Factors influencing antibiotic duration in culture-negative neonatal early-onset sepsis.

Abstract

Little guidance exists on the treatment duration of culture-negative early-onset sepsis (CN-EOS) in neonates, which may lead to prolonged antimicrobial therapy and adverse outcomes. Our objective was to identify risk factors associated with prolonged antibiotic therapy in CN-EOS in neonates. This was a retrospective, matched cohort study of neonates treated with empiric antibiotic therapy for EOS. Infants were sampled with matching of gestational age (GA) into short (≤3days) and prolonged (>3days) antibiotic course. Primary outcomes were to identify predictive factors that may be associated with prolonged therapy and compare rates of late-onset sepsis (LOS) and mortality. Secondary outcomes included necrotizing enterocolitis, feeding intolerance, and early development assessment. Predictors associated with prolonged antibiotic therapy were identified using multivariable-adjusted logistic regression. Three hundred infants were included with 150 infants in each group. Mean GA and birthweights were 34.2±4.7weeks and 2293±991g, respectively. Male gender, 5-min Apgar <7, immature-to-total neutrophil ratio ≥0.2, C-reactive protein (CRP) ≥10mg/L, need for vasopressors, and mechanical ventilation were identified as significant predictors for prolonged antibiotics in all infants. Independent of GA, elevated CRP (OR 40.84, 95% CI 15.28-109.15, p<0.001), need for vasopressors (OR 13.48, 95% CI 3.86-47.15, p<0.001), and mechanical ventilation (OR 12.98, 95% CI 4.91-34.35, p<0.001) remained significant predictors of prolonged antibiotic use. Infants in the prolonged courses experienced significant delays in achieving independent oral feeding compared with infants receiving short-course antibiotics (median 17.5 vs. 8days, p=0.002, respectively). There were no significant differences in LOS, mortality, or other neonatal comorbidities. Elevated CRP levels, need for vasopressors, and mechanical ventilation were associated with prolonged antibiotic use in neonates presumptively treated for CN-EOS. Further research is warranted in identifying selective biomarkers for EOS and evaluating whether early antibiotic discontinuation for CN-EOS, despite abnormal laboratory tests/illness severity, is safe and justified.