Abstract: Objective: To analyze the influencing factors of moderate to severe radiation-induced oral mucositis (RIOM) in patients with head and neck tumors after radiotherapy and to construct a risk prediction model, thereby providing a basis for the identification of high-risk patients and the formulation of prevention strategies. Methods: The retrospective study was conducted on 395 patients with head and neck tumors who received radiotherapy between January 2022 and June 2025. According to the oral mucosa grading standard, the patients were divided into the mild RIOM group and the moderate-to-severe RIOM group. Univariate analysis and logistic regression analysis were used to screen for independent influencing factors of moderate to severe RIOM. The R 4.1.3 software was used to conduct a nomogram prediction model for the accuracy and predictive efficacy. Results: Among the 395 patients, 199 developed moderate to severe RIOM, with an incidence rate of 50.38%. Logistic regression analysis revealed that cumulative radiation dose, nutritional status, pre-radiotherapy oral frailty, diabetes, targeted therapy, and concurrent chemotherapy were influencing factors for RIOM in patients with head and neck tumors receiving radiotherapy (P<0.05). The area under the curve (AUC) for predicting moderate to severe RIOM was 0.839, with a sensitivity of 0.79 and a specificity of 0.83. The optimal threshold value was 0.487. The consistency test of the calibration curve shows that the predicted probability converges to the actual probability. Conclusion: Patients with head and neck tumors receiving radiotherapy are at a high risk of developing moderate to severe RIOM. Those with higher cumulative radiation doses, poorer nutritional status, pre-radiotherapy oral frailty, diabetes, targeted therapy, and concurrent chemotherapy are more likely to develop moderate to severe RIOM. The constructed risk prediction model demonstrates good predictive performance and can provide a reference for clinical healthcare professionals to predict the risk of moderate to severe RIOM in patients undergoing radiotherapy for head and neck tumors.

Key words: Head and neck tumors, Radiotherapy, Oral mucositis, Risk prediction model

摘要： 目的：分析头颈部肿瘤患者放疗后发生中重度放射性口腔黏膜炎（Radiation-Induced Oral Mucositis，RIOM）的影响因素并构建风险预测模型，为高风险患者的识别和预防策略的制定提供依据。方法：回顾性选取 2022 年 1 月至 2025 年 6 月收治的 395 例接受放疗的头颈部肿瘤患者，按口腔黏膜分级标准将患者分为轻度 RIOM 组和中重度 RIOM 组，采用单因素分析和 Logistic 回归分析筛选中重度 RIOM 的独立影响因素，使用 R 4.1.3 构建列线图，评估预测模型的准确性和预测效能。结果：395 例患者发生中重度 RIOM199 例，发生率为50.38%。Logistic 回归分析结果显示，累计放疗剂量、营养状况、放疗前口腔衰弱、糖尿病、靶向治疗、 同步化疗是头颈部肿瘤放疗患者发生 RIOM 的影响因素（P<0.05）。预测模型受试者工作特征曲线下面积为 0.839，灵敏度为 0.79，特异度为 0.83；最佳阈值为 0.487。校正曲线一致性检验显示，预测概率与实际概率趋于一致。结论：头颈部肿瘤放疗患者 发生中重度 RIOM 的风险较高，累计放疗剂量越高、营养状况越差、放疗前存在口腔衰弱、合并糖尿病、靶向治疗、同步化疗的患者更易发生中重度 RIOM。构建的风险预测模型预测性能良好，可为临床医护人员预测头颈部肿瘤放疗患者中重度 RIOM 的发生风险提供参考。

关键词: 头颈部肿瘤, 放射治疗, 口腔黏膜炎, 风险预测模型