Gamechanger: Do we really need a lumbar puncture for all febrile infants 0-28 days old?

A new international pooled analysis challenges the age-old dogma that all febrile infants 0-28 days require a lumbar puncture (LP). Can the PECARN febrile infant prediction rule safely identify a low-risk subset for invasive bacterial illnesses (bacterial meningitis and bacteremia) [1]?

Bottom Line

For more than  four decades, the standard of care for febrile infants in the first month of life has been aggressive: full sepsis workup (including an LP), admission, and IV antibiotics. A new study in JAMA suggests this paradigm may be shifting [2, 3].

• In an international pooled analysis of more than  1,500 febrile infants aged ≤28 days, the updated PECARN febrile infant prediction rule missed zero cases of bacterial meningitis.
• Inclusion Criteria: Non-ill-appearing, full time (≥37 weeks) infants aged 0–28 days with fever (≥38.0°C), who underwent blood and urine testing including procalcitonin (PCT).
• Exclusion Criteria: Critically ill appearance, prematurity, comorbidities, or antibiotic use in preceding days.
• Implication: Cerebrospinal fluid analysis is unnecessary for a subset of non-ill-appearing febrile infants ≤28 days old.
• What now? The current data provides a solid, practice-changing, evidence-based foundation for a shared decision-making conversation that wasn’t possible before.

Study

To answer this question, the authors performed 2 distinct analyses:

1. Primary Analysis (The “External” Test): To test the rule’s validity in new, diverse populations, the primary analysis pooled data from 4 prospective international cohort studies (Canada, Spain, Europe, UK/Ireland).

• • Population: 1,537 non-ill-appearing, full-term (≥37 weeks) infants aged 0–28 days with fever (≥38C)
  • Why no US data? This was done to validate the PECARN rule externally, avoiding the bias of testing it on the same US population from which it was derived.

2. Secondary Analysis (The “Maximize Power” Test): To generate the most precise safety estimates possible, the authors then pooled the 4 international cohorts PLUS the 2 original US-based PECARN cohorts.

• • Population: 2,531 infants total
  • Result: Even with the added US data, the rule missed zero cases of bacterial meningitis.

What is the updated PECARN febrile infant prediction?

An infant ≤28 days old is low risk if they meet all 3 criteria:

The Findings

The prevalence of Invasive Bacterial Infections (IBI) in all studied patients was 4.5%.

• 3.8% bacteremia
• 0.7% meningitis

Performance of the PECARN Rule

Number needed to tap calculation

One of the most compelling arguments for using this rule is the statistical trade-off required to find a single missed case. The authors provide estimated Negative Predictive Values (NPV) across a range of disease prevalences.

If we assume a 1.00% prevalence of bacterial meningitis (which is conservative; the study observed 0.7%), the NPV for bacterial meningitis is 99.95% [2].

This means that for every 10,000 PECARN low-risk infants, 9,995 do not have bacterial meningitis, and 5 might. We can translate this into a “Number Needed to Tap” (NNT) to find one missed case:

• Risk of Missed Case = 1 – 0.9995 = 0.0005
• NNT = 1 / 0.0005 = 2000

Bottom Line: You would hypothetically need to perform 2,000 lumbar punctures on low-risk infants to find ONE case of bacterial meningitis that the rule missed.

Important guardrails: Who is this rule for?

Before applying these findings, we need to understand the strict inclusion criteria. This study—and the PECARN rule itself—was only validated on a specific population.

The “Must-Have” Checklist:

• ✅ Non-ill-Appearing: The infant cannot appear ill. The study defined this strictly, excluding infants with abnormal appearance, work of breathing, or circulation findings (often using the Pediatric Assessment Triangle or other illness indicators). If the baby looks sick, the rule does not apply.
• ✅ Full-Term: Infants must be ≥37 weeks gestation. Preterm infants have different immunological risks and were excluded.
• ✅ Age 0–28 Days: This specific analysis focused exclusively on the first 28 days of life.
• ✅ Proven Fever: Documented temperature ≥38C

The SBI vs. IBI distinction

If you are already using the PECARN rule for older infants (29–60 days), you likely use it to rule out Serious Bacterial Infections (SBIs), which includes urinary tract infections (UTIs) [1].

This study is different – it focused purely on invasive bacterial infections (IBIs), which is defined as bacteremia and/or bacterial meningitis.

What did the PECARN rule miss?

The rule had perfect sensitivity for bacterial meningitis, but it did miss 5 cases of bacteremia out of more than 2,500 infants ≤28 days old despite a low-risk stratification. Let’s look at the 5 cases classified as “missed bacteremia:

• 1 case: H. influenzae bacteremia
• 1 case: E. coli bacteremia (without UTI)
• 1 case: E. coli bacteremia (with E. coli UTI)
• 2 cases: S. aureus bacteremia (One of these also had a concurrent E. coli UTI).

The authors note that S. aureus in blood cultures can be a contaminant rather than a true pathogen. If these S. aureus cases were indeed contaminants, the true sensitivity of the rule would be even higher than reported.

Notably, all 5 cases of missed bacteremia occurred in infants aged 8-21 days. There were 0 missed bacteremia cases in the 22-28 day age group.

How do we reconcile this with the most current 2021 AAP guidelines?

To understand why this study is a big deal, we have to look at what the American Academy of Pediatrics (AAP) guidelines currently tells us to do. The new data exposes a potential practice shift specifically for infants in the third week of life (8–21 days).

Additional Considerations

1. Procalcitonin is mandatory: This rule relies on serum procalcitonin. If your facility only uses CRP and WBC, you cannot use this reduction strategy safely.

2. Consider herpes simplex virus (HSV) meningoencephalitis: This PECARN rule is to identify young febrile infants with bacterial infections and not HSV. You thus must still risk-stratify for HSV separately (seizures, vesicles, maternal history, etc) and perform a LP if HSV is suspected, independent of the PECARN prediction rule.

Summary

For the first time, we have high-quality, multi-national data suggesting that a routine LP may not be necessary for every febrile infant ≤28 days old. While guidelines have not officially changed, this study provides the evidence needed to support shared decision-making with caregivers.

We can now honestly tell parents: “Based on these blood and urine tests, the chance of your baby having bacterial meningitis is extremely low—likely less than 1 in 2,000. We can safely hold off on the spinal tap and antibiotics right now and admit for observation.”

That is a conversation we couldn’t have yesterday.

References

1. Kuppermann N, Dayan PS, Levine DA, et al; Febrile Infant Working Group of the Pediatric Emergency Care Applied Research Network (PECARN). A clinical prediction rule to identify febrile infants 60 days and younger at low risk for serious bacterial infections. JAMA Pediatr. 2019;173(4):342-351. doi:10.1001/jamapediatrics.2018.5501. PMID 30776077
2. Burstein B, Waterfield T, Umana E, Xie J, Kuppermann N. Prediction of Bacteremia and Bacterial Meningitis Among Febrile Infants Aged 28 Days or Younger. JAMA. Published online December 8, 2025. doi: 10.1001/jama.2025.21454
3. Searns JB, O’Leary ST. Moving the Field Forward to Safely Do Less With Febrile Neonates. JAMA. Published online December 8, 2025. doi: 10.1001/jama.2025.23133
4. Pantell RH, Roberts KB, Adams WG, et al; Subcommittee on Febrile Infants. Evaluation and management of well-appearing febrile infants 8 to 60 days old. Pediatrics. 2021;148(2):e2021052228. PMID 34281996