Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Transcription

1 Original article 199 Predictors of occult bacteremia in young febrile children in the era of heptavalent pneumococcal conjugated vaccine Santiago Mintegi, Javier Benito, Jesus Sanchez, Beatriz Azkunaga, Itziar Iturralde and Silvia Garcia Background Changes in the prevalence of pneumococcal occult bacteremia (PnOB) because of pneumococcal conjugated vaccine (PCV-7) may have altered the predictive value of the complete blood count (CBC) for selected patients at risk of having occult bacteremia (OB). Objective To analyze the rate of OB in infants with high fever without source related to pneumococcal vaccination status and to study the yield of the CBC to identify children with OB in the era of PCV-7. Methods Retrospective review of 1586 children of age 3 36 admitted to a Paediatric Emergency Department with a CBC and a blood culture performed as part of their evaluation for fever. Patients with altered urinalysis were excluded. Results Blood culture was positive in 15 (0.9%; pneumococcus 10, 0.6%). Of the 1586 children, 1040 (65.6%) showed less than leukocytes/mm 3 (2 PnOB, 0.19%). Of the 546 children with more than leukocytes/mm 3, eight had a PnOB (positive predictive value = 1.46%, negative predictive value = 99.8%). Of the 1586 children, 1177 (74.2%) showed absolute neutrophil count less than /mm 3 (3 PnOB, 0.25%). Of those 409 with more than neutrophil/mm 3, seven had a PnOB (positive predictive value = 1.71%, negative predictive value = 99.7%). Among the 429 children with at least two doses of PCV-7, one (0.23%) had a PnOB (vs. nine of % with no dose or one dose of PCV-7). Conclusion In the era of PCV-7, rate of PnOB is related to the pneumococcal vaccination status. The yield of the CBC is lower than in the prevaccinal era. Decisions based on CBC must be reconsidered. European Journal of Emergency Medicine 16: c 2009 Wolters Kluwer Health Lippincott Williams & Wilkins. European Journal of Emergency Medicine 2009, 16: Keywords: fever, infant, occult bacteremia Pediatric Emergency Department Cruces Hospital, Basque Country, Spain Correspondence to Santiago Mintegi, MD, Paediatric Emergency Department, Cruces Hospital, Barakaldo, Basque Country 48903, Spain Tel: ; smintegi@hotmail.com Received 29 March 2008 Accepted 4 October 2008 Introduction Fever without source (FWS) is one of the most frequent reasons for visits to the pediatric emergency department [1,2]. In our institution, a tertiary-care teaching hospital in the Basque Country, 3 4% of annual attendances are children below 3 years of age with fever without focal signs [3,4]. Most FWS are because of benign viral infections, which do not require treatment, but, sometimes, they may be because of a severe underlying bacterial infection. Streptococcus pneumoniae is the cause of many invasive diseases in children such as occult bacteremia (OB), meningitis, and bacteremic pneumonia, and it is recognized to be the main cause of OB in young children with FWS. The heptavalent pneumococcal conjugate vaccine (PCV-7) was introduced in the Basque Country in 2001 and resulted in a decline in S. pneumoniae occult bacteremia (PnOB) rate in our area [5], attributable to a decrease in the rate of disease caused by PCV-7 serotypes as had been broadly reported in the United States [6 9]. The introduction of PCV-7 has also altered the distribution of serotypes of S. pneumoniae causing invasive disease in children, including bacteremia [10]. This decline has altered the recommendations of management of infants with FWS [11]. Around 2 5% of infants with OB develop a serious bacterial infection (SBI), including bacterial meningitis [12], if not treated with empirical antibiotics. One of the challenges pediatricians face is to identify wellappearing febrile children aged 3 36 at risk of OB. To identify children with higher risk of OB, clinicians obtained complete blood counts (CBC), as it was known that the PnOB risk is increased with elevated white blood cell counts (WBC) and absolute neutrophil counts (ANC) and, in the prevaccinal era, 80 90% of OB was because of pneumococcus. In this way, it was classically reported that 6% of children with high FWS and more than WBC/mm 3 may have a PnOB [2], and 8% of those with an ANC higher than /mm 3 may also have a PnOB [13], and, in fact, most practice guidelines recommendations c 2009 Wolters Kluwer Health Lippincott Williams & Wilkins DOI: /MEJ.0b013e32831cefc9

