Application of Local Estimation Methods for Structural Equation Model in Stroke Patients Reported Outcome Evaluation
10.11783/j.issn.1002-3674.2024.05.005
- VernacularTitle:结构方程模型局部估计在脑卒中患者报告结局评价中的应用
- Author:
Yao ZHANG
1
;
Jie YANG
;
Juanjuan YAN
Author Information
1. 山西中医药大学健康服务与管理学院(030024)
- Keywords:
Latent variable;
pSEM;
PRO;
Scale
- From:
Chinese Journal of Health Statistics
2024;41(5):658-661
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the psychological causes in stroke patients by comparing the global and local estimation methods of structural equation model with the survey data of reported outcomes.Methods 475 stroke patients were randomly investigated with the Stroke(patient reported outcome,PRO)scale.The actual data were analyzed with two methods,and the sensitivity of different random effects was analyzed.Results In the process of dimension reduction,the loss information of data was less than 5%.Sensitivity analysis shows that there are no random effects of categorical variables in this study,and there is a certain difference between the results of global and local estimation with structural equation model,and the standardized path coefficients reflected by the latter were slightly greater than the former.Based on the actual data of this study,it was found that social communication can reduce depression and avoidance,and verbal communication can alleviate anxiety to some extent.Conclusion The data involving latent variables and random effects were effectively analyzed by the traditional structural model combined with local estimation method.The causes of psychological problems in stroke patients are complex.
