气管内抽吸物细菌培养对呼吸机相关性肺炎诊断价值的荟萃分析

The role of endotracheal aspirate culture in the diagnosis of ventilator-associated pneumonia： a meta analysis

目的探讨气管内抽吸物（EA）细菌培养是否能够替代BALF定量培养用于呼吸机相关性肺炎（VAP）的微生物学诊断和指导抗菌药物的使用。方法以“呼吸机相关性肺炎”、“诊断”和“成人”为检索词，检索中国学术期刊全文数据库（CNKI）和万方数据库，以“ventilatore associated pneumonia”、“diagnosis”和“adult”为检索词检索PubMed和EMBASE数据库，收集1990--2011年发表的有关VAP微生物学诊断方法的中英文文章；以BALF定量培养作为参考标准，计算EA培养微生物学诊断VAP的阳性、假阳性、阴性和假阴性率，应用MetaDiSc 1．4软件进行荟萃分析，计算各研究中EA诊断VAP的敏感度、特异度以及总体敏感度和特异度。结果共检索到中英文文献1665篇，其中8篇文献纳入本研究，共561例患者，每例患者均同时做了EA和BALF细菌培养。如果将BALF定量培养作为参考标准，EA诊断VAP的敏感度为72％，特异度为71％。如果将EA定量和半定量培养分别进行分析，EA定量培养诊断VAP的敏感度为90％，特异度为65％，阳性预测值为68％，阴性预测值为89％；EA半定量培养诊断VAP的敏感度为50％，特异度为80％，阳性预测值为77％，阴性预测值为58％。结论EA定量培养诊断VAP的敏感度较高而特异度较低，EA半定量培养的敏感度和特异度均较低，因此EA不能替代BALF作为独立的微生物学指标诊断VAP，但具有一定的临床参考价值。

Objective To investigate the role of endotracheal aspirate （EA） culture in the diagnosis and antibiotic management in ventilator-associated pneumonia （VAP）. Methods We searched CNKI, Wanfang, PUBMED and EMBASE databases published from January 1990 to December 2011 ,to find relevant literatures on VAP microbiological diagnostic techniques including EA and bronehoalveolar lavage （BALF）. The following key words were used： ventilator associated pneumonia, diagnosis and adult. Meta- analysis was performed and the sensitivity and specificity of EA on VAP diagnosis were calculated. Results Our literature search identified 1665 potential articles, 8 of which fulfilled our selection criteria including 561 patients with paired cultures. Using BALF quantitative culture as reference standard, the sensitivity and specificity of EA were 72% and 71%. When considering quantitative culture of EA only, the sensitivity and specificity improved to 90% and 65% , while the positive and the negative predictive values were 68% and 89% respectively. However, the sensitivity and specificity of semi-quantitative culture of EA were only 50% and 80% , with a positive predictive value of 77% and a negative predictive value of 58% respectively. Conclusions EA culture had relatively poor sensitivity and specificity, although quantitative culture of EA only could improve the sensitivity. Initiating therapy on the basis of EA quantitative culture may still result in excessive antibiotic usage. Our data suggested that EA could provide some information for clinical decision but could not replace the role of BALF quantitative culture in VAP diagnosis.