Original Studies
Prevalence and Predictors of Bacterial Meningitis in Young Infants With Fever Without a Source Elena Martinez, MD,* Santiago Mintegi, MD, PhD,*† Begoña Vilar, MD,‡ Maria Jesus Martinez, MD,§ Amaia Lopez, MD,* Estibaliz Catediano, MD, and Borja Gomez, MD*† Background: Classical criteria differ when performing cerebrospinal fluid (CSF) analysis in infants younger than 90 days with fever without a source (FWS). Our objectives were to analyze the prevalence and microbiology of bacterial meningitis in this group and its prevalence in relation to clinical and laboratory risk factors. Methods: This is a substudy of a prospective registry including all infants of this age with FWS seen between September 2003 and August 2013 in a Pediatric Emergency Department of a Tertiary Teaching Hospital. Results: Lumbar puncture was performed in 639 (27.0%) of the 2362 infants with FWS seen, the rate being higher in not well-appearing infants [60.9% vs. 25.7%; odds ratio (OR), 4.49] and in those ≤21 days old (70.1% vs. 20.4%; OR, 9.14). Eleven infants were diagnosed with bacterial meningitis: 9 were ≤21 days old (prevalence 2.8% vs. 0.1%; OR, 30.42) and 5 were not well-appearing infants (5.7% vs. 0.2%; OR, 23.06). Bacteria isolated were Streptococcus agalactiae (3), Escherichia coli (3), Listeria monocytogenes (3), Streptococcus pneumoniae (1) and Neisseria meningitidis (1). None of the 1975 well-appearing infants >21 days old were diagnosed with bacterial meningitis, regardless of whether biomarkers were altered. Conclusions: In infants younger than 90 days with FWS, performing CSF analysis for ruling out bacterial meningitis must be strongly considered in not well-appearing infants and in those ≤21 days old. The recommendation of systematically performing CSF analysis in well-appearing infants 22–90 days of age on the basis of analytical criteria alone must be reevaluated. Key Words: fever, young infants, bacterial meningitis, biomarkers (Pediatr Infect Dis J 2015;34:494–498)
I
nfants younger than 90 days with fever without a source (FWS) have a greater risk of serious bacterial infections (SBIs) than did older infants, and this has traditionally meant a more interventionist approach in these patients, including hospital admission and performance of tests.1,2 In recent decades, however, the prevalence of SBIs has decreased, thanks to better intrapartum antibiotic prophylaxis3,4 and the early diagnosis of urinary malformations, as a result of prenatal ultrasound scanning.5,6 Recent studies have also demonstrated a shift in the bacterial species involved. Listeria monocytogenes, classically considered as one of the main pathogens in young infants, has lost importance, and Streptococcus agalactiae and Escherichia coli are now being responsible for 75% of cases with a microbiological diagnosis of bacteremia in this age group.7,8 Accepted for publication November 7, 2014. From the *Pediatric Emergency Department, Cruces University Hospital, Barakaldo, Spain; †University of the Basque Country, Bilbao, Spain; ‡Department of Microbiology and §Department of Neuropediatrics, Cruces University Hospital, Barakaldo, Spain. The authors have no funding or conflicts of interest to disclose. Address for correspondence: Borja Gomez, MD, Pediatric Emergency Department, Cruces University Hospital, Plaza de Cruces s/n, Barakaldo, Spain. E-mail: [email protected]. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0891-3668/15/3405-0494 DOI: 10.1097/INF.0000000000000629
494 | www.pidj.com
Regarding the current epidemiology of meningitis in this age group, recent publications have reported a significant decrease in the prevalence of L. monocytogenes as the causal agent of neonatal meningitis,9 in parallel with the trend observed in bacteremia. Cerebrospinal fluid (CSF) analysis to rule out meningitis as the cause of the fever is performed more frequently in these pediatric patients given their milder clinical manifestations of invasive infections. Interestingly, 2 of the classical sets of risk criteria most used in practice differ regarding when to indicate this type of analysis. Specifically, although the Philadelphia criteria recommend the systematic use of lumbar puncture (LP) to classify patients as low risk,10 Rochester criteria do not include this test.11 Since the publication of these sets of criteria, the indications for CSF analysis have gradually changed, and currently, the decision is taken on a caseby-case basis, depending on the general condition of the patient, their age and the results of other tests. For instance, the American College of Emergency Physicians has recommended its use on a systematic basis only in children younger than 1 month.1 Moreover, regardless of the guidelines adopted, there is a mixed adherence to any of these criteria in practice,12,13 even in the specific subgroup of febrile neonates.14 With regard to the recommendation of performing CSF analysis based on blood test results alone, various authors have demonstrated that white blood cell count is not useful for identifying infants who are suitable candidates for this test.15 On the other hand, new markers, such as C-reactive protein (CRP)16,17 and more recently procalcitonin (PCT),17–19 have proven to be more accurate for identifying young infants with SBIs, especially, in cases of invasive bacterial infection (IBI). Their use in the management of febrile infants could also modify the recommendations regarding the use of LP. The objectives of this study were (a) to assess the prevalence of bacterial meningitis in infants younger than 90 days with FWS seen in a Pediatric Emergency Department (PED); (b) to identify the bacterial species involved and (c) to analyze this IBI in relation to clinical and laboratory findings classically considered to indicate high risk.
METHODS We performed a secondary analysis of data from an observational prospective study that included all the infants younger than 90 days with FWS who attended our PED over a 10-year period (September 2003–August 2013). Our PED is in a tertiary hospital, and each year receives around 55,000 children younger than 14 years, with approximately 2300 of them infants younger than 90 days.
Protocol for the Management of Infants Younger Than 90 Days With FWS Our protocol for the management of infants younger than 90 days with FWS recommends the following in all cases: collection of sterile urine sample, urine dipstick test, complete blood cell count, measurement of CRP and PCT levels (the latter since November 2007) and blood and urine cultures. During influenza epidemics, we also perform a rapid test for influenza.
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
We recommend LP to collect CSF for analysis including cell count, Gram staining and bacterial and viral cultures in •• all infants who are not well-appearing or have clinical manifestations suggestive of bacterial meningitis, •• infants younger than 21 days and •• infants with abnormal blood test results (leukocytes 15,000/µL, neutrophils >10,000/µL, CRP >20 mg/L or PCT ≥0.5 ng/mL).
Microbiologic Processing of the CSF Samples Gram staining was performed on CSF samples prepared by cytocentrifugation. Samples were then centrifuged, and 2–3 drops of the sediment were cultured in chocolate agar and incubated at 35°C for 48 hours with a CO2 concentration of 5–7%; in addition, 0.5 mL of the sediment was added to a liquid medium (brain–heart infusion broth). In infants younger than 1 month, we also cultured a sample in a blood agar for the identification of pathogens, such as S. agalactiae and L. monocytogenes.
Data Collection In 2003, we started an electronic registry in our PED, including all the infants younger than 90 days seen with FWS. The database includes demographic characteristics (age, sex and month of visit), medical history, duration of fever in hours, maximum body temperature measured at home and on arrival at the PED, general appearance of the child, findings on physical examination, laboratory tests performed and their results, treatment administered, time under observation, diagnosis on discharge from the PED, site of care (hospital admission or outpatient management) and diagnosis after culture results. In addition, patients’ evolution was monitored: those who were admitted to a ward, through the review of medical records, and those who were sent home, through telephone interviews conducted by medical residents after a period of training. The monitoring phone call is made within a month after the PED visit to know the evolution of the patient after being discharged and confirm the resolution of the fever. If several phone calls were unsuccessful, medical reports from posterior visits to our PED during the following months for any complaint were reviewed, to confirm that no mention to any possible complication of that febrile episode was recorded.
