Age-Dependent Clinical Patterns of Primary Epstein-Barr Virus Infection in Children: Insights for Diagnostic Accuracy.

Primary Epstein-Barr virus (EBV) infection in children exhibits substantial clinical heterogeneity, often complicating early diagnosis and leading to unnecessary antibiotic use. This retrospective study evaluated 695 children (0-18 years) diagnosed with primary EBV infection at a tertiary pediatric center between 2010 and 2015, defined by positive viral capsid antigen (VCA) IgM and negative Epstein-Barr nuclear antigen (EBNA) IgG. Clinical, laboratory, and ultrasonographic findings were compared according to age group (≤4 vs. >4 years) and clinical setting (inpatient vs. outpatient). The median age was 3.75 years (IQR: 2-6.25), and more than half of the patients were ≤4 years. Younger children more frequently presented with nonspecific respiratory and gastrointestinal symptoms, whereas older children more commonly exhibited the classic infectious mononucleosis (IM) phenotype, including sore throat, dysphagia, lymphadenopathy, and hepatosplenomegaly (p < 0.001). Antibiotics were prescribed in 64.2% of patients, while 21.7% required hospitalization. Multivariable logistic regression analyses demonstrated that age was not an independent predictor of hospitalization, classic IM phenotype, or antibiotic use. Instead, specific clinical and laboratory findings-such as lymphopenia, lymphadenopathy, vomiting, thrombocytosis, and tonsillar hypertrophy-emerged as the key determinants of clinical outcomes. To enhance diagnostic discrimination, receiver operating characteristic (ROC) analysis of ANC/ALC and AST/ALT ratios was performed, and a composite risk score (0-2) was derived. Although both markers showed modest discriminative ability (AUC 0.607 and 0.575), their high negative predictive values (>90%) suggest potential utility as rule-out tools. The composite score demonstrated a stepwise increase in the probability of classic IM presentation across age groups. In conclusion, primary EBV infection demonstrates a clear age-related clinical spectrum; however, clinical and laboratory features rather than age alone drive key outcomes. These findings highlight the need for age-specific diagnostic strategies and improved antimicrobial stewardship, while the proposed risk score provides a foundation for future validation studies.