Residency · Residency · Pediatrics

Fever Without a Source in Infants

Overview

Fever without a source (FWS) in young infants under 90 days is one of the most common and anxiety-provoking presentations in pediatric emergency medicine. The risk of serious bacterial infection (SBI), including UTI, bacteremia, and meningitis, is highest in this age group. The clinical challenge is identifying the small percentage of infants with SBI among the majority who have self-limited viral infections. Traditional approaches used age-stratified risk criteria, while newer prediction rules (PECARN, Step-by-Step) incorporate biomarkers to refine risk assessment. Overall SBI rates in febrile infants 0-90 days are approximately 8-12%, with UTI being the most common SBI.

Definitions

Fever is defined as a rectal temperature of 38.0C (100.4F) or higher. Serious bacterial infection (SBI) encompasses UTI, bacteremia, or bacterial meningitis. Invasive bacterial infection (IBI) refers specifically to bacteremia or bacterial meningitis, excluding UTI.

Age-Stratified Approach

0-28 Days (Neonates)

Neonates represent the highest-risk group due to their immature immune system, limited clinical signs, and exposure to maternal organisms. The SBI rate is 10-20%, and the bacterial meningitis rate is 1-2%. The standard approach is a full sepsis workup including blood culture, urinalysis with urine culture, and lumbar puncture with CSF analysis and culture. Empiric antibiotics are ampicillin plus gentamicin, which covers Group B Streptococcus, E. coli, Listeria, and enterococci. An alternative is ampicillin plus cefotaxime if available; ceftriaxone is avoided in neonates due to bilirubin displacement risk. All febrile neonates should be admitted pending culture results for a minimum of 24-48 hours. HSV evaluation should be considered with the addition of acyclovir if risk factors are present (maternal HSV, vesicles, seizures, elevated LFTs, CSF pleocytosis), along with HSV PCR of CSF and surface cultures.

29-60 Days

The SBI rate is approximately 8-10%, and the bacterial meningitis rate is 0.5-1%. The traditional approach involved a full sepsis workup including LP, empiric antibiotics, and admission. Newer approaches using PECARN and Step-by-Step risk stratification use clinical appearance plus biomarkers to identify low-risk infants who may avoid LP and hospitalization. Empiric antibiotics, when indicated, are ceftriaxone (50-75 mg/kg IV/IM), with or without ampicillin for Listeria coverage (which is debatable after 28 days).

61-90 Days

The SBI rate is approximately 5-7%, and the bacterial meningitis rate is less than 0.5%. Well-appearing infants with reassuring labs may be managed as outpatients with close follow-up. UTI is the most common SBI, and blood culture plus urine culture are recommended. LP should be considered if the infant appears ill or has abnormal labs but may be deferred in well-appearing infants with normal inflammatory markers.

<image>Age-stratified approach to febrile infant evaluation showing workup and management recommendations for neonates (0-28 days: full sepsis workup, admission, empiric ampicillin + gentamicin), young infants (29-60 days: traditional full workup vs. biomarker-guided risk stratification), and older infants (61-90 days: selective workup based on clinical appearance and labs)</image>

Prediction Rules and Risk Stratification

PECARN Febrile Infant Rule (2019)

This multicenter study of febrile infants aged 29-60 days uses sequential screening for IBI. First, ill appearance (abnormal Young Infant Observation Scale) identifies high-risk infants. Second, a positive urinalysis (leukocyte esterase, nitrite, or 5 or more WBC/HPF) indicates UTI is likely and urine culture should be obtained, but it does not predict IBI. Third, an ANC of 4,090/mcL or greater indicates higher IBI risk, prompting consideration of LP and empiric antibiotics. Fourth, a procalcitonin of 1.71 ng/mL or greater indicates higher IBI risk. If the infant is well-appearing with a negative UA, ANC below 4,090, and procalcitonin below 1.71, the risk of IBI is very low (less than 0.5%), and outpatient management without LP may be considered. Sensitivity for IBI is 97-99%, and the rule has been validated externally in multiple cohorts.

