FactsheetNeonatal infectionEarly-onset sepsis
A summary of six guideline documents from four bodies on early-onset sepsis in newborns: the AAP (three reports, 2018–2019), the Canadian Paediatric Society (2017), NICE NG195 (updated May 2026) and the American College of Critical Care Medicine (2017). It covers risk assessment, investigations, antibiotics, prevention and newborn septic shock.
The AAP defines early-onset sepsis microbiologically: a blood or cerebrospinal fluid (CSF) culture obtained within 72 hours after birth growing a pathogenic bacterial species.1
How the sources express strength. The AAP clinical reports give numbered summary points without formal grades. NICE conveys strength through its wording (“use”, “give”, “consider”) and shows the year each recommendation was made or amended in brackets. The CPS statement uses a modified GRADE system (strong or weak recommendation; quality of evidence).
Boxes labelled “guideline text” report discussion or data from the source, not recommendations.
01Definition and burden
Definitions of the early-onset window differ. The AAP uses culture within 72 hours after birth.1 NICE uses “early-onset neonatal infection” for infection less than 72 hours after birth.6 The CPS statement defines early-onset sepsis as sepsis occurring within the first seven days of life.4
The current overall incidence of EOS in the United States is approximately 0.8 cases per 1000 live births. Incidence is approximately 0.5 per 1000 at ≥37 weeks, 1 per 1000 at 34 to 36 weeks, 6 per 1000 at <34 weeks, 20 per 1000 at <29 weeks and 32 per 1000 at 22 to 24 weeks.1
Because gestational age is the strongest predictor of EOS, and about two-thirds of preterm births are associated with preterm labour, premature rupture of membranes (PROM) or clinical chorioamnionitis, the AAP states that risk stratification strategies cannot be applied to preterm infants in the same manner as for term neonates.1
02Assessing risk at 35 weeks and over
The AAP describes three approaches and states that each is a reasonable approach to risk assessment for infants born at ≥35 weeks. It also states that no strategy can be used to immediately identify all infants who will develop EOS, or avoid the treatment of a substantial number of infants who are uninfected; so every strategy must include measures to monitor infants not initially identified, and to minimise the duration of antibiotics in uninfected infants.2
Evidence supports the predictive value of gestational age, maternal intraamniotic infection (represented either by intrapartum fever or the obstetric clinical diagnosis of chorioamnionitis), duration of rupture of membranes, maternal GBS colonisation, administration of appropriate intrapartum antibiotic therapy, and the newborn clinical condition.2
Twin gestation, fetal tachycardia and meconium-stained liquor are associated with EOS but are not independent predictors in multivariable analyses.2
Approach 1 · Categorical risk-factor assessment
Threshold values for intrapartum risk factors sort babies into categories. The AAP describes the categories used in these algorithms, and the action attached to each.2
- 1. Any newborn infant who is ill appearing
- Laboratory testing and empirical antibiotic therapy.2
- 2. Mother with a clinical diagnosis of chorioamnionitis
- Laboratory testing and empirical antibiotic therapy.2
- 3. Mother colonised with GBS who received inadequate IAP, with ROM >18 hours or birth before 37 weeks
- Laboratory testing.2
- 4. Mother colonised with GBS who received inadequate IAP, no additional risk factors
- Observation in hospital for ≥48 hours.2
The thresholds the categories turn on. Two definitions do most of the work here.
