Positron Emission Tomography with Computed Tomography in Evaluations of Classical Fever of Unknown Origin and Length of Hospitalization: A 10-Year Medical Record Review of a Tertiary Hospital.

Abstract

Positron emission tomography with computed tomography (PET/CT) has proven its value for the differential diagnosis of fever of unknown origin (FUO). However, the extent to which PET/CT during FUO evaluation can shorten the length of hospital stay (LOS) remains unclear. A retrospective review of the medical records over a 10-year period from January 2009 to December 2018 of a tertiary university hospital was performed. The inclusion criteria were symptoms with fever persisting for >3 weeks before admission, as defined in classical FUO. Medical records in which PET/CT was performed after the final diagnosis, such as neoplastic causes, were excluded. Moreover, in the neoplasm category evaluated using PET/CT, only diagnostic PET/CT cases were enrolled; PET/CT cases for confirming metastasis or staging were excluded. Final diagnoses were categorized as infection, neoplasm, noninfectious non-neoplastic inflammatory disorder, miscellaneous, and uncategorizable. Each category was separated into evaluation with and without PET/CT for statistical analyses. In total, 91 patients underwent evaluation for FUO and about one in three underwent PET/CT. Overall LOS was not different between the PET/CT and non-PET/CT groups; however, there were differences in LOS within the categories. For infectious causes, the mean LOS was 21.1 and 11.1 days in the PET/CT and non-PET/CT groups, respectively (P = 0.022). For neoplastic causes, the mean LOS was 11.4 and 36.0 days in the PET/CT and non-PET/CT groups, respectively (P = 0.02). Most patients with FUO were aged 50 - 60 years, and their family and work roles were crucial. A lower LOS may benefit both the patients' families and society at large. Interestingly, PET/CT may contribute to shortening the LOS during FUO evaluation when the causes are neoplastic, by approximately 24 days.