Establishment of Risk Prediction Model for Cardiovascular Outcomes in Patients with Connective Tissue Disease
10.3969/j.issn.1005-5185.2025.10.013
- VernacularTitle:结缔组织病不良心血管预后超声心动图预测模型的构建
- Author:
Yilu SHI
1
;
Yaxi WANG
1
;
Dan ZHANG
1
;
Xiaoxiao LIU
1
;
Shurong YUN
1
;
Ting SONG
1
;
Xiaoshan ZHANG
1
Author Information
1. 内蒙古医科大学附属医院超声医学科,内蒙古 呼和浩特 010050
- Publication Type:Journal Article
- Keywords:
Connective tissue diseases;
Echocardiography;
Cardiovascular system;
Forecasting
- From:
Chinese Journal of Medical Imaging
2025;33(10):1104-1112
- CountryChina
- Language:Chinese
-
Abstract:
Purpose To assess the utility of echocardiographic parameters in predicting adverse cardiovascular in patients with connective tissue disease.Materials and Methods A retrospective analysis was conducted on the clinical and echocardiographic records of patients with connective tissue disease from the Affiliated Hospital of Inner Mongolia Medical University(June 1st,2020 to June 1st,2023)who had complete medical data and at least twelve months of follow-up.Variables were screened based on univariate analysis,combined with clinical expertise and XGBoost feature weight analysis;this information was used to construct a Cox proportional hazards regression model designed to predict composite endpoint events of adverse cardiovascular outcomes.Internal validation was performed using the Bootstrap resampling method,and the model's performance was evaluated.Results The study included 123 participants.The incidence of positive events reached 39.02%(48/123).Mitral valve early diastolic annular velocity(reflecting left heart function)(HR=0.79,P=0.041)and tricuspid annular plane systolic excursion(reflecting right heart function)(HR=0.92,P=0.044)emerged as significant predictors for adverse cardiovascular outcomes.Compared with the clinical model,the model combined with left heart function parameters showed significant improvements in both risk classification and absolute accuracy for short-term and medium-term adverse prognosis(NRI365=0.054,IDI365=0.060,NRI730=0.064,IDI730=0.079,all P<0.05)and optimized risk classification for long-term adverse prognosis(NRI1 095=0.256,P<0.05).In contrast,the model combined with right heart function parameters improved risk classification at all time points(NRI365=0.054,NRI730=0.000,NRI1 095=0.135,all P<0.05).Conclusion Mitral valve early diastolic annular velocity and tricuspid annular plane systolic excursion,which reflect cardiac function,are factors for predicting adverse cardiovascular outcomes among individuals diagnosed with connective tissue disease.Multi-parameter combined models incorporating echocardiographic variables can provide incremental predictive value compared with clinical models.