Intraamniotic infection (ACOG, via AAP)
- Confirmed: positive amniotic fluid Gram stain and/or culture, or placental histopathology.2
- Suspected: maternal intrapartum fever — either a single temperature ≥39.0 °C, or 38.0–38.9 °C persisting >30 minutes — plus one or more of maternal leucocytosis, purulent cervical drainage, fetal tachycardia.2
- ACOG recommends intrapartum antibiotics whenever intraamniotic infection is diagnosed or suspected, and when otherwise unexplained maternal fever occurs in isolation.2
Adequate IAP (CPS)
- At least one dose, given at least 4 hours before birth, of IV penicillin G (initial dose 5 million units) or ampicillin (initial dose 2 g).4
- Or IV cefazolin (initial dose 2 g) if the mother is allergic to penicillin but at low risk for anaphylaxis.4
- Clindamycin and vancomycin regimens should be considered inadequate IAP when managing the neonate, because their efficacy has not been confirmed in clinical trials.4
The CPS statement works the same thresholds into a ward-level protocol. Its risk factors are maternal GBS colonisation in the current pregnancy, GBS bacteriuria at any time in the current pregnancy, a previous infant with invasive GBS disease, prolonged rupture of membranes ≥18 hours, and maternal fever ≥38 °C.4
- Clinical signs of sepsis: respiratory distress, temperature instability, tachycardia, seizures, hypotonia, lethargy, poor peripheral perfusion, hypotension, acidosis
- Prompt investigation including CBC, blood culture and lumbar puncture, and empirical IV antibiotics (ampicillin and an aminoglycoside). Chest x-ray if respiratory signs. Strong, moderate quality4
- Early respiratory signs only, no risk factors
- May be observed for up to 6 hours before starting investigations and antibiotics. Strong, low quality4
- GBS-positive with adequate IAP and no other risk factors (or GBS-negative/unknown with one other risk factor and adequate IAP)
- No investigation or treatment. May go home after 24 hours if well, other discharge criteria are met, and parents understand the signs of sepsis and when to seek care. Strong4
- GBS-positive with inadequate IAP and no other risk factors (or GBS-negative/unknown with one other risk factor and inadequate IAP)
- Examine at birth, observe closely in hospital with vital signs every 3 to 4 hours, reassess before discharge; home after 24 hours if well, with ready access to care. Routine investigation or treatment is not required. Strong, low quality4
- Multiple risk factors and/or chorioamnionitis
- Individualised approach considering the severity of risk factors and maternal antibiotics. At minimum, close observation in hospital for at least 24 hours with vital signs every 3 to 4 hours and reassessment before discharge. A CBC after 4 hours of age may help: WBC <5 × 10⁹/L and ANC <1.5 × 10⁹/L have the highest positive predictive value. Some infants may warrant investigation and antibiotics. Weak, low quality4
- Well late-preterm infants, 35 to 36 weeks
- If stable enough to stay with their mother, manage as for ≥37 weeks, but observe in hospital for at least 48 hours. Weak, low quality4
The AAP lists these limitations: a lack of clear definitions for newborn clinical illness, difficulty establishing the clinical diagnosis of maternal chorioamnionitis, inconsistent consideration of intrapartum antibiotics, and absence of guidance on what defines an abnormal laboratory result in the newborn.2
In a review of nearly 400 000 newborn infants, approximately 450 term infants exposed to chorioamnionitis would have to be treated per case of confirmed EOS. The AAP says these data provide a strong argument against using the clinical diagnosis of chorioamnionitis as a sole indicator of risk in term infants.2
Approach 2 · Multivariate risk assessment (sepsis calculator)
An individualised synthesis of established risk factors and the newborn clinical condition, estimating each infant’s risk. The predictive models were developed from a cohort of 608 000 newborn infants, using objective data known at birth and the evolving newborn condition during the first 6 to 12 hours.2
- Objective data entered
- Gestational age · highest maternal intrapartum temperature · maternal GBS colonisation status · duration of ROM · type and duration of intrapartum antibiotic therapies.2