2 200 European Journal of Emergency Medicine 2009, Vol 16 4 have incorporated information derived from WBC or ANC. CBC is used by many physicians to screen febrile children for underlying bacterial infection, especially those who are young and have a highly elevated temperature, prolonged fever, or no localizing signs of infection. Changes in PnOB prevalence may have altered the yield of CBC to select patients at risk of having OB [14]. In this way, the value of CBC may be lower than reported and this fact may also alter recommendations of management of infants with fever [15]. The aims of the study were to analyze the rate of OB in children aged 3 36 with high FWS related to pneumococcal vaccinal status and to study the yield of the CBC and C-reactive protein (CRP) to identify children with OB in the era of PCV-7. Patients and methods A retrospective standardized chart review of all children aged 3 36 presenting to the pediatric emergency department (PED) of a tertiary teaching hospital (Hospital Cruces,BasqueCountry,SouthernEurope)wasundertaken between 1 October 2004 and 30 September 2006 who had a CBC and a blood culture (BC) performed as part of their evaluation for fever. In our PED, we strongly recommend to practice CBC, CRP, and BC in children of age 3 36 with FWS with no doses of PCV-7 or incomplete vaccination, and we recommend to our pediatric emergency physicians not to practice those tests if a child has received at least three doses of PCV-7. Final decision (to perform or not to perform the tests) is taken by the physician attending the child (mostly a pediatric emergency physician). We decided not to change the practice of our PED when we underwent thestudy.infact,in25%ofthechildrenwithoutapcv-7 complete vaccination status admitted in our PED, CBC, CRP, and BC are not usually practiced [11]. From our computerized database we identified infants visiting our emergency department diagnosed with FWS and recorded their medical record number, sex, date of birth, date of visit, vaccination status (including PCV-7. Vaccination status is reflected on the card of each patient), registered temperature in degrees centigrade (at home and triage temperature), the absence of a focal bacterial infection (including acute otitis media, pharyngitis, skin infection, or bone and joint infection), ancillary tests practiced (CBC, ANC, and CRP), and organism identified in BC. There were two data abstractors not blinded to the study hypotheses after developing a coding manual. Patient cards were reviewed and data were extracted by a pediatric emergency resident. A pediatric emergency physician then reviewed the patient cards and the extracted data. disagreement was detected between the two physicians. Definitions FWS: axillary or rectal temperature registry at home or rectal registry in the PED of a temperature Z 391C, without catarrhal or respiratory symptoms/signs (such as tachypnea) or a diarrhoeal process, in a patient with a normal physical exam (no signs of acute otitis media, osteoarticular or soft tissues infection, and normal pulmonary auscultation), according to the diagnosis codification of the Spanish Society of Paediatric Emergencies [16]. In our PED, we used to register rectal temperature in young children with FWS, but, in our environment, families usually register axillary temperatures [17] and, in this way, we decided to include children with an axillary temperature Z 391C as well. S. pneumoniae OB: only well-appearing infants aged 3 36 diagnosed with FWS and positive BC with S. pneumoniae were considered as having an S. pneumoniae OB. ANC was computed as the total WBC 10 3 cells/mm 3 multiplied by the sum of the percentage of bands and polymorphonuclear cell count divided by 100. In cases in which an automated differential was computed, ANC was calculated as the WBC multiplied by the percentage of granulocytes divided by 100. BCs characterized as true pathogens included all cultures yielding S. pneumoniae, Neisseria meningitidis, enterococcus, group A streptococcus, or Salmonella species. Growth of Staphylococcus epidermidis, Propionibacterium acnes, and diphtheroids from previously healthy immunocompetent infants (with no history of cardiac disease, ventriculoperitoneal shunts, indwelling catheters, or other prosthetic devices) were categorized as contaminants. Inclusion criteria and exclusion criteria are exposed in Fig. 1. Well-appearing was defined by a Pediatric Emergency physician and