一种ROC曲线数据提取方法并用于miRNA-122诊断慢性病毒性肝炎的Meta分析

    A Method of Data Extraction and Meta-Analysis for miRNA-122 in The Diagnosis of Chronic Viral Hepatitis from Healthy

    • 摘要: 目的 提供一种数据提取方法并用于microRNA-122对慢性病毒性肝炎的诊断性Meta分析。 方法 计算机检索CBM、CNKI、VIP、EMBASE、PubMed数据库,对检索文献进行筛查,纳入符合标准的文献,并汇总文献中有关microRNA-122对丙型肝炎病毒/乙型肝炎病毒的受试者工作特征曲线诊断的结果;对于未正式报道敏感性数值和特异性数值的文献,采用数据提取软件GetData Graph Digitizer 2.24获取文献中相关敏感性和特异性数值,用于microRNA-122鉴别健康人和慢性肝炎病毒(乙型/丙型)感染患者,通过Stata12软件等进行分析,结果以敏感性、特异性以及对受试者工作特征曲线进行汇总。 结果 最终纳入6篇文献,其中3篇是关于丙型肝炎病毒感染患者的诊断,另外3篇为乙型肝炎病毒感染患者的诊断;共有321名慢性病毒性肝炎感染者(包括112名丙型肝炎患者、209名乙型肝炎患者)以及191名健康对照者被纳入研究;应用我们提取数据的方法表明,通过软件提取的敏感性和特异性数据与原文数据保持一致。数据提取汇总后的敏感性、特异性分别为0.91(95%可信区间 0.83~0.96)、0.87(95%可信区间0.80~0.92),受试者工作特征曲线下面积为0.94(95%可信区间0.92~0.96)。 结论 以GetData Graph Digitizer软件创建的提取受试者工作特征曲线中数据的方法具有很好的真实性和可重复性,以提取的数据汇总分析表明,血清microRNA-122可很好地用于鉴别健康人与慢性肝炎感染患者,并且对慢性乙型肝炎病毒感染的诊断具有更高的灵敏性。

       

      Abstract: Objective To provide method of data extraction and a meta-analysis for microRNA-122 in the diagnosis of chronic viral hepatitis from healthy. Methods Combination of computer and manual was used to search for literatures in CBM、CNKI、VIP、EMBASE、PubMed database, and the key words HCV, HBV and microRNA-122 were related to inclusion serum levels of microRNA-122 for HBV/HCV. And the software of GetData Graph Digitizer 2.24 was exploited to extract data. Values of sensitivity, specificity and symmetrical receiver operating characteristic curvefor microRNA-122 in the diagnosis of chronic viral hepatitis B/C from healthy were analyzed by stata12 software. Result Six articles were included, and three for diagnosis healthy from HCV, and three for healthy people from HBV. Which involved 321 chronic viral hepatitis patients (112 HCV patients and 209 HBV patients) and 191 healthy controls in our study; The abstracted data of sensitivity and specificity were pooled 0.91 (95%CI 0.83~0.96), 0.87(95%CI 0.80~0.952) respectively, the area under roc curve of symmetrical receiver operating characteristic curve was 0.93(95%CI 0.90~0.95). Conclusion We created a repeatability and facticity method of using GetData Graph Digitizer software to extract the date of receiver operating characteristic curve,the extracted data were pooled in our meta-analysis. Meta-analysis showed that serum microRNA-122 can be perfect bio-marker in the diagnosis of chronic viral hepatitis from healthy,especially in the diagnosis of HBV from healthy.

       

    /

    返回文章
    返回