- Newborn clinical condition
- Infants are classified as clinically ill, equivocal or well appearing. The AAP notes this classification requires ongoing clinical assessment over the first 12 hours after birth.2
- Risk ≥1 per 1000 live births
- Blood culture and enhanced clinical observation.2
- Risk ≥3 per 1000 live births
- Blood culture and empirical antibiotic therapy.2
- Local thresholds
- Institutions may set different risk thresholds for specific actions than those validated, but should consider quantifying the effect of the chosen thresholds to affirm safety and efficacy.2
In a prospective validation in 204 685 infants, blood culture testing declined by 66% and empirical antibiotic administration by 48%, compared with the previous CDC categorical algorithm. No adverse effects were noted during birth hospitalisation. Readmissions for culture-confirmed infection in the week after discharge were rare (approximately 5 in 100 000 births) and did not differ by approach.2
The AAP lists three advantages: it gives differential information on an individual infant’s risk rather than placing infants in categories with a wide range of risk; it includes only objective data and not a clinical diagnosis of maternal chorioamnionitis; and it results in relatively few well-appearing infants being treated empirically. The concern is the workload: it requires increased clinical surveillance in the well nursery or postpartum unit, and changes to the frequency of vital signs and assessments for infants at moderate risk.2
Approach 3 · Risk assessment based primarily on the newborn’s clinical condition
Regardless of any estimation of neonatal or maternal risk factors, infants who appear ill at birth, and those who develop signs of illness over the first 48 hours after birth, are either treated empirically with antibiotics or further evaluated by laboratory screening.2
- What counts as a sign of illness (CPS)
- Respiratory distress, temperature instability, tachycardia, seizures, hypotonia, lethargy, poor peripheral perfusion, hypotension, acidosis. The CPS notes that initial signs may be subtle.4
- How often to look
- In the cohort the AAP cites, 7611 infants were managed with serial physical examinations every 4 to 6 hours through 48 hours of age, with significant decreases in laboratory tests, blood cultures and empirical antibiotics compared with 7628 infants managed categorically. Two infants who developed EOS were identified as they developed signs of illness.2 The CPS protocol uses vital signs every 3 to 4 hours for at least 24 hours in its observation groups.4
- Where to start
- Institutions adopting this approach need to decide whether to use a categorical or multivariate approach to identify at-risk infants for serial monitoring, or to apply it to all newborn infants.2
- How well examination performs
- Good clinical condition at birth is associated with a reduction in risk for EOS of approximately 60% to 70%.2 In an analysis of 2785 infants with birth weights ≥2000 g, the sensitivity and negative predictive value of clinical signs were 92% and 99%, compared with sensitivities below 50% for a low ANC and an increased I:T ratio.4
- An NICHD panel position
- Flag infants for EOS risk on the basis of the obstetric diagnosis of suspected intraamniotic infection, but rely primarily on clinical observation alone for well-appearing term and late-preterm infants.2
The AAP states that physicians and families must understand that identifying initially well-appearing infants who go on to develop clinical illness is, in its words, “not a failure of care but rather an anticipated outcome” of this approach to EOS risk management.2
The NICE framework
- Any red flag, or 2 or more non-red-flag risk factors or clinical indicators
- Follow the recommendations on investigations before starting antibiotics, start antibiotic treatment, and do not wait for the test results before starting antibiotics. NICE 20216
- No red flags and only 1 risk factor or 1 clinical indicator
- Use clinical judgement to decide whether it is safe to withhold antibiotics, and whether vital signs and clinical condition need monitoring; if monitoring is needed, continue for at least 12 hours using a newborn early warning system. NICE 20216
- No risk factors or clinical indicators
