# Neonatal Sepsis: Early-Onset and Late-Onset
## Overview
Neonatal sepsis remains a leading cause of morbidity and mortality in newborns. It is defined by timing of onset: early-onset sepsis (EOS) occurs within the first 72 hours of life (some definitions extend to 7 days), while late-onset sepsis (LOS) presents after 72 hours. The incidence of EOS is approximately 0.5-1 per 1000 live births in term infants but rises dramatically to 10-20 per 1000 in very low birth weight infants. The nonspecific nature of clinical signs in neonates makes diagnosis particularly challenging, and a high index of suspicion is essential for timely treatment.
## Microbiology
### Early-Onset Sepsis
The organisms responsible for EOS are acquired through vertical transmission from the maternal genital tract, either by ascending infection through ruptured membranes or during passage through the birth canal. Group B Streptococcus (GBS) remains the most common cause in term infants despite widespread intrapartum antibiotic prophylaxis. In preterm infants, Escherichia coli has become the predominant pathogen, with an increasing prevalence of ampicillin-resistant strains. Listeria monocytogenes is uncommon but clinically important. Other organisms include Enterococcus, Staphylococcus aureus, and various gram-negative bacilli.
### Late-Onset Sepsis
LOS is acquired through horizontal transmission, typically nosocomial in origin, from skin flora or gastrointestinal translocation. Coagulase-negative staphylococci (CoNS) are the most common isolates in the NICU, predominantly associated with central line infections. Staphylococcus aureus (including MRSA), gram-negative bacilli (Klebsiella, Enterobacter, Pseudomonas, Serratia), and Candida species are also important pathogens. Candida is particularly concerning in extremely low birth weight infants on prolonged antibiotics. GBS can also cause late-onset disease, most commonly presenting as meningitis.
## Risk Factors
### EOS Risk Factors
The most important modifiable risk factor for EOS is maternal GBS colonization. Other significant risk factors include prolonged rupture of membranes (greater than 18 hours), clinical chorioamnionitis (manifesting as maternal fever, uterine tenderness, and fetal tachycardia), prematurity itself, inadequate intrapartum antibiotic prophylaxis, and a history of a previous infant with invasive GBS disease.
### LOS Risk Factors
For late-onset sepsis, prematurity and low birth weight are the strongest risk factors, compounded by prolonged NICU hospitalization. Central venous catheters create a direct portal of entry for skin organisms, making central line-associated bloodstream infections (CLABSI) a major contributor. Mechanical ventilation, total parenteral nutrition, and intralipid administration each independently increase risk. Prolonged antibiotic exposure promotes colonization with resistant organisms and Candida. Abdominal pathology, particularly necrotizing enterocolitis, provides a source for bacterial translocation.
## Clinical Presentation
The clinical signs of neonatal sepsis are notoriously nonspecific and overlap with many other neonatal conditions. Respiratory findings include apnea (often the first sign in preterm infants), tachypnea, grunting, and increasing oxygen requirements. Cardiovascular signs range from tachycardia to bradycardia, poor perfusion, and hypotension (a late and ominous finding). Temperature instability is common, with hypothermia more frequent than fever in preterm infants and fever more common in term infants. Neurologic signs include lethargy, irritability, poor feeding, and seizures when meningitis is present. Gastrointestinal manifestations include feeding intolerance, abdominal distension, and bilious emesis. Hematologic abnormalities such as jaundice, petechiae, and overt bleeding from disseminated intravascular coagulation may develop. Metabolic derangements including hypoglycemia and metabolic acidosis round out the clinical picture.
## Diagnostic Workup
### Blood Culture
The blood culture is the gold standard for diagnosing sepsis and should be obtained before starting antibiotics. A minimum of 1 mL of blood should be collected, with two separate cultures being ideal. Sensitivity is limited by the low-level bacteremia common in neonates and the small blood volumes available for sampling. With modern automated detection systems, most true-positive cultures are identified within 24-36 hours of incubation.
### Complete Blood Count
The complete blood count with differential provides supportive but not definitive diagnostic information. Concerning findings include leukopenia (less than 5000), leukocytosis (greater than 30,000 in term infants), and a left shift with an immature-to-total neutrophil ratio greater than 0.2. However, individual CBC parameters have low sensitivity and specificity for sepsis. Thrombocytopenia is a late and nonspecific marker. Serial CBCs obtained over the first 24-48 hours may improve diagnostic utility.
