A Clinical Prediction Rule to Identify Febrile Infants 60 Days and Younger at Low Risk for Serious Bacterial...

Area of Science:

• Pediatrics
• Infectious Diseases
• Clinical Diagnostics

Background:

• Serious bacterial infections (SBIs) in young febrile infants can lead to severe complications.
• Current diagnostic methods like lumbar punctures and hospitalizations carry risks and costs.
• Biomarkers beyond white blood cell count (WBC) may improve risk stratification for SBIs in infants.

Purpose of the Study:

• To develop and validate a clinical prediction rule for identifying febrile infants aged 60 days or younger who are at low risk for SBIs.
• To establish a reliable tool to potentially avoid invasive procedures and reduce healthcare costs.

Main Methods:

• Prospective observational study involving 1821 febrile infants (≤60 days old) across 26 emergency departments.
• Data collected included clinical findings, WBC, absolute neutrophil count (ANC), serum procalcitonin, and urinalysis.
• Binary recursive partitioning analysis was used to derive and validate the prediction rule.

Main Results:

• The prediction rule identified low-risk infants using a negative urinalysis, ANC ≤4090/µL, and procalcitonin ≤1.71 ng/mL.
• In the validation cohort, the rule demonstrated high sensitivity (97.7%) and negative predictive value (99.6%) for SBIs.
• The rule correctly identified most infants at low risk, with only a few misclassifications and no missed cases of bacterial meningitis.

Conclusions:

• A validated prediction rule incorporating urinalysis, ANC, and procalcitonin can accurately identify febrile infants at low risk for SBIs.
• Clinical implementation of this rule could significantly decrease unnecessary lumbar punctures, antibiotic use, and hospitalizations.
• Further validation in independent cohorts is recommended before widespread clinical adoption.