ABSTRACT
Fever of unknown origin (FUO) is generally considered a major diagnostic challenge because many diseases may cause this well-defined, rather rare, clinical syndrome. The diagnostic criteria of classic FUO were delineated in the landmark article of R. Petersdorf and P. Beeson in 1961 and subsequently modified by D. Durack and A. Street in 1991 (1,2). Recurrent FUO is probably the most perplexing and intriguing presentation that can be defined as a subtype of FUO, meeting the classic criteria of FUO and characterized by at least two episodes of fever with fever-free intervals of at least two weeks and seeming remission of the underlying illness (3). We realize that the proposed duration of the fever-free interval is somewhat arbitrary, but, in cases of rare clinical syndromes, adherence to standardized definitions is needed to permit comparison of groups of patients over time and in countries and between different hospital settings. This symptom-free period may vary from weeks to years, and we suggest this fever-free interval of at least two weeks for several reasons. First, this time window allows exclusion from the category of recurrent FUO those diseases that recur due to interruption or tapering of an inadequate empiric therapy. Typical examples are incompletely treated endocarditis (too short-and/or too low-dosed antibiotic therapy) and noninfectious inflammatory disorders treated with nonsteroidal anti-inflammatory agents or corticosteroids. Second, as long as fever persists, patients presenting with prolonged fever remain prepared to undergo the whole battery of less or more costly, invasive tests in order to reach a final diagnosis and get appropriate treatment. However, when fever subsides spontaneously, they become reluctant for further investigations in a few days, a week to 10 days, in our experience. Third, physicians who are familiar with the good prognosis of unexplained FUO stop the investigations when fever and symptoms subside and propose a watchful, waiting outpatient follow-up (4).
