经导管主动脉瓣置换术后谵妄发生的危险因素及其预测模型构建

Delirium after transcatheter aortic valve replacement:risk factors and development of a predictive model

目的探讨经导管主动脉瓣置换术(transcatheter aortic valve replacement,TAVR)后患者谵妄发生危险因素,构建风险预测模型并分析其预测效能。方法选择本院2021年1月至2022年11月159例拟行TAVR治疗的患者为研究对象,收集比较谵妄组和未发生谵妄组患者临床资料和疾病特征,分析TAVR术后谵妄发生的危险因素,并利用Logistic回归方程构建预测模型,通过绘制ROC曲线及Nomogram图分析该预测模型预测效能。结果159例TAVR术后患者谵妄发生率36.48%;年龄≥65岁(OR=1.241)、糖尿病史(OR=1.455)、手术时间长(OR=1.182)、术后机械通气时间长(OR=1.189)是TAVR术后谵妄发生的危险因素(P<0.05),高氧合指数PaO 2/FiO 2(OR=0.731)是TAVR术后谵妄发生的保护因素(均P<0.05),风险Nomogram图分析显示,基于上述危险因素构建的预测模型预测TAVR术后谵妄发生的C-index为0.868(95%CI:0.737~0.926),区分度良好;列线图模型预测可能性绝对误差为0.014,一致性良好;ROC曲线分析显示曲线下面积为为0.868,特异度为83.17%,灵敏度为77.59%。结论高龄、糖尿病史、手术时间、术后机械通气时间及PaO 2/FiO 2水平是TAVR术后谵妄发生的影响因素,构建模型预测TAVR术后谵妄发生的效果较好。

Objective To explore the risk factors of delirium after transcatheter aortic valve replacement(TAVR)so as to establish a prediction model based on the identified factors and further analyse the efficiency of the prediction model.Methods From January 2021 to November 2022,a total of 159 patients who were scheduled to receive TAVR surgery in our hospital were included as the research subjects.The patients were observed and discharged from hospital after TAVR surgery and the postoperative delirium was recorded.According to postoperative delirium,the patients were divided into delirium group and control group(without postoperative delirium).Clinical data of patients in delirium group were compared with those in control group.Risk factors of postoperative delirium were then analysed.Logistic regression equation was used to establish a model for predicting the risk factors.The efficiency of the prediction model was analysed by plotting ROC and Nomogram curve.Results Among 159 patients,58 developed with delirium after TAVR during the observation period,accounting for 36.48%.The risk factors after TAVR were:age≥65 years(OR=1.241),diabetes(OR=1.455),long operation time(OR=1.182)and long postoperative mechanical ventilation time(OR=1.189)(P<0.05).A high level PaO 2/FiO 2(OR=0.731)was identified as a protective factor(P<0.05).The risk Nomogram analysis revealed that the C-index of the prediction model based on above risk factors for delirium after TAVR was 0.868(95% CI:0.737~0.926),with a good discrimination.The absolute error of Nomogram model was 0.014 with a good consistency.ROC curve analysis showed an area under the curve at 0.868,with the specificity at 83.17% and the sensitivity at 77.59%.Conclusions Elder age,diabetes,operation time,time of postoperative mechanical ventilation and PaO_(2)/FiO_(2) level are the influencing factors of delirium after TAVR.The developed prediction model based on above factors is effective in predicting the occurrence of delirium after TAVR.