Fever of unknown origin: a new definition and proposal for diagnostic work-up

Abstract

Fever of unknown origin (FUO) is regarded as a the experience of the investigator, the diagnostic facilities, formidable diagnostic challenge. Petersdorf and Beeson [1] and differences in management between countries or even made it possible to perform epidemiological studies by hospitals. A quality criterion, specifying the necessary introducing a definition of FUO, namely, an illness charexaminations, is preferable. We propose criteria for classiacterized by a rectal temperature of 38.38C or higher on at cal FUO as given in Table 1. Defining the necessary least three occasions, evolving during at least 3 weeks, ‘initial’ examinations will remain a matter of debate with no diagnosis reached after 1 week of ‘intelligent’ among experts, but it should be possible to reach an inpatient examination. This definition has been modernized agreement about a minimum set of diagnostic tests. by excluding immunocompromised patients and by shortThe work-up of the immunocompetent patient with ening the 1 week inpatient examination period to 3 days or prolonged fever should start with the standard examination three outpatient visits [2]. protocol for prolonged fever as proposed in Table 1. For Subgroups, such as HIV-associated, neutropenic, bone clinicians with little experience in this field, the use of a marrow and organ transplant recipients, and nosocomial pre-coded history and physical examination form is advisFUO, should be distinguished because the spectrum of Table 1 underlying diseases is different. In addition, the immune Revised definition for classical fever of unknown origin dysfunction of such patients does not allow the watchful 1 Illness of more than 3 weeks duration waiting diagnostic approach in classical undiagnosed FUO, but rather may require early empirical, antimicrobial 2 Temperature of at least 38.38C (1018F) or lower temperature therapy. with signs of inflammation on several (three or more) occasions The criterion of 1 week inpatient examination needs revision indeed. This criterion represents the time required 3 No diagnosis or reasonable (eventually confirmed) diagnostic hypothesis after performing a standard initial diagnostic in the Petersdorf study for completion of routine laboraa investigation protocol tory, bacteriological and serological studies, a tuberculin skin test, and radiographs of chest, kidneys, and gallb 4 Exclusion of immunocompromised patients bladder. Present-day medicine is characterized by more a Standardized thorough history and physical examination, routine outpatient examination and a much higher pace of examiblood tests (erythrocyte sedimentation rate or CRP, hemoglobin, leukocyte nation, resulting in an enormous number of studies (radioand differential count, creatinine, sodium, potassium, protein electrographs, ultrasonography, CT scans, MRI, scintigraphy) that phoresis) enzymes (alkaline phosphatase, aminotransferase, lactate decan be performed during a 1-week hospital stay. A hydrogenase, creatine phosphokinase), urinalysis, antinuclear and antineutrophil cytoplasmic antibodies, cultures of blood and urine, tuberculin quantitative criterion of ‘1 week examination’ is biased by skin test, X-ray of the chest, abdominal ultrasonography, examinations indicated by clues obtained by the aforementioned tests. b 9 9 *Corresponding author. Tel.: 131-24-361-5215; fax: 131-24-354,1.0310 WBC/l, polymorphonuclears ,0.5310 / l, HIV0788. seropositivity, use of $10 mg prednisone for at least 2 weeks, severe E-mail address: e.dekleijn@onco.azn.nl (E.M.H.A. de Kleijn) hypogammaglobulinemia (IgG,50%).