The diagnostic utility of computed tomography scans performed for febrile neutropenia in a single centre

Abstract

BackgroundComputed tomography (CT) imaging has become a first line investigation for most cases of febrile neutropenia (FN) which can be the only sign of infection in oncology patients undergoing active chemotherapy and bone marrow transplants. The utility of routine non-targeted imaging remains unclear. ObjectiveTo assess and compare the diagnostic rate between targeted, non-targeted and pan-scan CT in identifying an acute source of infection in adult oncology patients with FN. Materials and MethodsA retrospective observational study was conducted between February 2019 and March 2023 on 417 consecutive CT examinations for the clinical indication of source identification in FN. Scans were noted for the anatomical regions that were imaged and reports were classified as positive, negative or equivocal for infection. Pre-existing pathology was also noted. Results were tabulated and statistical analyses for comparison between groups of scans was performed using chi-square test. ResultsAll targeted regional scans had statistically significant difference in positive rate compared to non-targeted scans of the respective region; chest (Χ²(1)=18.11, P<.001); sinus (Χ²(1)=15.36, P<.001); abdomen and pelvis (Χ²(1)=5.95, P=.01). Pneumonia (41.3 %) was much more likely to be the diagnosis compared to sinusitis (16.2 %) in concomitant CT chest to sinus examinations (Χ²(1)=45.3, P<.001). Pan-scans had a higher incidence of positive diagnosis compared to all-targeted scans (Χ²(1)=4.91, P=.03) but when compared to higher yield targeted scans (abdomen and chest), there was no statistical difference (Χ²(1)=2.43, P=.12). 20/54 patients had pan-scans despite having localising symptoms. ConclusionImaging guided by presenting signs and symptoms can help to reduce unnecessary imaging and promote more judicious use of non-targeted and pan-scan CT in current practices.