Limitations of the risk factor based approach in early neonatal sepsis evaluations

Orna Flidel-Rimon, Sophie Galstyan, Ada Juster-Reicher, Ilya Rozin, Eric S. Shinwell

Research output: Contribution to journalArticlepeer-review

30 Scopus citations


Aim: Guidelines for detection of early neonatal sepsis employ a risk factor approach combined with laboratory parameters. In an era of increasing intrapartum antibiotic prophylaxis (IAP), we re-assessed the approach as a whole and each of the risk factors individually. Method: This retrospective study included infants with risk factors for sepsis or those treated with antibiotics or who had documented early sepsis. Safety of the protocol was assessed by the number of cases of either missed or partially treated late sepsis or meningitis and the sepsis-related mortality rate. Predictive value of each clinical and laboratory factor was calculated. Results: Of the 22 215 neonates, 2096 were assessed. IAP among infants with risk factors rose from 68% in 2005 to 78% in 2008 (p = 0.001). A total of 1662 asymptomatic infants had risk factors, 635 received antibiotics and one (0.06%) had sepsis. A total of 434 symptomatic infants were treated with antibiotics and of these 234 had risk factors and 20 (4.6%) had sepsis. No cases of partially treated or missed sepsis were detected. Poor predictive value was found for all risk factors except prematurity and leukopenia. Conclusion: The risk factor based approach in asymptomatic infants cannot be justified. In-hospital observation of asymptomatic infants for 2-3 days with antibiotic treatment being reserved only for symptomatic infants may be a reasonable alternative.

Original languageEnglish
Pages (from-to)e540-e544
JournalActa Paediatrica, International Journal of Paediatrics
Issue number12
StatePublished - Dec 2012
Externally publishedYes


  • Early neonatal sepsis
  • Intrapartum antibiotic prophylaxis
  • Risk factors


Dive into the research topics of 'Limitations of the risk factor based approach in early neonatal sepsis evaluations'. Together they form a unique fingerprint.

Cite this