- Continue routine postnatal care. NICE 20216
- Sepsis calculator
- The Kaiser Permanente neonatal sepsis calculator can be used as an alternative to this framework for babies born after 34+0 weeks cared for in a neonatal unit, transitional care or postnatal ward. It should only be used if it is part of a prospective audit. NICE 20216
- Timing
- If a baby needs antibiotic treatment, give this as soon as possible and always within 1 hour of the decision to treat. NICE 20216
The NICE risk factors, grouped
Red flag risk factor
- In a multiple pregnancy, suspected or confirmed infection in another baby
Group B streptococcus
- A previous baby with invasive group B streptococcal infection
- Maternal group B streptococcal colonisation, bacteriuria or infection in the current pregnancy
Prematurity and membranes
- Pre-term birth following spontaneous labour before 37 weeks’ gestation
- Confirmed rupture of membranes: more than 18 hours before a pre-term birth, or more than 24 hours before a term birth 24 h added 2026
Maternal infection
- Suspected or confirmed chorioamnionitis amended 2026
- Suspected or confirmed maternal sepsis in the intrapartum or early postpartum period amended 2026
The NICE clinical indicators, grouped
Red flag clinical indicators
- Signs of shock; seizures; apnoea; need for mechanical ventilation; need for cardiopulmonary resuscitation
Breathing and oxygenation
- Signs of respiratory distress, including grunting, recession and tachypnoea
- Hypoxia — for example central cyanosis or a reduced oxygen saturation level
- Persistent pulmonary hypertension of newborns
Circulation, tone and behaviour
- Abnormal heart rate, either bradycardia or tachycardia
- Altered behaviour or responsiveness; altered muscle tone, for example floppiness
- Signs of neonatal encephalopathy
Feeding
- Feed intolerance, including vomiting, excessive gastric aspirates and abdominal distension
- Feeding difficulties, for example feed refusal
Temperature, blood and metabolic
- Metabolic acidosis, with a base deficit of 10 mmol/litre or greater
- Altered glucose homeostasis, meaning hypoglycaemia or hyperglycaemia
- Unexplained excessive bleeding, thrombocytopenia, or abnormal coagulation
- Jaundice within 24 hours of birth
- Temperature abnormality — lower than 36 °C or higher than 38 °C — unexplained by environmental factors
Both lists carry every item NICE lists, using NICE’s own clinical terms, but grouped here by system rather than in NG195’s order. NICE’s boxes are the definitive version: see NG195, recommendation 1.9 and box 2, and recommendation 1.10.3 for the framework that acts on them.6
03Assessing risk below 35 weeks
The AAP states that the Neonatal Early-Onset Sepsis Risk Calculator does not apply to infants born before 34 0/7 weeks, and that preterm infants can be categorised by level of risk by the circumstances of their preterm birth.1
Highest risk
- Circumstances
- Maternal cervical incompetence, preterm labour, PROM, clinical concern for intraamniotic infection (IAI), or acute onset of unexplained nonreassuring fetal status.1
- AAP summary point
- Such neonates should undergo EOS evaluation with a blood culture and empirical antibiotic treatment.1
- CSF
- Obtaining CSF for culture before antibiotics should be considered if the infant will tolerate the procedure and it will not delay antibiotic therapy.1
Lower risk (all three criteria)
- Criteria
- (1) Obstetric indications for preterm birth, such as maternal preeclampsia or other noninfectious medical illness or placental insufficiency; (2) birth by caesarean delivery; (3) absence of labour, attempts to induce labour, or any rupture of membranes before delivery.1
- Acceptable initial approaches
- (1) No laboratory evaluation and no empirical antibiotic therapy, or (2) a blood culture and clinical monitoring.1
- If not improving
- For infants who do not improve after initial stabilisation and/or have severe systemic instability, empirical antibiotics may be reasonable but are not mandatory.1
04Investigations
- Diagnosis
- The diagnosis of EOS is made by using blood or CSF cultures. EOS cannot be diagnosed by using laboratory tests, such as a complete blood cell count or CRP, or by surface cultures, gastric aspirate analysis or urine culture. AAP summary point2
