预测维持性血液透析患者心血管事件发生风险列线图模型的建立

Development of a clinical nomogram for predicting cardiovascular events in maintenance hemodialysis patients

目的探讨维持性血液透析患者发生心血管事件的危险因素,并构建相应列线图模型。方法纳入2014年5月~2017年3月在温州市中西医结合医院行维持性血液透析治疗的229例终末期尿毒症患者,采用门诊或电话进行定期随访,每季度1次,随访终点为主要心血管疾病不良事件,根据COX多因素回归结果建立列线图模型。结果年龄≥60岁(HR=2.80;95%CI:1.56~5.01;P<0.001)、心血管疾病史(HR=2.08;95%CI:1.07~4.07;P=0.032)、透析龄≥36月(HR=2.36;95%CI:1.25~4.44;P=0.008)以及QT间期离散度≥63 ms(HR=2.14;95%CI:1.20~3.82;P=0.010)是血液透析患者发生心血管事件的独立危险因素。该列线图模型初始C-index为0.719(95%CI:0.652~0.786),采用重复Bootstrap自抽样方法进行1000次内部验证后C-index为0.722(95%CI:0.656~0.788),且经Hosmer-Lemeshow检验该模型拥有良好的区分度和一致性(χ~2=8.124,P=0.632)。结论基于上述4个影响因素构建的列线图能较为准确预测维持性血液透析患者心血管事件发生风险。

Objective To investigate the influencing factors for cardiovascular events in maintenance hemodialysis(MHD)patients,and to construct a clinical nomogram risk predictive model.Methods A total of 229 patients with maintenance hemodialysis,who admitted into Wenzhou Integrated Traditional Chinese and Western Medicine Hospital during May 2014 to March 2017,were analyzed retrospectively.The major adverse events of the cardiovascular disease were tracked.A nomogram was developed on the results of the multivariate Cox model.Results Four risk factors were found by COX regression model,including age≥60 years(HR=2.80;95%CI:1.56~5.01;P<0.001),the history of cardiovascular disease(HR=2.08;95%CI:1.07~4.07;P=0.032),dialysis age≥36 months(HR=2.36;95%CI:1.25~4.44;P=0.008),and QT interval dispersion.≥63 ms(HR=2.14;95%CI:1.20~3.82;P=0.010).The initial concordance index(C-index)of nomogram was 0.719(95%CI:0.652~0.786),and C-index of nomogram was 0.722(95%CI:0.656~0.788)after1000 times of internal validation.Hosmer-Lemeshow statistic(χ~2=8.124,P=0.632)Conclusion The nomogram based on the related factors could accurately predict cardiovascular events in maintenance hemodialysis(MHD)patients.