Written by Kimiko Dunbar
Spoon Feed
PECARN derived and internally validated two prediction rules to identify infants 61–90 days at low risk for invasive bacterial infection (IBI): one with urinalysis (UA) + Tmax ≤ 38.9°C; another with procalcitonin (PCT) ≤ 0.24 ng/mL + absolute neutrophil count (ANC) ≤ 10,710.
New rules are nice, but proceed with caution
To avoid unnecessary testing in febrile infants, clinical prediction rules can risk-stratify those who can safely be managed without a full sepsis rule-out. AAP guidelines cover febrile infants < 60 days but not those 61–90 days.
In this multicenter retrospective cohort study of 4,952 non-critically ill febrile infants 61–90 days, they used PECARN data to derive and internally validate two prediction rules to identify those at low risk for IBI. Using recursive partitioning, a rule based on negative urinalysis plus temperature ≤ 38.9°C had 86.0% sensitivity (95%CI 77.6–92.1), specificity 58.9%, negative predictive value 99.5%. A separate rule based on procalcitonin (PCT) ≤ 0.24 ng/mL plus ANC ≤ 10,710 cells/mm³ had 100% sensitivity and 65.8% specificity, but performance was worse on cross-validation. Limitations include retrospective design and only internal validation. Most patients lacked procalcitonin data, making the sample size much smaller for this subset. Additionally, over 70% of febrile infants were excluded because they lacked blood culture or urinalysis, potentially introducing selection bias.
It’s important to note that while the non-PCT based rule had excellent NPV (99.5%), 14 infants were misclassified as low risk for IBI. All but 1 had isolated bacteremia; one infant had GBS meningitis, with CSF obtained on initial presentation, possibly indicating ill-appearance on arrival.
How does this change my practice?
Current guidelines don’t include older babies, which makes management highly variable. These potential prediction rules are helpful. I’ll pay more attention to the height of the fever and consider getting a procalcitonin in this age group, particularly if presenting with a fever without a clear cause. Ultimately, these prediction rules should be used with caution, especially prior to prospective external validation.
Source
PECARN Registry Working Group. Prediction Rule to Identify Febrile Infants 61-90 Days at Low Risk for Invasive Bacterial Infections. Pediatrics. 2025 Sep 1;156(3):e2025071666. doi: 10.1542/peds.2025-071666. PMID: 40854562; PMCID: PMC12432541.
