Rao Yanjun, Wei Jihong, Liu Shuang, Liao Bo
Department of Rehabilitation Medicine, Mianyang Hospital Affiliated to Medical College of University of Electronic Science and Technology (Mianyang Central Hospital), Mianyang 621000, Sichuan, China. Corresponding author: Wei Jihong, Email:
Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2023 Apr;35(4):371-375. doi: 10.3760/cma.j.cn121430-20220525-00512.
To establish a predictive model for severe swallowing disorder after acute ischemic stroke based on nomogram model, and evaluate its effectiveness.
A prospective study was conducted. The patients with acute ischemic stroke admitted to Mianyang Central Hospital from October 2018 to October 2021 were enrolled. Patients were divided into severe swallowing disorder group and non-severe swallowing disorder group according to whether severe swallowing disorder occurred within 72 hours after admission. The differences in general information, personal history, past medical history, and clinical characteristics of patients between the two groups were compared. The risk factors of severe swallowing disorder were analyzed by multivariate Logistic regression analysis, and the relevant nomogram model was established. The bootstrap method was used to perform self-sampling internal validation on the model, and consistency index, calibration curve, receiver operator characteristic curve (ROC curve), and decision curve were used to evaluate the predictive performance of the model.
A total of 264 patients with acute ischemic stroke were enrolled, and the incidence of severe swallowing disorder within 72 hours after admission was 19.3% (51/264). Compared with the non-severe swallowing disorder group, the severe swallowing disorder group had a higher proportion of patients aged of ≥ 60 years old, with severe neurological deficits [National Institutes of Health stroke scale (NIHSS) score ≥ 7], severe functional impairments [Barthel index, an activity of daily living functional status assessment index, < 40], brainstem infarction and lesions ≥ 40 mm (78.43% vs. 56.81%, 52.94% vs. 28.64%, 39.22% vs. 12.21%, 31.37% vs. 13.62%, 54.90% vs. 24.41%), and the differences were statistically significant (all P < 0.01). Multivariate Logistic regression analysis showed that age ≥ 60 years old [odds ratio (OR) = 3.542, 95% confidence interval (95%CI) was 1.527-8.215], NIHSS score ≥ 7 (OR = 2.741, 95%CI was 1.337-5.619), Barthel index < 40 (OR = 4.517, 95%CI was 2.013-10.136), brain stem infarction (OR = 2.498, 95%CI was 1.078-5.790) and lesion ≥ 40 mm (OR = 2.283, 95%CI was 1.485-3.508) were independent risk factors for severe swallowing disorder after acute ischemic stroke (all P < 0.05). The results of model validation showed that the consistency index was 0.805, and the trend of the calibration curve was basically consistent with the ideal curve, indicating that the model had good prediction accuracy. ROC curve analysis showed that the area under the ROC curve (AUC) predicted by nomogram model for severe swallowing disorder after acute ischemic stroke was 0.817 (95%CI was 0.788-0.852), indicating that the model had good discrimination. The decision curve showed that within the range of 5% to 90%, the nomogram model had a higher net benefit value for predicting the risk of severe swallowing disorder after acute ischemic stroke, indicating that the model had good clinical predictive performance.
The independent risk factors of severe swallowing disorder after acute ischemic stroke include age ≥ 60 years old, NIHSS score ≥ 7, Barthel index < 40, brainstem infarction and lesion size ≥ 40 mm. The nomogram model established based on these factors can effectively predict the occurrence of severe swallowing disorder after acute ischemic stroke.
基于列线图模型建立急性缺血性卒中后严重吞咽障碍的预测模型，并评估其有效性。
进行一项前瞻性研究。纳入2018年10月至2021年10月在绵阳市中心医院住院的急性缺血性卒中患者。根据入院后72小时内是否发生严重吞咽障碍，将患者分为严重吞咽障碍组和非严重吞咽障碍组。比较两组患者的一般资料、个人史、既往病史及临床特征的差异。采用多因素Logistic回归分析急性缺血性卒中后严重吞咽障碍的危险因素，并建立相关列线图模型。采用自抽样内部验证法对模型进行验证，应用一致性指数、校准曲线、受试者工作特征曲线（ROC曲线）及决策曲线评估模型的预测性能。
共纳入264例急性缺血性卒中患者，入院后72小时内严重吞咽障碍的发生率为19.3%（51/264）。与非严重吞咽障碍组相比，严重吞咽障碍组年龄≥60岁、伴有严重神经功能缺损[美国国立卫生研究院卒中量表（NIHSS）评分≥7分]、严重功能障碍[Barthel指数，日常生活活动功能状态评估指标，<40分]、脑干梗死及病灶≥40 mm的患者比例更高（78.43% 对56.81%，52.94%对28.64%，39.22%对12.21%，31.37%对13.62%，54.90%对24.41%），差异均有统计学意义（均P<0.01）。多因素Logistic回归分析显示，年龄≥60岁[比值比（OR）=3.542，95%置信区间（95%CI）为1.527 - 8.215]、NIHSS评分≥7分（OR = 2.741，95%CI为1.337 - 5.619）、Barthel指数<40分（OR = 4.517，95%CI为2.013 - 10.136）、脑干梗死（OR = 2.498，95%CI为1.078 - 5.790）及病灶≥40 mm（OR = 2.283，95%CI为1.485 - 3.508）是急性缺血性卒中后严重吞咽障碍的独立危险因素（均P<0.05）。模型验证结果显示，一致性指数为0.805，校准曲线趋势与理想曲线基本一致，表明模型预测准确性良好。ROC曲线分析显示，列线图模型预测急性缺血性卒中后严重吞咽障碍的ROC曲线下面积（AUC）为0.817（95%CI为0.788 - 0.852），表明模型具有良好的区分度。决策曲线显示，在5%至90%的范围内，列线图模型预测急性缺血性卒中后严重吞咽障碍风险的净效益值更高，表明模型具有良好的临床预测性能。
急性缺血性卒中后严重吞咽障碍的独立危险因素包括年龄≥60岁、NIHSS评分≥7分、Barthel指数<40分、脑干梗死及病灶大小≥40 mm。基于这些因素建立的列线图模型可有效预测急性缺血性卒中后严重吞咽障碍的发生。