Step-by-Step Approach (European)

This sequential evaluation of febrile infants aged 29-90 days first assesses for ill appearance (if present, full workup and empiric antibiotics). For infants aged 29-60 days, any one of the following indicates higher risk: procalcitonin of 0.5 or greater, CRP of 20 mg/L or greater, ANC of 10,000 or greater, or positive UA. If all are negative, the infant is low risk and outpatient management may be considered. This approach has been validated across European centers with high sensitivity for IBI.

Traditional Criteria (Historical Context)

The Rochester criteria, Philadelphia criteria, and Boston criteria have been largely supplanted by PECARN and Step-by-Step due to improved biomarker use.

Biomarkers

Procalcitonin

Procalcitonin rises earlier than CRP (within 2-4 hours of bacterial infection) and is a better discriminator of IBI than WBC or CRP in young infants. Cut-off values vary by study: 0.5 ng/mL in Step-by-Step and 1.71 ng/mL in PECARN. Limitations include physiologic elevation in the first 48 hours of life, elevation in some viral infections, and lack of universal availability.

C-Reactive Protein (CRP)

CRP rises 6-12 hours after infection onset and may be falsely negative early. A CRP below 20 mg/L indicates low risk in the Step-by-Step algorithm. It is less discriminatory than procalcitonin for early infection.

Complete Blood Count

WBC and ANC are traditionally used but have imperfect sensitivity and specificity. ANC thresholds of 4,090 (PECARN) or 10,000 (Step-by-Step) are associated with higher IBI risk. Bandemia has low sensitivity and high interobserver variability.

Urinalysis and UTI

UTI is the most common SBI in febrile infants at all age groups. A proper specimen must be obtained by catheterization or suprapubic aspiration in non-toilet-trained children, as bag specimens have unacceptable contamination rates. Enhanced urinalysis (Gram stain plus hemocytometer WBC count) is more accurate than standard UA. A positive UA shows leukocyte esterase, nitrites, or 5 or more WBC/HPF (standard) or 10 or more WBC/mm3 (hemocytometer). All infants with a positive UA should have a urine culture sent.

Lumbar Puncture

LP is mandatory in neonates (0-28 days) with fever. It is recommended in infants aged 29-60 days who appear ill or have abnormal inflammatory markers. It may be deferred in well-appearing 29-60 day infants meeting low-risk criteria on validated prediction rules. CSF interpretation in neonates uses different normal ranges: WBC up to 20-25 cells/mm3 (higher than in older children), protein up to 100-150 mg/dL (higher than in older children), and glucose greater than 50% of serum glucose. Traumatic LP is common in neonates and complicates interpretation; the correction ratio of 1 WBC per 500-1000 RBCs is unreliable. CSF enterovirus PCR is helpful during enterovirus season, and a positive result may allow early discharge if blood and urine cultures are negative.

<image>PECARN febrile infant algorithm for infants 29-60 days showing sequential assessment: clinical appearance (ill = high risk), urinalysis (positive = UTI risk, does not predict IBI), ANC threshold (>= 4090), and procalcitonin threshold (>= 1.71 ng/mL), with management recommendations for low-risk (outpatient possible) and higher-risk (LP, empiric antibiotics, admission) categories</image>

Empiric Antibiotic Selection

0-28 Days

Ampicillin is given at 50-75 mg/kg/dose every 6-8 hours IV (covering GBS, Listeria, enterococci). Gentamicin is given at 4-5 mg/kg/dose IV every 24 hours (providing synergy with ampicillin for GBS and gram-negative coverage). Acyclovir (20 mg/kg/dose every 8 hours IV) is added if HSV risk factors or clinical concern exist.

