Background: Escherichia coli bacteremia is a major cause of severe sepsis in children. Little is known about predictors of severity.
Methods: We analyzed 84 children ≤18 years of age with E. coli bacteremia from the prospective COLIBAFI study performed during 2005–2007. The severity of bacteremia was defined as occurrence of death or transfer to intensive care unit. Studied characteristics included age, gender, birth weight, history of prematurity, immunodepression, nosocomial infection, portal of entry, phylogenetic group and subgroup belonging, O-type, virulence gene content and antimicrobial susceptibility. We compared bacterial characteristics in urinary- versus digestive-source bacteremia, in children ≤3 versus >3 month of age, and in children versus adults. We also searched for risk factors of severity.
Results: Median age was 2.4 months, 57% males. Most frequent portals of entry were urinary (66.2%) and digestive (19.5%) tracts. Most isolates (63.1%) belonged to B2 phylogroup. Strains in children ≤3 months of age exhibited more virulence genes, especially neuC and fyuA/irp2, and were less resistant to antibiotics than in children >3 months of age. Comparing community-acquired urinary-source bacteremia between children and adults, we found that bacteremia were less severe in children, whose strains exhibited a specific virulence gene repertoire and had a higher resistance score than in adults. Seventeen children (20.2%) had a severe bacteremia and 8 died. Non-urinary portal of entry and age ≤3 months of age were the only risk factors associated with severity.
Conclusions: E. coli strains responsible for bacteremia exhibit specific characteristics according to age of children. However, host characteristics and portal of entry are the main determinants of severity of E. coli bacteremia in children, as observed in adults.
From the *AP-HP, Hôpital Bichat, Service de Biostatistiques; †IAME, UMR 1137, INSERM, Univ Paris Diderot, Sorbonne Paris Cité; ‡EA 3105, Univ Paris Diderot, Sorbonne Paris Cité; §AP-HP, Hôpital Robert-Debré, Laboratoire de Microbiologie; ¶AP-HP, Hôpital Robert Debré, Service de Néonatologie, Paris; and ‖AP-HP, Hôpital Beaujon, Service de Médecine Interne, Clichy, France.
C.B., O.C. and S.B. contributed equally to the work.
E.B. is deceased.
Accepted for publication February 2, 2014.
Authors’ contribution: C.B., S.B. and E.D. wrote the article. O.C. performed and analyzed the bacterial data. C.B., C.L. and F.M. performed the statistical analyses. A.L., S.B., E.B., Y.A. and E.D. conceived the study.
The COLIBAFI study was supported by grants from “Réseau de Recherche Clinique” (INSERM: RBM-03-58) and “Projet Hospitalier de Recherche Clinique” (Assistance Publique-Hôpitaux de Paris: AOR 04 053). The authors have no funding or conflicts of interest to disclose.
This work has partly been reported at the 32nd Réunion interdisciplinaire de chimiothérapie anti-infectieuse (RICAI), held in Paris, France on November 21–22, 2012 (poster 548).
Address for correspondence: Erick Denamur, MD, PhD, IAME, UMR 1137 INSERM, Universités Paris Diderot et Paris Nord, Sorbonne Paris Cité, Faculté de Médecine, Site Xavier Bichat, 16 rue Henri Huchard, 75018 Paris, France. E-mail: email@example.com.