首页|期刊导航|中国实用内科杂志|基于列线图模型的血液透析患者透析中低血压风险预测与评估

基于列线图模型的血液透析患者透析中低血压风险预测与评估OA

Risk prediction and assessment of intradialytic hypotension in hemodialysis patients based on a nomogram model

中文摘要英文摘要

目的 探讨维持性血液透析(MHD)患者透析中低血压(IDH)的危险因素,构建列线图模型来预测IDH的发生风险并评估其价值.方法 回顾性分析2022-04-01-2022-06-30首都医科大学附属北京潞河医院血液净化中心200例MHD患者资料.采用患者2021-10-01-2021-12-31时段的数据进行时间验证.通过单因素及多因素logistic回归分析来确定1DH的独立危险因素,并建立列线图预测模型,以校准曲线及受试者工作特征曲线(ROC)评估模型的校准度及区分度,建立决策曲线(DCA)评估其临床使用价值.结果 透析前收缩压(OR=0.974,95%CI 0.953~0.995,P=0.018),尿素氮(O R=0.927,95%CI 0.862~0.997,P=0.042)和中性粒细胞与淋巴细胞比值(NLR)(OR=1.504,95%CI 1.204~1.880,P<0.001)是1DH发生的独立危险因素.基于上述因素和白蛋白构建MHD患者发生IDH的列线图预测模型,训练集和验证集ROC曲线下面积(AUC)分别为0.751(95%CI 0.675~0.827)和0.729(95%CI 0.642~0.817).验证集校准曲线Hosmer-Lemeshow检验结果显示拟合良好(x2=7.694,P=0.464);DCA分析结果显示当阈值概率在0.17~0.80时模型能产生更好的临床效益.结论 基于透析前收缩压、尿素氮、NLR及白蛋白构建的预测IDH的列线图模型简单实用,准确性较高,具有一定的临床应用价值.

Objective To investigate the risk factors for intradialytic hypotension(IDH)in maintenance hemodialysis(MHD)patients,and to develop a nomogram model to predict the risk of IDH and evaluate the model's predictive value.Methods A retrospective analysis was conducted on clinical data from 200 MHD patients at the Blood Purification Center of Beijing Luhe Hospital,Capital Medical University,between April 1 and June 30,2022.Temporal validation was performed using data of these patients from October 1 to December 31,2021.Univariate and multivariate logistic regression analyses were used to identify independent risk factors for IDH,and a nomogram prediction model was established.The calibration and discrimination of the model were evaluated using calibration curves and ROC curves,and a decision curve analysis(DC A)was performed to assess its clinical utility.Results Pre-dialysis systolic blood pressure(OR=0.974,95%CI:0.953-0.995;P=0.018),blood urea nitrogen(OR=0.927,95%CI:0.862-0.997;P=0.042),and neutrophil-to-lymphocyte ratio(NLR;OR=1.504,95%CI:1.204-1.880;P<0.001)were identified as independent risk factors for intradialytic hypotension(IDH).A nomogram prediction model for intradialytic hypotension(IDH)occurrence in MHD patients was constructed based on the aforementioned factors and albumin levels.The area under the ROC curve(AUC)was 0.751(95%CI:0.675-0.827)in the training set and 0.729(95%CI:0.642-0.817)in the validation set.The calibration curve of the validation set demonstrated good fit by the Hosmer-Lemeshow test(x2=7.694,P=0.464);decision curve analysis(DCA)indicated enhanced clinical net benefit when the threshold probability ranged from 0.17 to 0.80.Conclusion The nomogram model for predicting IDH constructed based on predialysis systolic blood pressure,blood urea nitrogen,NLR,and albumin is simple and practical,and has high accuracy,demonstrating certain clinical application value.

连晓英;刘小菁;马也娉;曹倩颖;史亚男;杨帆;邵枫;匡泽民;李忠心

首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京潞河医院肾内科,北京 101100首都医科大学附属北京安贞医院心内科高血压中心,北京 100029首都医科大学附属北京潞河医院肾内科,北京 101100

医药卫生

血液透析透析中低血压中性粒细胞与淋巴细胞比值预测模型

hemodialysisintradialytic hypotensionneutrophil-to-lymphocyte ratioprediction model

《中国实用内科杂志》 2026 (3)

208-214,7

北京市自然科学基金面上专项(M22009)

10.19538/j.nk2026030105

评论