【作者】 邹俊辉；

【导师】 肖祖克；

【作者基本信息】 南昌大学 ， 内科学， 2010， 硕士

【摘要】 目的应用系统分析方法评价当前酶联免疫斑点技术快速诊断结核病研究文献的质量。评价酶联免疫斑点技术快速诊断结核病的临床价值。方法通过检索PubMed、EMBASE、CNKI等数据库和其他方式收集文献。获得酶联免疫斑点技术诊断结核病的文献,筛选文献、提取数据和文献质量评价由由两位研究者分别独立完成。采用Meta-DiSc 1.4版软件检验研究间的异质性,对诊断准确度指标进行系统分析,计算合并灵敏度、合并特异度、合并似然比和ROC曲线。并依据QUADAS条目中对纳入的文献进行分析。结果最终有8篇文献符合设定的纳入标准,共涉及确诊结核病患者492例,非结核感染对照组人数769例。文献在偏倚方面控制较好,偏倚主要来自参考标准结果解释时未实施盲法,变异和报告质量较差。异质性检验无阈值效应,但存在其他异质性。Meta分析结果显示酶联免疫斑点技术诊断结核病的总灵敏度是94.0% [95% CI(92.0%-96.0%)],总特异度为62.0% [95% CI(58.0%-65.0%)],总的诊断优势比为39.18 [95% CI(13.48-113.91)],合并阳性似然比为3.30[95% CI(1.93-5.65)],合并阴性似然比为0.10[95% CI(0.05-0.17)],综合受试者工作曲线下面积为0.9785,Q值为0.9345。结论本研究显示:酶联免疫斑点技术是快速诊断结核病的较好方法,但是由于诊断试验研究的严密性对结果的影响较大,从所纳入的文献来看,其临床试验方法学质量不高,存在受试对象构成不同,盲法使用不足,报告方式不够规范,未进行临床决策分析等多种方法学局限性,限制其在临床上广泛应用。下一步研究的重点是依据STARD标准设计更为严格的试验,并结合卫生经济学展开经济-效果比研究,为指导酶联免疫斑点技术检查方法提供更为全面、可靠的资料。更多还原

【Abstract】 Objective To evealuate the quality of the current studies on the value of the diagnosis of enzyme link immunospot (ELISPOT) assay for rapid detection of tuberculosis. To estimate the accuracy of diagnosting tuberculosis using ELISPOT assay.Methods PubMed, EMBASE, CNKI were searched to retrieve relevant studies on detecting tuberculosis using ELISPOT assay. Articles selection and data extraction were carried out by two reviewers independently. Meta-disc software was used to handle data from included studies. The sensitivity (SEN), specificity (SPE),likelihood ratio(LR )and summary receiver operating characteristics (SROC) were used to assess the diagnostic value of ELISPOT assay. The included studies were analyzed by QUADAS.Results Eight studies were included according to the criteria, which included 492 patients with tuberculosis and 769 samples of non-tuberculosis as control.These eight studies had well controlled the bias of partial verification , differential verification , incorporation and withdrawals .The reference standard review could have a greater bias. The reporting quality were poorly reported. The heterogeneity (except for threshold effect) was found among these studies. The summarized accuracy indicators like sensitivity, specificity, and diagnostic odds ratio (DOR) were 94.0% [95% CI(92.0% -96.0%)], 62.0% [95% CI(58.0%-65.0%)],and 39.18 [95% CI(13.48-113.91)] respectively.The positive likelihood ratio(PLR)and negative likelihood ratio (NLR) were 3.30[95% CI(1.93-5.65)] and 0.10[95% CI(0.05-0.17)]. The area under of SROC curve was 0.9785 . The Q index was 0.9345.Conclusion The study demonstrated that the ELISPOT assay possibilities is a relatively sensitive method for rapid detection of tuberculosis. Due to the limitations such as bad design, low quality of experimental methods, different subjects, insufficient blind method, inconsistent reported standard, and not carrying on clinical decision analysis et al, the popularization of the new method is restricted. In view of these facts, as a new technology, the ELISPOT assay needs to be further tested by studies designed based on the STARD. In order to supply more general and reliable data, it is necessary to examine the economic effects of ELISPOT assay.更多还原