AIP、TyG指数和HALP评分对脑卒中患者非计划再入院的联合预测价值研究

The Combined Predictive Value of AIP, TyG Index and HALP Score for Unplanned Readmission in Stroke Patients

摘要: 目的：探讨动脉粥样硬化指数、甘油三酯葡萄糖乘积指数和血红蛋白、白蛋白、淋巴细胞和血小板综合评分对脑卒中患者非计划再入院的独立及联合预测价值。方法：回顾性选取1050例脑卒中出院患者，根据30天内是否发生非计划再入院分为病例组(n = 280)和对照组(n = 770)。首先进行单因素分析筛选相关因素，再采用多因素Logistic回归分析明确非计划再入院的独立危险因素，并通过受试者工作特征(ROC)曲线评估各指标及联合模型的预测效能。结果：AIP高、TyG指数高和HALP评分低是脑卒中患者非计划再入院的独立预测因素。ROC曲线分析显示，AIP、TyG指数和HALP评分的曲线下面积(AUC)分别为0.743、0.608和0.593，而三者联合预测模型的AUC提升至0.814，预测价值显著优于任一单一指标。结论：AIP、TyG指数和HALP评分是脑卒中患者非计划再入院的有效预测指标。由三者构建的联合预测模型具有更高的预测效能，有助于临床早期识别再入院高风险患者并实施针对性干预。

Abstract: Objective: To investigate the independent and combined predictive value of the Atherogenic Index (AIP), the Triglyceride-Glucose (TyG) Index, and the Hemoglobin, Albumin, Lymphocyte, and Platelet (HALP) Score for unplanned readmission in stroke patients. Methods: This retrospective study included 1050 stroke patients discharged from the hospital. Based on the occurrence of unplanned readmission within 30 days, patients were categorized into a case group (n = 280) and a control group (n = 770). Univariate analysis was first conducted to screen for associated factors. Subsequently, multivariate logistic regression analysis was performed to identify independent risk factors for unplanned readmission. The predictive performance of individual indicators and a combined model was assessed using Receiver Operating Characteristic (ROC) curve analysis. Results: High AIP, high TyG index, and low HALP score were identified as independent predictive factors for unplanned readmission in stroke patients. ROC curve analysis revealed that the areas under the curve (AUC) for AIP, TyG index, and HALP score were 0.743, 0.608, and 0.593, respectively. The combined prediction model incorporating all three indicators achieved an AUC of 0.814, demonstrating significantly superior predictive value compared to any single indicator alone. Conclusion: AIP, TyG index, and HALP score are effective predictors of unplanned readmission in stroke patients. The combined prediction model constructed from these three indicators exhibits higher predictive efficacy and can aid in the early clinical identification of high-risk patients for readmission, enabling targeted interventions.