维持性血液透析患者死亡的独立危险因素分析及列线图模型的构建

Analysis of independent risk factors for death in patients on maintenance hemodialysis and construction of a predictive nomogram model

摘要:

目的 分析维持性血液透析（MHD）患者死亡的独立危险因素，并构建列线图模型以预测死亡风险，为临床干预提供依据。方法 选取2019年4月—2024年4月在天津市北辰医院进行血液透析的280例患者为研究对象。根据患者临床结局分为死亡组（74例）与存活组（206例）。收集患者一般资料（年龄、性别、体质量指数、高血压史、糖尿病史、冠心病史、心力衰竭史、脑血管疾病史、外周血管疾病史、透析时间等）、首次透析前48 h内实验室指标［血钙、血磷、血钾、血钠、血氯、血镁、血铁、血红蛋白、血清白蛋白、前白蛋白、血清甲状旁腺激素、血尿素氮、血肌酐、β2微球蛋白、估计肾小球滤过率（eGFR）、C反应蛋白（CRP）、甘油三酯（TG）、总胆固醇（TC）、脑钠肽（BNP）］、首次血液透析住院时超声心动图检查指标［左心房直径（LAD）、左心室后壁厚度（LCPW）、左心室舒张末期内径（LVDd）］。采用多因素逐步Cox回归模型分析MHD患者死亡的影响因素并构建列线图模型。结果 死亡组的年龄、血氯、TC、β2微球蛋白、eGFR、CRP、BNP、主动脉钙化、LAD均高于存活组（P <0.05），血磷、血钾、血铁、转铁蛋白饱、血清白蛋白、低密度脂蛋白胆固醇、血肌酐、血尿素氮均低于存活组（P <0.05）。多因素逐步Cox回归分析结果表明，血钾高［H^R=0.530（95% CI：0.319，0.880）］、血铁高［H^R=0.854（95% CI：0.752，0.971）］、血肌酐高［H^R=0.994（95% CI：0.991，0.997）］均为MHD患者死亡的保护因素（P <0.05）；eGFR［H^R=1.112（95% CI：1.009，1.226）］、BNP［H^R=1.004（95% CI：1.001，1.007）］均为MHD患者死亡的危险因素（P <0.05）。结论 确定了MHD患者的死亡危险因素，并通过构建列线图模型有效预测了患者的生存情况，帮助临床医生更好地理解死亡风险，从而优化治疗策略，改善患者预后。

Abstract:

Objective To identify independent risk factors for mortality in patients undergoing maintenance hemodialysis (MHD) and develop a predictive nomogram model.Methods A retrospective analysis included 280 MHD patients treated between April 2019 and April 2024. Patients were categorized into death (n = 74) and survival (n = 206) groups. Data collected included demographics, comorbidities, dialysis vintage, pre-dialysis laboratory parameters (electrolytes, hemoglobin, albumin, prealbumin, PTH, urea, creatinine, β2-microglobulin, eGFR, CRP, lipids, BNP), and initial dialysis echocardiography (LAD, LCPW, LVDd). Cox proportional hazards regression identified mortality risk factors, and a nomogram was constructed.Results The death group had higher age, chloride, TC, β2-microglobulin, eGFR, CRP, BNP, aortic calcification prevalence, and LAD (P < 0.05), and lower phosphate, potassium, iron, transferrin saturation, albumin, LDL-C, creatinine, and urea nitrogen (P < 0.05). Multivariate Cox analysis identified high potassium [H^R = 0.530 (95% CI: 0.319, 0.880) ], high iron [H^R = 0.854 (95% CI: 0.752, 0.971) ], and high creatinine [H^R = 0.994 (95% CI: 0.991, 0.997) ] as protective factors, while high eGFR [H^R = 1.112 (95% CI: 1.009, 1.226) ] and high BNP [H^R = 1.004 (95% CI: 1.001, 1.007) ] were independent risk factors for death (P < 0.05).Conclusion Key factors influencing MHD mortality were identified. The nomogram model effectively predicts survival, aiding clinicians in risk assessment and treatment optimization.