New approach to EOS reduces testing and antibiotics

August 6, 2020

Examining how an intensive clinical examination can change the approach to early-onset sepsis.

Using only intensive clinical examination for initial identification of early-onset sepsis (EOS) in late preterm and term neonates substantially decreases laboratory testing and antibiotic use with no adverse outcomes. These were the ­findings of investigators who analyzed the effect of such an approach at their California hospital during a 5-year period.

As part of this approach, late preterm and preterm neonates who appeared well at birth remained with their mothers, even when the infants had perinatal risk factors for EOS. Nurses provided all these babies with enhanced clinical monitoring, and a neonatal hospitalist performed the initial clinical assessment of infants born to mothers with chorioamnionitis. Early in the study period, such chorioamnionitis-exposed infants were admitted to the NICU, but later on these babies instead remained with their mothers for the entire hospitalization if they appeared well.

Antibiotic treatment was administered only if the treating physician decided it was appropriate because the neonate was clinically ill at birth or later developed clinical signs of illness that posed a risk of sepsis. In addition, neonates were not routinely screened using sepsis laboratory testing, and when such testing was performed it was only once antibiotic treatment was started. Finally, routine evaluation for EOS did not include lumbar puncture for cerebrospinal fluid studies and culture. These tests generally were performed only in the presence of neurologic signs, such as mental status changes or seizures, at presentation or when the blood culture became positive.

Analysis of outcomes in 20,394 neonates ≥35 weeks’ gestation found that the percentage of neonates exposed to ampicillin decreased from 11.1% in the period before the clinical-examination approach was instituted to 4.1% during the intervention period. In addition, C-reactive protein testing declined from 15.3% to 6.3%. Overall, during an almost 5-year period, implementation of the clinical-examination approach reduced the use of antibiotics by 63% and sepsis laboratory testing by 59% (Frymoyer A, et al. J Pediatr. 2020. E pub ahead of print).

Thoughts from Dr. Farber

This paper is a potentially seminal article in my view. I was taught that at-risk newborns required work-ups and antibiotics; this is a drastic change. I expect it will take time to become standard; more seasoned physicians will not necessarily adjust their approach based on articles. In time, the next generation of pediatricians can show how it can be done safely.

download issueDownload Issue : Vol 37 No 8