我国疟疾诊断试验准确性的Meta分析

Meta-analysis of the accuracy of malaria diagnostic tests

目的通过诊断试验Meta分析,定量综合评价我国聚合酶链反应技术(PCR)、快速诊断试剂盒(RDTs)、环介导等温扩增技术(LAMP)等疟疾诊断试验的诊断效果。方法根据检索策略,检索CNKI、万方数据、维普、PUBMED、PMC等数据库,初筛出公开发表的关于疟疾诊断试验的文献,按照纳入与排除标准,纳入符合条件的文献,进行文献质量评价并提取相关数据,采用SPSS20、RevMan5.3和MetaDiSc1.4绘制漏斗图、森林图和SROC曲线。结果有57篇文献符合纳入标准,效应量合并分析结果显示PCR法检测疟疾的综合灵敏度为0.99(0.98-0.99),特异度为0.96(0.95-0.97),SROC曲线下面积为0.994 3,诊断优势比为1056.20(408.08-2 733.68)。RDTs法检测恶性疟的综合灵敏度为0.93(0.92-0.94),特异度为0.99(0.98-0.99),SROC曲线下面积为0.995 5,诊断优势比为657.30(365.22-1 182.95);检测间日疟的综合灵敏度为0.91(0.89-0.92),特异度为0.99(0.98-0.99),SROC曲线下面积为0.990 0,诊断优势比为656.17(348.21-1 236.51)。LAMP法检测恶性疟的综合灵敏度为0.93(0.88-0.97),特异度为0.98(0.94-1.00),SROC曲线下面积为0.951 5,诊断优势比为293.72(77.27-1 116.55);检测间日疟的综合灵敏度为0.98(0.96-0.99),特异度为0.94(0.88-0.97),SROC曲线下面积为0.996 3,诊断优势比为1 072.50(237.16-4 850.09)。结论聚合酶链反应技术(PCR)、快速诊断试剂盒(RDTs)、环介导等温扩增技术(LAMP)对疟疾的诊断准确性均较高,但各有差异,其中用RDTs检测恶性疟的诊断准确性比检测间日疟高。

Objective To quantitatively evaluate the effectiveness with which a polymerase chain reaction （PCR）, rap id diagnostic kits （RDTs）, and loop-mediated isothermal amplification （LAMP） diagnosed malaria according to a meta-a- nalysis. Methods In accordance with a search strategy, CNKI, the Wan Fang database, the vip database, PubMed, and the PMC database were searched for articles about malaria diagnostic tests. In accordance with inclusion and exclusion criteria, articles that met inclusion criteria were qualitatively evaluated and relevant information was gleaned. SPSS20, RevMan5.3, and MetaDiScl. 4 were used to produce funnel plots, forest plots, and SROC curves. Results Fifty-seven articles met the inclusion criteria. A combined effect analysis indicated that PCR detected malaria with an overall sensitivity of 0.99 （0.98-0.99）, a specificity of 0.96 （0.95-0.97）, an area under the SROC curve of 0. 9943, and a diagnostic odds ratio of 1 056.20 （408.08--2 733.68）. RDTs detected P. falciparum malaria with an overall sensitivity of 0.93 （0.92-0.94）, a specificity of 0.99 （0.98-0.99）, an area under the SROC curve of 0. 9955, and a diagnostic odds ratio of 657.30 （365.22--1 182.95）. RDTs detected P. vivax malaria with an overall sensitivity of 0.91 （0. 89-0.92）, a specificity was 0.99 （0.98-0.99）, an area under the SROC curve of 0. 9900, and a diagnostic odds ratio of 656.17 （348. 21-1 236.51）. LAMP detected P. falciparum malaria with an overall sensitivity of 0.93 （0.88-0. 97）, a specificity was 0.98 （0.94-1.00）, an area under the SROC curve of 0. 9515, and a diagnostic odds ratio of 293.72 （77.27-1 116. 55）. LAMP detected P. vivace malaria with an overall sensitivity of 0.98 （0.96-0.99）, a specificity was 0.94 （0.88- 0.97）, an area under the SROC curve of 0. 9963, and a diagnostic odds ratio of 1 072.50 （237.16-4 850.09）. Conclusion PCR, RDTs, and LAMP were highly accurate at diagnosing malaria, but the techniques did different. RDTs were more accurate at detecting P. falciparum malaria than P. vivax malaria.