目的：基于糖尿病性白内障(DC)患者白内障超声乳化术后黄斑水肿(ME)风险因素，构建ME风险列线图预测模型。方法：回顾性收集2022年1月至2024年12月于我院接受白内障超声乳化术的1 751例1 751眼DC患者资料，根据术后是否发生ME，将患者分为ME组(138例)、N-ME组(1 613例)。通过单因素和Logistic多元回归分析，确定DC患者白内障超声乳化术后ME的危险因素，构建列线图风险预测模型，绘制受试者工作特征曲线(ROC)、校正曲线、Hosmer-Lemeshow拟合优度检验评估模型的区分度和校准度，采用决策曲线评估模型的临床收益率。结果：ME组患者年龄、糖尿病病程、胰岛素治疗占比、视网膜病变占比、最佳矫正视力(BCVA)、黄斑中心凹视网膜厚度(CSMT)、黄斑容积、糖化血红蛋白(HbA1c)、血管内皮生长因子(VEGF)均高于N-ME组(均P<0.05)。经Logistic多元回归分析，糖尿病病程、视网膜病变、BCVA、CSMT、黄斑容积、HbA1c、VEGF是DC患者白内障超声乳化术后ME发生的危险因素(均P<0.05)。基于危险因素构建列线图风险预测模型，ROC曲线提示模型区分度良好[训练集的AUC为0.998(95%CI：0.997-1.000)，验证集的AUC为0.999(95%CI：0.997-1.000)]，校正曲线、Hosmer-Lemeshow拟合优度检验提示模型的拟合度较高(训练集：R2=0.917， χ2=0.806，P=0.999； 验证集：R2=0.900， χ2=0.675，P=1.000)。决策曲线显示，模型在0.00-1.00阈值概率范围内具有较高的净收益率。结论：糖尿病病程、视网膜病变、BCVA、CSMT、黄斑容积、HbA1c、VEGF是DC患者白内障超声乳化术后ME的危险因素，基于此构建的列线图风险预测模型在预测DC患者白内障超声乳化术后ME风险有较好的区分度和一致性。

AIM: To construct a risk nomogram prediction model of macular edema(ME)based on the risk factors of ME after phacoemulsification in diabetic cataract(DC)patients.METHODS: A retrospective collection of data was conducted on 1 751 DC patients(1 751 eyes)who underwent cataract phacoemulsification surgery in the hospital from January 2022 to December 2024. Based on whether they developed ME after surgery, the patients were divided into the ME group(n=138)and the N-ME group(n=1 613). By conducting univariate and Logistic multiple regression analysis, the risk factors for postoperative ME in DC patients undergoing phacoemulsification were identified. A nomogram of risk prediction model was constructed, and the receiver operating characteristic(ROC)curve, calibration curve, Hosmer-Lemeshow goodness of fit test were plotted to evaluate the discrimination and calibration of the model. The decision curve was used to evaluate the clinical return on investment of the model.RESULTS: Age, course of diabetes, proportion of insulin treatment, proportion of retinopathy, best corrected visual acuity(BCVA), central subfield macular thickness(CSMT), macular volume, glycated hemoglobin(HbA1c), vascular endothelial growth factor(VEGF)in the ME group were higher than those in the N-ME group(all P<0.05). Multivariate Logistic regression analysis showed that diabetes course, retinopathy, BCVA, CSMT, macular volume, HbA1c and VEGF were the risk factors for ME after phacoemulsification in DC patients(all P<0.05). A nomogram of risk prediction model was constructed based on risk factors, and the ROC curve suggested good model differentiation [AUC of training set was 0.998(95% CI: 0.997-1.000), and AUC of validation set was 0.999(95% CI: 0.997-1.000)], set: R2=0.917, χ2=0.806, P=0.999; verification set: R2=0.900, χ2=0.675, P=1.000). The decision curve showed that the model had a high net return rate within the probability range of 0.00-1.00 threshold.CONCLUSION: Diabetes course, retinopathy, BCVA, CSMT, macular volume, HbA1c and VEGF are risk factors for ME after cataract phacoemulsification in DC patients. The nomogram of risk prediction model based on this construction has good differentiation and consistency in predicting the risk of ME after cataract phacoemulsification in DC patients.