Diagnostic value of basic and extended procedures in pediatric fever of unknown origin - results of a nation-wide surveillance study.

Abstract

Background: Pediatric fever of unknown origin (FUO) remains a major diagnostic challenge encompassing infectious, autoimmune and autoinflammatory as well as malignant diseases. Despite improved diagnostical procedures, no underlying disease can be identified in approximately one third of patients. Current guidelines recommend stepwise diagnostical approaches based on potential diagnostic clues (PDC+). This prospective nation-wide surveillance study aims to evaluate basic and advanced diagnostic procedures in the diagnostic of FUO.

Methods and results: Via the German Pediatric Surveillance Unit, children were pseudonymized enrolled based on the following inclusion criteria: fever ≥ 38.5 °C for at least five out of ten consecutive days and no identifiable cause after standard diagnostic workup. The questionnaire collected patient data, symptoms, performed diagnostic workup, and the final diagnosis, if applicable. Among 113 included children, an underlying disease was identified in 72 cases (63.7%). 26 patients (23.1%) had infections, 31 children (27.4%) had systemic onset juvenile idiopathic arthritis/Still's disease (sJIA/SD), and 15 children (13.3%) were classified as "other diagnoses" including autoinflammatory diseases, vasculitides, and malignancies. Children with sJIA/SD presented with more clinical symptoms (mean = 4.6, SD = 2.17) compared to those with infections (mean = 2.7, SD = 1.40, p=.002). Basic diagnostic workup revealed at least one PDC + in every case (mean = 6.6). Most PDC+ (mean = 3.1, SD = 1.82) were identified by history taking and physical examination as well as by laboratory testing and blood count (mean = 2.5 PDC+, SD = 1.32). A total of 370 basic imaging investigations, e.g. abdominal ultrasound and echocardiography, provided in total 50 PDCs+ demonstrating their high diagnostical yield. Primarily in children with unclear clinical presentation advanced imaging revealed PDC+. Overall, the number of misleading PDCs (PDC-) was low (mean = 1.0, SD = 1.14) with autoantibody testing being the basic diagnostic procedure accounting for the majority of PDC-.

Conclusion: The results of this nation-wide surveillance study highlight the value of basic and advanced diagnostical approaches in the management of pediatric FUO. Careful history taking, physical examination, and targeted basic diagnostic testing can identify multiple PDC+ with only few misleading PDC-, which may contribute to diagnostic accuracy and avoid unnecessary testing. The findings can be introduced into guidelines in order to standardize and improve the diagnostic approaches in children with FUO.