Abstract: Objective  To analyze the correlation between frailty index, nutritional risk, and the risk of post-stroke nosocomial infection in elderly patients with acute ischemic stroke (AIS), and to construct a risk prediction model. 
Methods  This prospective observational study enrolled elderly patients (≥60 years) with AIS who were admitted to the Department of Neurology of Wuhan No.1 Hospital and received intravenous thrombolysis between January 2022 and December 2024. Patients were divided into an infection group and a non-infection group based on the occurrence of nosocomial infection after intravenous thrombolysis. Differences in general clinical data, characteristics on admission, treatment-related features, pre-treatment frailty status, nutritional risk, and inflammatory factor levels were compared between the two groups. Multivariate logistic regression analysis was used to identify independent risk factors associated with infection in elderly AIS patients and to construct a prediction model. The predictive performance of the model for infection was evaluated using the ROC curve analysis.
Results  A total of 201 elderly AIS patients were enrolled, with an age range of 60-76 years and a mean age of (66.3±3.3) years. The infection group comprised 68 patients (infection rate 33.8%), and the non-infection group comprised 133 patients. Compared to the non-infection group, the infection group had significantly higher values for the following parameters (P＜0.05): onset-to-needle time, age, proportion of patients with diabetes mellitus history, NIHSS score on admission, proportion with conscious disturbance, pre-treatment frailty index, proportion of frail patients, nutritional risk screening 2002 (NRS 2002) score, proportion of nutritional risk, and levels of C-reactive protein, procalcitonin, and IL-6. Multivariate logistic regression analysis identified the following as independent risk factors for nosocomial infection in elderly AIS patients: high frailty index (OR 5.812, 95%CI 1.959-35.207, P=0.015), high NRS 2002 score (OR 1.370, 95%CI 1.106-1.696, P=0.004), high procalcitonin level (OR 25.358, 95%CI 3.074-209.209, P=0.003), and high IL-6 level (OR 1.052, 95%CI 1.009-1.098, P=0.017). ROC curve analysis indicated that the prediction model, constructed based on these four factors, had an AUC of 0.834 (95%CI 0.771-0.896), with a sensitivity of 79.03% and a specificity of 81.00%, indicating good predictive performance.
Conclusions  A significant correlation exists between pre-thrombolysis frailty status, nutritional risk, and the risk of post-stroke nosocomial infection in elderly patients with AIS. The prediction model based on frailty index, nutritional risk score, and inflammatory factors shows high predictive efficacy for nosocomial infection risk in these patients.

Key words: Elderly; Acute ischemic stroke; Frailty; Nutritional risk; Nosocomial infection

摘要： 目的 分析老年急性缺血性卒中患者衰弱指数、营养风险与卒中后院内感染风险的相关性并构建风险预测模型。
方法 本研究为前瞻性观察性研究，纳入2022年1月—2024年12月于武汉市第一医院神经内科就诊并接受静脉溶栓治疗的老年（≥60岁）急性缺血性卒中患者。根据静脉溶栓治疗后是否发生院内感染，将患者分为感染组和无感染组。对比两组患者的一般临床资料、入院时临床特征、治疗相关指标、治疗前衰弱情况、营养风险及炎症因子水平的差异。采用多因素logistic回归分析筛选与老年急性缺血性卒中患者感染相关的独立危险因素，并构建预测模型。采用ROC曲线分析模型对老年急性缺血性卒中患者感染的预测效能。
结果 共入组201例老年急性缺血性卒中患者，年龄范围60～76岁，平均（66.3±3.3）岁。其中感染组68例，感染率为33.8%，无感染组133例。与无感染组相比，感染组的发病至溶栓时间更长，年龄、糖尿病病史比例、入院时NIHSS评分、出现意识障碍比例、治疗前衰弱指数、衰弱比例、营养风险筛查2002（nutritional risk screening 2002，NRS 2002）评分、有营养风险比例，以及C反应蛋白、降钙素原和IL-6水平更高，差异均有统计学意义。多因素logistic回归分析显示，衰弱指数高（OR 5.812，95%CI 1.959～35.207，P=0.015）、NRS 2002评分高（OR 1.370，95%CI 1.106～1.696，P=0.004）、降钙素原水平高（OR 25.358，95%CI 3.074～209.209，P=0.003）和IL-6水平高（OR 1.052，95%CI 1.009～1.098，P=0.017）是老年急性缺血性卒中患者院内感染的独立危险因素。ROC曲线分析显示，基于上述4种影响因素构建的老年急性缺血性卒中患者院内感染风险预测模型的AUC为0.834（95%CI 0.771～0.896），敏感度为79.03%，特异度为81.00%，具有较高的预测效能。
结论 老年急性缺血性卒中患者静脉溶栓前衰弱情况、营养风险与卒中后院内感染风险之间存在显著的相关性。基于衰弱指数、营养风险评分及炎症因子构建的预测模型对老年急性缺血性卒中患者院内感染风险的预测效能较高。

关键词: 老年; 急性缺血性卒中; 衰弱; 营养风险; 院内感染