Definitions •• FWS: axillary or rectal temperature measured at home ≥38°C or rectal temperature measured in the PED ≥38°C in an infant in whom after anamnesis and physical examination it is not possible to identify the source of the fever (including normal chest auscultation and an absence of signs of acute otitis media, bone and joint infection and soft tissue infection) in accordance with the diagnostic codes published by the Spanish Society of Pediatric Emergencies.20 If an infant was attended more than once within the same febrile episode, only the first visit was registered, although the final diagnosis registered could have been identified in another visit. •• Well-appearing: normal findings according to the pediatric assessment triangle (PAT) (appearance, work of breathing and circulation to skin),21 as assessed by the doctor attending the child within an hour of arriving at the PED. •• Bacterial meningitis: detection of a bacterial pathogen in the CSF (with or without associated pleocytosis) or in the blood culture with pleocytosis associated. © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Bacterial Meningitis in Infants
•• Pleocytosis: the cut-off points used in our PED are >25 cells/ mm3 in infants younger than 1 month, >10 cells/mm3 in infants 1–2 months and >5 cells/mm3 in those older than 2 months.22 •• Bacteremia: growth of a true bacterial pathogen in a blood culture. Growth of Staphylococcus epidermidis, Propionibacterium acnes, diphtheroids or any other bacterium classically considered a contaminant if isolated from previously healthy immunocompetent infants (with no history of heart disease, ventriculoperitoneal shunt, indwelling catheter or other prosthetic devices) was not considered as bacteremia for the purpose of this study.
Statistical Analysis The statistical analysis was performed using the IBM SPSS Statistics for Windows (version 19, Armonk, NY). Data are expressed as means, confidence intervals (CIs) and standard deviations for quantitative variables and as numbers and percentage for categorical variables. Continuous variables were compared with the Student’s t test and categorical variables with the χ2 test or Fisher exact test. The level of significance was set at P ≤ 0.05. The study was approved by the Research Committee of the PED. Given that all the data were extracted from a database in which the patient identities were anonymous and inclusion in the registry did not implynany additional intervention, informed consent was not required.
RESULTS During the study period, 2362 infants younger than 90 days with FWS were seen in the department. Table 1 reports descriptive statistics for main epidemiological variables and complementary tests performed. An LP was performed on the first visit in 639 (27.0%) infants, and samples of CSF analysis were successfully obtained in 603 cases (94.3%). The rate of LPs was higher in infants who were classified as not well-appearing on arrival to the PED [60.9% vs. 25.7% among those classified as well-appearing infants; odds ratio (OR), 4.49 (95% CI: 2.83–7.15)] and in those ≤21 days of age [70.1% vs. 20.4% in those >21 days of age; OR, 9.14 (95% CI: 6.95–12.02)]. The LP was more frequently performed during the first 2 years of the study period (36.5% vs. 24.7% during the past 8 years; P < 0.001), according to the changes in the recommendations of clinical guidelines, with no statistical variations in relation to the annual rates of not well-appearing infants and of infants ≤21 days of age. TABLE 1. Epidemiological and Clinical Characteristics and Complementary Tests Performed to the 2362 Infants Attended Age (days) Sex (male) Previously healthy Evolution time of fever (hours) Temperature upon arrival to the PED (°C) Well-appearing Urine dipstick Blood culture PCR WBC count Urine culture PCT (since November 2007) Lumbar puncture CSF culture
50 ± 23 1356 (57.4%) 2058 (87.1%) 5 (2–12) 38.3 ± 0.7 2275 (96.3%) 2319 (98.1%) 2182 (92.3%) 2179 (92.2%) 2166 (91.7%) 1711 (72.4%) 1248 (52.8%) 639 (27.0%) 603 (25.5%)
Evolution time is expressed as median and interquartile range; age and temperature are expressed as mean ± standard deviation. PCR indicates polymerase chain reaction; WBC, white blood cells.
www.pidj.com | 495
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
Martinez et al
Eleven children (0.46%; 95% CI: 0.19–0.73%) were diagnosed with bacterial meningitis, 3 of them meeting criteria for sepsis. All but one were born at term. No significant variation was observed in relation to the annual prevalence of bacterial meningitis throughout the study period. The bacteria isolated were S. agalactiae (3), E. coli (3), L. monocytogenes (3), S. pneumoniae (1) and Neisseria meningitidis (1). Table 2 lists the characteristics of these 11 patients: 9 were ≤21 days of age [prevalence of 2.8% vs. 0.1% in those who are older than 21 days; OR, 30.42 (95% CI: 6.13–204.60)] and 5 children were not wellappearing on arrival to the PED [prevalence of 5.7% vs. 0.2% in those classified as well-appearing infants; OR, 23.06 (95% CI: 5.97–87.44)]. Description of the anterior fontanelle was recorded in 10 of the 11 patients; it was described as normotensive in all of them. Out of the 1975 well-appearing infants >21 days of age, blood tests were performed in 1810 (91.6%) and LP for CSF analysis in 378 (19.1%), the rate being higher among the 793 infants who presented any abnormal blood test result [29.1% vs. 14.4% in those with all normal blood test results; OR, 2.44 (95% CI: 1.92–3.09)]. None of these infants were diagnosed with bacterial meningitis, regardless of whether they had abnormal blood test results. Figure 1 illustrates the distribution of the cases of bacterial meningitis by general condition and age. As bacteremia may lead to meningitis, a similar flowchart was created to illustrate the distribution of the cases of bacteremia by general condition and age (Fig. 2). There were 28 positive blood cultures among the 1975 well-appearing infants >21 days of age (1.4%; 95% CI: 0.8–1.9%): 17 occult bacteremias and 11 urinary tract infections with bacteremia associated. Eight of them were initially managed as outpatients without performing a LP. All of them did well.
Globally, 1302 well-appearing infants >21 days of age were initially managed as outpatients without performing a LP. The 114 infants who were presented with leukocyturia (suspected urinary tract infection) and 20 of the 1188 infants with a normal urine dipstick (1.6%) received antibiotic treatment. None of them were subsequently diagnosed with bacterial meningitis in later visits, and previously referred patients’ monitoring ruled out a bad outcome. Of them, 135 (10.4%) presented a posterior visit and 23 of them (1.7%) were admitted. All of them did well. One of the 11 patients with bacterial meningitis (S. agalactiae) presented seizures within the first 24 hours. During admission, head ultrasounds were performed in all these 11 patients; in 1, with meningitis by N. meningitidis, grade I bilateral subependymal hemorrhage was detected, but this was considered to be an incidental finding unrelated to the infection and was observed to resolve in subsequent ultrasounds. Auditory-evoked potential analysis was also performed in 10 infants, and the findings were normal in all cases. None of the children died. To date, after discharge, we have identified 1 case of motor developmental delay in a patient with meningitis by E. coli.
