Abstract
BACKGROUND: Empiric administration of ampicillin and gentamicin is recommended for newborns at risk of early-onset sepsis (EOS). There are limited data on antimicrobial susceptibility of all EOS pathogens. METHODS: Retrospective review of antimicrobial susceptibility data from a prospective EOS surveillance study of infants born ≥22 weeks' gestation and cared for in Neonatal Research Network centers April 2015-March 2017. Nonsusceptible was defined as intermediate or resistant on final result. RESULTS: We identified 239 pathogens (235 bacteria, 4 fungi) in 235 EOS cases among 217,480 live-born infants. Antimicrobial susceptibility data were available for 189/239 (79.1%) isolates. Among 81 Gram-positive isolates with ampicillin and gentamicin susceptibility data, all were susceptible in vitro to either ampicillin or gentamicin. Among Gram-negative isolates with ampicillin and gentamicin susceptibility data, 72/94 (76.6%) isolates were nonsusceptible to ampicillin, 8/94 (8.5%) were nonsusceptible to gentamicin, and 7/96 (7.3%) isolates were nonsusceptible to both. Five percent or less of tested Gram-negative isolates were nonsusceptible to each of third or fourth generation cephalosporins, piperacillin-tazobactam, and carbapenems. Overall, we estimated that 8% of EOS cases were caused by isolates nonsusceptible to both ampicillin and gentamicin; these were most likely to occur among preterm, very-low birth weight infants. CONCLUSIONS: The vast majority of contemporary EOS pathogens are susceptible to the combination of ampicillin and gentamicin. Clinicians may consider the addition of broader-spectrum therapy among newborns at highest risk of EOS, but we caution that neither the substitution nor the addition of 1 single antimicrobial agent is likely to provide adequate empiric therapy in all cases.