- Blood culture volume
- Paediatric blood culture bottles generally require a minimum of 1 mL of blood for optimal recovery of organisms.1
- White cell indices and CRP
- White blood cell indices or a single C-reactive protein should not be used routinely as screening or diagnostic tests for EOS, nor to routinely exclude EOS. CPS: strong, moderate quality4
- CRP alone
- Do not use C-reactive protein results alone to make decisions about starting, continuing or stopping antibiotics. NICE 20266
After 4 hours of age, the likelihood ratio for sepsis was 80.5 for a white cell count <5 × 10⁹/L, 115 for an absolute neutrophil count <1 × 10⁹/L, and 10.7 for an I:T ratio >0.6 (CPS).4
Single values of CRP or procalcitonin obtained after birth are neither sensitive nor specific to guide EOS care. Consistently normal values over the first 48 hours are associated with the absence of EOS, but serial abnormal values alone should not be used to decide whether to give antibiotics in the absence of culture-confirmed infection (AAP).2
8% to 40% of infants with early-onset meningitis are reported to have negative blood cultures. In term infants, a CSF white cell count >20–25 cells/mm³ is considered abnormal, with a sensitivity of 79% and specificity of 81% for bacterial meningitis (CPS).4
05Empirical antibiotics
AAP (US)
- First choice
- The combination of ampicillin and gentamicin is the appropriate empirical antibiotic regimen for most infants at risk for EOS.2
NICE (UK)
- First choice
- Use intravenous benzylpenicillin sodium with gentamicin as the first-choice regimen for empirical treatment of suspected early-onset infection, unless microbiological surveillance data show local resistance patterns that indicate the need for a different antibiotic. NICE 20126
- Doses
- Benzylpenicillin sodium 25 mg/kg every 12 hours; consider every 8 hours based on clinical judgement (for example, if the baby appears very ill). Gentamicin starting dose 5 mg/kg; a second dose, if given, should usually be 36 hours after the first. NICE 20126
Although two-thirds of E. coli EOS isolates are ampicillin-resistant, most remain susceptible to gentamicin, and the AAP states that routine empirical use of broader-spectrum agents is typically not justified and may be harmful. Nonetheless, approximately 7% of E. coli cases (1.7% of all EOS cases) were resistant to both ampicillin and gentamicin in recent CDC surveillance.2
06Stopping and duration
- AAP: the rule
- When blood cultures are sterile, antibiotic therapy should be discontinued by 36 to 48 hours of incubation, unless there is clear evidence of site-specific infection.1
- AAP: the common trap
- Persistent cardiorespiratory instability is common among preterm infants with VLBW, and is not alone an indication for prolonged empirical antibiotics. Laboratory test abnormalities alone rarely justify prolonged empirical antibiotics, particularly among preterm infants at lower risk.1
- NICE: stop at 36 hours
- For babies treated because of risk factors or clinical indicators, stop antibiotics at 36 hours if: the blood culture is negative; the initial clinical suspicion of infection was not strong; the baby’s clinical condition is reassuring with no clinical indicators of possible infection; and the levels and trends of CRP are reassuring. NICE 2012, amended 20266
- NICE: total course
- Give intravenous antibiotic treatment for a total of 7 days for babies with a positive blood culture, or a negative blood culture if sepsis has been strongly suspected. NICE 20126
- NICE: oral switch
- For babies born from 35+0 weeks with a negative blood culture who need antibiotics for more than 36 hours because initial suspicion was strong, consider switching from intravenous to oral antibiotics at 36 hours or thereafter to complete the course (up to 7 days in total) if the clinical condition is reassuring, CRP levels and trends are reassuring, and the baby is tolerating oral feeds. Use amoxicillin unless local resistance data indicate otherwise. NICE 20266
- Confirmed GBS (AAP)
- Treatment is generally 10 days for bacteraemia without focus and 14 days for uncomplicated meningitis, given intravenously for the entire course.3
