Tan Juntao, He Yuxin, Li Zhanbiao, Xu Xiaomei, Zhang Qinghua, Xu Qian, Zhang Lingqin, Xiang Shoushu, Tang Xuewen, Zhao Wenlong
Operation Management Office, Affiliated Banan Hospital of Chongqing Medical University, Chongqing, China.
Department of Medical Administration, Affiliated Banan Hospital of Chongqing Medical University, Chongqing, China.
Front Cardiovasc Med. 2022 Apr 7;9:875702. doi: 10.3389/fcvm.2022.875702. eCollection 2022.
Heart failure (HF) is an end-stage manifestation of and cause of death in coronary heart disease (CHD). The objective of this study was to establish and validate a non-invasive diagnostic nomogram to identify HF in patients with CHD.
We retrospectively analyzed the clinical data of 44,772 CHD patients from five tertiary hospitals. Univariate logistic regression analyses and least absolute shrinkage and selection operator (LASSO) regression analyses were used to identify independent factors. A nomogram based on the multivariate logistic regression model was constructed using these independent factors. The concordance index (C-index), receiver operating characteristic (ROC) curves, calibration curves, decision curve analysis (DCA), and clinical impact curves (CIC) were used to evaluate the predictive accuracy and clinical value of this nomogram.
The predictive factors in the multivariate model included hypertension, age, and the total bilirubin, uric acid, urea nitrogen, triglyceride, and total cholesterol levels. The area under the curve (AUC) values of the nomogram in the training set, internal validation set, external validation set1, and external validation set2 were 0.720 (95% CI: 0.712-0.727), 0.723 (95% CI: 0.712-0.735), 0.692 (95% CI: 0.674-0.710), and 0.655 (95% CI: 0.634-0.677), respectively. The calibration curves indicated that the nomogram had strong calibration. DCA and CIC indicated that the nomogram can be used as an effective tool in clinical practice.
The developed predictive model combines the clinical and laboratory factors of patients with CHD and is useful in individualized prediction of HF probability for clinical decision-making during treatment and management.
心力衰竭（HF）是冠心病（CHD）的终末期表现和死亡原因。本研究的目的是建立并验证一种非侵入性诊断列线图，以识别冠心病患者中的心力衰竭。
我们回顾性分析了来自五家三级医院的44772例冠心病患者的临床资料。采用单因素逻辑回归分析和最小绝对收缩和选择算子（LASSO）回归分析来识别独立因素。使用这些独立因素构建基于多因素逻辑回归模型的列线图。采用一致性指数（C指数）、受试者工作特征（ROC）曲线、校准曲线、决策曲线分析（DCA）和临床影响曲线（CIC）来评估该列线图的预测准确性和临床价值。
多因素模型中的预测因素包括高血压、年龄以及总胆红素、尿酸、尿素氮、甘油三酯和总胆固醇水平。列线图在训练集、内部验证集、外部验证集1和外部验证集2中的曲线下面积（AUC）值分别为0.720（95%CI：0.712 - 0.727）、0.723（95%CI：0.712 - 0.735）、0.692（95%CI：0.674 - 0.710）和0.655（95%CI：0.634 - 0.677）。校准曲线表明列线图具有良好的校准性。DCA和CIC表明该列线图可作为临床实践中的有效工具。
所建立的预测模型结合了冠心病患者的临床和实验室因素，有助于在治疗和管理过程中对心力衰竭概率进行个体化预测，以辅助临床决策。
