OBJECTIVETo investigate the correlated factors and clinical features of fungal rhinosinusitis.

METHODSThe clinical data of 110 patients with fungal rhinosinusitis treated by surgery and another group of 110 patients with chronic rhinosinusitis who were sampled randomly between January 1999 and June 2004 in our hospital were retrospectively compared. The correlated factors and the clinical features of fungal rhinosinusitis were investigated by using the multiple factor Logistic regression analysis and chi-square test. The pathological types of 110 fungal rhinosinusitis were classified by using Gomori methenamine silver staining which was special for fungi.

RESULTSThe logistic regression predictive equation for fungal rhinosinusitis was : y = -8.713 + 0.496x1 + 4.575x2 + 1. 190x3 + 4.119x4 + 1.199x5 + 2. 698x6, P = exp (y)/[1 + exp(y)], in which the concomitant variables were course of the disease (x1), haem-nasal discharge (x2), headache (x3), calcified plaque in CT scan (x4), age (x5) and unilateral/bilateral sinus lesion (x6), respectively. The P value meant the probability of suffering fungal rhinosinusitis. Compared with chronic rhinosinusitis, the clinical features of fungal rhinosinusitis were female, over 40-year-old, course of disease < 3 years, headache, haem-nasal discharge, unilateral sinus lesion and calcified plaque in CT scan. Among the 110 patients with fungal rhinosinusitis, 34 cases were chronic invasive and 76 were non-invasive.

CONCLUSIONSThe clinical features of fungal rhinosinusitis are significant for the diagnosis, and it can be predicted by using the suitable logistic predictable equation.