RESUMEN
The present meta-analysis aims to assess the evidence regarding the diagnostic accuracy and performance characteristics of the colorimetric redox indicator [CRI] assay with a special emphasis on the use of the resazurin microtiter assay [REMA] for determination of primary anti-tuberculosis drug resistance. By updating previous literature searches in Medline PubMed, ISI Web, Web of Science and Google academic databases of the REMA test for determination of primary anti-tuberculosis drug resistance, this meta-analysis includes 14 studies for isoniazid [INH]; 15 studies for rifampicin [RIF]; 6 studies for streptomycin [STR]; and 5 studies for ethambutol [EMB]. SROC curve analysis was performed for meta-analysis and diagnostic accuracy was summarized.: Pooled sensitivity was 96% [94-98%] for INH, 97% [95-98%] for RIF, 92% [87-96%] for EMB and 92% [88-95%] for STR. Pooled specificity for INH, RIF, EMB and STR was 96% [95-98%], 99% [98-99%], 86% [81-89%] and 90% [87-93%], respectively. Susceptibility testing results had been obtained in 8-9 days. In conclusion, REMA seems to be a reliable test for the determination of multidrug resistant [MDR] isolates in laboratories with limited resources. However, few studies for STR and EMB have been found, and costeffectiveness studies need to be determined to recommend its widespread use