REVIEW ARTICLE A Meta-analysis of the Rates of Listeria monocytogenes and Enterococcus in Febrile Infants Rianna Leazer, MD, Amy M. Perkins, MS, Kyrie Shomaker, MD, Bryan Fine, MD, MPH ABSTRACT CONTEXT: A change in the epidemiology of pathogens causing serious bacterial infection (SBI) has been noted since original recommendations were made for the empirical antibiotic choices for young infants with fever. OBJECTIVE: To assess the prevalence of SBI caused by Listeria monocytogenes and Enterococcus species. DATA SOURCES: A literature search was conducted on keywords related to SBI, L. monocytogenes, and Enterococcus spp. infections. STUDY SELECTION: Eligible studies were those conducted in the United States and published between January 1998 and June 2014 focusing on SBI in infants #90 days of age. DATA EXTRACTION: The rates of urinary tract infection, bacteremia, and meningitis for each pathogen were recorded for each study. Meta-analysis was performed to calculate the prevalence for each pathogen in a random effects model with 0.5 continuity correction added to studies with zero events. RESULTS: Sixteen studies were included. A total of 20 703 blood cultures were included, with weighted prevalences for L. monocytogenes and Enterococcus spp. bacteremia of 0.03% and 0.09%, respectively. A total of 13 775 cerebrospinal fluid cultures were included with event rates (unweighted prevalences) for L. monocytogenes and Enterococcus spp. meningitis of 0.02% and 0.03%, respectively. A total of 18 283 urine cultures were included, with no cases of L. monocytogenes and a weighted prevalence for Enterococcus spp. urinary tract infection of 0.28%. LIMITATIONS: There may have been reporting bias or incomplete retrieval or inadvertent exclusion of relevant studies. CONCLUSIONS: SBI caused by L. monocytogenes and Enterococcus spp. in febrile infants is rare, and therefore clinicians may consider a change in empirical antibiotic choices. www.hospitalpediatrics.org DOI:10.1542/hpeds.2015-0187 Copyright © 2016 by the American Academy of Pediatrics Address correspondence to Rianna Leazer, MD, Children’s Hospital of the King’s Daughters, 601 Children’s Lane, Norfolk, VA 23507. E-mail: [email protected] HOSPITAL PEDIATRICS (ISSN Numbers: Print, 2154-1663; Online, 2154-1671). Department of Pediatrics, Children’s Hospital of the FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. ’ Kings Daughters and FUNDING: No external funding. Eastern Virginia Medical School, Norfolk, Virginia POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose. HOSPITAL PEDIATRICS Volume 6, Issue 4, April 2016 187 Downloaded from www.aappublications.org/news by guest on September 25, 2021 Fever often prompts evaluation for serious METHODS sets, only the study with the larger data set bacterial infection (SBI) in infants, and a The Meta-Analysis of Observational Studies was included for analysis. Because the subset may be admitted to the hospital and in Epidemiology8 and Preferred Reporting purpose of this review is to evaluate placed on empirical antibiotics while Items for Systematic Reviews and Meta- rates of occult infections caused by awaiting bacterial culture results. For Analyses9 guidelines were used in L. monocytogenes and Enterococcus spp., otherwise healthy infants outside the formulating this review. only cases of urinary tract infection (UTI), intensive care setting, empirical antibiotic bacteremia, and meningitis were included coverage has classically included the Search Strategies as SBIs. combination of ampicillin and either The PubMed database was searched by the Data Collection Process gentamicin or a third-generation primary author with guidance from 1,2 cephalosporin. These regimens were institutional librarians using combinations The primary outcomes for this review were adopted when group B Streptococcus (GBS) of the terms “infant,”“fever,”“serious the prevalence rates of UTI, bacteremia, and was the predominant pathogen causing SBI bacterial infection,”“Enterococcus,” meningitis caused by L. monocytogenes and and Listeria monocytogenes was more “Listeria monocytogenes,”“sepsis,”“urinary Enterococcus spp. in febrile infants # commonly encountered as a neonatal tract infection,”“bacteremia,” and 90 days of age. The results from each pathogen. Since these regimens were “meningitis.” To include recent and relevant included study were reviewed, and initially determined, there have been epidemiologic data, search limitations prevalence rates of UTI, bacteremia, and changes in the epidemiology of pathogens included studies published over ∼15 years, meningitis caused by each pathogen were causing SBI in young infants. The most between January 1998 and June 2014. recorded. notable changes