1: 785–786.
- 25. Rossi LN, Brunelli G, Duzioni N, Rossi G (1986) Lumbar puncture and febrile convulsions. Helv Paediatr Acta 41: 19–24.
- 26. Rutter N, Smales OR (1977) Role of routine investigations in children presenting with their first febrile convulsion. Arch Dis Child 52: 188–191.
- 27. Surpure JS (1980) Febrile convulsions. What happens to the infant admitted to the hospital. Clin Pediatr (Phila) 19: 361–362.
- 28. Anderson AB, Desisto MJ, Marshall PC, Dewitt TG (1989) Duration of fever prior to onset of a simple febrile seizure: a predictor of significant illness and neurologic course. Pediatr Emerg Care 5: 12–15.
- 29. Lee P, Verrier Jones K (1991) Urinary tract infection in febrile convulsions. Arch Dis Child 66: 1287–1290.
- 30. Dawson KP, Capaldi N (1994) The initial hospital management of childhood febrile convulsions: an audit. J Qual Clin Pract 14: 111–114.
- 31. van Stuijvenberg M, Moll HA, Steyerberg EW, van Gijssel EN, Moons KG, et al. (1998) The duration of febrile seizures and peripheral leukocytosis. J Pediatr 133: 557–558.
- 32. Teach SJ, Geil PA (1999) Incidence of bacteremia, urinary tract infections, and unsuspected bacterial meningitis in children with febrile seizures. Pediatr Emerg Care 15: 9–12.
- 33. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) The PRISMA Group (2009) Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6: e1000097 doi:10.1371/journal.pmed.1000097.
- 34. Waruiru C, Appleton R (2004) Febrile seizures: an update. Arch Dis Child 89: 751–756.
- 35. Laman M, Manning L, Hwaiwhange I, Vince J, Aipit S, et al. (2010) Lumbar puncture in children from an area of malaria endemicity who present with a febrile seizure. Clin Infect Dis 51: 534–540.
- 36. Joshi Batajoo R, Rayamajhi A, Mahaseth C (2008) Children with first episode of fever with seizure: is lumbar puncture necessary? J Nepal Med Assoc 47: 109–112.
- 37. Silber E, Sonnenberg P, Ho KC, Koornhof HJ, Eintracht S, et al. (1999) Meningitis in a community with a high prevalence of tuberculosis and HIV infection. J Neurol Sci 162: 20–26.
- 38. Leary PM, Cremin BJ, Daubenton JD, Peter JC (1993) A Study of African Children with Prolonged Focal Seizure and a Specific CT Scan Finding. J Trop Pediatr 39: 176–178.
- 39. WHO (1997) Antimicrobial and support therapy for bacterial meningitis in children. Available: http://www.who.int/csr/resources/publications/meningitis/whoemcbac982.pdf. Accessed 15 August 2012.
- 40. Joint Working Group of the Research Unit of the Royal College of Physicians and the British Paediatric Association (1991) Guidelines for the management of convulsions with fever. BMJ 303: 634–636.
- 41. European Center for Disease Prevention and Control (2012) Vaccination schedules. Available: http://www.ecdc.europa.eu/en/activities/surveillance/euvac/schedules/Pages/schedules.aspx. Accessed 15 August 2012.
- 42. Australian Government Department of Human Services (2007) Current Immunisation Schedule: The National Immunisation Program (NIP) Schedule (0–4 Years). Available: http://www.medicareaustralia.gov.au/provider/patients/acir/schedule.jsp. Accessed 15 August 2012.
- 43. WHO (2012) Immunization surveillance, assessment and monitoring. Available: http://www.who.int/immunization_monitoring/en/. Accessed 15 August 2012.
- 44. Richardson WS, Wilson MC, Guyatt GH, Cook DJ, Nishikawa J (1999) Users' guides to the medical literature: XV. How to use an article about disease probability for differential diagnosis. Evidence-Based Medicine Working Group. JAMA 281: 1214–1219.
- 45. Juni P, Witschi A, Bloch R, Egger M (1999) The hazards of scoring the quality of clinical trials for meta-analysis. JAMA 282: 1054–1060.
- 46. Stuart A, Ord JK (1994) Kendall's Advanced Theory of Statistics, 6th ed. London: Edward Arnold.
- 47. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7: 177–188.
- 48. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327: 557–560.
- 49. Armon K, Stephenson T, MacFaul R, Hemingway P, Werneke U, et al. (2003) An evidence and consensus based guideline for the management of a child after a seizure. Emerg Med J 20: 13–20.
- 50. Adams WG, Deaver KA, Cochi SL, Plikaytis BD, Zell ER, et al. (1993) Decline of childhood Haemophilus influenzae type b (Hib) disease in the Hib vaccine era. JAMA 269: 221–226.
- 51. Whitney CG, Farley MM, Hadler J, Harrison LH, Bennett NM, et al. (2003) Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med 348: 1737–1746.
- 52. American Academy of Pediatrics: Subcommittee on Febrile Seizures (2011) Neurodiagnostic evaluation of the child with a simple febrile seizure. Pediatrics 127: 389–394.
- 53. Al-Eissa YA (1995) Lumbar puncture in the clinical evaluation of children with seizures associated with fever. Pediatr Emerg Care 11: 347–350.
- 54. Shinnar S, Glauser TA (2002) Febrile seizures. J Child Neurol 17 Suppl 1S44–52.
- 55. Atabaki S, Ochsenschlager D (1999) Post-lumbar puncture headache and backache in pediatrics: a case series and demonstration of magnetic resonance imaging findings. Arch Pediatr Adolesc Med 153: 770–773.
- 56. Fischer GW, Brenz RW, Alden ER, Beckwith JB (1975) Lumbar punctures and meningitis. Am J Dis Child 129: 590–592.
- 57. Nigrovic LE, Malley R, Macias CG, Kanegaye JT, Moro-Sutherland DM, et al. (2008) Effect of antibiotic pretreatment on cerebrospinal fluid profiles of children with bacterial meningitis. Pediatrics 122: 726–730.
- 58. Verburgh ME, Bruijnzeels MA, van der Wouden JC, van Suijlekom-Smit LW, van der Velden J, et al. (1992) Incidence of febrile seizures in The Netherlands. Neuroepidemiology 11: 169–172.