Clinical microbiology and infection vol:9 issue:S1 pages:252
European Congress of Clinical Microbiology and Infectious Diseases edition:13th location:Glasgow date:10-13 May 2003
Objectives: Several studies in oncological patients have suggested that the antipyretic response to naproxen is sufficiently specific to distinguish tumorrelated
fever from infectious fever.We sought to determine the sensitivity and specificity of the ‘naproxen test’ in unselected patients presenting with a prolonged febrile illness of unknown etiology.
Methods: From a prospectively collected cohort of 290 patients admitted to a single university center with a prolonged unexplained feverish illness, we reviewed the charts of a group of 77 patients who received a non-steroidal anti-inflammatory drug (NSAID), mainly naproxen, during the diagnosticwork-up. A complete reponse to the NSAID (‘positive naproxen test’) was defined as a decrease of maximal body temperature to below 37.2 8C. The
final diagnosis was established at the time of discharge from the hospital or during follow-up and the response to the NSAID was compared in the different diagnostic categories.
Results: A complete antipyretic response to the NSAID occurred in 6 of 11 (54.5%) patients with neoplastic fever vs. in 25 of 66 (37.9%) patients with nonneoplastic fever (P¼0.48). The sensitivity, specificity, and accuracy of the
naproxen test for neoplastic fever were 55, 62, and 61%, respectively.
Conclusion: The naproxen test is not specific for tumor-related fever and has no differential diagnostic role in the work-up of a patient with prolonged unexplained fever.