Background: Data regarding current antimicrobial resistance in neonatal early-onset sepsis (EOS) and possible resistance risk factors are essential for determining the optimal empiric regimen.
Methods: EOS data retrospectively collected nationwide in Israel in 18 hospitals for 2010-2015. EOS defined as bacteremia or meningitis during the first 7 days of life for full-terms and 3 days for pre-terms.
Results: Participating hospitals represented 75% of the total birth cohort in Israel (~180,000 live births (LB)/year). Of 488 EOS cases (0.63/1,000LB), 271 (56%) were full-term and 217 (44%) were pre-term (0.35 and 0.28/1,000LB, respectively). Gram-negative pathogens predominated over Gram-positive (0.35 vs. 0.28/1000LB, respectively) (OR 1.28 [95%CI 1.07-1.52], p<0.01). E.coli, Group B Streptococcus, and K.pneumoniae were most prevalent (0.23, 0.21, and 0.06/1000LB, respectively). mong the 277 Gram-negatives, 16% were gentamicin-resistant, 14% were extended-spectrum beta-lactamase (ESBL)-positive, 10% were gentamicin-resistant and ESBL-positive (50% of gentamicin-resistant isolates), and 3.5% were amikacin-resistant. Higher resistance rates were found among pre-terms vs. full-terms:18% vs. 14% gentamicin-resistant (OR=1.38 [95%CI 0.6-3.1],p=0.44), 16% vs. 5% ESBL, 8% vs. 5% ESBL and gentamicin-resistant, and 4.5% vs. 3% amikacin-resistant, respectively.
Gentamicin resistance or ESBL’s in Gram-negatives were not associated with previous maternal hospitalizations or antimicrobials during pregnancy.
Mortality was documented in 22% of neonates with Gram-negatives and 8% with Gram-positives (p<0.01); Gentamicin resistance was associated with increased mortality (OR=2.82; [95%CI 1.14-6.95], p<0.05).
Conclusions: In this nationwide survey, EOS caused predominantly by Gram-negatives with high gentamicin-resistance rates, without identifiable resistance risk factors. As EOS is life threatening we suggest considering modification of the current empiric therapy for amikacin-based regimens.
