Department of Cardiac Surgery, the First Affiliated Hospital of Anhui Medical University, Hefei, Anhui Province, China.
Scand Cardiovasc J. 2024 Dec;58(1):2373084. doi: 10.1080/14017431.2024.2373084. Epub 2024 Jul 4.
Despite advancements in surgical techniques, operations for infective endocarditis (IE) remain associated with relatively high mortality. The aim of this study was to develop a nomogram model to predict the early postoperative mortality in patients undergoing cardiac surgery for infective endocarditis based on the preoperative clinical features.
We retrospectively analyzed the clinical data of 357 patients with IE who underwent surgeries at our center between January 2007 and June 2023. Independent risk factors for early postoperative mortality were identified using univariate and multivariate logistic regression models. Based on these factors, a predictive model was developed and presented in a nomogram. The performance of the nomogram was evaluated through the receiver operating characteristic (ROC) curve, calibration plot, and decision curve analysis (DCA). Internal validation was performed utilizing the bootstrapping method.
The nomogram included nine predictors: age, stroke, pulmonary embolism, albumin level, cardiac function class IV, antibotic use <4weeks, vegetation size ≥1.5 cm, perivalvular abscess and preoperative dialysis. The area under the ROC curve (AUC) of the model was 0.88 (95%CI:0.80-0.96). The calibration plot indicated strong prediction consistency of the nomogram with satisfactory Hosmer-Lemeshow test results (χ2 = 13.490, = 0.142). Decision curve analysis indicated that the nomogram model provided greater clinical net benefits compared to "operate-all" or "operate-none" strategies.
The innovative nomogram model offers cardiovascular surgeons a tool to predict the risk of early postoperative mortality in patients undergoing IE operations. This model can serve as a valuable reference for preoperative decision-making and can enhance the clinical outcomes of IE patients.
尽管外科技术不断进步，但感染性心内膜炎（IE）的手术治疗仍然与相对较高的死亡率相关。本研究旨在基于术前临床特征，建立预测 IE 患者心脏手术后早期术后死亡率的列线图模型。
我们回顾性分析了 2007 年 1 月至 2023 年 6 月期间在我们中心接受手术治疗的 357 例 IE 患者的临床资料。使用单因素和多因素逻辑回归模型确定早期术后死亡率的独立危险因素。基于这些因素，建立了预测模型，并以列线图形式呈现。通过接受者操作特征（ROC）曲线、校准图和决策曲线分析（DCA）评估列线图的性能。内部验证采用 bootstrap 方法进行。
该列线图包括 9 个预测因素：年龄、中风、肺栓塞、白蛋白水平、心功能 IV 级、抗生素使用<4 周、赘生物大小≥1.5cm、瓣周脓肿和术前透析。模型的 ROC 曲线下面积（AUC）为 0.88（95%CI：0.80-0.96）。校准图表明列线图具有很强的预测一致性，Hosmer-Lemeshow 检验结果良好（χ2=13.490，P=0.142）。决策曲线分析表明，与“全部手术”或“全部不手术”策略相比，列线图模型提供了更大的临床净收益。
创新的列线图模型为心血管外科医生提供了一种预测 IE 手术患者早期术后死亡率风险的工具。该模型可作为术前决策的有价值参考，并可改善 IE 患者的临床结局。
