Development and evaluation of a clinical feature-based diagnostic model for aortic dissection

CHEN Zhuang-shuo; HUANG Guo-hai; WANG Wan; LU Sen-rong; ZHAO Xian-wu; ZHENG Run-fa

doi:10.13820/j.cnki.gdyx.20250849

Abstract：Objective To analyze the clinical differences between patients with and without aortic dissection(AD),and to construct a simple and practical risk prediction model.Methods A total of 166 patients presenting with chest pain were included and categorized into the AD group and the non-AD group.Clinical data were compared between the two groups.Multivariate logistic regression was used to identify independent risk factors for AD,which were then used to construct a nomogram for survival risk prediction.The model's performance was evaluated using a calibration curve and the area under the receiver operating characteristic(ROC)curve(AUC).Results There were 73 patients with AD,and 93 patients without AD.Compared to non-AD patients,AD patients had a lower prevalence of diabetes but a higher pro-portion of comorbid back pain,abdominal pain,severe pain,sweating,and asymmetric blood pressure.In laboratory find-ings,AD patients exhibited higher white blood cell counts and D-dimer levels,while platelet counts,creatine kinase(CK),creatine kinase isozyme MB(CK-MB)levels,and the proportion of troponin I(TNI)＜40 ng/L were lower.Additionally,electrocardiogram(ECG)analysis showed a lower incidence of ST-T segment changes in AD patients(P＜0.05).Multivariate logistic regression analysis identified a history of diabetes(OR=0.131,P=0.026),severe pain(OR=29.748,P＜0.001),troponin I＞40 ng/L(OR=0.093,P=0.009),and D-dimer＞0.5 μg/mL(OR=30.997,P＜0.001)as independent predictors of AD.The nomogram constructed based on these factors demonstrated high accuracy,with an AUC of 0.956.Conclusion This study provides a valuable diagnostic tool for emergency physicians,which may help reduce the misdiagnosis and missed diagnosis rates of AD,and improve early diagnosis rates.