### Inflammatory Markers
C-reactive protein (CRP) rises 6-12 hours after the onset of infection, meaning a single value at the time of initial evaluation may be falsely reassuring. Serial CRP measurements at presentation and again at 24-48 hours achieve approximately 90% sensitivity for ruling out infection and are useful for guiding antibiotic duration. Procalcitonin has a physiologic surge in the first 24-48 hours of life that limits its utility for EOS evaluation, but it is more useful in LOS. The NeoPINS trial demonstrated that procalcitonin-guided therapy can safely reduce antibiotic duration in neonates.
### Lumbar Puncture
Lumbar puncture is indicated in all infants with a positive blood culture, when there is clinical suspicion for meningitis (seizures, irritability, bulging fontanelle), and in all evaluations for late-onset sepsis. Its role in well-appearing term infants undergoing EOS evaluation remains controversial. CSF interpretation in neonates differs from older children: normal values include up to 20-25 WBC/mm3 and protein up to 150 mg/dL in preterm infants. Culture remains the gold standard, though CSF PCR panels are increasingly available and can provide rapid results.
### Urinalysis and Urine Culture
Urine studies are not useful in the evaluation of EOS because bacteriuria in this setting represents hematogenous seeding rather than ascending urinary tract infection. However, a catheterized urine specimen for culture is important in the evaluation of LOS in infants beyond 7 days of life.
### Chest X-ray
Chest radiography should be obtained when respiratory symptoms are present. Notably, GBS pneumonia can be radiographically indistinguishable from respiratory distress syndrome, making clinical correlation essential.
## Approach to EOS in Term and Late-Preterm Infants
### Traditional Risk Factor-Based Approach
The AAP 2010 algorithm relies on maternal risk factors including GBS colonization status, duration of rupture of membranes, presence of chorioamnionitis, and adequacy of intrapartum prophylaxis to determine which infants require evaluation. While this approach has high sensitivity, it has notably low specificity, leading to laboratory evaluation and empiric antibiotic treatment of many uninfected infants. Under this paradigm, approximately 10-15% of all newborns receive blood cultures and empiric antibiotics.
### Kaiser Neonatal EOS Calculator
The Kaiser Neonatal EOS Calculator represents a more refined approach, using a multivariate prediction model that incorporates gestational age, duration of membrane rupture, highest maternal temperature, GBS colonization status, and type of intrapartum antibiotics. It calculates an individualized EOS risk per 1000 live births and then stratifies infants into three clinical categories: well-appearing, equivocal, and clinical illness. This approach significantly reduces sepsis evaluations and antibiotic use by 40-60% compared to the traditional algorithm and has been validated in multiple large cohorts. Its limitations include inapplicability to infants below 34 weeks gestation, dependence on accurate maternal data, and the rare possibility of missed cases.
### Serial Physical Examination Approach
Well-appearing term infants with risk factors can alternatively be monitored with serial vital signs and clinical assessments performed every 4 hours for 24-48 hours. No routine laboratory testing is obtained unless clinical signs develop. This approach reduces unnecessary testing and avoids mother-infant separation, but it requires reliable nursing assessment and ensured follow-up.
## Empiric Antibiotic Therapy
### EOS
The standard first-line regimen for early-onset sepsis is ampicillin plus gentamicin, which provides coverage against GBS, Listeria, and most E. coli strains. Extended-interval (once daily) gentamicin dosing is gaining favor based on pharmacokinetic optimization. Duration of therapy depends on results: if blood cultures remain negative at 36-48 hours and the infant appears clinically well, antibiotics should be discontinued. Culture-positive sepsis requires 10-14 days of treatment, extended to 14-21 days for meningitis depending on the organism. The diagnosis of "culture-negative sepsis" should be used sparingly, as it is frequently overdiagnosed.
