目的 分析急性脑出血患者合并吞咽障碍的危险因素，并通过决策树模型建立急性脑出血患者合并吞咽障碍的风险预警模型。方法 回顾性选取南阳市第二人民医院2022年10月至2023年10月收治的100例急性脑出血患者为研究对象，根据患者吞咽障碍发生情况分为吞咽障碍组和非吞咽障碍，采用logistic回归分析急性脑出血患者合并吞咽障碍的危险因素，运用IBM SPSS Modeler建立预测急性脑出血患者合并吞咽障碍的决策树模型。结果 吞咽障碍组41例，非吞咽障碍组59例，吞咽障碍发生率为41。00％。吞咽障碍组和非吞咽障碍组患者年龄、血肿体积、美国国立卫生研究院卒中量表(national institutes of health stroke scale，NIHSS)评分、脑干出血、气管插管时间等资料对比(P＜0。05)。Logistic回归分析结果显示，年龄≥60岁、血肿体积≥30 mL、NIHSS评分升高、脑干出血、气管插管时间≥48h均是急性脑出血患者合并吞咽障碍的独立危险因素(P＜0。05)。决策树模型显示，气管插管时间是急性脑出血患者合并吞咽障碍最为重要的影响因素，而NIHSS评分对急性脑出血患者合并吞咽障碍的影响较小。受试者工作特征(receiver operating characteristic，ROC)曲线结果显示，两种模型的预测效能均为中等，多因素的ROC曲线下面积(ares under the curve，AUC)值高于决策树(Z=1。853，P=0。064)。结论 年龄≥60岁、血肿体积≥30 mL、NIHSS评分升高、脑干出血、气管插管时间≥48 h均是急性脑出血患者合并吞咽障碍的影响因素，本研究构建的决策树模型对吞咽障碍发生风险具有较好的预测效能。
Building a predictive model for acute stroke patients with combined dysphagia based on decision tree
Objective To analyze the risk factors for swallowing disorders in patients with acute cerebral hemorrhage and establish a risk prediction model for swallowing disorders in patients with acute cerebral hemorrhage using a decision tree model.Methods A retrospective study was conducted from October 2022 to October 2023,including 100 patients with acute cerebral hemorrhage as the research subjects.Based on the occurrence of swallowing disorders,the patients were divided into a swallowing disorder group and a non-swallowing disorder group.Logistic regression analysis was used to analyze the risk factors for acute cerebral hemorrhage patients with swallowing disorders.IBM SPSS Modeler was utilized to establish a decision tree model for predicting the occurrence of swallowing disorders in acute cerebral hemorrhage pa-tients.Results There were 41 cases in the swallowing disorder group and 59 cases in the non-swallowing disorder group,with a prevalence of swallowing disorders of 41.00％.There were significant differences in age,hematoma vol-ume,NIHSS score,brainstem hemorrhage,and tracheal intubation time between the swallowing disorder group and the non-swallowing disorder group(P＜0.05).The logistic regression analysis showed that age ≥60 years,hematoma vol-ume ≥30 mL,elevated NIHSS score,brainstem hemorrhage,and tracheal intubation time ≥48 h were independent risk factors for swallowing disorders in patients with acute cerebral hemorrhage(P＜0.05).The decision tree model showed that tracheal intubation time was the most important factor influencing the occurrence of swallowing disorders in patients with acute cerebral hemorrhage,while the NIHSS score had a smaller impact.The receiver operating characteristic(ROC)results showed that both models had moderate predictive efficiency,and the ares under the characteristic(AUC)value of the multivariate model was higher than that of the decision tree model(Z=1.853,P=0.064).Conclusion Age ≥60 years,hematoma volume ≥30 mL,elevated NIHSS score,brainstem hemorrhage,and tracheal intubation time ≥48h are factors influencing the occurrence of swallowing disorders in patients with acute cerebral hemorrhage.The decision tree model constructed in this study has good predictive efficiency for the risk of swallowing disorders.
Acute cerebral hemorrhageSwallowing disorderInfluencing factorsDecision tree
NETL
NSTL
万方数据
