Abstract: Objective  To analyze the influencing factors of pre-hospital delay in acute ischemic stroke (AIS) and construct a risk prediction model. 
Methods  AIS patients admitted to The First People’s Hospital of Lianyungang from March 2021 to January 2024 were selected as research subjects. They were divided into the model group and the internal validation group. Based on pre-hospital delay conditions, the model group was further divided into the non-delayed (<3.5 h) group and the delayed (≥3.5 h) group. The data from the model group was used to establish a nomogram model, and the data from the internal validation group was used to assess the model’s generalization performance. Multivariate logistic regression was used to analyze the influencing factors of pre-hospital delay in AIS patients. The rms package in R software was used to construct a nomogram prediction model for the pre-hospital delay of AIS patients. The ROC curve, calibration curve, and Hosmer-Lemeshow test were used to evaluate the predictive efficacy of the model, and the clinical decision curve was used to evaluate its clinical application value. 
Results  A total of 268 patients with AIS were enrolled, among whom 141 (52.61%) had pre-hospital delay. In patients of the model group, the delayed group was older on average (P<0.001) and had significantly higher proportions of the following characteristics: living in rural areas (P=0.004), nocturnal onset (P=0.018), without disturbance of consciousness (P=0.004), and not receiving disease knowledge education (P=0.001) compared with the non-delayed group. Multivariate logistic regression analysis showed that advanced age (OR 1.082, 95%CI 1.038-1.128, P<0.001), living in rural areas (OR 3.201, 95%CI 1.402-7.307, P=0.006), nocturnal onset (OR 6.873, 95%CI 2.809-16.815, P<0.001), without disturbance of consciousness (OR 4.599, 95%CI 1.934-10.940, P=0.001), and not receiving disease knowledge education (OR 4.134, 95%CI 1.927-8.866, P<0.001) were independent risk factors for pre-hospital delay in AIS patients. Based on these factors, a nomogram model was developed, with higher total scores indicating an increased risk of pre-hospital delay. The ROC curve showed that the AUC of the model’s prediction of pre-hospital delay was 0.822 (95%CI 0.763-0.880) in the model group and 0.844 (95%CI 0.755-0.932) in the internal validation group. In addition, both groups of data passed the Hosmer-Lemeshow test. The calibration curve showed that the predicted value of the model was relatively consistent with the true value. The clinical decision curve showed that the clinical application value of the model was reasonable.
Conclusions  The risk of pre-hospital delay of AIS patients is influenced by age, residence, time of onset, presence or absence of disturbance of consciousness, and whether they received disease knowledge education. The model constructed on this basis has good predictive efficacy.

Key words: Acute ischemic stroke; Pre-hospital delay; Influencing factor; Prediction model

摘要： 目的 分析急性缺血性卒中（acute ischemic stroke，AIS）院前延迟的影响因素并构建风险预测模型。
方法 选取2021年3月—2024年1月于连云港市第一人民医院就诊的AIS患者为研究对象，将其分为模型组和内部验证组，并根据院前延迟情况将模型组分为未延迟（＜3.5 h）组和延迟（≥3.5 h）组。模型组的数据用以建立列线图模型，内部验证组的数据用以评估模型的泛化性能。采用多因素logistic回归分析AIS患者院前延迟的影响因素，采用R软件中rms程序包构建AIS患者院前延迟的列线图预测模型，并采用ROC曲线、校准曲线及Hosmer-Lemeshow检验评估模型预测效能，采用临床决策曲线评估其临床应用价值。
结果 共纳入268例AIS患者，其中141例（52.61%）患者发生院前延迟。在模型组患者中，与未延迟组相比，延迟组平均年龄较高（P＜0.001），在农村居住（P=0.004）、夜间发病（P=0.018）、无意识障碍（P=0.004）及未接受疾病知识宣教（P=0.001）的占比显著较高。多因素logistic回归分析显示，高龄（OR 1.082，95%CI 1.038～1.128，P＜0.001）、在农村居住（OR 3.201，95%CI 1.402～7.307，P=0.006）、夜间发病（OR 6.873，95%CI 2.809～16.815，P＜0.001）、无意识障碍（OR 4.599，95%CI 1.934～10.940，P=0.001）以及未接受疾病知识宣教（OR 4.134，95%CI 1.927～8.866，P＜0.001）为AIS患者院前延迟的独立危险因素。基于以上因素构建列线图模型，模型总得分越高，院前延迟风险越高。ROC曲线显示，模型预测院前延迟在模型组中的AUC为0.822（95%CI 0.763～0.880），在内部验证组中的AUC为0.844（95%CI 0.755～0.932）。此外，两组数据均通过Hosmer-Lemeshow检验。校准曲线显示，模型预测值与真实值较为一致。临床决策曲线显示，模型临床应用价值尚可。
结论 AIS患者院前延迟风险受年龄、居住地、发病时间、有无意识障碍及是否接受疾病知识宣教影响，在此基础上构建的模型预测效能良好。

关键词: 急性缺血性卒中; 院前延迟; 影响因素; 预测模型