Analysis of the Construction of a Predictive Model for Eosinophilic Chronic Rhinosinusitis

PURPOSE: This study aimed to determine indices to diagnose and predict eosinophilic chronic rhinosinusitis (ECRS) during the initial clinic visit.

PATIENTS AND METHODS: We retrospectively analyzed 116 patients with chronic rhinosinusitis who underwent endoscopic sinus surgery and were classified according to the postoperative pathological diagnosis.

General data and various clinical indicators were analyzed, and indicators with statistically significant differences between groups were further incorporated into a multivariate logistic regression to establish a comprehensive prediction model.

The receiver operating characteristic (ROC) curve was used to compare the two significant valuable single factors from previous studies, the difference in CT scores between the ethmoid sinus and the sum difference of the maxillary sinus (EM difference) and the absolute value of peripheral blood eosinophil (bEOS), with a comprehensive prediction model.

RESULTS: There were significant differences in history of allergic asthma (p < 0.001), visual analog scale (VAS) score (p=0.005), sino-nasal outcome test-22(SNOT-22) scale score (p=0.004), Lund-Mackay scale score (p=0.017), EM difference (p=0.002), percentage of bEOS (%)(p=0.001), and absolute value of bEOS (×109/L) (p=0.000) between the two groups (p< 0.05).

The history of allergic disease, VAS and bEOS were screened out and included in the comprehensive prediction model.

The area under the curve (AUC) of the comprehensive prediction model (0.804)> the AUC of the absolute value of the bEOS (0.764)>the AUC of the EM difference (0.655).

The AUC of the EM difference and the comprehensive prediction model were statistically different (P=0.025).

There was no statistical difference between the absolute value of bEOS and the AUC of the comprehensive prediction model.

CONCLUSION: The comprehensive prediction model covering the three aspects of allergic asthma history, VAS score, and bEOS count had the highest AUC compared to the other predictors and had good predictive power for the diagnosis of ECRS.