Ensuring a Locally Tailored Response to Early Onset Sepsis Screening Meets or Exceeds the Performance of Published Approaches

Hosp Pediatr. 2020 Oct;10(10):877-883. doi: 10.1542/hpeds.2020-0153.


Background: Evaluation of well-appearing neonates for early-onset sepsis (EOS) remains controversial. Multiple risk stratification approaches are currently used for the evaluation of EOS. Our aim was to quantify and compare frequency of laboratory evaluation and empirical antibiotics between published and local EOS approaches.

Methods: This retrospective cohort study included 8240 infants born ≥35 + 0/7 weeks' gestation at an institution from October 1, 2014, to March 1, 2018. Excluded from analysis were 156 patients who exhibited either major congenital anomalies or required antibiotics for surgical issues. A total of 1680 patient charts with risk factors for EOS were reviewed for further demographic data, clinical presentation, laboratory results, and probable recommendations from 4 EOS risk assessment approaches.

Results: Laboratory evaluation recommendation was 7.1% for Centers for Disease Control and Prevention 2010 guidelines and local 2016 EOS algorithm, 6% for local 2019 EOS algorithm, and 5.9% for Kaiser Permanente neonatal EOS calculator (neonatal EOS calculator). Antibiotic recommendation was 6% for 2010 Centers for Disease Control and Prevention guidelines, 4.3% for neonatal EOS calculator, and 3.3% for local 2016 and 2019 EOS algorithms.

Conclusions: Of the 4 approaches reviewed, the local 2019 EOS algorithm and the neonatal EOS calculator were similar in recommending the lowest frequency of laboratory evaluation and the local 2016 and 2019 EOS algorithms had the lowest recommended antibiotic usage in this population.