with a normal Pediatric Assessment Triangle (Fig. 2) after being attended by the physician. In this way, to be well appearing, the appearance, respiratory, and circulatory items had to be assessed to be normal when arriving in the PED and reflected in the chart of the patient. Infants with abnormalities in the urine dipstick (presence of leucocyturia or nitrituria), where this screening test was utilized, were excluded. In these patients we routinely undergo WBC and BC. Statistical and data analysis Data were expressed as mean and standard deviation for quantitative variables or numbers and percentages for categorical variables. Categorical data were examined by using the w 2 test or the Fisher s test. The SPSS 14.0

3 Predictors of occult bacteremia Mintegi et al. 201 Fig. 1 Child 3 36 of age axillary or rectal registry at home or rectal registry in the PED of a T a 39 C Immunodeficiency Catarrhal or respiratory symptoms/signs (such as tachypnea) or a diarrhoeal process rmal physical exam (no signs of acute otitis media, osteoarticular or soft tissues infection and normal pulmonary auscultation) Pediatric assessment triangle of the APLS (figure 2) assessed by a pediatric emergency faculty physician or a pediatric resident rmal Abnormal CBC and blood culture obtained as part of their evaluation for fever Flow-chart exposing inclusion and exclusion criteria. APLS, Advanced Pediatric Life Support; CBC, complete blood counts; PED, pediatric emergency department. for Windows (SPSS Inc., Chicago, Illinois, USA) was used for all statistical calculations. Statistical significance was defined as P value less than The study was approved by the Research Committee of The Emergency Department. Results Between 1 October 2004 and 30 September 2006, well-appearing infants aged 3 36 were admitted with FWS. Of these, 1586 were previously healthy well-appearing children of age 3 36 with high

4 202 European Journal of Emergency Medicine 2009, Vol 16 4 Fig. 2 Fig. 3 Appearence Respiration Circulation Age distribution Pediatric Assessment Triangle of the Advanced Pediatric Life Support. FWS, non-altered urine dipstick, and in whom physicians decided to practice a CBC and BC to identify patients at risk of having OB. Eight hundred and ninety-seven (56.6%) were male and age distribution and PCV-7 status can be seen in Figs 3 and 4. Of the 1586 episodes, BC was positive in 15. The causative organisms are represented in Table 1. Data of the infants with OB are presented in the Table 2. Of the 1586 infants, 1040 (65.6%) showed less than leukocytes/mm 3 (two had a PnOB, 0.19%). Of those 546 with more than leukocytes/mm 3, eight had a PnOB [positive predictive value (PPV) 1.46%, negative predictive value (NPV) 99.8%, likelihood ratio = 2.4]. Of the 1586 infants, 1177 (74.2%) showed an ANC less than neutrophil/mm 3 (three had a PnOB, 0.25%). Of those 409 with more than neutrophil/mm 3, seven had a PnOB (PPV = 1.71, NPV = 99.7, likelihood ratio = 2.74). Of the 1090 infants who had received no dose or one dose of PCV-7, nine had a PnOB (0.82%, 95% confidence interval: ; vs. 0.23%, 95% confidence interval: , among those with two or more doses of PCV-7, P = 0.17). PPV of WBC greater than /mm 3 in infants with no dose or one dose of PCV-7 was 1.8% (vs. 0.6% among infants with at least two doses). PPV of ANC greater than /mm 3 in infants with no dose or one dose of PCV-7 was 2.0% (vs. 0.9% among infants with at least two doses). CRP was performed in 1572 of the cases (99.1%). CRP values were similar in PnOB group (44.1 ± 47.6 mg/l) and in children with negative BC (37.7 ± 51.5 mg/l, P = 0.68). Forty percent of children with PnOB had a CRP value higher than 20 mg/l. Discussion The results of this study provide additional evidence that the rate of OB in children of age 3 36 with high FWS in the era of PCV-7 is decreasing, it seems to be different in vaccinated children and nonvaccinated children and the PPV of WBC and ANC have decreased because of the decline of PnOB incidence in wellappearing children with FWS. Although bacteremia rates are now lower in wellappearing febrile infants, the evaluation of these patients continues being a challenge for the physician [5,6,8,9,14]. S. pneumoniae is the main cause of OB in young children with FWS. The PCV-7 has resulted in a dramatic decrease in invasive pneumococcal infections, especially meningitis and pneumococcal bacteremia, particularly in children below 2 years of age [6,8,9]. Although PCV-7 