### LOS
Empiric therapy for late-onset sepsis typically consists of vancomycin plus gentamicin (or an anti-pseudomonal beta-lactam), providing coverage for CoNS, MRSA, and gram-negative organisms. The regimen should be tailored based on the local antibiogram. Amphotericin B should be added when there is high risk for invasive candidiasis, particularly in extremely low birth weight infants with prolonged antibiotic exposure and central venous lines. Duration is guided by the identified organism, source, and clinical response.
| Feature | Early-Onset Sepsis (EOS) | Late-Onset Sepsis (LOS) |
|---------|-------------------------|------------------------|
| Timing | 0-72 hours of life | >72 hours of life |
| Transmission | Vertical (maternal) | Horizontal (nosocomial) |
| Common organisms | GBS, E. coli, Listeria | CoNS, S. aureus, Gram-negatives, Candida |
| Key risk factors | Chorioamnionitis, GBS colonization, PROM | Central lines, prolonged antibiotics, TPN |
| Empiric antibiotics | Ampicillin + Gentamicin | Vancomycin + Gentamicin (± Amphotericin B) |
| Duration (culture-positive) | 10-14 days (14-21 for meningitis) | Organism/source-dependent |
| Duration (culture-negative) | Stop at 36-48 hours | Stop at 36-48 hours |
## Antibiotic Stewardship in the NICU
Prolonged empiric antibiotic exposure exceeding 5 days in culture-negative infants has been associated with increased risk of necrotizing enterocolitis, increased late-onset sepsis, increased mortality, and alteration of the developing microbiome. Stewardship strategies include utilizing sepsis calculators for risk stratification, following serial biomarkers to guide treatment duration, preferring narrow-spectrum agents when possible, and establishing clear stop criteria for discontinuing antibiotics in well-appearing infants with negative cultures.
A flowchart comparing three approaches to evaluating well-appearing term infants for early-onset sepsis: the traditional risk-factor based algorithm (AAP 2010), the Kaiser Neonatal EOS Calculator pathway, and the serial physical examination approach. Show decision points, lab testing triggers, and antibiotic initiation criteria for each, with outcomes data on number needed to treat and missed cases. Color-coded clinical decision-support diagram.
An illustration depicting the pathogenesis of early-onset neonatal sepsis, showing ascending bacterial infection from the maternal vaginal flora (GBS, E. coli) through ruptured membranes to the amniotic fluid, with subsequent fetal aspiration and bacteremia. Include the placental-fetal interface and the neonate's immature immune system components (reduced complement, immature neutrophil function, low immunoglobulin levels). Medical cross-sectional illustration.
A timeline diagram showing the key differences between early-onset and late-onset neonatal sepsis, including timing (0-72 hours vs. >72 hours), common organisms (GBS/E. coli vs. CoNS/S. aureus/Candida), risk factors (maternal vs. nosocomial), and empiric antibiotic choices. Horizontal timeline format with branching information boxes above and below.
## Clinical Pearls
Apnea may be the first and only sign of neonatal sepsis, and any new apnea in a preterm infant warrants a sepsis evaluation. A negative blood culture at 36-48 hours with modern automated systems makes bacteremia highly unlikely, and empiric antibiotics should be discontinued at that point. The Kaiser EOS calculator significantly reduces unnecessary antibiotic exposure without increasing the rate of missed infections. "Culture-negative sepsis" is overdiagnosed in clinical practice; most culture-negative infants who receive full antibiotic courses do not actually have sepsis. Prolonged empiric antibiotics exceeding 5 days in culture-negative infants cause measurable harm and should be avoided through rigorous antibiotic stewardship. GBS meningitis requires higher-dose penicillin or ampicillin and a longer treatment duration of 14-21 days. Lumbar puncture should not be deferred if meningitis is clinically suspected; while CSF sterilization occurs rapidly after antibiotic administration, CSF cellular and biochemical parameters remain abnormal and diagnostically useful.
## References
- Puopolo KM, et al. Management of neonates born at >=35 0/7 weeks' gestation with suspected or proven early-onset bacterial sepsis. *Pediatrics*. 2018;142(6):e20182894.
- Puopolo KM, et al. Management of neonates born at <=34 6/7 weeks' gestation with suspected or proven early-onset bacterial sepsis. *Pediatrics*. 2018;142(6):e20182896.
- Kuzniewicz MW, et al. A quantitative, risk-based approach to the management of neonatal early-onset sepsis. *JAMA Pediatr*. 2017;171(4):365-371.
- Stoll BJ, et al. Early-onset neonatal sepsis 2015 to 2017: the rise of Escherichia coli. *JAMA Pediatr*. 2022;176(10):e223157.
- Stocker M, et al. Procalcitonin-guided decision making for duration of antibiotic therapy in neonates (NeoPINS). *JAMA*. 2017;318(12):1165-1175.