Qin Cuijie, Li Chuang, Luo Yunpeng, Li Zhen, Cao Hui
Department of Endovascular Surgery, The First Affiliated Hospital of Zhengzhou University, Zhengzhou University, Zhengzhou, China.
Front Cardiovasc Med. 2023 Sep 7;10:1096020. doi: 10.3389/fcvm.2023.1096020. eCollection 2023.
Coronary artery stenosis occurs frequently in patients with carotid artery stenosis. We developed a clinical predictive model to investigate the clinical risk of asymptomatic obstructive coronary artery stenosis in patients with carotid artery stenosis ≥ 50%.
From January 2018 to January 2022, carotid stenosis patients hospitalized at the First Affiliated Hospital of Zhengzhou University's Department of Endovascular Surgery were subjected to a retrospective analysis of their clinical information and imaging results. Excluded criteria were patients with lacking data, symptomatic coronary stenosis, prior coronary artery bypass grafting, and coronary stent implantation. Patients were separated into case and control groups according to whether or not they had obstructive coronary stenosis. Independent predictors were screened using univariate and multivariate logistic regression, and their accuracy was confirmed using least absolute shrinkage and selection operator (LASSO) regression. A Nomogram prediction model was developed using the aforementioned filtered factors. The model's discrimination and specificity were evaluated using the receiver operating characteristic curve (ROC) and Hosmer-Lemeshow goodness-of-fit test. Internal validation employed the Bootstrap procedure. The clinical decision curve analysis (DCA) of the prediction model was developed to assess the clinical applicability of the model.
The investigation included a total of 227 patients, of whom 132 (58.1%) had coronary artery stenosis. Hypertension, Grade I plaque, HbA1c ≥ 7.0%, MPV ≥ 9.2fl, and Fib ≥ 3.0 g/L were independent predictors, with OR values of (2.506, 0.219, 0.457, 1.876, 2.005), according to multivariate logistic regression. Risk factor screening and validation using lasso regression. The predictors chosen based on the optimal value are consistent with the predictors identified by multiple regression. The area under the ROC curve (AUC) of the model based on the above predictors was 0.701 (0.633-0.770), indicating that the model discriminated well. The calibration curve of the model closely matched the actual curve, and > 0.05 in the Hosmer-Lemeshow goodness-of-fit test indicated the model's accuracy. The results of the DCA curve demonstrate the clinical applicability of the prediction model.
Hypertension, grade I plaque, HbA1c ≥ 7.0%, MPV ≥ 9.2 fl, and Fib ≥ 3.0 g/L are predictors of asymptomatic coronary stenosis in patients with carotid stenosis ≥50%. The diagnostic model is clinically applicable and useful for identifying patients at high risk.
冠状动脉狭窄在颈动脉狭窄患者中频繁发生。我们开发了一种临床预测模型，以研究颈动脉狭窄≥50%的患者无症状性阻塞性冠状动脉狭窄的临床风险。
对2018年1月至2022年1月在郑州大学第一附属医院血管外科住院的颈动脉狭窄患者的临床信息和影像学结果进行回顾性分析。排除标准为数据缺失、有症状的冠状动脉狭窄、既往冠状动脉搭桥术和冠状动脉支架植入术患者。根据是否存在阻塞性冠状动脉狭窄将患者分为病例组和对照组。使用单因素和多因素逻辑回归筛选独立预测因素，并使用最小绝对收缩和选择算子（LASSO）回归确认其准确性。使用上述筛选出的因素建立列线图预测模型。使用受试者工作特征曲线（ROC）和Hosmer-Lemeshow拟合优度检验评估模型的辨别力和特异性。内部验证采用Bootstrap程序。建立预测模型的临床决策曲线分析（DCA）以评估模型的临床适用性。
该研究共纳入227例患者，其中132例（58.1%）有冠状动脉狭窄。多因素逻辑回归显示，高血压、I级斑块、糖化血红蛋白≥7.0%、平均血小板体积≥9.2fl和纤维蛋白原≥3.0g/L是独立预测因素，OR值分别为（2.506、0.219、0.457、1.876、2.005）。使用LASSO回归进行危险因素筛选和验证。基于最佳值选择的预测因素与多元回归确定的预测因素一致。基于上述预测因素的模型的ROC曲线下面积（AUC）为0.701（0.633-0.770），表明该模型辨别力良好。模型的校准曲线与实际曲线密切匹配，Hosmer-Lemeshow拟合优度检验中P>0.
