目的  构建并验证接受血液透析治疗的人类免疫缺陷病毒(human immunodeficiency virus，HIV)/获得性免疫缺陷综合征(acquired immunodeficiency syndrome，AIDS)患者病死率预测模型，评估其预测效能和临床实用性。 方法  采用回顾性队列研究，纳入2013年7月─2024年6月于成都公共卫生临床医学中心接受规律血液透析(≥3个月)的81例HIV感染患者(年龄≥18岁，CKD G5)，按1:1随机分为训练集(40例)和验证集(41例)。单因素COX分析筛选预后变量，逐步回归构建多因素模型并绘制列线图，通过受试者工作特征(receiver operating characteristic，ROC)分析、校正曲线及决策曲线分析评估模型性能。  结果  训练集单因素COX回归分析筛选出血红蛋白(HR=0.942，95%CI：0.904～0.981， P=0.004)、白蛋白(HR=0.848，95%CI：0.739～0.973，P=0.019)和年龄(HR=1.090，95%CI：1.010～1.170， P=0.023)作为候选预后变量，经多因素逐步COX回归确认三者均为独立预后因素。训练集列线图对1年、3年和5年生存预测的AUC值分别为0.846(95%CI：0.692～1.000)、0.750(95%CI：0.566～0.934)和0.791(95%CI：0.618～0.964)，验证集分别为0.806(95%CI：0.644～0.968)、0.723(95%CI：0.540～0.906)和0.640(95%CI：0.446～0.834)。校正曲线斜率接近1。决策曲线分析显示在阈概率10%～80%范围内模型具有显著的临床净收益。 结论  基于血红蛋白、白蛋白和年龄构建的病死率预测列线图具有良好的判别力、校准度和临床实用性，可为HIV/AIDS血液透析患者个体化风险评估和临床决策提供简洁、稳健的量化工具。

Objective  To construct and validate a mortality prediction nomogram for patients with human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) undergoing hemodialysis, and to evaluate its predictive performance and clinical utility. Methods  A retrospective cohort study was conducted, enrolling 81 HIV-infected patients (age≥18 years, CKD stage G5) who received regular hemodialysis (≥3 months) at Chengdu Public Health Clinical Medical Center from July 2013 to June 2024. Patients were randomly divided into a training set (n=40) and a validation set (n=41) at a 1:1 ratio. Univariate COX regression analysis was used to screen prognostic variables, and stepwise regression was applied to build a multivariate model, from which a nomogram was constructed. Model performance was assessed using receiver operating characteristic (ROC) analysis, calibration curves, and decision curve analysis (DCA). Results  Univariate Cox regression in the training set identified hemoglobin (HR= 0.942, 95% CI: 0.904～0.981, P=0.004), albumin (HR=0.848, 95% CI:0.739～0.973, P=0.019), and age (HR=1.090, 95% CI:1.010～1.170, P=0.023) as candidate prognostic variables. Multivariate stepwise Cox regression confirmed all three as independent prognostic factors. The AUC values of the nomogram for predicting 1-, 3-, and 5-year survival in the training set were 0.846 (95% CI: 0.692～1.000), 0.750 (95% CI: 0.566～0.934), and 0.791 (95% CI: 0.618～0.964), respectively; the corresponding values in the validation set were 0.806 (95% CI: 0.644～0.968), 0.723 (95% CI: 0.540～0.906), and 0.640 (95% CI: 0.446～0.834). The calibration curve slopes were close to 1. Decision curve analysis demonstrated significant clinical net benefit across threshold probabilities ranging from 10% to 80%. Conclusion The mortality prediction nomogram based on hemoglobin, albumin, and age exhibits good discriminative ability, calibration, and clinical utility, providing a concise and robust quantitative tool for individualized risk assessment and clinical decision‑making in HIV/AIDS patients on hemodialysis.