【摘要】 目的 探讨急性脑梗死（ACI）患者早期应用替罗非班及静脉溶栓（IVT）治疗后预后不良的影响因素，并构建决策树模型。方法 选取2022年6月至2024年10月三六三医院收治的ACI患者，依据治疗后患者的预后情况分为预后不良组和预后良好组。收集所有受试者的临床资料和实验室指标等，通过单因素分析和多因素logistic回归分析筛选出预后不良的影响因素，应用SPSS Modeler软件构建预后不良的决策树模型，并分析决策树模型的预测效能。结果 最终纳入420例患者，其中预后不良92例，占21.90%。经1∶1的倾向性评分匹配后，预后不良组和预后良好组各92例患者。多因素logistic回归分析显示，发病至溶栓时间、责任大动脉病变、美国国立卫生院卒中量表（NIHSS）评分、D-二聚体（D-D）和空腹血糖（FBG）为ACI患者治疗后预后不良的影响因素（P<0.05）；构建的决策树模型共筛选出5个解释变量，其中发病至溶栓时间是最重要的预测因子，决策树模型的AUC为0.860(95%CI:0.801～0.907)，优于logistic回归模型的AUC[0.795(95%CI:0.730～0.851)]。结论 影响ACI患者早期应用替罗非班及IVT治疗后预后不良因素为发病至溶栓时间、责任大动脉病变、NIHSS评分、D-D和FBG，构建的决策树模型具有较好的预测效能。更多还原

【Abstract】 Objective To investigate the influencing factors of poor prognosis of early application of tirofiban and intravenous thrombolysis(IVT) treatment in patients with acute cerebral infarction(ACI) and construct a prediction model.Methods ACI patients admitted to 363 Hospital from June 2022 to October 2024 were selected as the study subjects. Based on post-treatment prognosis, the patients were divided into a poor prognosis group and a good prognosis group. Clinical data and laboratory indicators of all participants were collected. Risk factors for poor prognosis were screened through univariate and multivariate analyses, and a decision tree model for poor prognosis was constructed using SPSS Modeler software. The predictive performance of the decision tree model was analyzed. Results A total of 420 patients were finally included in the results, among which 92 cases had a poor prognosis, accounting for 21. 90%. After 1∶1 propensity score matching, there were 92 patients in each of the poor prognosis group and the good prognosis group. Multivariate logistic regression analysis showed that the time from onset to thrombolysis, responsible large artery lesion, NIHSS score, D-D, and FBG were risk factors for poor prognosis in ACI patients after treatment(P<0. 05). The constructed decision tree model identified five explanatory variables, with the time from onset to thrombolysis being the most important predictor. The AUC of the decision tree model was 0. 860(95%CI: 0. 801 to 0. 907), which was superior to that of the logistic regression model(0. 795, 95%CI: 0. 730 to 0. 851). Conclusion The factors influencing poor prognosis in ACI patients after early treatment with tirofiban and IVT were the time from onset to thrombolysis, responsible large artery lesion, NIHSS score, D-D, and FBG. The decision tree model constructed in this study demonstrated good predictive performance.更多还原