摘要
目的系统评价计算机断层扫描对创伤性膈疝及膈肌损伤的诊断价值。方法检索PubMed、OVID和中文科技期刊全文数据库、维普中文科技期刊数据库,检索时间截至2008年8月。按照文内所述纳入标准收集计算机断层扫描诊断创伤性膈疝及膈肌损伤的研究文献,采用QUADAS工具进行质量评价,Metadisc1.4、RevMan4.2软件进行异质性分析同时利用随机效应模型对诊断准确性指标进行Meta分析并绘制综合受试者工作特征曲线。结果共纳入8项研究(n=728),8项研究的合并诊断比值比为45.74(95%可信区间23.39~89.44),各研究间异质性较小(P=0.25,I2=22.1%);合并灵敏度0.82(95%可信区间0.78~0.86),各研究间有中度异质性(P=0.169,I2=32.5%);合并特异度为0.92(95%可信区间0.89~0.95),各研究间有高度异质性(P=0,I2=82%);综合受试者工作特征曲线下面积0.9350,SE=0.0156,Q*=0.8712。结论综合考虑目前影响诊断的各种因素,CT仍显示较好的诊断效能,可以作为诊断创伤性膈疝及膈肌损伤的一种相对可靠而稳定的无创性检测工具。
Objective To systematically assess the diagnostic performance of CT for traumatic diaphragmatic hernia and diaphragmatic injury. Methods A comprehensive search was performed to retrieve relevant studies on CT in the diagnosis of traumatic diaphragmatic hernia and diaphragmatic injury according to the included criteria, using PubMed,OVID,CNKI and VIP up to August 2008. QUADAS as a tool was applied for the quality assessment of the included studies. MetaDiScl.4 software and RevMan4.2 were allowed for exploration of heterogeneity, implemented for meta-analysis of accuracy estimates by random effects model and performed for charting summary receiver operating characteristics (SROC). Results Eight studies were selected, with 728 patients involved. The pooled accuracy estimates for the 8 studies like diagnostic odds ratio (DOR), sensitivity and specificity were 45.74 (95%CI 23.39-89.44 ),0.82 (95%CI 0.78-0.86) and 0.92 (95% CI 0.89-0.95) respectively, with different levels of heterogeneity from P= 22.1% to /2=82%. The area under SROC curve was 0.935 0 and the Q* index was 0.871 2 (SE=0.015 6). Conclusions CT has a high diagnostic efficacy after considering all the factors which can influence the performance and it can be used as a reliable and stable non-invasive tool for diagnosing traumatic diaphragmatic hernia and diaphragmatic injury.
出处
《循证医学》
CSCD
2009年第2期91-97,共7页
The Journal of Evidence-Based Medicine