卒中后吞咽障碍患者卒中相关性肺炎预测模型的构建

Establishment of a predictive model for stroke⁃associated pneumonia in post⁃stroke patients with swallowing disorders

目的分析卒中后吞咽障碍患者卒中相关性肺炎(stroke-associated pneumonia,SAP)的影响因素,构建SAP风险预测模型并验证其预测效果。方法前瞻性连续纳入2022年11月至2023年5月,重庆市3所医院神经内科住院的卒中后吞咽障碍患者为对象,调查患者一般情况、临床资料、吞咽及口腔健康状况。应用弹性网络筛选预测因素,多因素Logistic回归建模并进行验证,绘制列线图展示模型结果。结果共纳入211例患者,46例(21.80%)发生卒中相关性肺炎。弹性网络共筛选出5个预测因素,即美国国立卫生院卒中量表(national institutes of health stroke scale,NIHSS评分)、洼田饮水试验、唾液分泌状况、口腔健康评估量表(oral health assessment tool,OHAT)评分、口腔卫生状况。似然比检验结果提示,模型在统计上显著(χ^(2)=104.17,P<0.01),判别指数为0.960。模型受试者操作特征曲线下面积(AUC)为0.93(95%CI:0.87~0.93),最佳风险阈值为0.10,准确度为83.87%,灵敏度为84.62%,特异度为83.67%。结论构建的模型能较好地预测卒中后吞咽障碍患者SAP风险,对尽早识别高风险患者、制定个体化预防策略具有临床意义。

Objective Investigate the influencing factors of stroke⁃associated pneumonia(SAP)in patients with post⁃stroke dysphagia,and to develop a risk prediction model for SAP,as well as verify its predictive effectiveness.Methods Prospective continuous inclusion of patients with post⁃stroke dysphagia admitted to the neurology departments of three hospitals in Chongqing from November 2022 to May 2023.Investigated the general characteristics,clinical data,swallowing function,and oral health status of the patients.An elastic net method was applied to select predictive factors,followed by constructing a multiple⁃factor logistic regression model and validating it.The results were presented using nomogram to display the outcomes of the model.Results A total of 211 patients were included,56(21.80%)of whom developed SAP.The SAP prediction model included 5 items,including national institutes of health stroke scale(NIHSS)score,Water swallow test,salivation status,OHAT score and oral hygiene status.The likelihood ratio test results showed that the model was statistically significant(χ^(2)=132.62,P<0.001)and the discrimination index was 0.960.The area under the ROC curve of this model was 0.93(95%CI:0.87⁃0.93),with the optimal critical value of 0.10,accuracy of 83.87%,sensitivity of 84.62%and the specificity of 83.67%.Conclusion The constructed model can effectively predict the risk of stroke⁃associated pneumonia(SAP)in patients with post⁃stroke dysphagia.This has clinical significance in identifying high⁃risk patients early and developing personalized preventive strategies.