摘要
目的探讨老年心源性呼吸衰竭患者机械通气后早期拔管失败的危险因素及建立预测模型。方法纳入本院心源性呼吸衰竭接受有创机械通气后于24 h内拔除气管插管的年龄≥60岁患者311例,将30 d内再次插管的41例为失败组,成功撤机的270例为撤机组。用多因素logistic回归分析30 d内再次气管插管的危险因素,构建预测模型,用ROC曲线和校准曲线及决策曲线分析验证模型的准确性和获益率。结果失败组院内感染发生率和30 d内病死率较撤机组显著升高(23.4%vs 10.9%,P=0.016;23.4%vs 8.3%,P=0.001)。多因素logistic回归分析示,年龄、N末端B型钠尿肽前体、慢性阻塞性肺疾病3~4级和重度肾功能不全是再次插管的独立危险因素(P<0.05),以此构建预测模型后,ROC曲线分析模型的曲线下面积为0.854,敏感性73.3%,特异性83.0%。校准曲线显示,模型具有良好的校准能力。决策曲线分析显示,阈概率3%~76%内具有临床获益。结论心源性呼吸衰竭机械通气后早期拔管失败的危险因素较多,以此构建的风险预测模型效能较好,可为临床提供指导。
Objective To study the risk factors for early extubation failure(EF)in elderly cardiogenic respiratory failure(CRF)patients after mechanical ventilation,and establishment of its prediction model.Methods Three hundred and eleven CRF patients aged≥60 years with their tracheal intubation extubated within 24 h after mechanical ventilation were divided into EF group(n=41)and extubation success group(n=270).The prediction model of early EF was established by multivariate logistic regression analysis of the risk factors for reintubation within 30 days.The accuracy and efficacy of the model were tested by ROC curve analysis,adjusted curve analysis and DCA respectively.Results The in-hospital infection rate and death rate within 30 days were significantly higher in EF group than in extubation success group(23.4%vs 10.9%,P=0.016;23.4%vs 8.3%,P=0.001).Multivariate logistic regression analysis showed that age,NT-proBNP,grade 3-4 COPD and severe renal insufficiency were independent risk factors for reintubation(P<0.05).The AUC of early EF prediction model for predicting the risk of EF was 0.854 with a sensitivity of 73.3%and a specificity of 83.0%.Adjusted curve analysis showed that the adjusting performance of the model was good.DCA displayed that the model was effective in predicting the risk factors for early EF when its threshold probability was 3%-76%.Conclusion Age≥70 years,NT-proBNP<50%,grade 3-4 COPD and severe renal insufficiency are the risk factors for early EF in CRF patients after mechanical ventilation.The performance of early EF prediction model is good and can thus provide guidance for clinical extubation.
作者
关晓楠
李艳兵
刘文婷
宗敏
马桂伶
张建军
Guan Xiaonan;Li Yanbing;Liu Wenting;Zong Min;Ma Guiling;Zhang Jianjun(Department of Cardiology,Affiliated Beijing Chaoyang Hospital of Capital Medical University,Beijing 100043,China)
出处
《中华老年心脑血管病杂志》
北大核心
2022年第3期247-250,共4页
Chinese Journal of Geriatric Heart,Brain and Vessel Diseases
关键词
呼吸功能不全
呼吸
人工
气管插管拔除
预测
respiratory insufficiency
respiration,artificial
airway extubation
forecasting
