Should we give antibiotics to neonates with mild non-progressive symptoms? A comparison of serial clinical observation and the neonatal sepsis risk calculator

OBJECTIVE: To compare two strategies [the neonatal sepsis risk calculator (NSC) and the updated serial clinical observation approach (SCO)] for the management of asymptomatic neonates at risk of early-onset sepsis (EOS) and neonates with mild non-progressive symptoms in the first hours of life. METHODS: This was a single-center, retrospective cohort study conducted over 15 months (01/01/2019–31/03/2020). All live births at ≥34 weeks of gestation were included. Infants were managed using SCO and decisions were compared with those retrospectively projected by the NSC. The proportion of infants recommended for antibiotics or laboratory testing was compared in both strategies. McNemar's non-parametric test was used to assess significant differences in matched proportions. RESULTS: Among the 3,445 neonates (late-preterm, n = 178; full-term, n = 3,267) 262 (7.6%) presented with symptoms of suspected EOS. There were no cases of culture-proven EOS. Only 1.9% of the neonates were treated with antibiotics (median antibiotic treatment, 2 days) and 4.0% were evaluated. According to NSC, antibiotics would have been administered in 5.4% of infants (absolute difference between SCO and NSC, 3.51%; 95% CI, 3.14–3.71%; p <0.0001) and 5.6% of infants would have undergone “rule out sepsis” (absolute difference between SCO and NSC, 1.63%, 95% CI 1.10–2.05; p <0.0001). CONCLUSION: SCO minimizes laboratory testing and unnecessary antibiotics in infants at risk of EOS or with mild non-progressive symptoms, without the risk of a worse neonatal outcome. The NSC recommends almost three times more antibiotics than the SCO without improving neonatal outcomes.