ICU患者人工气道相关压力性损伤风险列线图预测模型的构建  

Construction of a nomogram prediction model for the risk of pressure-related injury associated with artificial airway in ICU patients

作　　者：周祥龙 张国琴[2] 张帅 姚惠萍[3] 陈小芳[4] 蒋雪梅 高丽 陈群 ZHOU Xianglong;ZHANG Guoqin;ZHANG Shuai;YAO Huiping;CHEN Xiaofang;JIANG Xuemei;GAO Li;CHEN Qun(The Third People's Hospital of Taizhou,Taizhou,Jiangsu 225321;Taizhou People's Hospital,Taizhou 225300,Jiangsu,China;Zhejiang People's Hospital,Hangzhou 310000,Zhejiang,China;Huzhou First People's Hospital,Huzhou 313099,Zhejiang,China)

机构地区：[1]泰州市第三人民医院,江苏泰州225321 [2]泰州市人民医院,江苏泰州225300 [3]浙江省人民医院,浙江杭州310000 [4]湖州市第一人民医院,浙江湖州313099

出　　处：《护士进修杂志》2023年第14期1254-1260,共7页Journal of Nurses Training

基　　金：浙江省医药卫生科技项目(编号:2020369770)。

摘　　要：目的探讨ICU患者人工气道相关压力性损伤的危险因素并构建预测风险列线图模型。方法选取2020年12月-2021年7月,浙江省某三级甲等医院ICU收治的留置人工气道患者650例,将压力性损伤发生组(n=131)和未发生压力性损伤组(n=519)的各项指标进行对比,通过单因素分析和多因素logistic回归分析,探索ICU患者人工气道相关压力性损伤发生的独立危险因素,构建列线图预测模型,应用Bootstrapping法评估模型预测效果。结果多因素logistic回归分析显示:人工气道的留置时间(OR=1.035)、使用声门下吸引导管(OR=2.878)、气管导管在口中的位置(OR=4.937)、插管时潮湿评分(OR=0.215)、插管时移动力评分(OR=0.225)、插管时摩擦力/剪切力评分(OR=0.170)是ICU患者人工气道相关压力性损伤发生的独立危险因素。根据上述6项危险因素构建的ICU患者人工气道相关压力性损伤风险列线图预测模型,一致性指数为0.842(95%CI为80.4%~87.9%),灵敏度为77.9%,特异度为78.4%,最大约登指数为0.563,对应的风险预测值为0.170,列线图预测临界值是177分。内部验证结果,一致性指数为0.835,灵敏度为79.2%,特异度为75.3%,模型Brier值为0.112,校正曲线表明,预测曲线和实际观察曲线基本吻合,决策曲线分析表明,模型有良好的临床获益。结论本研究构建的列线图能有效预测ICU患者人工气道相关压力性损伤的发生,具有一定临床指导意义。Objective To investigate the risk factors of pressure injury associated with artificial airway in ICU patients and construct a columnar graph model to predict the occurrence of pressure injury.Methods Six hundred and fifty patients with an indwelling artificial airway admitted to the ICU of a tertiary care hospital in Zhejiang province were selected,and the indicators of the group with pressure injury(n=131)and the group without pressure injury(n=519)were compared to explore the independent risk factors for the occurrence of pressure injury associated with artificial airway in ICU patients through single-factor analysis and multi-factor logistic regression analysis.The prediction model of column line graph was constructed and the predictive effect of the model was assessed by Bootstrapping method.Results Multi-factor logistic regression analysis showed that the duration of retention of the artificial airway(OR=1.035),the use of a subsonic suction catheter(OR=2.878),the position of the tracheal tube in the mouth(OR=4.937),the humidity rating in the Braden score during intubation(OR=0.215),the mobility score(OR=0.225),the friction and shear force score(OR=0.170)were independent risk factors for the occurrence of pressure-related injuries associated with the artificial airway.A line graph prediction model for the risk of pressure-related injury associated with an artificial airway in ICU patients,constructed from the six risk factors mentioned above,had a consistency index of 0.842(95%CI was 80.4%to 87.9%)and a sensitivity of 77.9%.The specificity was 78.4%and the maximum Jorden index was 0.563,corresponding to a risk prediction value of 0.170 and a threshold value of 177 points for columnar graph prediction.The internal validation results,with a consistency index of 0.835,sensitivity of 79.2%,specificity of 75.3%,and a model Brier value of 0.112,showed that the prediction curve and the actual observed curve largely matched,and the decision curve analysis indicated that the model had good clinical benefit.Conclusion

关 键 词：重症患者 人工气道 医疗器械相关压力性损伤 风险预测模型 列线图 

分 类 号：R473[医药卫生—护理学]