No one would argue with the need to be a good steward of antibiotics. This makes newborn care challenging because of the need to have a low threshold to begin antibiotics while awaiting cultures for infants with suspected sepsis. Is there a better way to predict which infants are at low risk for early onset sepsis (EOS) to avoid antibiotic use? Flannery et al (10.1542/peds.2021-052900) from Children’s Hospital of Philadelphia share with us an analysis of a retrospective cohort study of 7,549 preterm and term infants born between 2009 and 2014 who had blood cultures drawn because of a concern for sepsis. The authors separated these infants into high- and low-risk for EOS based on delivery characteristics. They then determined which infants received antibiotics or developed sepsis.
The “low-risk” criteria were identified in prior studies, and included birth by caesarean section, no labor or membrane rupture until the time of delivery, and no antepartum concerns for intraamniotic infection or non-reassuring fetal status. 1,121 infants met these low-risk criteria. None of these infants categorized as low-risk, which included those with a gestational ages of 28-39 weeks, developed EOS. Despite their low risk, 80.4% of infants in the low-risk category received antibiotics for a similar duration as those at higher risk. Among the higher risk infants, 41 (0.6%) had EOS.
Are these results convincing enough to consider not using antibiotics on extremely or even moderately preterm infants who meet low-risk delivery characteristics? Should we opt to just observe these babies who are low-risk for signs of clinical instability and start antibiotics then? While we may do this for term infants who are stable, we may be less likely to follow suit for preterm infants for reasons well explained by Drs. Ivana Culic and Amy O’Connell from Harvard in an accompanying commentary (10.1542/peds.2021-054221). They highlight the strengths and limitations of this study, including the small number of extremely preterm infants and some concerns about generalizability issues, and the importance of finding strategies that will reduce unnecessary use of antibiotics. Please link to the study and the commentary and see if your threshold for not using antibiotics is influenced—and if not, stay tuned for prospective studies that will likely follow as a result of the road paved by Dr. Flannery et al.