The term “fever of unknown origin” (LDL) is used to denote prolonged fever of unclear etiology. The controversial question is how long a fever should be and how much diagnostic testing needs to be done before a diagnosis of fever of unknown origin is made. Criteria are to some extent arbitrary and vary among authors.
In his classic work on this issue, Petersdorf in 1961 identified fever of unknown origin as body temperature above 38.3 ° C for more than 3 weeks with a diagnosis that remains unclear after staying in the hospital for 1 week.
From the Petersdorf article it follows that he did not seek to define fever of unknown origin, but wanted to establish criteria for studying this condition. This is just a working definition, intended primarily for the selection of 100 patients with fever of unknown origin in the group for a prospective study.
The author chose a period of 3 weeks in order to exclude most of the acute, self-sustaining infectious diseases. The temperature value of 38.3 ° С should exclude, in his words, “habitual hyperthermia”, characteristic of healthy individuals with normal body temperature above average values. Finally, he proposed as a diagnostic criterion for 1 week of hospitalization, in order to select cases that were not identified during the “routine laboratory examination.”
Of course, the set of generally accepted laboratory methods in 1961 was very far from the scope of the survey today. In addition, pediatricians should consider that the work of Petersdorf is dedicated only to adult patients.
When studying undiagnosed fever in children, Brewis in 1965 noted that definition of fever of unknown origin in clinical practice is far from being as accurate as described by Petersdorf.