Abstract
Objectives: Fever of Unknown Origin (FUO) in children remains difficult to diagnose. This prospective, observational study aimed to determine the clinico-etiological profile and the role of clinical procedures in establishing the diagnosis.<br>Methods: Children (1 month-16 years) with prolonged fever meeting the criteria of FUO were included. A detailed medical history, physical examination and laboratory tests were conducted according to an algorithm (phase I-III). Clinical procedures that were useful in obtaining final etiology were evaluated.<br>Results: In total, 69 children were enrolled, categorized into five groups based on etiology of FUO. Infections (31.8%) were the leading cause followed by malignancy (20.2%), connective tissue disorder and miscellaneous conditions (14.5%); undiagnosed patients (18.8%) had either a spontaneous resolution or incomplete workup. Duration of fever was 20-60 days, longer in malignancies, least in infections and mean ± SD duration between admission and a definitive diagnosis was 5.65 ± 1.06 days. Tuberculosis was the most common among the infections, systemic onset juvenile idiopathic arthritis and Common acute lymphoblastic leukemia antigen positive B cell acute lymphoblastic leukemia were common causes of connective tissue disorder and malignancy, respectively.<br>The diagnosis was established by Phase II (33.3%) and Phase III (34.8%) investigations. Non-invasive and invasive investigations were useful in 56.3%, and 43.7 %, respectively.<br>Conclusions: Infections are the most common cause of FUO in children followed by malignancy. A stepwise approach is useful in diagnosis. Infections and connective tissue disorders are diagnosed by non-invasive investigations while malignancy and miscellaneous conditions by invasive investigations. Focused clinical and diagnostic approach is mandated. J Microbiol Infect Dis 2019; 9(3):137-143.

Key words: Acute lymphoblastic leukemia, algorithm, children, fever of unknown origin, children