DISCUSSION In the light of our results, the systematic analysis of CSF on the basis of blood test results alone does not seem appropriate beyond the neonatal period in well-appearing infants younger than 90 days with FWS. The management of infants younger than 90 days with FWS has been changing over recent decades, slowly moving toward a less interventionist approach. Changes in the prevalence
TABLE 2. Clinical Characteristics and Laboratory Tests Results of Patients With Bacterial Meningitis Isolated Bacteria Positive Cultures Escherichia coli, blood and CSF E. coli, blood and CSF Listeria monocytogenes, CSF Streptococcus agalactiae, blood L. monocytogenes, CSF L. monocytogenes, CSF E. coli, blood and urine S. agalactiae, blood and CSF S. agalactiae, CSF Streptococcus pneumoniae, blood and CSF Neisseria meningitidis, blood and CSF
Sex
Previously Age Healthy (days)
Duration Rectal of Fever Temperature General Number of Cells (hours) at PED (°C) Appearance in CSF
PCT (ng/mL)
CRP (mg/L)
Leukocyte Count (cells/µL)
Absolute Neutrophil Count (cells/µL)
Male
Yes
4
5
38.2
W-a
7380
4.3
53
2200
800
Female
Yes
6
0‡
38.5
W-a*†
0.3
4
6600
4400
Male
Yes
8
1
38.1
W-a
5 → 5760 after 48 h 3690
1
25,000
16,000
Female
Yes
9
1
38.3
W-a
266
1.9
7
20,100
16,884
Female
Yes
11
1
38.8
W-a
6750
7
32,200
22,540
Female
Yes
13
8
39.2
W-a
6859
3
15,600
7100
Male
Yes
15
1
38.3
Not W-a*
70
70
13,900
9035
Male
Yes
15
0‡
39.6
Not W-a*
620
89.3
25
4100
2400
Female
Yes
17
12
39.9
Not W-a
18
2.5
26
5400
2600
Female
Yes
23
0‡
38.4
Not W-a
4
272
26,700
21,627
Male
Preterm infant
49
2
39.1
Not W-a
3
9
6400
2500
0.1
1.1
*Patients who met the criteria for sepsis. †Clinical worsening after admission. ‡Fever first detected at PED, chief complaint other than fever (fussiness, …) W-a indicates well-appearing.
496 | www.pidj.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
FIGURE 1. Prevalence of bacterial meningitis by general appearance and age.
FIGURE 2. Prevalence of bacteremia by general appearance and age. of different bacterial species involved in invasive infections in this age group,4,7,8 the poor performance of the laboratory parameters traditionally used (such as leukocytosis)15 and the introduction of new biomarkers16–19 have seen the classical recommendations of systematic admission, and empirical antibiotic therapy gives way to a more individualized management depending on several factors, such as age and other blood test results. There is no consensus on the indications for LP, either in the classical risk classification criteria or in the most recent management protocols. Specifically, the Philadelphia criteria recommend CSF analysis in all infants younger than 90 days with FWS,10 whereas the Rochester criteria do not include it among the recommended tests.11 Regardless of the guidelines adopted, given the milder clinical manifestations of young infants, in clinical practice it is more common to request this test the younger the infant is and in patients with altered blood tests, even without clinical manifestations suggestive of meningitis. Our group previously reported that management of previously healthy well-appearing infants younger than 3 months with FWS in an Emergency Department is possible to be done without routine LP.23 In our PED, the first step we take as part of the triage is to apply the PAT. This tool has proven to be useful for identifying unstable patients and physiopathologically classifying pediatric patients.24 Out of the 11 infants with bacterial meningitis in our sample, 5 were classified on arrival as not well-appearing (2 of them with clinical signs of sepsis). It is essential to have an adequate method for identifying patients who, before any laboratory tests, give the impression of being clinically unstable. The PAT has been found to be an easy-to-learn useful tool,21 which can be used by clinicians with different levels of experience. With regard to age, 9 of the cases of bacterial meningitis involved infants ≤21 days of age. The 2 cases identified in infants >21 days of age were a 23-day-old infant with a meningitis by S. pneumoniae7 (close to the age cut-off point) and a 49-day-old infant with a meningitis by N. meningitidis, but who was born preterm—after 35 © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Bacterial Meningitis in Infants
weeks of gestation (a corrected age of 14 days). Both were not wellappearing on arrival to the PED. Other recent studies agree that the secondary cut-off point in terms of age to identify a subgroup of patients with greater risk of IBI should be 21 days. In another substudy using this registry, Garcia et al26 found that the rate of IBI in infants aged 15–21 days was 33.3% (95% CI: 23.7–42.9%), similar to that among those aged 7–14 days (31.9%, 95% CI: 21.1–42.7%) and greater than that of infants >21 days old (18.3%, 95% CI: 16.3–20.3%). Out of the 2362 infants studied, 1975 (83.6%) were wellappearing and >21 days old. The decision to perform LP in these patients was influenced by the results of other blood tests used in our center (leukocyte count, absolute neutrophil count, CRP and/or PCT levels). None of these children were diagnosed with bacterial meningitis. These results point to a need for more individualized management, reconsidering the recommendation for systematic LPs in these children based on blood test results alone. Although recent studies do show that the most recently discovered biomarkers, such as CRP and especially PCT, identify infants at greater risk of developing bacteremia, even in well-appearing infants,19 it seems that they are not so useful for detecting bacterial meningitis when the clinical picture is not suggestive of this form of infection. Recently published papers have described a fall in the prevalence of L. monocytogenes in this age group.7–9 In our population, we diagnosed the same number of cases of meningitis due to this species as due to S. agalactiae and to E. coli. However, 2 of the 3 cases of meningitis due to L. monocytogenes occurred during an epidemic outbreak in an external maternity ward, there being only 1 case in the 6 years since then. Our results based on this registry agree with other recent studies,7,8 in that this bacterial species currently plays a relatively minor role in cases of bacteremia diagnosed in this age group. Furthermore, none of the 51 positive blood cultures identified in our registry were classified as positive due to L. monocytogenes, the most commonly isolated bacterial species being E. coli (21), followed by S. agalactiae (9) and Enterococcus faecalis (5). The outcome of patients with bacterial meningitis was better than that classically described in the literature both in terms of mortality and short-term sequelae. However, given the relatively small sample size, we should be cautious about drawing conclusions on this matter. In our series, none of the infants died, a mortality rate due to bacterial meningitis of 10–15% have been reported in this age group.9 Just 1 patient had seizures within the first 24 hours after arrival, a finding which has been reported in up to 34% of cases of neonatal meningitis.9 Although cranial ultrasound was performed in all 11 patients, there were abnormal findings in only 1 case, and these findings were considered unrelated to the meningitis. Specifically, the imaging detected a grade I bilateral subependymal hemorrhage, which did not require a specific treatment and resolved during subsequent monitoring. Previous studies also noted a high rate of intracranial complications in the imaging tests.27 Ten of the 11 patients underwent auditory-evoked potential analysis, all the results being normal. In the medium term, 1 case of a motor developmental delay has been detected. Although the assessment of complications was not the objective of the study, and hence, the follow-up period varied between patients, we have not found any other complications in any of the 11 patients to date. It is likely that the low rate of complications identified during the imaging tests and subsequent check-ups is due to the fact that most of the other studies have included patients diagnosed with meningitis in the maternity ward and in the neonatal unit. In our study, we only included patients diagnosed in the PED, these being infants having been discharged after birth. We should recognize some limitations of our study. First, we did not carry out CSF analysis in all the infants studied. Although this could mean that the prevalence of bacterial meningitis is underestimated, the subsequent monitoring undertaken suggests that it is not likely that cases were missed. Furthermore, it would not currently www.pidj.com | 497
