A Decade Later: Serial Clinical Examinations for Early Onset Sepsis in Neonates of at Least 35 Weeks’ Gestation

Culture-positive early onset sepsis (EOS) in neonates of at least 35 weeks’ gestation is rare (0.2–0.4 per 1000), and identifying those infected is challenging, with almost one-half of cases being well-appearing at birth.1–4 Clinical vigilance is therefore required, yet uncertainty remains on how to best identify those infected. American Academy of Pediatrics (AAP) EOS guidelines recommend 3 potential approaches: “categorical risk assessment,” the Neonatal Sepsis Calculator (NSC), or serial clinical examinations (SCE).5 Ongoing surveillance of these approaches in clinical practice is important.5 We previously reported our 5-year experience with SCE6; herein we share our cumulative 10-year experience with SCE.In March 2015, we implemented an SCE approach within a quality improvement (QI) framework for EOS in neonates of at least 35 weeks’ gestation, as previously reported.6–8 In brief, the approach entailed clinical monitoring of all neonates regardless of risk factors, including vital signs and nurse assessments every 30 minutes for 2 hours after birth, then every 4 hours until 24 hours, followed by every 8 hours thereafter. Well newborn care included a 3:1 mother-infant dyad to nurse ratio. Level II through IV neonatal intensive care unit care was available on site, and a neonatal hospitalist was available in-house 24/7. Antibiotic treatment was not recommended unless a neonate developed clinical signs concerning for sepsis. No formal criteria for starting antibiotics were in place, leaving to clinician discretion. Sepsis laboratory evaluation (complete blood count and C-reactive protein [CRP]) was recommended only once antibiotic treatment was started. Sepsis laboratory evaluation to “screen” neonates was discouraged.As part of ongoing QI surveillance, we tracked yearly inborn neonates of at least 35 weeks’ gestation who had antibiotic treatment (defined by ampicillin receipt) or sepsis laboratory evaluation (defined by CRP measurement) within the first 3 days after birth. All cases of culture-positive EOS were reviewed. The local institutional review board determined this project did not meet the definition of human subject research.From March 2015 to March 2025, 43 772 neonates of at least 35 weeks’ gestation underwent SCE for EOS risk assessment. Overall, 3.4% (n = 1481) received ampicillin and 5.4% (n = 2370) had CRP testing. During the last 5 years, ampicillin exposure (Figure 1A) ranged from 2.4% to 3.5%, and CRP testing (Figure 1B) ranged from 4.3% to 5.4%.Over the 10 years, 12 neonates had culture-positive EOS (Table 1), representing an incidence of 0.27 per 1000. Clinical signs of illness were present at birth in 4 neonates, with the remaining 8 developing clinical signs of illness after birth. One neonate died due to Esherichia coli sepsis after presenting with fever at 20 hours. No other neonate required invasive ventilation or inotropic support. Retrospectively applying the NSC, all neonates who developed clinical signs of illness after birth, including the one who died, would have had “routine vitals” recommended (EOS risk ≤0.32 per 1000). Similarly, categorical risk assessment recommendations were “routine care” in all but 1 of the neonates who presented after birth.Over a decade of implementing SCE to guide EOS risk stratification in over 43 000 neonates of at least 35 weeks’ gestation at our institution, antibiotic use and sepsis laboratory testing remained low and similar to centers implementing the NSC.1,9The importance of the clinical examination in identifying neonates with EOS cannot be overstated regardless of stratification approach employed. Two-thirds of our EOS cases were initially well-appearing. All had “low risk” NSC scores at birth (≤0.32 per 1000), and only 1 would have been empirically treated using the categorical risk assessment approach. The clinical examination was central to identifying these neonates. Similarly, over one-half of EOS cases in the Kaiser experience had no perinatal risk factors, and NSC recommendations, accordingly, depend heavily on the clinical examination.1 We had 1 neonate die who was initially well-appearing at birth; none of the 3 AAP-recommended approaches would have recommended empirical antibiotics.EOS risk stratification is challenging given the need to balance the risks of a low-incidence, high-morbidity disease alongside the risks of overtreatment and antibiotic exposure to a large group of neonates. Although our experience is limited given its single-center nature, it is in line with a prior meta-analysis.4 In short, serial clinical examinations—such as focused assessments every 4 hours during the first 24 hours—should be the cornerstone of any EOS approach to provide a comprehensive safety net for all neonates regardless of risk factors.The authors thank the nurses and physicians in our newborn nurseries who were instrumental in making this practice change successful and sustained.