have been the decline in Additional limitations included studies Study Quality Assessment cases of infections with GBS and performed in humans and those published L. monocytogenes and a rise in infections Studies were assessed for risk of bias via in English. References from studies included a 13-item assessment tool developed from caused by Staphylococcus aureus and in data analysis were reviewed by 3 of the 3–5 recommendations published by the Agency Enterobacteriacae. The etiology of these authors for additional relevant citations. changes is probably multifactorial, with for Healthcare Research and Quality for observational studies.10 From the Agency for proposed influences being changes in food Study Selection handling in the United States,6,7 the Healthcare Research and Quality question Studies were screened by the primary bank, applicable questions were chosen in implementation of intrapartum GBS author through title or abstract review to prophylaxis, vaccination for other previously the domains of Selection, Performance, include studies focusing on the intended Attrition, Reporting, and Conflict of Interest. common pathogens such as Haemophilus population of otherwise healthy febrile fl Each study was independently assessed by in uenzae, and trends in antibiotic use in infants in the United States. Studies were the community. 3 authors (R.L., B.F., and K.S.) and discussed therefore excluded if performed outside the until consensus was reached for each fi Although changes in the epidemiology of SBI United States, if there was speci c focus on question. in febrile infants have been demonstrated in preterm infants, if cultures drawn in the ICU several studies, meningitis and bacteremia setting or the newborn nursery were Data Synthesis and Analysis are rare events, and the epidemiology of included, or there was focus on 1 pathogen Heterogeneity between the studies was infections is often regional. With small or disease process (eg, SBI associated with assessed via the Cochran Q test and I2. sample sizes in previous studies, it has pneumonia). The remaining studies were Negative values for I2 or lower confidence therefore been difficult to recommend a then excluded upon abstract or full-text limits were set to 0. Because the included change in the empirical antibiotic choice for review if the specific data for infants studies had low event rates and zero event febrile infants, and practitioners often #90 days of age were not available, or if rates were common, heterogeneity testing continue to include empirical antibiotic pathogens causing SBI or the total number may not be a reliable assessment to coverage for L. monocytogenes and of positive blood, urine, or cerebrospinal determine the most appropriate method for a Enterococcus spp. Stronger evidence fluid (CSF) cultures were not included in meta-analysis. In addition, study populations supporting a low rate of invasive infections reported results. For studies that otherwise were assumed to differ from each other, and caused by Enterococcus spp. and met inclusion criteria but did not describe therefore a random effects model was L. monocytogenes in febrile infants is specific age group characteristics or consistently used with the assumption of at needed to inform practitioners in choosing culture data in the text, the primary author least moderate heterogeneity between empirical antibiotic regimens. This meta- was contacted and the study was included if studies. Meta-analysis was performed to analysis was performed to determine the the necessary information was calculate the prevalence for each pathogen. prevalence of SBI caused by subsequently obtained. Study dates, To adjust for the occurrence of zero events, a L. monocytogenes and Enterococcus spp. in authors, and institutions were used to continuity correction of 0.5 was added to otherwise healthy febrile infants #90 days cross-reference studies for duplicate data studies with zero events. Weighted prevalence of age in the United States. sets. For publications with duplicate data rates (including lower and upper 95% 188 LEAZER et al Downloaded from www.aappublications.org/news by guest on September 25, 2021 confidence limits) and the number needed to restricted to studies published in the last prevalence of bacteremia of 2.68% (95% screen (NNS) to find 1 positive culture caused 15 years, the dates during which the studies confidence interval [CI], 2.00%–3.36%). by L. monocytogenes or Enterococcus spp. were performed vary and are listed in the There were 2 cases of L. monocytogenes were calculated for each pathogen. The data tables. The included studies are and 22 cases of Enterococcus spp. causing were analyzed in Microsoft Excel 2010 observational,