[Fever of unknown origin: diagnostic strategies and tactical approaches]

Diagnosis of fever of unknown origin (FUO) is a major challenge for internists, as emphasized by the high rate of diagnostic failure, despite the fast-moving progress in medical technology. Numerous clues are available in clinical and standard biological data; a better use of the available tests is warranted.

Improvement in diagnostic accuracy might be expected by developing strategies targeted toward a more systematic search of diagnosis clues. Intuition and the hypothetic-deductive method that are the most common clinical strategies are the most perfectible. It implies to enjoy the fun of clinical examination, to have a large experience in bedside training, to be confident in his/her own semiological skills, to refer frequently to heuristics, and to use carefully Occam's razor principle. Laboratory tests might be revisited; immunological and serological tests are of little value; standard biological tests provide many insufficiently exploited clues. Imaging procedures depend on objectives: whole body CT scan should be performed early within the first days of hospitalisation, preceded by standard chest radiograph and abdomen ultrasonography; followed by either indium-111 or technetium-99m, labelled leukocytes if deep abscesses are suspected or 18-FDG PET scan in the case of suspected inflammatory disease involving tissues, lymph nodes or arteries. Early identification of the best tissue to be the site of biopsy is one of the most decisive procedures.

Strategies and tactical approaches for the diagnosis of FUO might be driven by the search of significant clues. Self-clinical experience driven by a wide bedside training is of major concern. Standard laboratory tests might be better used and the choice of imaging depends on objectives. Identification of the most appropriate tissue to be sampled for histological examination is one of the most beneficial step.