NICE also says decisions on switching to oral antibiotics or going home should be agreed by a senior neonatologist or paediatrician (consultant or similar level), and that parents of babies on oral antibiotics should have at least 2 follow-up consultations, including 1 at the end of treatment.6
07Prevention: intrapartum antibiotic prophylaxis
The American College of Obstetricians and Gynecologists currently recommends universal antenatal testing for GBS colonisation using vaginal-rectal cultures at 36 0/7 to 37 6/7 weeks. Penicillin G has a narrower antimicrobial spectrum than ampicillin and remains the preferred agent, but ampicillin is acceptable. In one study, intrapartum prophylaxis given ≥4 hours before delivery was most effective in preventing early-onset GBS disease.3
Cefazolin is recommended for women with a penicillin allergy at low risk of anaphylaxis; clindamycin for those at high risk of anaphylaxis whose GBS is known to be susceptible to clindamycin.3
The CPS statement notes that because the efficacy of clindamycin and vancomycin regimens has not been confirmed in clinical trials, they should be considered inadequate intrapartum prophylaxis when managing the neonate, and that prophylaxis is not recommended when a caesarean section is performed before the onset of labour with intact membranes, regardless of GBS status.4
08Septic shock in the newborn
The ACCM 2017 parameters give separate recommendations for newborns; those below are from the newborn section.5
- Access
- Placement of an umbilical arterial and venous catheter is preferred. Intraosseous access, particularly in preterm newborns, is not the preferred route of drug administration. Level 1D5
- Fluids
- Fluid boluses of 10 mL/kg can be administered, observing for hepatomegaly and increased work of breathing. Up to 60 mL/kg may be required in the first hour. Level 1C5
- Glucose
- A 10% dextrose-containing isotonic solution at maintenance rate provides age-appropriate glucose to prevent hypoglycaemia.5
- Inotropes
- Low-dose dopamine (<8 µg/kg/min) with dobutamine (up to 10 µg/kg/min) is initially recommended; if the response is inadequate, adrenaline (epinephrine) 0.05–0.3 µg/kg/min can be infused. Level 1C5
- PPHN
- Inhaled nitric oxide should be administered as the first treatment when available; milrinone and inhaled iloprost are back-up therapies. Level 1B5
References
- Puopolo KM, Benitz WE, Zaoutis TE; AAP Committee on Fetus and Newborn, AAP Committee on Infectious Diseases. Management of neonates born at ≤34 6/7 weeks’ gestation with suspected or proven early-onset bacterial sepsis. Pediatrics. 2018;142(6):e20182896. doi:10.1542/peds.2018-2896
- Puopolo KM, Benitz WE, Zaoutis TE; AAP Committee on Fetus and Newborn, AAP Committee on Infectious Diseases. Management of neonates born at ≥35 0/7 weeks’ gestation with suspected or proven early-onset bacterial sepsis. Pediatrics. 2018;142(6):e20182894. doi:10.1542/peds.2018-2894
- Puopolo KM, Lynfield R, Cummings JJ; AAP Committee on Fetus and Newborn, AAP Committee on Infectious Diseases. Management of infants at risk for group B streptococcal disease. Pediatrics. 2019;144(2):e20191881. doi:10.1542/peds.2019-1881
- Jefferies AL; Canadian Paediatric Society, Fetus and Newborn Committee. Management of term infants at increased risk for early onset bacterial sepsis. Paediatr Child Health. 2017;22(4):223–228. doi:10.1093/pch/pxx023
- Davis AL, Carcillo JA, Aneja RK, et al. American College of Critical Care Medicine clinical practice parameters for hemodynamic support of pediatric and neonatal septic shock. Crit Care Med. 2017;45(6):1061–1093. doi:10.1097/CCM.0000000000002425
- National Institute for Health and Care Excellence. Neonatal infection: antibiotics for prevention and treatment. NICE guideline NG195. Published April 2021; last updated 13 May 2026. nice.org.uk/guidance/ng195
Tiny Taught · Neonatal & Paediatric Education
Recommendations and summary points are adapted from the cited sources and follow their wording as closely as possible. Items marked “guideline text” are not recommendations. Doses and settings are for orientation only; local guidelines apply.
Original layout and graphics © Tiny Taught. Clinical content is summarised from the cited sources, which remain the copyright of their authors and publishers. NICE guidance remains © NICE and is summarised here subject to NICE’s reuse terms.