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
Martinez et al
be ethical to perform this test on all infants only for research purposes. Second, we had a very low incidence of bacterial meningitis in our population (0.46%), so our results and recommendations cannot be extrapolated to communities that have a higher incidence of this bacterial infection. In the same way, our PED includes an observation unit that allows us to observe these well-appearing young febrile infants for up to 24 hours. Probably, performing a LP in these well-appearing patients is more frequent than in those PEDs without an observation unit in which the patient must be either discharged or admitted to the pediatric inpatient ward. Third, as we have described, the prevalence of bacterial meningitis is clearly influenced by the general appearance. In our study, we classified patients using the PAT, and it may not be possible to generalize our findings to centers where this tool is not used or where the staff has less experience in the initial assessment of pediatric patients. Finally, in this study, we did not analyze cases of viral meningitis. Based on the same patient registry, we know that this diagnosis was more common in our population (5.5%). We have previously reported that the symptoms of patients with confirmed viral meningitis are similar to those of patients with no intracranial infection, and that leukocyte count and CRP are not good predictors of this condition.28
CONCLUSIONS When managing infants younger than 90 days with FWS, we should seriously consider the use of CSF analysis to rule out bacterial meningitis in the subsets of infants that are ≤21 days old or are not well-appearing. On the other hand, the systematic performance of LPs in well-appearing infants older than 21 days on the basis of laboratory test results alone should be reevaluated. REFERENCES 1. ACEP Clinical Policies Committee and the Clinical Policies Subcommittee on Pediatric Fever. Clinical policy for children younger than three years presenting to the Emergency Department with fever. Ann Emerg Med. 2003;42:530–545. 2. Ishimine P. Fever without source in children 0 to 36 months of age. Pediatr Clin North Am. 2006;53:167–194. 3. Schrag SJ, Zywicki S, Farley MM, et al. Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis. N Engl J Med. 2000;342:15–20. 4. Phares CR, Lynfield R, Farley MM. Epidemiology of invasive group B streptococcal disease in the United States, 1999–2005. JAMA. 2008;299:2055–65. 5. Raynor BD. Routine ultrasound in pregnancy. Clin Obstet Gynecol. 2003;46:882–889. 6. Sairam S, Al-Habib A, Sasson S, et al. Natural history of fetal hydronephrosis diagnosed on mid-trimester ultrasound. Ultrasound Obstet Gynecol. 2001;17:191–196. 7. Greenhow TL, Hung YY, Herz AM. Changing epidemiology of bacteremia in infants aged 1 week to 3 months. Pediatrics. 2012;129:e590–e596. 8. Biondi E, Evans R, Mischler M, et al. Epidemiology of bacteremia in febrile infants in the United States. Pediatrics. 2013;132:990–996. 9. Gaschignard J, Levy C, Romain O, et al. Neonatal Bacterial Meningitis: 444 Cases in 7 Years. Pediatr Infect Dis J. 2011;30:212–217.
498 | www.pidj.com
10. Baker MD, Bell LM, Avner JR. Outpatient management without antibiotics of fever in selected infants. N Engl J Med. 1993;329:1437–1441. 11. Jaskiewicz JA, McCarthy CA, Richardson AC, et al. Febrile infants at low risk for serious bacterial infection – an appraisal of the Rochester criteria and implications for management. Febrile Infant Collaborative Study Group. Pediatrics. 1994;94:390–396. 12. Goldman RD, Scolnik D, Chauvin-Kimoff L, et al; Fever in Infants Group Research, Pediatric Emergency Research of Canada. Practice variations in the treatment of febrile infants among pediatric emergency physicians. Pediatrics. 2009;124:439–445. 13. Meehan WP 3rd, Fleegler E, Bachur RG. Adherence to guidelines for managing the well-appearing febrile infant: assessment using a case-based, interactive survey. Pediatr Emerg Care. 2010;26:875–880. 14. Jain S, Cheng J, Alpern ER, et al. Management of febrile neonates in US pediatric emergency departments. Pediatrics. 2014;133:187–195. 15. Bonsu BK, Harper MB. Utility of the peripheral blood white blood cell count for identifying sick young infants who need lumbar puncture. Ann Emerg Med. 2003;41:206–214. 16. Andreola B, Bressan S, Callegaro S, et al. Procalcitonin and C-reactive protein as diagnostic markers of severe bacterial infections in febrile infants and children in the emergency department. Pediatr Infect Dis J. 2007;26:672–677. 17. Olaciregui I, Hernández U, Muñoz JA, et al. Markers that predict serious bacterial infection in infants under 3 months of age presenting with fever of unknown origin. Arch Dis Child. 2009;94:501–505. 18. Maniaci V, Dauber A, Weiss S, et al. Procalcitonin in young febrile infants for the detection of serious bacterial infections. Pediatrics. 2008;122:701–710. 19. Gomez B, Bressan S, Mintegi S, et al. Diagnostic value of procalcitonin in well-appearing young febrile infants. Pediatrics. 2012;130:815–822. 20. Grupo de trabajo de codificación diagnóstica de la Sociedad de Urgencias de Pediatría de la AEP. Codificación diagnóstica en Urgencias de Pediatría. Available at: http://seup.org/pdf_public/gt/mejora_codificacion.pdf. Accessed January 7, 2014. 21. Dieckmann RA, Brownstein D, Gausche-Hill M. The pediatric assessment triangle: a novel approach for the rapid evaluation of children. Pediatr Emerg Care. 2010;26:312–315. 22. Meehan WP 3rd, Bachur RG. Predictors of cerebrospinal fluid pleocytosis in febrile infants aged 0 to 90 days. Ped Emerg Care. 2008 May;24:287–93. 23. Mintegi S, Benito J, Astobiza E, et al. Well appearing young infants with fever without known source in the emergency department: are lumbar punctures always necessary? Eur J Emerg Med. 2010;17:167–169. 24. Horeczko T, Enriquez B, McGrath NE, et al. The pediatric assessment triangle: accuracy of its application by nurses in the triage of children. J Emerg Nurs. 2013;39:182–189. 25. Poehling KA, Talbot TR, Griffin MR, et al. Invasive pneumococcal disease among infants before and after introduction of pneumococcal conjugate vaccine. JAMA. 2006;295:1668–1674. 26. Garcia S, Mintegi S, Gomez B, et al. Is 15 days an appropriate cut-off age for considering serious bacterial infection in the management of febrile infants? Pediatr Infect Dis J. 2012;31:455–458. 27. Schmidt H, Cohrs S, Heinemann T, et al. Sleep disorders are long-term sequelae of both bacterial and viral meningitis. J Neurol Neurosurg Psychiatry. 2006;77:554–558. 28. Gomez B, Mintegi S, Rubio MC, et al. Clinical and analytical characteristics and short-term evolution of enteroviral meningitis in young infants presenting with fever without source. Pediatr Emerg Care. 2012;28:518–523.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
