Kaiser Escobar 2014 — Approximation

Kaiser EOS Calculator (Approximate)

Estimate neonatal early-onset sepsis risk in newborns ≥34 weeks using a simplified scoring model based on the Kaiser Permanente framework.

This is a simplified scoring approximation, NOT the proprietary Escobar regression. For the validated published probability, use the canonical Kaiser tool: neonatalsepsiscalculator.kaiserpermanente.org

Infant & Maternal Risk Factors

Range: 34–43 weeks. Enter decimal weeks (e.g. 38.4).
Default 0.5. Range 0.1–4.0.
No fever default: 37.0°C (98.6°F)
Hours from rupture to delivery. Range 0–96.
Your data never leaves this device. No accounts. No servers. No tracking.

Frequently Asked Questions About Neonatal Early-Onset Sepsis & the Kaiser EOS Calculator

Neonatal early-onset sepsis (EOS) is a bloodstream infection occurring in the first 72 hours of life (some definitions extend to 7 days). It is caused by bacteria acquired from the maternal genital tract during labor and delivery, most commonly Group B Streptococcus (GBS) and Escherichia coli.

Key risk factors include maternal GBS colonization, inadequate intrapartum antibiotic prophylaxis, prolonged rupture of membranes (≥18 hours), maternal fever or chorioamnionitis, and lower gestational age. EOS in term and late-preterm infants (≥34 weeks) is relatively rare, occurring in approximately 0.5 per 1000 live births in North America, but carries significant morbidity and mortality when it occurs.

The Kaiser Permanente EOS Calculator is a validated clinical decision support tool based on a multivariate logistic regression model developed by Escobar GJ and colleagues (Pediatrics 2014). It estimates the probability of early-onset sepsis using maternal risk factors and is refined by the infant's post-delivery clinical exam.

The tool was validated in a large cohort of over 600,000 newborns ≥34 weeks gestation in Kaiser Permanente Northern California. The original regression model incorporating maternal risk factors was published by Puopolo KM et al. (Pediatrics 2011). The canonical online tool is freely available at neonatalsepsiscalculator.kaiserpermanente.org.

This page provides a simplified approximation, not the validated regression. Always use the canonical tool for the published probability.

The Escobar 2014 logistic regression uses proprietary coefficients that are not fully published in the open literature in a form that allows complete reproduction. The Kaiser Permanente online tool implements the validated regression. To avoid presenting a number that could be mistaken for the validated probability, this page uses a simplified multiplicative scoring scheme as an approximation and prominently links to the canonical tool.

The approximate model captures the same qualitative direction of risk (higher maternal temperature, longer ROM duration, GBS-positive with no prophylaxis all increase risk; adequate broad-spectrum prophylaxis decreases it) but the exact numeric output will differ from the canonical tool. Do not use this page's probability output for clinical decisions — use the canonical Kaiser tool for that.

The Kaiser EOS calculator is validated for newborns ≥34 weeks gestation (late-preterm and term). This includes late-preterm infants (34–36 weeks), early-term (37–38 weeks), and full-term (39–41 weeks) newborns.

It is not intended for very preterm infants (<34 weeks), who have different sepsis risk epidemiology, different pathogen profiles, and are typically managed under separate neonatal intensive care unit protocols. Very preterm infants routinely receive empirical antibiotics due to prematurity itself, independent of the EOS calculator.

The Kaiser EOS framework integrates pre-delivery maternal risk factors with the infant's post-delivery clinical exam to determine management. The three exam categories are:

  • Well-appearing: No signs of cardiopulmonary or neurological distress. With low pre-test risk, routine care without workup is appropriate.
  • Equivocal: Transient signs of distress (e.g., tachypnea, temperature instability) that are improving. Typically warrants close observation: vital signs every 4 hours for 24 hours and reassessment.
  • Clinical illness: Persistent respiratory distress, hemodynamic instability, or signs of systemic illness. Warrants blood culture and empirical antibiotic therapy regardless of pre-test probability.

Clinical exam is the dominant factor: a well-appearing infant with high pre-test risk may not require immediate workup, while a clinically ill infant should be treated regardless of pre-test risk tier.

Adequate GBS intrapartum prophylaxis per CDC/ACOG guidelines requires:

  • Penicillin G (preferred), ampicillin, or cefazolin administered ≥4 hours before delivery
  • Alternative agents (clindamycin, vancomycin) for penicillin-allergic patients may be considered adequate when given sufficiently early, but require susceptibility testing confirmation

Prophylaxis is indicated for GBS-positive mothers, those with unknown GBS status at <37 weeks or with specific risk factors, or those who had a prior infant with GBS disease. The Kaiser calculator specifically differentiates between broad-spectrum antibiotics ≥4 hours before delivery (most protective), GBS-specific antibiotics 2–3.9 hours before delivery, GBS-specific antibiotics <2 hours before delivery (insufficient), and no antibiotics.

Maternal intrapartum fever (≥38.0°C / 100.4°F) is one of the strongest risk factors for neonatal EOS and is a key component of the Kaiser calculator. Higher temperatures confer incrementally greater risk.

Maternal fever during labor may indicate chorioamnionitis (infection of the amniotic fluid, membranes, and/or placenta), which dramatically increases EOS risk. Chorioamnionitis diagnosis is clinical and may be based on fever alone or combined with other signs (uterine tenderness, foul-smelling amniotic fluid, fetal tachycardia, maternal leukocytosis).

Even epidural-associated fever (not due to infection) has been associated with increased neonatal evaluations and antibiotic exposure, highlighting the clinical challenge in distinguishing infectious from non-infectious maternal fever.