Infants in the first months of life are frequently evaluated for fever in the emergency department (ED) and are at risk of urinary tract infections (UTIs) and invasive bacterial infections (IBIs), specifically bacteremia and bacterial meningitis. Approximately 10% of febrile young infants have a history of prematurity.1,2 Data regarding UTI and IBI risk among febrile preterm infants are limited, with studies reporting both higher3,4 and lower infection rates.2 Moreover, febrile preterm infants are excluded from management guidelines by the American Academy of Pediatrics (AAP)5 and studies to identify infants at low-risk of UTIs and IBIs, such as the Pediatric Emergency Care Applied Research Network (PECARN) prediction rule.6,7 Clinicians must manage febrile preterm infants without established clinical guidance or data to inform risk-stratification. In this study, we assessed the diagnostic accuracy of AAP recommendations5 and the updated PECARN prediction rule7 for identifying low-risk febrile preterm infants.This was a secondary analysis of prospective quality improvement data for all infants aged 60 days or fewer who were evaluated for fever at a tertiary pediatric ED.8,9 Standardized clinical, laboratory, and follow-up data were collected for all infants with a rectal temperature of at least 38.0°C (home or ED triage), and infants requiring hospitalization were admitted to a general pediatric ward, regardless of any history of prematurity history. Systematic procalcitonin (PCT) testing was introduced in June 2020. This analysis included all infants evaluated between June 2020 and December 2024, with a history of prematurity (<37 weeks gestation) but aged 8 to 60 days at ED presentation, who were well-appearing and otherwise met AAP inclusion criteria.5 Infants were excluded if they were ill-appearing (highest triage acuity, emergent interventions such as endotracheal intubation or vasopressors, requiring intensive care, or abnormal appearance according to the Pediatric Assessment Triangle). Final UTI, bacteremia, and bacterial meningitis status were determined by a review of microbiologic cultures and telephone follow-up after 14 days post-discharge for all infants8,9 and classified according to recognized definitions.6We evaluated the diagnostic accuracy (negative predictive value NPV, sensitivity, and specificity) of AAP and PECARN risk-stratification strategies. Infants were classified as low-risk of both UTI/IBI using updated PECARN criteria: PCT (≤0.5 ng/mL) and absolute neutrophil count (ANC) (≤4000/mm3) and negative urinalysis (no leukocyte esterase, nitrites, and ≤5 white blood cells per high powered field).7 Infants were classified as low-risk of IBI if they met AAP-recommended thresholds: 1) PCT (≤0.5 ng/mL) and C-reactive protein (CRP) (≤20.0 mg/L); 2) PCT (≤0.5 ng/mL) and ANC (≤4000/mm3); and 3) CRP (≤20.0 mg/L) and ANC (≤5200/mm3) and maximal temperature (≤38.5 °C).5 This study received approval from the institutional Research Ethics Board and followed the 2015 Standard of Reporting of Diagnostic Accuracy guidelines.Among 2254 febrile infants aged 8 to 60 days evaluated during the study period, 103 (4.6%) were born premature (Table 1). The median chronological age at presentation was 46 (IQR 33–53) days, with 92 (89.3%) infants aged 22 to 60 days old. Gestational ages ranged from 31 to 36 weeks (89.3% born 34–36 weeks). There were 10 (9.7%) infants with isolated UTIs, and 3 (2.9%) infants with IBIs, including 1 infant with bacteremia and bacterial meningitis. All AAP strategies performed with 100% sensitivity and NPV for ruling out IBIs (Table 2). The AAP-recommended combination of PCT and CRP had the highest specificity (89.8%; 95% CI 82.0–95.0; P < .01 vs PCT + ANC). The PECARN rule misclassified 1 infant with a UTI as low-risk (Supplemental Table 1) but missed no IBIs (NPV 98.3%, 95% CI 90.6–99.9; sensitivity 92.3%, 95% CI 64.0–99.8; specificity 67.5%, 95% CI 56.3–77.4).This is the first prospective assessment of risk-stratification for febrile infants aged 60 days or younger who were born prematurely. Nearly 5% of all febrile young infants evaluated in the ED were born premature, carrying a risk of UTI around 10% and IBI around 3%. Although 1 infant with a UTI but negative urinalysis was misclassified by the PECARN rule, neither AAP nor PECARN strategies missed any cases of IBI. The diagnostic accuracy of all strategies was similar to performance described previously among term-born infants in the same cohort8,9 and among other AAP-ineligible febrile infants (N = 141, including 37 preterm).10 Results of this exploratory analysis suggest that it may be appropriate to manage premature febrile infants according to risk-stratification strategies developed for term-born infants; however, larger multicentered prospective studies are needed to best inform clinical practice.This study has limitations. This was a single-center study. All infants aged 8 to 60 days were analyzed together, although around 90% were aged older than 21 days, limiting age-specific inferences. With only 3 IBIs, the independent contributions of gestational age at birth vs chronologic age at evaluation could not be assessed.3 Importantly, the small number of IBIs also limited precision of sensitivity estimates; however, specificities and NPVs could be estimated with higher precision.
Itzkovitz et al. (Mon,) studied this question.