Prevalence and Predictors of Bacterial Meningitis in Young Infants With Fever Without a Source Elena Martinez, MD,* Santiago Mintegi, MD, PhD,*† Begoña Vilar, MD,‡ Maria Jesus Martinez, MD,§ Amaia Lopez, MD,* Estibaliz Catediano, MD, and Borja Gomez, MD*† Background: Classical criteria differ when performing cerebrospinal fluid (CSF) analysis in infants younger than 90 days with fever without a source (FWS). Our objectives were to analyze the prevalence and microbiology of bacterial meningitis in this group and its prevalence in relation to clinical and laboratory risk factors. Methods: This is a substudy of a prospective registry including all infants of this age with FWS seen between September 2003 and August 2013 in a Pediatric Emergency Department of a Tertiary Teaching Hospital. Results: Lumbar puncture was performed in 639 (27.0%) of the 2362 infants with FWS seen, the rate being higher in not well-appearing infants [60.9% vs. 25.7%; odds ratio (OR), 4.49] and in those ≤21 days old (70.1% vs. 20.4%; OR, 9.14). Eleven infants were diagnosed with bacterial meningitis: 9 were ≤21 days old (prevalence 2.8% vs. 0.1%; OR, 30.42) and 5 were not well-appearing infants (5.7% vs. 0.2%; OR, 23.06). Bacteria isolated were Streptococcus agalactiae (3), Escherichia coli (3), Listeria monocytogenes (3), Streptococcus pneumoniae (1) and Neisseria meningitidis (1). None of the 1975 well-appearing infants >21 days old were diagnosed with bacterial meningitis, regardless of whether biomarkers were altered. Conclusions: In infants younger than 90 days with FWS, performing CSF analysis for ruling out bacterial meningitis must be strongly considered in not well-appearing infants and in those ≤21 days old. The recommendation of systematically performing CSF analysis in well-appearing infants 22–90 days of age on the basis of analytical criteria alone must be reevaluated. Key Words: fever, young infants, bacterial meningitis, biomarkers (Pediatr Infect Dis J 2015;34:494–498)
I
nfants younger than 90 days with fever without a source (FWS) have a greater risk of serious bacterial infections (SBIs) than did older infants, and this has traditionally meant a more interventionist approach in these patients, including hospital admission and performance of tests.1,2 In recent decades, however, the prevalence of SBIs has decreased, thanks to better intrapartum antibiotic prophylaxis3,4 and the early diagnosis of urinary malformations, as a result of prenatal ultrasound scanning.5,6 Recent studies have also demonstrated a shift in the bacterial species involved. Listeria monocytogenes, classically considered as one of the main pathogens in young infants, has lost importance, and Streptococcus agalactiae and Escherichia coli are now being responsible for 75% of cases with a microbiological diagnosis of bacteremia in this age group.7,8 Accepted for publication November 7, 2014. From the *Pediatric Emergency Department, Cruces University Hospital, Barakaldo, Spain; †University of the Basque Country, Bilbao, Spain; ‡Department of Microbiology and §Department of Neuropediatrics, Cruces University Hospital, Barakaldo, Spain. The authors have no funding or conflicts of interest to disclose. Address for correspondence: Borja Gomez, MD, Pediatric Emergency Department, Cruces University Hospital, Plaza de Cruces s/n, Barakaldo, Spain. E-mail: [email protected]. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0891-3668/15/3405-0494 DOI: 10.1097/INF.0000000000000629
494 | www.pidj.com
Regarding the current epidemiology of meningitis in this age group, recent publications have reported a significant decrease in the prevalence of L. monocytogenes as the causal agent of neonatal meningitis,9 in parallel with the trend observed in bacteremia. Cerebrospinal fluid (CSF) analysis to rule out meningitis as the cause of the fever is performed more frequently in these pediatric patients given their milder clinical manifestations of invasive infections. Interestingly, 2 of the classical sets of risk criteria most used in practice differ regarding when to indicate this type of analysis. Specifically, although the Philadelphia criteria recommend the systematic use of lumbar puncture (LP) to classify patients as low risk,10 Rochester criteria do not include this test.11 Since the publication of these sets of criteria, the indications for CSF analysis have gradually changed, and currently, the decision is taken on a caseby-case basis, depending on the general condition of the patient, their age and the results of other tests. For instance, the American College of Emergency Physicians has recommended its use on a systematic basis only in children younger than 1 month.1 Moreover, regardless of the guidelines adopted, there is a mixed adherence to any of these criteria in practice,12,13 even in the specific subgroup of febrile neonates.14 With regard to the recommendation of performing CSF analysis based on blood test results alone, various authors have demonstrated that white blood cell count is not useful for identifying infants who are suitable candidates for this test.15 On the other hand, new markers, such as C-reactive protein (CRP)16,17 and more recently procalcitonin (PCT),17–19 have proven to be more accurate for identifying young infants with SBIs, especially, in cases of invasive bacterial infection (IBI). Their use in the management of febrile infants could also modify the recommendations regarding the use of LP. The objectives of this study were (a) to assess the prevalence of bacterial meningitis in infants younger than 90 days with FWS seen in a Pediatric Emergency Department (PED); (b) to identify the bacterial species involved and (c) to analyze this IBI in relation to clinical and laboratory findings classically considered to indicate high risk.
METHODS We performed a secondary analysis of data from an observational prospective study that included all the infants younger than 90 days with FWS who attended our PED over a 10-year period (September 2003–August 2013). Our PED is in a tertiary hospital, and each year receives around 55,000 children younger than 14 years, with approximately 2300 of them infants younger than 90 days.
Protocol for the Management of Infants Younger Than 90 Days With FWS Our protocol for the management of infants younger than 90 days with FWS recommends the following in all cases: collection of sterile urine sample, urine dipstick test, complete blood cell count, measurement of CRP and PCT levels (the latter since November 2007) and blood and urine cultures. During influenza epidemics, we also perform a rapid test for influenza.
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
We recommend LP to collect CSF for analysis including cell count, Gram staining and bacterial and viral cultures in •• all infants who are not well-appearing or have clinical manifestations suggestive of bacterial meningitis, •• infants younger than 21 days and •• infants with abnormal blood test results (leukocytes 15,000/µL, neutrophils >10,000/µL, CRP >20 mg/L or PCT ≥0.5 ng/mL).
Microbiologic Processing of the CSF Samples Gram staining was performed on CSF samples prepared by cytocentrifugation. Samples were then centrifuged, and 2–3 drops of the sediment were cultured in chocolate agar and incubated at 35°C for 48 hours with a CO2 concentration of 5–7%; in addition, 0.5 mL of the sediment was added to a liquid medium (brain–heart infusion broth). In infants younger than 1 month, we also cultured a sample in a blood agar for the identification of pathogens, such as S. agalactiae and L. monocytogenes.
Data Collection In 2003, we started an electronic registry in our PED, including all the infants younger than 90 days seen with FWS. The database includes demographic characteristics (age, sex and month of visit), medical history, duration of fever in hours, maximum body temperature measured at home and on arrival at the PED, general appearance of the child, findings on physical examination, laboratory tests performed and their results, treatment administered, time under observation, diagnosis on discharge from the PED, site of care (hospital admission or outpatient management) and diagnosis after culture results. In addition, patients’ evolution was monitored: those who were admitted to a ward, through the review of medical records, and those who were sent home, through telephone interviews conducted by medical residents after a period of training. The monitoring phone call is made within a month after the PED visit to know the evolution of the patient after being discharged and confirm the resolution of the fever. If several phone calls were unsuccessful, medical reports from posterior visits to our PED during the following months for any complaint were reviewed, to confirm that no mention to any possible complication of that febrile episode was recorded.
