Construction of risk prediction model of heart failure in patients with coronary heart disease based on LASSO regression
10.3969/j.issn.1673-9701.2024.28.001
- VernacularTitle:基于LASSO回归的冠心病患者发生心力衰竭的风险预测模型构建
- Author:
Yikang XU
1
;
Jingru MA
;
Yang YANG
;
Lei LIU
;
Zhifeng ZHANG
;
Siqi SUN
;
Manman LI
;
Kaiwen ZHAN
Author Information
1. 沈阳医学院附属第二医院心内科,辽宁沈阳 110000
- Keywords:
Coronary heart disease;
Heart failure;
Nomograms;
Risk prediction;
LASSO regression
- From:
China Modern Doctor
2024;62(28):1-5,46
- CountryChina
- Language:Chinese
-
Abstract:
Objective To analyze the risk factors of heart failure in patients with coronary heart disease(CHD),and to construct and verify a nomogram prediction model for the risk of heart failure in patients with CHD.Methods The clinical data of 453 patients with CHD who were hospitalized in the Second Affiliated Hospital of Shenyang Medical College from January to December 2022 were retrospectively analyzed,including 278 patients with CHD combined with heart failure and 175 patients without heart failure.The patients were divided into training group(318 cases)and validation group(135 cases)according to the ratio of 7:3.R software was applied to perform LASSO regression to screen the risk factors,and Logistic regression to establish a prediction model and construct a nomogram.The calibration curve and receiver operating characteristic(ROC)curve were used to evaluate the calibration and discrimination of the model.Results LASSO regression analysis ultimately screened five risk factors from 22 variables,and Logistic regression results showed that age,smoking,history of myocardial infarction,New York Heart Association(NYHA)cardiac function class Ⅳ,and left ventricular ejection fraction(LVEF)were all independent risk factors for heart failure in CHD patients(P＜0.05).The model formula was Z=-2.927+0.045 × age+0.886 × smoking+0.808 × history of myocardial infarction-2.829 × NYHA cardiac function class Ⅳ+0.037×LVFF.Internal validation of the model showed that area under the curve was 0.727(95％CI:0.588-0.752),the sensitivity was 40.4％,the specificity was 84.3％,and the Youden index was 0.247.According to the calibration curve,the predicted value of the calibration curve was highly consistent with the actual value,and the Brier score was 0.106.Conclusion The risk prediction model for heart failure in patients with CHD based on LASSO regression has good discrimination and prediction efficiency,which can be used as an evaluation tool for medical staff to predict the risk of patients.
