The values of (1,3)-β-D-glucan and galactomannan in cases of invasive fungal rhinosinusitis.

OBJECTIVE The purpose of the present study was to investigate the values of the serum (1,3)-β-D-glucan test (G test) alone, the galactomannan test (GM test) alone, and their combination in the diagnosis of invasive fungal rhinosinusitis (IFRS). METHODS The present study retrospectively analysed the clinical data of 98 patients who were preliminarily diagnosed with "space-occupying lesions in nose". Of these 98 patients, 88 received the G test, 55 received the GM test, and 45 received both. A pathology analysis was used as the gold standard to diagnose IFRS. All data were analysed using SPSS 19.0. RESULTS The sensitivities (Se) of the G and GM tests alone were 60.0% and 28.6%, respectively, whereas the specificities (Sp) were 92.3% and 93.8%, respectively. Moreover, the positive predictive values (PPV) of the G and GM tests alone were 50.0% and 40.0%, respectively, and the negative predictive values (NPV) were 94.7% and 90.0%, respectively. In addition, the diagnostic odds ratios (DOR) were 18.0 and 6.0, respectively, and the Kappa values were 0.48 (P   0.05), respectively. When the G and GM tests were parallel combined, the Se was 66.7%, the Sp was 92.3%, the PPV was 57.1%, the NPV was 94.7%, the DOR was 24.0, and the Kappa value was 0.55 (P < 0.05). The present study was unable to evaluate the serial diagnosis due to the lack of patients testing positive. CONCLUSIONS The G/GM tests exhibited low Se and PPV when used to diagnose IFRS, while high Sp and NPV. Parallel diagnosis improved the diagnostic Se and DOR values.