Journal 2026 Vol.27 No.1
External Validation of the Febrile Infants Risk Score at Triage (FIRST) for Early Detection of Serious Bacterial Infections in Infants Less than Three Months Old: A Retrospective Cohort Study in a Philippine Tertiary Emergency Department
Marione Thea B. Rodriguez, M.D. , Soraya B. Sarapuddin, M.D.

Abstract
Objective:: Early identification of serious bacterial infection (SBI) in febrile infants younger than three months remains challenging. This study externally validated the diagnostic performance of the Febrile Infants Risk Score at Triage (FIRST) and the modified FIRST+ score for detecting SBI in a Filipino pediatric population.
Methodology: A single-center retrospective cohort study was conducted among febrile infants less than 3 months old presenting to a tertiary emergency department in the Philippines between January 1, 2020 and December 31, 2024. Medical records were reviewed to identify SBI, defined as urinary tract infection (UTI), bacteremia, or meningitis based on the final clinical diagnosis. The discrimination and calibration of FIRST and FIRST+ were evaluated using the original pre-specified cut-offs described in the derivation study. ensitivity, specificity, likelihood ratios, calibration metrics, and decision curve analysis were also determined.
Results: A total of 248 infants were included, of whom 114 (46%) had SBI. UTI was the most common diagnosis (31%), followed by bacteremia (13%) and meningitis (1%). Several clinical and laboratory characteristics associated with SBI included male sex, higher admission temperature, maternal group B Streptococcus positivity, elevated absolute neutrophil count, and increased Creactive protein levels. FIRST demonstrated poor discrimination (AUC 0.59) despite high sensitivity, while FIRST+ demonstrated good discrimination (AUC 0.84) and balanced sensitivity and specificity at their respective original cut-offs. Exploratory analysis suggested that a higher threshold (37.5) improved specificity and positive likelihood ratio, although this finding requires further validation.
Conclusion: FIRST+ demonstrated better discrimination and calibration than the triage-only FIRST score in this cohort of Filipino febrile infants. While FIRST+ may serve as a useful adjunctive risk stratification tool when laboratory testing is available, further prospective multicenter validation is necessary before broader clinical implementation.

KEYWORDS: Febrile Infants, Serious Bacterial Infection, Urinary Tract Infection, Bacteremia, Risk Assessment
https://doi.org/10.56964/pidspj20262701007
View Full Article in PDF format