have not been included in the official vaccination schedule in our area, the current coverage with PCV-7 in children below 2 years of age is approximately 50% in Basque Country [11]. Despite this small PCV-7 coverage, we have recently reported a 58% decrease in the rate of PnOB among children younger than 36 with FWS attended in our PED [5]. The results of this study confirm this earlier finding. We have found that the rate of PnOB in infants with FWS is higher in children not vaccinated or with only one dose of vaccine than in children who had received two or more doses of the PCV-7. In a recent study, Carstairs et al. [14] found similar results. They did not find any case of PnOB in infants who had received PCV-7 and the PnOB rate of 2.4% among infants who had not received this vaccine. In our study, the difference in PnOB rate between these two groups was smaller mainly because the PnOB rate among nonvaccinated infants was much lower, 0.83%. This significant difference could be explained by the starting date of both studies. Cartairs carried out the study during the first 2 years after PCV-7 was licensed in the United States. Our study was performed 5 years after the introduction of PCV-7 in

5 Predictors of occult bacteremia Mintegi et al. 203 Fig nvaccinated One dose Two doses Three doses Four doses Unknown Heptavalent pneumococcal vaccination status. Table 1 Causative organisms Causative organisms of occult bacteremia Streptococcus pneumoniae 10 Meningococcus B 2 Group A Streptococcus 1 Enterococcus faecium 1 Salmonella enteritidis 1 Spain. A more lasting, indirect protective effect could explain a decline of PnOB also among infants who have not received PCV-7. In contrast, the age distribution of PnOB also seems to be changing, probably because of the PCV-7 effect. In an earlier study, the risk of occult pneumococcal bacteremia was significantly lower in 3 6- month-old age group than in older age groups [12]. In our study, the age distribution of PnOB has changed and 30% of PnOB below 7 of age were detected. Despite the reduction in the rate of OB because of the widespread use of the conjugate pneumococcal vaccine [5,14], the evaluation of febrile children younger than 36 at risk of OB and SBI continues to consume health-care resources, causing transient discomfort to patients and contributing to resistance to antibiotics. In a cost-effectiveness study, Lee et al. [18] demonstrated that practicing CBC and selective BC and treatment using a WBC cutoff of WBC/mm 3 should be cost-effective at the rate of pneumococcal bacteremia of 1.5%. This recommendation was previously made by other authors as well [19,20]. In their article, Lee et al. [18] also affirm that if the rate of OB falls below 0.5% with the widespread use of the conjugate pneumococcal vaccine, then strategies n that use empiric testing and treatment should be eliminated. Taking into consideration this cost-effectiveness analysis and the OB rates found in our study, the practice of routinely obtaining a BC and WBC count may no longer be necessary when evaluating well-appearing infants with FWS, particularly when they have received at least two doses of PCV-7. In contrast, it has been traditionally reported that 6% of infants with FWS greater than 391C and more than WBC/mm 3 and 8% of those with an ANC higher than /mm 3 should have a PnOB [2,13]. The yield of WBC and ANC, however, seems to be much lower than the one reported earlier in the literature [2], probably because of the decline of PnOB in the group of children with high FWS. Stoll et al. [6] have recently reported a 3.2% PPV of a WBC count greater or equal to /mm 3 and a 3.8% PPV of an ANC greater or equal to /mm 3 in the era of the PCV-7. Our data very well demonstrate the fact that PPVs observed in this study were lower than the PPVs observed in the earlier study. We even found the 1% PPV of both cell count limits in infants who have received PCV-7. This evidence adds more doubts about the utility of CBC in the management of febrile infants in the PCV-7 era. Quantitative CRP concentration has been found to be a valuable laboratory test in the evaluation of febrile young children who are at risk of OB and SBI, with a better predictive value than the WBC or ANC [21]. The addition of CRP to ANC, however, seems to add a little diagnostic utility and does not add substantial diagnosis accuracy to the WBC count or ANC [22]. In our study, 60% of patients with PnOB had a CRP value less than

