Risk Factors and Predictive Model for Otitis Media With Effusion After Radiotherapy in Nasopharyngeal Carcinoma:A Retrospective StudyOA

Objective:To investigate the impact of different radiation fields on the risk of otitis media with effusion(OME)in nasopharyngeal carcinoma(NPC)patients and to develop a nomogram to identify high-risk patients after radiotherapy.Methods:Clinical data were collected from 170 NPC patients who underwent radiotherapy at Peking Union Medical College Hospital(January 2016-December 2021,development group)and 73 patients(January 2022-June 2024,validation group).Multivariate logistic regression analyses were used to identify risk factors for OME,and a nomogram was constructed based on these findings.Results:Multivariate analysis revealed that gender,invaded eustachian tube,and radiation to the parotid and periauricular field and parapharyngeal space were independent risk factors for OME.The nomogram demonstrated good discrimination,with AUC values of 0.813(development group)and 0.798(validation group).Calibration was confirmed with Hosmer-Lemeshow test results(p>0.33),and decision curve analysis indicated significant clinical utility within a threshold probability range of 0.09-0.94.When the risk threshold exceeded 0.38,the predicted high-risk population matched well with actual OME cases,highlighting high clinical predictive efficiency.Conclusion:In conclusion,radiation fields are significant risk factors for OME in NPC patients.The developed nomogram enables early identification of high-risk patients,supporting timely intervention.Summary Parotid-periauricular and parapharyngeal space radiation fields are independent risk factors for otitis media with effusion(OME)in nasopharyngeal carcinoma(NPC)patients.A clinically applicable nomogram integrating gender,eustachian tube invasion,parotid-periauricular radiation,and parapharyngeal space radiation enables pretreatment risk stratification.Decision curve analysis confirms significant net benefit for interventions based on the model within a wide threshold probability range of 0.09-0.94,with optimal risk prediction accuracy above 38%probability.Findings highlight the critical need for field-specific optimization in radiotherapy planning to minimize OME risk,particularly for high-risk anatomical regions.