不同类型简化肺栓塞严重指数预测急性肺栓塞患者预后的对照研究  

A comparative study of different types of simplified Pulmonary Embolism Severity Index in predicting the prognosis of patients with acute pulmonary embolism

作　　者：夏梦梅 徐平[1,2,3] XIA Mengmei;XU Ping(Department of Emergency,the Fourth People’s Hospital of Zigong City,Zigong,Sichuan 643000,P.R.China;Sichuan Vocational College of Health and Rehabilitation,Zigong,Sichuan 643000,P.R.China;Institute of Medical Big Data,Zigong Academy of Artificial Intelligence and Big Data for Medical Science,Zigong,Sichuan 643000,P.R.China)

机构地区：[1]自贡市第四人民医院急诊科,四川自贡643000 [2]四川卫生康复职业学院,四川自贡643000 [3]自贡市医学大数据与人工智能研究院医学大数据研究所,四川自贡643000

出　　处：《华西医学》2022年第11期1618-1622,共5页West China Medical Journal

基　　金：自贡市重点科技计划项目(2018CZ10);四川卫生康复职业学院2022年院级立项课题(CWKY-2022Y-05)。

摘　　要：目的比较不同类型的简化肺栓塞严重指数(simplified Pulmonary Embolism Severity Index,sPESI)对急性肺栓塞患者预后的预测价值,从而选择最优的评分系统应用于临床。方法回顾性收集自贡市第四人民医院2014年1月1日-2019年1月1日收治的急性肺栓塞患者资料进行分析,以入院后1个月内全因死亡为观察终点。采用动脉血氧分压/吸氧浓度替换sPESI中的动脉血氧饱和度,命名为psPESI模型;采用脉搏血氧饱和度/吸氧浓度替换sPESI中的动脉血氧饱和度,命名为ssPESI模型。进行受试者操作特征曲线下面积(area under the curve,AUC)、净重新分类改善指数(net reclassification improvement,NRI)、综合判别改善指数(integrated discrimination improvement,IDI)、校准和决策曲线分析。结果共纳入280例患者,低危109例,中危155例,高危16例;男性165例(58.93%);入院后1个月全因死亡率为10.71%(30/280)。sPESI、psPESI和ssPESI的AUC分别为0.756、0.822和0.807,且ssPESI的AUC高于sPESI(P=0.038)而不低于psPESI(P=0.388)。ssPESI相对于sPESI的NRI为0.928(P<0.001),IDI为0.084(P<0.001);ssPESI相对于psPESI的NRI为0.041(P=0.227),IDI为–0.028(P=0.060)。psPESI(Hosmer-Lemeshow检验χ^(2)=12.591,P=0.182)和ssPESI(HosmerLemeshow检验χ^(2)=4.204,P=0.897)在内部验证队列中得到很好的校准,决策曲线分析提示在较宽的阈概率内ssPESI、psPESI的净受益率都比sPESI高。结论鉴于脉搏血氧饱和度为无创操作、临床易获得,且ssPESI与psPESI的预测能力相似,故推荐ssPESI作为新的评分系统应用于急性肺栓塞的预后评价。Objective To compare the prognostic value of different types of simplified Pulmonary Embolism Severity Index(sPESI)in patients with acute pulmonary embolism(APE),so as to select the best scoring system for clinical application.Methods We retrospectively collected the data of consecutive patients with APE in the Fourth People’s Hospital of Zigong City from January 1st,2014 to January 1st,2019.The endpoint was 1-month all-cause mortality.We tried to modify sPESI by replacing arterial oxyhaemoglobin saturation with arterial partial pressure of oxygen/fraction of inspired oxygen(new scoring system named psPESI),and modify sPESI by replacing arterial oxyhaemoglobin saturation with saturation of pulse oxygen/fraction of inspired oxygen(new scoring system named ssPESI),and analyzed the area under the receiver-operating characteristic curve(AUC),net reclassification improvement(NRI),integrated discrimination improvement(IDI),calibration and decision curve.Results A total of280 patients(109 with low-risk APE,155 with intermediate-risk APE,and 16 with high-risk APE)were enrolled in the study.Of these patients,165(58.93%)were male,and the 1-month all-cause mortality rate was 10.71%(30/280).The AUCs of sPESI,psPESI and ssPESI were 0.756,0.822 and 0.807,respectively,and the AUC of ssPESI was higher than that of sPESI(P=0.038)but not lower than that of psPESI(P=0.388).Comparing ssPESI with sPESI,the NRI was 0.928(P<0.001)and the IDI was 0.084(P<0.001);comparing ssPESI with psPESI,the NRI was 0.041(P=0.227)and the IDI was–0.028(P=0.060).The psPESI(Hosmer-Lemeshow testχ~2=12.591,P=0.182)and ssPESI(Hosmer-Lemeshow testχ~2=4.204,P=0.897)were well-calibrated in the internal validation cohort and obtained more net benefits within wide threshold probabilities than sPESI.Conclusion Since the saturation of pulse oxygen is non-invasive and easy to obtain,and the predictive ability of ssPESI is similar to that of psPESI,it is recommended that ssPESI be used as a new scoring system to evaluate the prognosis of APE.

关 键 词：简化肺栓塞严重指数 急性肺栓塞 动脉血氧饱和度 脉搏血氧饱和度 吸氧浓度 预后 

分 类 号：R563.5[医药卫生—呼吸系统]