目的 分析鼻咽癌同步放化疗患者营养不良的危险因素并建立列线图,为临床早期识别与干预提供实用工具。方法 以便利抽样法选取2023年2月至2025年2月某院收治的251例鼻咽癌同步放化疗患者为研究对象,以主观整体评估量表(patient-generated subjective global assessment,PG-SGA)评分≥4分为发生营养不良(113例),PG-SGA<4分为未发生营养不良(138例),分析不同临床特征患者营养不良发生情况;采用多因素Logistic回归分析营养不良的影响因素,并构建列线图模型。结果 与其他患者相比,体质量丢失>5%、食物摄入中及重度减少、中及重度应激状态、重度口腔黏膜炎、体质量指数<18.5 kg/m²、白蛋白<35 g/L、发生贫血的患者营养不良发生率更高,差异均有统计学意义(均P<0.05)。体质量、食物摄入、代谢应激状态、口腔黏膜炎、体质量指数、白蛋白水平、贫血是鼻咽癌患者同步放化疗后营养不良的独立影响因素(均P<0.05)。列线图预测模型的受试者工作特性曲线下的面积为0.986,C-index为0.979,校准曲线接近理想曲线。食物摄入、体质量及代谢应激状态的重要性度量分别为48.900%、43.900%、42.300%。结论 基于上述因素构建的列线图模型对营养不良有较高的预测价值。

Objective To analyze the risk factors for malnutrition in patients with nasopharyngeal carcinoma(NPC) undergoing concurrent chemoradiotherapy(CCRT),and to establish a nomogram for prediction,so as to provide a practical tool for early clinical identification and intervention. Methods A total of 251 patients with NPC undergoing CCRT admitted to a hospital from February 2023 to February 2025 were selected by convenience sampling.Malnutrition was assessed using the Patient-generated Subjective Global Assessment(PG-SGA).A total of 113 patients had PG-SGA≥4(malnutrition group) and 138 had PG-SGA<4(non-malnutrition group).Baseline characteristics were collected and malnutrition rates were compared across different clinical profiles.Multivariate Logistic regression analysis identified independent risk factors for malnutrition post-chemoradiotherapy,followed by construction of a nomogram prediction model incorporating these factors.Results Patients with weight loss>5%,moderate-to-severe reduction in food intake,moderate-to-severe metabolic stress,severe oral mucositis,BMI<18.5 kg/m²,albumin(ALB)<35 g/L,and anemia had significantly higher PG-SGA scores than other patients (all P<0.05).Body weight,food intake,metabolic stress,oral mucositis,BMI,ALB,and anemia were influencing factors of malnutrition after CCRT in NPC patients (all P<0.05).The nomogram constructed based on these factors demonstrated good predictive performance,with an AUC of 0.986,a C-index of 0.979,and a calibration curve close to the ideal curve. The %IncMse of food intake,body weight,and metabolic stress were 48.900%,43.900%,and 42.300%,respectively.Conclusions The nomogram model constructed based on the above factors has high predictive value.

R473.76

吴阶平医学基金(20200709300)；广东省基础与应用基础研究基金(2025A1515010068)