Definitions •• FWS: axillary or rectal temperature measured at home ≥38°C or rectal temperature measured in the PED ≥38°C in an infant in whom after anamnesis and physical examination it is not possible to identify the source of the fever (including normal chest auscultation and an absence of signs of acute otitis media, bone and joint infection and soft tissue infection) in accordance with the diagnostic codes published by the Spanish Society of Pediatric Emergencies.20 If an infant was attended more than once within the same febrile episode, only the first visit was registered, although the final diagnosis registered could have been identified in another visit. •• Well-appearing: normal findings according to the pediatric assessment triangle (PAT) (appearance, work of breathing and circulation to skin),21 as assessed by the doctor attending the child within an hour of arriving at the PED. •• Bacterial meningitis: detection of a bacterial pathogen in the CSF (with or without associated pleocytosis) or in the blood culture with pleocytosis associated. © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Bacterial Meningitis in Infants
•• Pleocytosis: the cut-off points used in our PED are >25 cells/ mm3 in infants younger than 1 month, >10 cells/mm3 in infants 1–2 months and >5 cells/mm3 in those older than 2 months.22 •• Bacteremia: growth of a true bacterial pathogen in a blood culture. Growth of Staphylococcus epidermidis, Propionibacterium acnes, diphtheroids or any other bacterium classically considered a contaminant if isolated from previously healthy immunocompetent infants (with no history of heart disease, ventriculoperitoneal shunt, indwelling catheter or other prosthetic devices) was not considered as bacteremia for the purpose of this study.
Statistical Analysis The statistical analysis was performed using the IBM SPSS Statistics for Windows (version 19, Armonk, NY). Data are expressed as means, confidence intervals (CIs) and standard deviations for quantitative variables and as numbers and percentage for categorical variables. Continuous variables were compared with the Student’s t test and categorical variables with the χ2 test or Fisher exact test. The level of significance was set at P ≤ 0.05. The study was approved by the Research Committee of the PED. Given that all the data were extracted from a database in which the patient identities were anonymous and inclusion in the registry did not implynany additional intervention, informed consent was not required.
RESULTS During the study period, 2362 infants younger than 90 days with FWS were seen in the department. Table 1 reports descriptive statistics for main epidemiological variables and complementary tests performed. An LP was performed on the first visit in 639 (27.0%) infants, and samples of CSF analysis were successfully obtained in 603 cases (94.3%). The rate of LPs was higher in infants who were classified as not well-appearing on arrival to the PED [60.9% vs. 25.7% among those classified as well-appearing infants; odds ratio (OR), 4.49 (95% CI: 2.83–7.15)] and in those ≤21 days of age [70.1% vs. 20.4% in those >21 days of age; OR, 9.14 (95% CI: 6.95–12.02)]. The LP was more frequently performed during the first 2 years of the study period (36.5% vs. 24.7% during the past 8 years; P < 0.001), according to the changes in the recommendations of clinical guidelines, with no statistical variations in relation to the annual rates of not well-appearing infants and of infants ≤21 days of age. TABLE 1. Epidemiological and Clinical Characteristics and Complementary Tests Performed to the 2362 Infants Attended Age (days) Sex (male) Previously healthy Evolution time of fever (hours) Temperature upon arrival to the PED (°C) Well-appearing Urine dipstick Blood culture PCR WBC count Urine culture PCT (since November 2007) Lumbar puncture CSF culture
50 ± 23 1356 (57.4%) 2058 (87.1%) 5 (2–12) 38.3 ± 0.7 2275 (96.3%) 2319 (98.1%) 2182 (92.3%) 2179 (92.2%) 2166 (91.7%) 1711 (72.4%) 1248 (52.8%) 639 (27.0%) 603 (25.5%)
Evolution time is expressed as median and interquartile range; age and temperature are expressed as mean ± standard deviation. PCR indicates polymerase chain reaction; WBC, white blood cells.
www.pidj.com | 495
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
Martinez et al
Eleven children (0.46%; 95% CI: 0.19–0.73%) were diagnosed with bacterial meningitis, 3 of them meeting criteria for sepsis. All but one were born at term. No significant variation was observed in relation to the annual prevalence of bacterial meningitis throughout the study period. The bacteria isolated were S. agalactiae (3), E. coli (3), L. monocytogenes (3), S. pneumoniae (1) and Neisseria meningitidis (1). Table 2 lists the characteristics of these 11 patients: 9 were ≤21 days of age [prevalence of 2.8% vs. 0.1% in those who are older than 21 days; OR, 30.42 (95% CI: 6.13–204.60)] and 5 children were not wellappearing on arrival to the PED [prevalence of 5.7% vs. 0.2% in those classified as well-appearing infants; OR, 23.06 (95% CI: 5.97–87.44)]. Description of the anterior fontanelle was recorded in 10 of the 11 patients; it was described as normotensive in all of them. Out of the 1975 well-appearing infants >21 days of age, blood tests were performed in 1810 (91.6%) and LP for CSF analysis in 378 (19.1%), the rate being higher among the 793 infants who presented any abnormal blood test result [29.1% vs. 14.4% in those with all normal blood test results; OR, 2.44 (95% CI: 1.92–3.09)]. None of these infants were diagnosed with bacterial meningitis, regardless of whether they had abnormal blood test results. Figure 1 illustrates the distribution of the cases of bacterial meningitis by general condition and age. As bacteremia may lead to meningitis, a similar flowchart was created to illustrate the distribution of the cases of bacteremia by general condition and age (Fig. 2). There were 28 positive blood cultures among the 1975 well-appearing infants >21 days of age (1.4%; 95% CI: 0.8–1.9%): 17 occult bacteremias and 11 urinary tract infections with bacteremia associated. Eight of them were initially managed as outpatients without performing a LP. All of them did well.
Globally, 1302 well-appearing infants >21 days of age were initially managed as outpatients without performing a LP. The 114 infants who were presented with leukocyturia (suspected urinary tract infection) and 20 of the 1188 infants with a normal urine dipstick (1.6%) received antibiotic treatment. None of them were subsequently diagnosed with bacterial meningitis in later visits, and previously referred patients’ monitoring ruled out a bad outcome. Of them, 135 (10.4%) presented a posterior visit and 23 of them (1.7%) were admitted. All of them did well. One of the 11 patients with bacterial meningitis (S. agalactiae) presented seizures within the first 24 hours. During admission, head ultrasounds were performed in all these 11 patients; in 1, with meningitis by N. meningitidis, grade I bilateral subependymal hemorrhage was detected, but this was considered to be an incidental finding unrelated to the infection and was observed to resolve in subsequent ultrasounds. Auditory-evoked potential analysis was also performed in 10 infants, and the findings were normal in all cases. None of the children died. To date, after discharge, we have identified 1 case of motor developmental delay in a patient with meningitis by E. coli.