6 204 European Journal of Emergency Medicine 2009, Vol 16 4 Table 2 Data of the children with occult bacteremia Age () Sex Number of doses of PCV-7 Temperature (1C) WBC/mm 3 ANC/mm 3 C-reactive protein (mg/l) Blood culture 4 Male Streptococcus pneumoniae serotype 8 5 Female Streptococcus pneumoniae serotype 1 6 Male Streptococcus pneumoniae serotype 17 7 Female Streptococcus pneumoniae serotype 1 9 Male Streptococcus pneumoniae serotype 1 10 Female Streptococcus pneumoniae serotype 6 10 Male Streptococcus pneumoniae serotype Male Streptococcus pneumoniae unknown serotype 16 Male Streptococcus pneumoniae serotype 5 22 Male Streptococcus pneumoniae serotype 5 4 Female Unknown Enterococcus faecium 35 Female Group A Streptococcus 11 Male Salmonella enteritidis 5 Male Meningococcus B 9 Male Meningococcus B ANC, absolute neutrophil counts; WBC, white blood cell counts. 20 mg/l. The arrival of new, more sensitive markers of bacterial infection, such as procalcitonin, could be valuable in the management of these patients in the future [23], as Andreola et al. [24] assessed in predicting SBI in febrile children in the emergency department. The epidemiology of outpatient bacteremia in month-old children is changing after the introduction of the PCV-7. In some series, the incidence of E. coli bacteremia is similar to the incidence of pneumococcal bacteremia [9]. In our study, 66% of OB was because of pneumococcus and we did not find any bacteremia because of E. coli, probably because children with altered dipstick were excluded. Our study has several limitations. The fact that this is a retrospective study limits the collection of data. In contrast, we only included those well-appearing infants aged 3 36 with FWS Z 391C in whom physicians obtained a CBC and BC as part of their evaluation for fever. Inclusion of all the well-appearing infants aged 3 36 with FWS Z 391C should have had more accuracy to study the data. This fact should have implied change to our practice in the management of young children with high fever, and it does not look to be acceptable to practice these tests to those young children with a complete PCV-7 status. In our PED, before this study was carried out, we strongly recommended practicing CBC and BC in children non-pcv-7 vaccinated or with an incomplete vaccination, but tests were not practiced in one-fourth of the children as we recently reported [11]. These children were studied and did not show any difference with children included in the reported study and tests were not performed by the decision of pediatrician attending the child [11]. This fact also reflects the pediatricians irregular compliance with the management guidelines for infants with FWS, as it has been reported by other authors [25]. We decided not to change the practice of our PED when this study was carried out, although this fact may represent a bias in the interpretation of the results. Nevertheless, none of the children in whom a CBC and BC were not obtained were admitted again in our PED with an OB or a SBI. Recently, a replacement invasive pneumococcal disease with serotypes not covered by PCV-7 has been reported among Alaska Native children despite high PCV-7 coverage [26]. This fact and the introduction of pneumococcal conjugate vaccines including more serotypes emphasize the importance of surveillance of the yield of WBC and ANC, while evaluating previously healthy infants with high fever. We conclude that pneumococcal bacteremia was found to be lower in well-appearing febrile infants who had received heptavalent pneumococcal conjugate vaccine. Our study supports that the strategy of routine laboratory screening and selective empirical antibiotic treatment of well-appearing febrile children below 3 years of age with two or more doses of PCV-7 must no longer be recommended. Acknowledgements Santiago Mintegi for conceiving the idea for the study, analyzing the data, drafting the manuscript, and reviewing the literature. Javier Benito, Jesus Sanchez, Beatriz Azkunaga, who drafted the manuscript, and reviewed the literature. Itziar Iturralde, who collected and analyzed the data Silvia Garcia who reviewed the literature. Competing interest: none declared. References 1 Slater M, Krug SE. Evaluation of the infant with fever source: an evidence based approach. Emerg Med Clin rth Am 1999; 17: Lee GM, Harper MB. Risk of bacteremia for febrile young children in the post-haemophilus influenzae type b era. Arch Pediatr Adolesc Med 1998; 152: Mintegi S, Benito J, García S, Corrales A, Bartolomé MJ, Trebolazabala N. Patient demand and management in a hospital pediatric emergency setting. An Pediatr (Barc) 2004; 61:

