| 多个评估系统对社区获得性肺炎严重度评估的荟萃分析 |
| |
| 引用本文： | 冯耘,程挺,刘嘉琳,万欢英,程齐俭. 多个评估系统对社区获得性肺炎严重度评估的荟萃分析[J]. 诊断学理论与实践, 2016, 15(6): 586-594. DOI: 10.16150/j.1671-2870.2016.06.009 |
| |
| 作者姓名： | 冯耘 程挺 刘嘉琳 万欢英 程齐俭 |
| |
| 作者单位： | 上海交通大学医学院附属瑞金医院呼吸科,上海 200025 |
| |
| 基金项目： | 上海申康医院发展中心,郊区医院临床能力建设项目（SHDC12015901） |
| |
| 摘 要： | 目的: 通过分析已有的临床数据,观察多个评估系统对社区获得性肺炎（community acquired pneumonia,CAP）患者30 d的死亡风险和ICU入住风险的预测价值。方法: 用关键词途径对2004年至2015年的PubMed和EMBASE数据库进行文献检索,并对符合条件的所有研究进行荟萃分析。结果: 共纳入文献48篇,共51 639例CAP患者采用各评价系统对其30 d死亡风险进行评估,共评估了10 590例CAP患者入住ICU的风险。通过汇总计算得出,对于30 d死亡风险的预测,CRB-65（confusion,respiratory,age 65 years）评分系统的灵敏度为98%,特异度为33%,综合受试者工作特征（summary receiver operating characteristic,sROC）曲线下面积为0.56;英国胸科协会改良肺炎评分(confusion, uremia, respiratory,blood pressure, age 65 years, CURB-65评分)系统的灵敏度为84%,特异度为55%,sROC曲线下面积为0.78;肺炎严重度指数(pneumonia severity index, PSI)系统的灵敏度为90%,特异度为57%,sROC曲线下面积为0.88;以符合2007美国感染病学会(Infectious Diseases Society of America,IDSA) /美国胸科学会（American Thoracic Society, ATS) 3项以上次要标准且不符合主要标准的评估灵敏度为76%,特异度为90%,sROC曲线下面积为0.89;采用SMART-COP评估的灵敏度为77%,特异度为65%,sROC曲线下面积为0.67。对于入住ICU风险,采用CURB-65系统评估的灵敏度为52%,特异度为77%,sROC曲线下面积为0.67;采用PSI系统评估的灵敏度为70%,特异度为61%,sROC曲线下面积为0.69;采用SMART-COP评分评估的 灵敏度为84%,特异度为70%,sROC曲线下面积为0.74。结论: PSI、CURB-65评分系统对患者30 d死亡风险的预测价值较大,可作为住院标准的重要参考。对于收住ICU的风险评估,SMART-COP的预测价值相对较大,而PSI、CURB-65评分价值相对较低。
|
| 关 键 词： | 社区获得性肺炎 预测价值 荟萃分析 |
| 收稿时间： | 2016-05-30 |
Validation and meta-analysis of various severity scale system for predicting mortality and ICU admission in community-acquired pneumonia patients |
| |
| FENG Yun,CHENG Ting,LIU Jialin,WAN Huanying,CHENG Qijian. Validation and meta-analysis of various severity scale system for predicting mortality and ICU admission in community-acquired pneumonia patients[J]. Journal of Diagnostics Concepts & Practice, 2016, 15(6): 586-594. DOI: 10.16150/j.1671-2870.2016.06.009 |
| |
| Authors: | FENG Yun CHENG Ting LIU Jialin WAN Huanying CHENG Qijian |
| |
| Affiliation: | Department of Pulmonary Disease, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200025, China |
| |
| Abstract: | Objective: To review and analyze the value of various severity scale system for predicting 30 day mortality and ICU admission in community acquired pneumonia (CAP) patients via meta-analysis. Methods: MEDLINE and EMBASE (2004-2015) were retrieved and 48 papers were collected. Altogether 51 639 CAP patients were enrolled for predicting 30 day mortality, and 10 590 CAP patients were enrolled for predicting ICU admission. The values of five severity scale system: CRB-65, CURB-65, PSI, 2007 IDSA/ATS, SMART-COP for predicting 30 day mortality and ICU admission were meta-analyzed. Results: For predicting 30 day mortality, CRB-65 had a sensitivity of 98%; specificity of 33% and area under summary receiver operator characteristic (sROC) curve (AUC) 0.56. For CURB-65, the above mentioned indices were 84%, 55% and 0.78, respectively. PSI(pneumonia severity index) had a sensitivity of 90%, specificity of 57% and area under sROC curve 0.88; the above mentioned indices for 2007 IDSA/ATS were 76%, 90% and 0.89, respectively. For predicting ICU admission. CURB-65 had a sensitivity of 52%, specificity of 77% and area under sROC 0.67; the above mentioned indices for PSI were 70%, 61%and 0.69, respectively. The SMART-COP had a sensitivity of 84%, specificity of 70%, and area under sROC curve 0.74. Conclusions: PSI, CURB-65 could be used as important references for predicting 30 day mortality in patients with CAP. For ICU admission, the prediction value of SMART-COP is relatively better than PSI and CURB-65 scale systems. |
| |
| Keywords: | Community-acquired pneumonia Prediction value Meta-analysis |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《诊断学理论与实践》浏览原始摘要信息 |
|
点击此处可从《诊断学理论与实践》下载全文 |
|