DISCUSSION In the light of our results, the systematic analysis of CSF on the basis of blood test results alone does not seem appropriate beyond the neonatal period in well-appearing infants younger than 90 days with FWS. The management of infants younger than 90 days with FWS has been changing over recent decades, slowly moving toward a less interventionist approach. Changes in the prevalence
TABLE 2. Clinical Characteristics and Laboratory Tests Results of Patients With Bacterial Meningitis Isolated Bacteria Positive Cultures Escherichia coli, blood and CSF E. coli, blood and CSF Listeria monocytogenes, CSF Streptococcus agalactiae, blood L. monocytogenes, CSF L. monocytogenes, CSF E. coli, blood and urine S. agalactiae, blood and CSF S. agalactiae, CSF Streptococcus pneumoniae, blood and CSF Neisseria meningitidis, blood and CSF
Sex
Previously Age Healthy (days)
Duration Rectal of Fever Temperature General Number of Cells (hours) at PED (°C) Appearance in CSF
PCT (ng/mL)
CRP (mg/L)
Leukocyte Count (cells/µL)
Absolute Neutrophil Count (cells/µL)
Male
Yes
4
5
38.2
W-a
7380
4.3
53
2200
800
Female
Yes
6
0‡
38.5
W-a*†
0.3
4
6600
4400
Male
Yes
8
1
38.1
W-a
5 → 5760 after 48 h 3690
1
25,000
16,000
Female
Yes
9
1
38.3
W-a
266
1.9
7
20,100
16,884
Female
Yes
11
1
38.8
W-a
6750
7
32,200
22,540
Female
Yes
13
8
39.2
W-a
6859
3
15,600
7100
Male
Yes
15
1
38.3
Not W-a*
70
70
13,900
9035
Male
Yes
15
0‡
39.6
Not W-a*
620
89.3
25
4100
2400
Female
Yes
17
12
39.9
Not W-a
18
2.5
26
5400
2600
Female
Yes
23
0‡
38.4
Not W-a
4
272
26,700
21,627
Male
Preterm infant
49
2
39.1
Not W-a
3
9
6400
2500
0.1
1.1
*Patients who met the criteria for sepsis. †Clinical worsening after admission. ‡Fever first detected at PED, chief complaint other than fever (fussiness, …) W-a indicates well-appearing.
496 | www.pidj.com
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
FIGURE 1. Prevalence of bacterial meningitis by general appearance and age.
FIGURE 2. Prevalence of bacteremia by general appearance and age. of different bacterial species involved in invasive infections in this age group,4,7,8 the poor performance of the laboratory parameters traditionally used (such as leukocytosis)15 and the introduction of new biomarkers16–19 have seen the classical recommendations of systematic admission, and empirical antibiotic therapy gives way to a more individualized management depending on several factors, such as age and other blood test results. There is no consensus on the indications for LP, either in the classical risk classification criteria or in the most recent management protocols. Specifically, the Philadelphia criteria recommend CSF analysis in all infants younger than 90 days with FWS,10 whereas the Rochester criteria do not include it among the recommended tests.11 Regardless of the guidelines adopted, given the milder clinical manifestations of young infants, in clinical practice it is more common to request this test the younger the infant is and in patients with altered blood tests, even without clinical manifestations suggestive of meningitis. Our group previously reported that management of previously healthy well-appearing infants younger than 3 months with FWS in an Emergency Department is possible to be done without routine LP.23 In our PED, the first step we take as part of the triage is to apply the PAT. This tool has proven to be useful for identifying unstable patients and physiopathologically classifying pediatric patients.24 Out of the 11 infants with bacterial meningitis in our sample, 5 were classified on arrival as not well-appearing (2 of them with clinical signs of sepsis). It is essential to have an adequate method for identifying patients who, before any laboratory tests, give the impression of being clinically unstable. The PAT has been found to be an easy-to-learn useful tool,21 which can be used by clinicians with different levels of experience. With regard to age, 9 of the cases of bacterial meningitis involved infants ≤21 days of age. The 2 cases identified in infants >21 days of age were a 23-day-old infant with a meningitis by S. pneumoniae7 (close to the age cut-off point) and a 49-day-old infant with a meningitis by N. meningitidis, but who was born preterm—after 35 © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Bacterial Meningitis in Infants
weeks of gestation (a corrected age of 14 days). Both were not wellappearing on arrival to the PED. Other recent studies agree that the secondary cut-off point in terms of age to identify a subgroup of patients with greater risk of IBI should be 21 days. In another substudy using this registry, Garcia et al26 found that the rate of IBI in infants aged 15–21 days was 33.3% (95% CI: 23.7–42.9%), similar to that among those aged 7–14 days (31.9%, 95% CI: 21.1–42.7%) and greater than that of infants >21 days old (18.3%, 95% CI: 16.3–20.3%). Out of the 2362 infants studied, 1975 (83.6%) were wellappearing and >21 days old. The decision to perform LP in these patients was influenced by the results of other blood tests used in our center (leukocyte count, absolute neutrophil count, CRP and/or PCT levels). None of these children were diagnosed with bacterial meningitis. These results point to a need for more individualized management, reconsidering the recommendation for systematic LPs in these children based on blood test results alone. Although recent studies do show that the most recently discovered biomarkers, such as CRP and especially PCT, identify infants at greater risk of developing bacteremia, even in well-appearing infants,19 it seems that they are not so useful for detecting bacterial meningitis when the clinical picture is not suggestive of this form of infection. Recently published papers have described a fall in the prevalence of L. monocytogenes in this age group.7–9 In our population, we diagnosed the same number of cases of meningitis due to this species as due to S. agalactiae and to E. coli. However, 2 of the 3 cases of meningitis due to L. monocytogenes occurred during an epidemic outbreak in an external maternity ward, there being only 1 case in the 6 years since then. Our results based on this registry agree with other recent studies,7,8 in that this bacterial species currently plays a relatively minor role in cases of bacteremia diagnosed in this age group. Furthermore, none of the 51 positive blood cultures identified in our registry were classified as positive due to L. monocytogenes, the most commonly isolated bacterial species being E. coli (21), followed by S. agalactiae (9) and Enterococcus faecalis (5). The outcome of patients with bacterial meningitis was better than that classically described in the literature both in terms of mortality and short-term sequelae. However, given the relatively small sample size, we should be cautious about drawing conclusions on this matter. In our series, none of the infants died, a mortality rate due to bacterial meningitis of 10–15% have been reported in this age group.9 Just 1 patient had seizures within the first 24 hours after arrival, a finding which has been reported in up to 34% of cases of neonatal meningitis.9 Although cranial ultrasound was performed in all 11 patients, there were abnormal findings in only 1 case, and these findings were considered unrelated to the meningitis. Specifically, the imaging detected a grade I bilateral subependymal hemorrhage, which did not require a specific treatment and resolved during subsequent monitoring. Previous studies also noted a high rate of intracranial complications in the imaging tests.27 Ten of the 11 patients underwent auditory-evoked potential analysis, all the results being normal. In the medium term, 1 case of a motor developmental delay has been detected. Although the assessment of complications was not the objective of the study, and hence, the follow-up period varied between patients, we have not found any other complications in any of the 11 patients to date. It is likely that the low rate of complications identified during the imaging tests and subsequent check-ups is due to the fact that most of the other studies have included patients diagnosed with meningitis in the maternity ward and in the neonatal unit. In our study, we only included patients diagnosed in the PED, these being infants having been discharged after birth. We should recognize some limitations of our study. First, we did not carry out CSF analysis in all the infants studied. Although this could mean that the prevalence of bacterial meningitis is underestimated, the subsequent monitoring undertaken suggests that it is not likely that cases were missed. Furthermore, it would not currently www.pidj.com | 497
