The Clinic Value of Multivariate Analysis for Detecting Risk Factors of Thromboembolism in Patients with Persistent Atrial Fibrillation

LI Gang,SHEN Yu-xiang,YUAN Jian-jun,ZHAI Ya-ping,SHANG Bao-jun
DOI: https://doi.org/10.3969/j.issn.1004-437x.2005.01.012
2005-01-01
Abstract:Objective: To study the clinic value of thromboembolism risk assessment in patients with persistent atrial fibrillation(PAF)by making the multivariate analysis for detecting risk factors of them. Methods: 54 patients with PAF were divided into 2 groups: thrombus group(n=26)and non-thrombus group(n=28). 35 patiants with heart diseases but without atrial fibrillation were selected as control. Firstly all subjects underwent echocardiography to detect LA dimension, then the plasma TATGMP-140vWF and D_D levels were assayed with ELISA. In the end,the multivariate analysis was made for the results. Results: In multivariate logistic regression, increase of LA dimension (OR=1.374,P=0.042) vWF (OR=1.031, P=0.030) D_D (OR=7.565, P=0.015) were independently associated with thromboembolism; Among single risk factor, the sensitivity of LA dimension to detect thrombus was the highest, and the specificity of D_D was the highest; Along with more independent risk factors, the specificity to diagnose thromboembolism in patients with PAF was higher, but the sensitivity was lower. Conclusion: Increase of LA dimensionvWFD_D are independent risk factors of thromboembolism in patients with PAF. The assessment value of D_D is greater than of others. Considering more risk factors, the specificity of detecting thrombus is higher and the risk assessment of thromboembolism is more reliable.
What problem does this paper attempt to address?