Construction and validation of a risk prediction model for acute myocardial infarction complicated by malignant ventricular arrhythmias
10.3760/cma.j.cn114656-20250106-00005
- VernacularTitle:急性心肌梗死并发恶性室性心律失常的风险预测模型构建及验证
- Author:
Dongli SONG
1
;
Shengnan LIU
;
Shuo WU
;
Jie GAO
;
Xiao ZHANG
;
Weikai CUI
;
Yifan WANG
;
Jiali WANG
;
Yuguo CHEN
Author Information
1. 山东大学齐鲁医院急诊科，济南 250012
- Keywords:
Myocardial infarction;
Arrhythmia;
Risk factor;
Prediction model;
Predictive factors
- From:
Chinese Journal of Emergency Medicine
2025;34(7):923-931
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the risk factors for in-hospital malignant ventricular arrhythmia (MVA) in acute myocardial infarction (AMI) and to construct and validate a risk prediction model.Methods:This study was a retrospective cohort study. Patients aged≥18 years who were admitted to Qilu Hospital of Shandong University with a diagnosis of AMI and underwent coronary angiography (CAG) from May 2016 to March 2023 were selected, and the patients' clinical routine test indicators and CAG results were collected. Univariate and bidirectional stepwise logistic regression were used to screen out the risk factors for constructing the best prediction model. The prediction model was constructed by combining the results of multivariate logistic regression. The Hosmer-Lemeshow test and ROC curve, calibration curve, and decision curve were drawn to evaluate the model. The nomogram was drawn to visualize the model, and the Bootstrap self-sampling method was used for internal validation. The ROC curve was drawn to evaluate the predictive performance of each risk factor and prediction model. Finally, a multicollinearity test was performed.Results:Among the 4 205 patients finally included in the study, 115 patients (2.735%) developed MVA during hospitalization. The predictive factors screened out included age (X1), diastolic blood pressure (X2), respiratory rate (X3), blood glucose (X4), serum potassium (X5), logarithmic NT-proBNP (X6), myocardial infarction type (NSTEMI=X7, unclassified=X8), J wave (X9), Killip grade (Ⅱ=X10, Ⅲ=X11, Ⅳ=X12), and the regression equation was ln(p/1-p)=-4.699+0.029×X1-0.012×X2+0.059×X3+0.148×X4-1.175×X5+0.866×X6-1.427×X7-0.475×X8+0.758×X9+0.294×X10+0.902×X11+1.815×X12. The area under the ROC curve (AUC) of the model was 0.855 (95% CI: 0.816-0.894), and the Hosmer-Lemeshow test ( χ2=14.178, P=0.077) and the calibration curve showed that the predicted probability was consistent with the actual probability. The probability threshold of 0% to 65% had a better clinical net benefit. The area under the internal validation ROC curve (AUC) was 0.855, 95% CI: 0.813-0.891. The prediction performance of the nine variables was stronger than that of any single variable. There was no multicollinearity between the variables. Conclusions:Age, diastolic blood pressure, respiratory rate, blood glucose, serum potassium, NT-proBNP, type of AMI, J wave, and Killip class are forecasting indicator for in-hospital MVA in AMI. The risk prediction model based on the above factors has good predictive performance.
