Background: Fever of unknown origin (FUO) was first defined by Petersdorf and Beeson in 1961 and later refined by Durack and Street (1991). Pediatric FUO affects 0.5–3% of febrile children (Antoon et al., 2015) and has traditionally been attributed mainly to infection, followed by connective tissue/autoimmune disease and malignancy (Chusid, 2017; Haidar and Singh, 2022). Recent series suggest a shift toward non-infectious causes: a seven-year adult Turkish cohort found collagen vascular disease had overtaken infection as the leading etiology (Kaya et al., 2025), and the only prior pediatric series from the region, a two-year Egyptian cohort, likewise found collagen vascular disease most common (31.8%), ahead of infection (28.2%) (Shoman et al., 2024). We aimed to describe the etiology, clinical profile, and imaging yield of pediatric FUO at a tertiary care center in Qatar.
Methods: We retrospectively reviewed children aged 3 months to 17 years admitted with FUO (temperature ≥38.0°C for ≥8 days without an identified source after standard screening) or fever without a source, at Sidra Medicine between September 2018 and December 2024. Demographic, clinical, laboratory, and imaging data were abstracted via a structured electronic database. Final diagnoses were grouped into infectious, inflammatory/autoimmune, miscellaneous, and undiagnosed categories. Descriptive statistics summarized the cohort.
Results: Sixty children met inclusion criteria (median age 8.0 years, IQR 4.0–12.0; range 3 months–16 years; 51.7% male; 18 nationalities). Prior antibiotic exposure (86.7%) and prior hospitalization (45.0%) were common. Median fever duration was 9 days (IQR 3–14); median length of stay was 8 days (IQR 4.75–15.25); 10.0% required intensive care and 3.3% died. The leading presenting symptoms and examination finding were gastrointestinal complaints (56.7%) and lymphadenopathy (28.3%), respectively.
A final diagnosis was reached in 48/60 children (80.0%). Infectious disease was the leading category (24/60, 40.0%), most commonly Epstein-Barr virus infection (4/24, 16.7%); no tuberculosis, brucellosis, or Kawasaki disease was confirmed. Inflammatory conditions accounted for 22/60 (36.7%), predominantly systemic-onset juvenile idiopathic arthritis (4/22, 18.2%) and systemic lupus erythematosus (3/22, 13.6%). No etiology was identified in 9/60 (15.0%); 5/60 (8.3%) had miscellaneous diagnoses. Inflammatory cases had higher median C-reactive protein (130.0 vs 80.6 mg/L), erythrocyte sedimentation rate (84 vs 50 mm/hr), and ferritin (355 vs 221.5 µg/L), and lower hemoglobin (9.4 vs 11.0 g/dL) than infectious cases. Magnetic resonance imaging (MRI) was performed in 28/60 (46.7%; abnormal in 78.6%) and positron emission tomography-computed tomography (PET-CT) in 12/60 (20.0%; abnormal in all cases studied); both were performed in 8 patients, among whom PET-CT detected abnormalities in 2 with a normal MRI.
Conclusion: In this first characterization of pediatric FUO in the Gulf region, infectious disease remained the leading cause, contrasting with the non-infectious predominance reported elsewhere, with inflammatory disease a close second and an 80% diagnostic yield. Elevated inflammatory markers with lower hemoglobin distinguished inflammatory from infectious causes; preliminary paired imaging data suggest PET-CT may add value over MRI alone, warranting confirmation in a larger cohort.