7 Predictors of occult bacteremia Mintegi et al Mintegi S, González M, Pérez A, Pijoán JI, Capapé S, Benito J. Lactante de 3-24 meses con fiebre sin foco en urgencias: características, tratamiento y evolución posterior. An Pediatr 2005; 62: Benito-Fernandez J, Mintegi-Raso S, Gonzalez-Balenciaga M, Capape-Zache S, Sanchez-Etxaniz J, Garcia-Gonzalez S. Pneumococcal bacteremia among infants with fever without source before and after introduction of pnemococcal conjugate vaccine in Basque Country (Spain). Pediatr Infect Dis J 2007; 19: Stoll ML, Rubin LG. Incidence of occult bacteremia among highly febrile young children in the era of the pneumococcal conjugate vaccine. A study from a Children s hospital emergency department and urgent care center. Arch Pediatr Adolesc Med 2004; 158: Black S, Shinefield H, Fireman B, Lewis E, Ray P, Hansen J, et al. Efficacy, safety and immunogenicity of heptavalent pneumococcal conjugate vaccine in children: northern California Kaiser Permanente Vaccine Study Center Group. Pediatr Infect Dis J 2000; 19: Whitney CG, Farley MM, Hadler J, Harrison LH, Bennett NM, Lynfield R, et al. Active Bacterial Core Surveillance of the Emerging Infections Program Network. Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med 2003; 348: Herz AM, Greenhow TL, Alcantara J, Hansen J, Baxter RP, Black SB, Shinefield HR. Changing epidemiology of outpatient bacteremia in 3- to-36-month-old children after the introduction of the heptavalentconjugated pneumococcal vaccine. Pediatr Infect Dis J 2006; 25: Steenhoff AP, Shah SS, Ratner AJ, Patil SM, McGowan KL. Emergence of vaccine-related pneumococcal serotypes as a cause of bacteremia. Clin Infect Dis 2006; 42: Mintegi S, Benito J, González M, Astobiza E, Sánchez J, Santiago M. Impact of the pneumococcal conjugate vaccine in the management of highly febrile children aged 6 24 in an emergency department. Pediatr Emerg Care 2006; 22: Kuppermann N. Occult bacteremia in young febrile children. Pediatr Clin rth Am 1999; 46: Kuppermann N, Fleisher GR, Jaffe DM. Predictors of occult pneumococcal bacteremia in young febrile children. Ann Emerg Med 1998; 31: Carstairs KL, Tanen DA, Johnson AS, Kailes SB, Riffenburgh RH. Pneumococcal bacteremia in febrile infants presenting to the emergency department before and after the introduction of the heptavalent pneumococcal vaccine. Ann Emerg Med 2007; 49: Antonyrajah B, Mukundan D. Fever without apparent source on clinical examination. Curr Opin Pediatr 2008; 20: (last visited 08-July-21). 17 Rodriguez Serna A, Astibiza Beobide E, González Balenciaga M, Azkunaga Santibañez B, Benito Fernandez J, Mintegi Raso S. Cambios de los hábitos poblacionales en el tratamiento de la fiebre en la infancia. An Pediatr (Barc) 2006; 64: Lee GM, Fleisher GR, Harper MB. Management of febrile children in the age of conjugate pneumococcal vaccine: a cost-effectiveness analysis. Pediatrics 2001; 108: Lieu T, Schwartz S, Jaffe D, Fleisher G. Strategies for diagnosis and treatment of children at risk for occult bacteremia: clinical effectiveness and cost-effectiveness. J Pediatr 1991; 118: Downs S, McNutt R, Margolis P. Management of infants at risk for occult bacteremia: a decision analysis. J Pediatr 1991; 118: Pulliam PN, Attia MW, Cronan KM. C-reactive protein in febrile children 1 to 36 of age with clinically undetectable serious bacterial infection. Pediatrics 2001; 108: Isaacman DJ, Burke BL. Utility of the serum C-reactive protein for detection of occult bacterial infection in children. Arch Pediatr Adolesc Med 2002; 156: Jones AE, Flechtl JF, Brown MD, Ballew JJ, Kline JA. Procalcitonin test in the diagnosis of bacteremia: a meta-analysis. Ann Emerg Med 2006; 50: Andreola B, Bressan S, Callegaro S, Liverani A, Plebani M, Da Dalt L, 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: Singleton RJ, Hennessy TW, Bulkow LR, Hammitt LL, Zulz T, Hurlburt DA, et al. Invasive pneumococcal disease caused by nonvaccine serotypes among alaska native children with high levels of 7-valent pneumococcal conjugate vaccine coverage. JAMA 2007; 297: Belfer RA, Gittelman MA, Muniz AE. Management of febrile infants and children by pediatric emergency medicine and emergency medicine: comparison with practice guidelines. Pediatr Emerg Care 2001; 17:83 87.