原发性高血压并左室肥厚的风险预测可视化模型构建及效能验证  

Construction and Validation of a Risk Prediction Visualization Model for Primary Hypertension with Left Ventricular Hypertrophy

作　　者：姬翔 王海虹 孙继英 JI Xiang;WANG Haihong;SUN Jiying(Department of Function,Henan Electric Power Hospital,Zhengzhou Henan 450000,China)

机构地区：[1]河南电力医院功能科,河南郑州450000

出　　处：《临床研究》2025年第3期6-10,共5页Clinical Research

摘　　要：目的针对原发性高血压患者,分析其合并左室肥厚的影响因素,进而建立风险预测可视化模型,并验证其预测效能。方法回顾性分析2020年11月至2023年11月河南电力医院收治的300例原发性高血压患者临床资料,根据2∶1的分组比例,将研究对象分为建模组和验证组,分别为200例和100例。另根据建模组患者是否合并左室肥厚将其进一步分为合并组(n=40)和未合并组(n=160)。针对原发性高血压患者,依赖多因素Logistic回归分析对患者合并左室肥厚的影响因素进行分析,列线图的绘制依托于R3.4.3软件,另列线图模型的内部验证、外部验证、校准度的检验、预测效能、临床净获益率分别通过Bootstrap法、验证组数据、校准曲线、受试者工作特征(ROC)曲线、临床决策曲线分析(DCA)进行。结果合并组年龄、收缩压、超敏C反应蛋白(hs-CRP)、血尿酸(UA)、胱抑素C(Cys C)、同型半胱氨酸(Hcy)均高于未合并组,差异均有统计学意义(P<0.05);多因素Logistic回归分析结果显示,年龄、收缩压、hs-CRP、UA、Cys C、Hcy均是原发性高血压并左室肥厚的危险因素,差异均有统计学意义(P<0.05);本研究列线图模型经Bootstrap法进行内部验证的一致性指数(C-index)为0.837,验证组进行外部验证的C-index为0.823,两组校准曲线与工作曲线拟合度均良好;ROC曲线结果显示,建模组、验证组评估列线图模型预测原发性高血压并左室肥厚的曲线下面积(AUC)分别为0.881、0.847;DCA曲线显示,建模组风险阈值概率为9%~99%,验证组风险阈值概率为9%~90%时,有较高的净获益值。结论原发性高血压患者年龄、收缩压、hs-CRP、UA、Cys C、Hcy较高均是合并左室肥厚的危险因素,据此构建的列线图预测模型具有良好的校准度、预测效能,可以指导临床早期筛查高危患者以及时进行防治。Objective To analyze the influencing factors for left ventricular hypertrophy in patients with primary hypertension and to establish a risk prediction visualization model,verifying its predictive efficacy.Methods A retrospective analysis was conducted on the clinical data of 300 patients with primary hypertension treated at Henan Electric Power Hospital from November 2020 to November 2023.Patients were divided into a modeling group and a validation group in a 2:1 ratio,with 200 cases in the modeling group and 100 in the validation group.Additionally,patients in the modeling group were further divided into a combined group(n=40)and a non-combined group(n=160)based on the presence of Left ventricular hypertrophy.Multivariate logistic regression analysis was utilized to identify the influencing factors for Left ventricular hypertrophy in primary hypertension patients.A nomogram was created using R 3.4.3 software,and the internal validation,external validation,calibration,predictive efficacy,and clinical net benefit of the nomogram model were assessed using the Bootstrap method,validation group data,calibration curves,receiver operating characteristic(ROC)curves,and decision curve analysis(DCA).Results The combined group had higher levels of age,systolic blood pressure,high-sensitivity C-reactive protein(hs-CRP),uric acid(UA),cystatin C(Cys C),and homocysteine(Hcy)compared to the non-combined group,with statistically significant differences(P<0.05).The multivariate logistic regression analysis indicated that age,systolic blood pressure,hs-CRP,UA,Cys C,and Hcy were significant risk factors for Left ventricular hypertrophy in primary hypertension patients(P<0.05).The consistency index(C-index)for the nomogram model after internal validation using the Bootstrap method was 0.837,while the C-index for external validation with the validation group was 0.823,with both groups showing good calibration between the calibration curves and the ideal curve.The ROC curve results showed the area under the curve(AUC)for predicting Le

关 键 词：原发性高血压 左室肥厚 危险因素 可视化模型 

分 类 号：R259[医药卫生—中西医结合]