Supplementary Material for: Predictive Value of Cardiac Biomarkers Combined with Clinical, Radiological Factors for Venous Thromboembolism in Patients with Spontaneous Intracerebral Hemorrhage

Introduction Venous thromboembolism (VTE) is an important complication after spontaneous intracerebral hemorrhage (ICH). However, it remains a clinical challenge to identify individuals at high risk for VTE in a population with ICH. This study aimed to develop a model integrating cardiac biomarkers with clinical-radiological factors for predicting VTE risk in patients with spontaneous ICH. Methods ICH patients were retrospectively enrolled between October 2019 and December 2022. Baseline clinical characteristics, laboratory data, and radiological features were collected. Patients with pulmonary embolism (PE) and deep vein thrombosis (DVT) were classified into the VTE group. Cox regression analysis was used to identify independent predictors of in-hospital VTE. A nomogram was developed based on the multivariate model, and its performance was evaluated using the concordance index (C-index), decision curve analysis (DCA), and net reclassification improvement (NRI). Results A total of 170 patients (mean age: 54.66 ± 13.6 years, 125 [73.5%] males) with ICH were included in the analysis. Thirty-six (21.2%) patients were assigned to the VTE group. Multivariate Cox analysis identified age (HR = 1.032, 95% CI: 1.002–1.062, p=0.033), baseline edema volume (HR = 1.034, 95% CI: 1.012–1.056, p=0.002), IVH (HR = 3.268, 95% CI: 1.635–6.530, p<0.001), Myo (HR = 1.002, 95% CI: 1.000–1.003, p=0.010), and BNP (HR = 1.003, 95% CI: 1.001–1.006, p=0.007) as independent predictors. The combined model showed better predictive performance than the clinical-radiological model alone (C-index: 0.791 vs 0.749). The nomogram demonstrated good calibration and clinical utility across a wide risk threshold range. Conclusion Myo and BNP provide incremental predictive value for VTE risk stratification in ICH patients beyond traditional factors. The developed nomogram offers a practical tool for individualized risk assessment, potentially guiding optimized VTE prophylaxis strategies.引言 静脉血栓栓塞症（Venous thromboembolism, VTE）是自发性脑出血（spontaneous intracerebral hemorrhage, ICH）后重要的并发症。然而，在ICH人群中识别VTE高危个体仍是临床难题。本研究旨在构建整合心脏生物标志物与临床-影像特征的模型，用于预测自发性ICH患者的VTE发生风险。 方法 本研究回顾性纳入2019年10月至2022年12月期间的ICH患者。收集患者的基线临床特征、实验室检查数据及影像学特征。将合并肺栓塞（pulmonary embolism, PE）与深静脉血栓形成（deep vein thrombosis, DVT）的患者归类为VTE组。采用Cox回归分析识别院内VTE的独立预测因子。基于多因素模型构建列线图（nomogram），并通过一致性指数（concordance index, C-index）、决策曲线分析（decision curve analysis, DCA）及净重新分类指数（net reclassification improvement, NRI）评估其预测性能。 结果 本研究共纳入170例ICH患者（平均年龄：54.66±13.6岁，男性125例[73.5%]）。其中36例（21.2%）被归入VTE组。多因素Cox回归分析显示，年龄（风险比HR=1.032，95%置信区间CI:1.002–1.062，p=0.033）、基线水肿体积（HR=1.034，95%CI:1.012–1.056，p=0.002）、脑室内出血（intraventricular hemorrhage, IVH）、肌红蛋白（Myo）以及B型钠尿肽（BNP）为独立预测因子。联合模型的预测性能优于单纯临床-影像模型（C-index：0.791 vs 0.749）。该列线图在广泛的风险阈值范围内展现出良好的校准度与临床实用性。 结论 肌红蛋白（Myo）与B型钠尿肽（BNP）可为ICH患者的VTE风险分层提供超越传统危险因素的增量预测价值。本研究构建的列线图可为个体化风险评估提供实用工具，有望指导优化VTE预防策略。