线性探针技术诊断耐药肺结核准确性的Meta分析

Diagnostic accuracy of line probe assays for drug-resistant tuberculosis： a Meta-analysis

目的 采用Meta分析方法评价线性探针技术对耐药肺结核诊断的准确性。方法 通过检索中文数据库（中国知网、万方数据知识服务平台、中国生物医学文献服务系统、维普期刊中文期刊服务平台）和英文数据库（PubMed、Embase、Cochrane Library）获得2000年1月1日至2017年9月1日发表的线性探针技术对中国耐药肺结核患者诊断准确性的研究。使用诊断准确性研究质量评价标准-2（QUADAS-2）进行质量评价。通过Meta分析（双变量或单变量模型）合并各研究（金标准为传统药敏试验或耐药相关基因测序）的灵敏度和特异度，然后对亚组（GenoType MTBDRplus、GenoType MTBDRsl和反向斑点杂交技术）进行分析，并进行敏感性分析。结果 最终纳入24篇文献，共包括82项研究。线性探针技术诊断耐利福平肺结核的灵敏度为0.91（0.88～0.94），特异度为0.98（0.97～0.99）；诊断耐异烟肼肺结核的灵敏度为0.80（0.77～0.83），特异度为0.98（0.96～0.99）；诊断耐多药肺结核的灵敏度为0.81（0.76～0.85），特异度为0.99（0.99～1.00）；诊断耐喹诺酮肺结核的灵敏度为0.92（0.88～0.95），特异度为0.94（0.91～0.97）；诊断耐二线注射类（包括卡那霉素、卷曲霉素、丁胺卡那霉素）肺结核的灵敏度为0.79（0.58～0.91），特异度为0.98（0.90～1.00）；诊断广泛耐多药肺结核的灵敏度为0.46（0.19～0.75），特异度为1.00（0.98～1.00）。亚组分析显示GenoType MTBDRplus和GenoType MTBDRsl诊断准确性整体上高于反向斑点杂交技术。敏感性分析显示本研究结果较为稳定。结论 线性探针技术对耐药肺结核的诊断准确性较高。

Objective To evaluate the diagnostic accuracy of line probe assays for drug-resistant tuberculosis （TB） in China. Methods Chinese databases （CNKI, Wanfang, SinoMed, VIP Information） and English databases （PubMed, Embase, Cochrane Library） were used to retrieve the literatures regarding the accuracy of line probe assays in the diagnosis of drug-resistant tuberculosis in China between January 1, 2000 and September 1, 2017. Quality Assessment of Diagnostic Accuracy Studies-2 was used to evaluate the quality of the included studies. Sensitivity and specificity in different studies （using drug sensitivity test or gene sequencing as gold standard） were combined by Meta-analysis using bivariate or univariate model. In addition, subgroup analysis （GenoType MTBDRplus, GenoType MTBDRsl and Reverse dot blot hybridization） and sensitivity analysis were also carried out. Results A total of 24 literatures involving 82 studies were included in the final analysis. The sensitivity and specificity of line probe assays for rifampicin resistant TB were 0.91（0.88-0.94） and 0.98 （0.97-0.99）, respectively. The sensitivity and specificity of line probe assays for isoniazid resistant TB were 0.80 （0.77-0.83） and 0.98 （0.96-0.99）, respectively. The sensitivity and specificity of line probe assays for multidrug-resistant TB were 0.81 （0.76-0.85） and 0.99 （0.99-1.00）, respectively. The sensitivity and specificity of line probe assays for quinolone resistant TB were 0.92（0.88-0.95） and 0.94 （0.91-0.97）, respectively. The sensitivity and specificity of line probe assays for second-line injectable drug resistant TB （including kanamycin, Capreomycin, amikacin） were 0.79（0.58-0.91） and 0.98 （0.90-1.00）, respectively. The sensitivity and specificity of line probe assays for extensively drug-resistant TB were 0.46 （0.19-0.75） and 1.00 （0.98-1.00）, respectively. Subgroup analysis showed that the overall diagnostic accuracy of GenoType MTBDRplus and GenoType MTBDRsl was higher than that of Reverse dot blot hybridization. According to the results of sensitivity analysis, the results of this study were robust. Conclusion The diagnostic accuracy of line probe assays for drug-resistant TB is high.