29-90 Days

Ceftriaxone at 50-75 mg/kg IV/IM is preferred, covering E. coli, GBS, and other gram-negatives. It may be given as a single IM dose for outpatient management of low-risk infants pending cultures. Ampicillin may be added for Listeria coverage, though the risk decreases significantly after 28 days and is often omitted after 30 days.

Duration

If cultures are negative at 24-36 hours and the infant is well-appearing, antibiotics can be discontinued. UTI requires 7-14 days (parenteral then oral step-down). Bacteremia without meningitis requires 10-14 days. Bacterial meningitis requires 14-21 days depending on the pathogen.

HSV Considerations

Neonatal HSV typically onsets in the first 2 weeks of life and can present with fever alone. The three forms are SEM (skin, eyes, mouth), CNS, and disseminated disease. Risk factors include maternal genital HSV (especially primary infection), vaginal delivery, and prolonged rupture of membranes. HSV evaluation should be initiated if vesicular rash, seizures, elevated LFTs or coagulopathy, CSF pleocytosis, or an ill-appearing neonate is present. Testing includes HSV PCR (CSF and blood), LFTs, and surface cultures (mouth, nasopharynx, conjunctivae, rectum). Empiric acyclovir (60 mg/kg/day IV divided every 8 hours) should be given while awaiting results if clinical suspicion exists.

Clinical Pearls

A well-appearing febrile neonate (0-28 days) still requires full sepsis workup, empiric antibiotics, and admission because clinical appearance alone cannot rule out SBI in this age group. Procalcitonin is the best single biomarker for predicting IBI in febrile infants, rising earlier than CRP with better discriminatory ability. UTI is the most common SBI, so a proper catheterized urine specimen must always be obtained; bag urine specimens are unreliable. CSF normal ranges are higher in neonates than in older children, with WBC up to 20-25 and protein up to 100-150 mg/dL; adult or older child thresholds should not be applied. A positive enterovirus PCR from CSF in a well-appearing infant with negative bacterial cultures can facilitate early discharge. Ceftriaxone must never be used in neonates under 28 days due to bilirubin displacement risk; cefotaxime or gentamicin should be used for gram-negative coverage instead.

Key Controversy: PECARN Rule and Reducing LP Rates

The PECARN febrile infant rule identifies a very low-risk group (less than 0.5% IBI risk) in whom LP may be safely deferred. This represents a paradigm shift from the traditional approach of performing LP on all febrile infants under 60 days. Adoption challenges include clinician discomfort with the possibility of missing rare meningitis, institutional protocol variability, and medicolegal concerns. Procalcitonin availability is not universal or rapidly resulted at all emergency departments, and delays reduce its clinical utility. Arguments for broader LP deferral include reducing painful procedures, radiation exposure from sedation alternatives, antibiotic exposure, and hospitalization costs. Arguments against include the devastating consequences of missed meningitis and the imperfect sensitivity of even validated prediction rules. The current trend is growing adoption of biomarker-guided risk stratification, especially at academic centers, with institutional protocols increasingly incorporating PECARN and Step-by-Step approaches.

References

• Kuppermann N, et al. A Clinical Prediction Rule to Identify Febrile Infants 60 Days and Younger at Low Risk for Serious Bacterial Infections (PECARN). JAMA Pediatr. 2019;173(4):342-351.
• Gomez B, et al. Validation of the Step-by-Step Approach in the Management of Young Febrile Infants. Pediatrics. 2016;138(2):e20154381.
• Pantell RH, et al. Evaluation and Management of Well-Appearing Febrile Infants 8-60 Days Old (AAP Clinical Practice Guideline). Pediatrics. 2021;148(2):e2021052228.
• Byington CL, et al. Serious Bacterial Infections in Febrile Infants 1 to 90 Days Old with and without Viral Infections. Pediatrics. 2004;113(6):1662-1666.
• Kimberlin DW, et al. Neonatal Herpes Simplex Virus Infection. N Engl J Med. 2004;352(5):466-475.

Read this lecture as Markdown