The Pediatric Infectious Disease Journal • Volume 34, Number 5, May 2015
Martinez et al
be ethical to perform this test on all infants only for research purposes. Second, we had a very low incidence of bacterial meningitis in our population (0.46%), so our results and recommendations cannot be extrapolated to communities that have a higher incidence of this bacterial infection. In the same way, our PED includes an observation unit that allows us to observe these well-appearing young febrile infants for up to 24 hours. Probably, performing a LP in these well-appearing patients is more frequent than in those PEDs without an observation unit in which the patient must be either discharged or admitted to the pediatric inpatient ward. Third, as we have described, the prevalence of bacterial meningitis is clearly influenced by the general appearance. In our study, we classified patients using the PAT, and it may not be possible to generalize our findings to centers where this tool is not used or where the staff has less experience in the initial assessment of pediatric patients. Finally, in this study, we did not analyze cases of viral meningitis. Based on the same patient registry, we know that this diagnosis was more common in our population (5.5%). We have previously reported that the symptoms of patients with confirmed viral meningitis are similar to those of patients with no intracranial infection, and that leukocyte count and CRP are not good predictors of this condition.28
CONCLUSIONS When managing infants younger than 90 days with FWS, we should seriously consider the use of CSF analysis to rule out bacterial meningitis in the subsets of infants that are ≤21 days old or are not well-appearing. On the other hand, the systematic performance of LPs in well-appearing infants older than 21 days on the basis of laboratory test results alone should be reevaluated. REFERENCES 1. ACEP Clinical Policies Committee and the Clinical Policies Subcommittee on Pediatric Fever. Clinical policy for children younger than three years presenting to the Emergency Department with fever. Ann Emerg Med. 2003;42:530–545. 2. Ishimine P. Fever without source in children 0 to 36 months of age. Pediatr Clin North Am. 2006;53:167–194. 3. Schrag SJ, Zywicki S, Farley MM, et al. Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis. N Engl J Med. 2000;342:15–20. 4. Phares CR, Lynfield R, Farley MM. Epidemiology of invasive group B streptococcal disease in the United States, 1999–2005. JAMA. 2008;299:2055–65. 5. Raynor BD. Routine ultrasound in pregnancy. Clin Obstet Gynecol. 2003;46:882–889. 6. Sairam S, Al-Habib A, Sasson S, et al. Natural history of fetal hydronephrosis diagnosed on mid-trimester ultrasound. Ultrasound Obstet Gynecol. 2001;17:191–196. 7. Greenhow TL, Hung YY, Herz AM. Changing epidemiology of bacteremia in infants aged 1 week to 3 months. Pediatrics. 2012;129:e590–e596. 8. Biondi E, Evans R, Mischler M, et al. Epidemiology of bacteremia in febrile infants in the United States. Pediatrics. 2013;132:990–996. 9. Gaschignard J, Levy C, Romain O, et al. Neonatal Bacterial Meningitis: 444 Cases in 7 Years. Pediatr Infect Dis J. 2011;30:212–217.
498 | www.pidj.com
10. Baker MD, Bell LM, Avner JR. Outpatient management without antibiotics of fever in selected infants. N Engl J Med. 1993;329:1437–1441. 11. Jaskiewicz JA, McCarthy CA, Richardson AC, et al. Febrile infants at low risk for serious bacterial infection – an appraisal of the Rochester criteria and implications for management. Febrile Infant Collaborative Study Group. Pediatrics. 1994;94:390–396. 12. Goldman RD, Scolnik D, Chauvin-Kimoff L, et al; Fever in Infants Group Research, Pediatric Emergency Research of Canada. Practice variations in the treatment of febrile infants among pediatric emergency physicians. Pediatrics. 2009;124:439–445. 13. Meehan WP 3rd, Fleegler E, Bachur RG. Adherence to guidelines for managing the well-appearing febrile infant: assessment using a case-based, interactive survey. Pediatr Emerg Care. 2010;26:875–880. 14. Jain S, Cheng J, Alpern ER, et al. Management of febrile neonates in US pediatric emergency departments. Pediatrics. 2014;133:187–195. 15. Bonsu BK, Harper MB. Utility of the peripheral blood white blood cell count for identifying sick young infants who need lumbar puncture. Ann Emerg Med. 2003;41:206–214. 16. Andreola B, Bressan S, Callegaro S, et al. Procalcitonin and C-reactive protein as diagnostic markers of severe bacterial infections in febrile infants and children in the emergency department. Pediatr Infect Dis J. 2007;26:672–677. 17. Olaciregui I, Hernández U, Muñoz JA, et al. Markers that predict serious bacterial infection in infants under 3 months of age presenting with fever of unknown origin. Arch Dis Child. 2009;94:501–505. 18. Maniaci V, Dauber A, Weiss S, et al. Procalcitonin in young febrile infants for the detection of serious bacterial infections. Pediatrics. 2008;122:701–710. 19. Gomez B, Bressan S, Mintegi S, et al. Diagnostic value of procalcitonin in well-appearing young febrile infants. Pediatrics. 2012;130:815–822. 20. Grupo de trabajo de codificación diagnóstica de la Sociedad de Urgencias de Pediatría de la AEP. Codificación diagnóstica en Urgencias de Pediatría. Available at: http://seup.org/pdf_public/gt/mejora_codificacion.pdf. Accessed January 7, 2014. 21. Dieckmann RA, Brownstein D, Gausche-Hill M. The pediatric assessment triangle: a novel approach for the rapid evaluation of children. Pediatr Emerg Care. 2010;26:312–315. 22. Meehan WP 3rd, Bachur RG. Predictors of cerebrospinal fluid pleocytosis in febrile infants aged 0 to 90 days. Ped Emerg Care. 2008 May;24:287–93. 23. Mintegi S, Benito J, Astobiza E, et al. Well appearing young infants with fever without known source in the emergency department: are lumbar punctures always necessary? Eur J Emerg Med. 2010;17:167–169. 24. Horeczko T, Enriquez B, McGrath NE, et al. The pediatric assessment triangle: accuracy of its application by nurses in the triage of children. J Emerg Nurs. 2013;39:182–189. 25. Poehling KA, Talbot TR, Griffin MR, et al. Invasive pneumococcal disease among infants before and after introduction of pneumococcal conjugate vaccine. JAMA. 2006;295:1668–1674. 26. Garcia S, Mintegi S, Gomez B, et al. Is 15 days an appropriate cut-off age for considering serious bacterial infection in the management of febrile infants? Pediatr Infect Dis J. 2012;31:455–458. 27. Schmidt H, Cohrs S, Heinemann T, et al. Sleep disorders are long-term sequelae of both bacterial and viral meningitis. J Neurol Neurosurg Psychiatry. 2006;77:554–558. 28. Gomez B, Mintegi S, Rubio MC, et al. Clinical and analytical characteristics and short-term evolution of enteroviral meningitis in young infants presenting with fever without source. Pediatr Emerg Care. 2012;28:518–523.
© 2015 Wolters Kluwer Health, Inc. All rights reserved.
