Methods: Children (<18 years of age) presenting with Enterobacteriaceae BSI at Tokyo Metropolitan Children’s Medical Center in Japan between March 2010 and March 2017 were enrolled. The clinical characteristics and outcomes of patients with MDR and non-MDR Enterobacteriaceae BSI were compared. The Centers for Disease Control and Prevention’s definition of MDR was used for this study.
Results: In total 134 blood cultures from 127 patients were analyzed. The median age was 3.6 years (IQR: 0.3-9.8 years), and boys accounted for 52.8% of the subject pool. Underlying diseases were noted in 92.1% of patients. In 79.1%, BSI developed >48 h after admission. The most common isolate was Escherichia coli (47.0%) followed by Klebsiella pneumoniae (23.1%). MDR was detected in 35.8% (48/134) of culture, with Escherichia coli occurring with the highest frequency (23.1%) followed by Klebsiella pneumoniae (7.5%). The MDR strains producing ESBL, AmpC, and carbapenemases comprised 17.9%, 7.5%, and 1.5% of the cases, respectively. Multivariate logistic regression analysis showed that the history of corticosteroid use within 30 days was independently associated with the development of MDR Enterobacteriaceae BSI (OR: 3.63; 95%CI: 1.35-9.73). Initial empiric therapy was less effective against MDR, than non-MDR, strains (51.1% vs. 83.8%, P<0.001). MDR was not significantly associated with an increased rate of all-cause mortality (MDR 10.4% vs. non-MDR 3.5%, P=0.134) or sepsis-related mortality (MDR 6.3% vs. non-MDR 2.3%, P=0.349).
Conclusion: A history of corticosteroid use within 30 days was an independent risk factor for the development of MDR Enterobacteriaceae BSI in children. MDR did not increase the mortality rate significantly.
H. Funakoshi, None
S. Ishii, None
K. Araki, None
T. Murai, None
K. Fukuoka, None
Y. Horikoshi, None